WO2024214220A1 - Load torque calculation device, encoder system, actuator, robot device, load torque calculation method, and program - Google Patents

Abstract

This load torque calculation device calculates a load torque value of an actuator comprising an input shaft, a motor that rotates the input shaft, a transmission to which power of the motor is input from the input shaft, an output shaft that rotates with the power output from the transmission, a first encoder that detects first rotation information about the input shaft, and a second encoder that detects second rotation information about the output shaft, the load torque calculation device comprising a computing unit that calculates a first load torque value of the actuator on the basis of the first rotation information, the second rotation information, and a first spring constant stored in a storage unit.

Description

LOAD TORQUE CALCULATION DEVICE, ENCODER SYSTEM, ACTUATOR, ROBOT DEVICE, LOAD TORQUE CALCULATION METHOD, AND PROGRAM

The present invention relates to a load torque calculation device, an encoder system, an actuator, a robot device, a load torque calculation method, and a program.

Actuators with reducers that are equipped with an encoder that detects the rotational position of the output member of the reducer are known (for example, Patent Document 1).

JP 2019-060477 A

One aspect is a load torque calculation device that calculates a load torque value of an actuator that includes an input shaft, a motor that rotates the input shaft, a transmission to which the power of the motor is input from the input shaft, an output shaft that rotates with the power output from the transmission, a first encoder that detects first rotation information of the input shaft, and a second encoder that detects second rotation information of the output shaft, and that includes a calculation unit that calculates the first load torque value of the actuator based on the first rotation information, the second rotation information, and a first spring constant stored in a storage unit.

One aspect is a load torque calculation device that calculates a load torque value of an actuator that includes an input shaft, a motor that rotates the input shaft, a transmission to which the power of the motor is input from the input shaft, an output shaft that rotates with the power output from the transmission, a first encoder that detects first rotation information of the input shaft, and a second encoder that detects second rotation information of the output shaft, and that includes a calculation unit that calculates a second load torque value of the actuator based on a motor drive current that drives the motor and a torque constant stored in a storage unit.

One aspect is an encoder system that includes the above-mentioned load torque calculation device, the first encoder, and the second encoder.

One aspect is an actuator that includes the above-mentioned encoder system, the input shaft, the motor, the transmission, and the output shaft.

One aspect is a robot device equipped with the above actuator.

One aspect is a load torque calculation method for calculating a load torque value of an actuator that includes an input shaft, a motor that rotates the input shaft, a transmission to which the power of the motor is input from the input shaft, an output shaft that rotates with the power output from the transmission, a first encoder that detects first rotation information of the input shaft, and a second encoder that detects second rotation information of the output shaft, the load torque calculation method including: calculating a first load torque value of the actuator based on the first rotation information, the second rotation information, and a first spring constant stored in a storage unit; and calculating a second load torque value of the actuator based on a motor drive current that drives the motor and a torque constant stored in the storage unit.

One aspect is a program that causes a computer to calculate a load torque value of an actuator that includes an input shaft, a motor that rotates the input shaft, a transmission to which the power of the motor is input from the input shaft, an output shaft that rotates with the power output from the transmission, a first encoder that detects first rotation information of the input shaft, and a second encoder that detects second rotation information of the output shaft, and causes the computer to execute the following operations: calculate a first load torque value of the actuator based on the first rotation information, the second rotation information, and a first spring constant stored in a storage unit; and calculate a second load torque value of the actuator based on a motor drive current that drives the motor and a torque constant stored in the storage unit.

FIG. 2 is a diagram illustrating an example of a functional configuration of an actuator according to the first embodiment. FIG. 2 is a diagram showing an example of a cross-sectional view of a side surface of the actuator according to the first embodiment. 5 is a diagram showing an example of the relationship between the rotation speed of the motor according to the first embodiment and the amount of change in each of the torsion amount and the effective current value. FIG. 5 is a diagram showing an example of a relationship between the elapsed time after the motor according to the first embodiment is driven and the amount of change in each of the torsion amount and the effective current value. FIG. FIG. 2 is a diagram illustrating an example of a functional configuration of a calculation unit according to the first embodiment. FIG. 11 is a diagram illustrating an example of a functional configuration of a calculation unit according to a modified example of the first embodiment. FIG. 11 is a diagram illustrating an example of a functional configuration of a calculation unit according to a modified example of the first embodiment. FIG. 11 is a diagram illustrating an example of a functional configuration of a calculation unit according to a modified example of the first embodiment. FIG. 4 is a diagram illustrating an example of a self-calibration process according to the first embodiment. FIG. 11 is a diagram illustrating an example of a relationship between an actual load torque and a first load torque calculated based on an amount of torsion according to the second embodiment. FIG. 11 is a diagram illustrating an example of a relationship between an actual load torque and a second load torque calculated based on a motor drive current according to the second embodiment. FIG. 11 is a diagram illustrating an example of a functional configuration of a calculation unit according to a second embodiment. FIG. 11 is a diagram illustrating an example of a torque calculation process according to the second embodiment. FIG. 13 is a diagram illustrating an example of a functional configuration of a calculation unit according to a third embodiment. FIG. 13 is a diagram illustrating an example of a relationship between a torsion amount and a load torque according to the third embodiment. FIG. 13 is a diagram illustrating an example of an error of an approximation equation of a load torque with respect to an actual measured value according to the third embodiment. FIG. 13 is a diagram illustrating an example of a torque calculation process according to the third embodiment. FIG. 13 is a diagram illustrating an example of a correction process according to the third embodiment. FIG. 13 is a diagram illustrating an example of a correction process according to the third embodiment. FIG. 13 is a diagram showing an example of the relationship between the amount of twist and the load torque for two different spring constants when pressurized and when depressurized according to the third embodiment. FIG. 13 is a diagram showing an example of actual measured values of load torque with respect to the amount of twist according to the third embodiment. FIG. 13 is a diagram illustrating an example of a value of a first load torque calculated from an amount of torsion by using a third-order polynomial based on coefficients generated by a correction unit according to the third embodiment. FIG. 13 is a diagram illustrating an example of a relationship between the amount of twist and a third-order coefficient of a third-order polynomial according to the third embodiment. FIG. 13 is a diagram illustrating an example of approximation to a third-order coefficient according to the third embodiment. FIG. 13 is a diagram illustrating an example of a relationship between the amount of twist and a second-order coefficient of a third-order polynomial according to the third embodiment. FIG. 13 is a diagram illustrating an example of approximation to a second-order coefficient according to the third embodiment. FIG. 13 is a diagram illustrating an example of a relationship between the amount of twist and a first-order coefficient of a third-order polynomial according to the third embodiment. FIG. 13 is a diagram illustrating an example of approximation to a first-order coefficient according to the third embodiment. FIG. 13 is a diagram illustrating an example of a relationship between the amount of twist and the zeroth-order coefficient of a third-order polynomial according to the third embodiment. FIG. 13 is a diagram illustrating an example of approximation to a zero-order coefficient when the amount of twist according to the third embodiment is positive. FIG. 13 is a diagram illustrating an example of approximation to a zero-order coefficient when the amount of twist according to the third embodiment is negative. FIG. 13 is a diagram showing an example of a spring constant reproduced by calculating coefficients of a third-order polynomial using an approximation formula according to the third embodiment. FIG. 33 is a diagram showing an example in which the graph shown in FIG. 32 according to the third embodiment is enlarged around the fourth quadrant. FIG. 13 is a diagram illustrating an example of a functional configuration of a calculation unit according to a fourth embodiment. FIG. 13 is a diagram illustrating an example of self-calibration and torque calculation processing according to the fourth embodiment.

(First embodiment)
Hereinafter, the first embodiment will be described in detail with reference to the drawings. Fig. 1 is a diagram showing an example of a functional configuration of an actuator 1 according to the present embodiment. Fig. 2 is a diagram showing an example of a cross-sectional view of a side surface of the actuator 1 according to the present embodiment.

First, an actuator 1 according to the present embodiment will be described with reference to FIG. 1 and FIG.
The actuator 1 includes a motor 2, a reducer (transmission) 3, an input shaft encoder 4, an output shaft encoder 5, a driver 6, a temperature sensor 10, an input shaft 31, and an output shaft 32. The actuator 1 includes two encoders, the input shaft encoder 4 and the output shaft encoder 5, so that the torque can be measured with high accuracy, as described below. The torque is generated when driving a driven part (load) connected to the actuator 1, and is also referred to as load torque. The load torque can also be said to be a torque applied between the output shaft 32 and the input shaft 31 when driving a driven part connected to the actuator 1. In other words, the load torque can also be said to be a required power. As an example, the output shaft 32 of the reducer 3 may be connected to an arm of a robot device to drive the arm. In this case, the load torque refers to the torque generated in the output shaft 32 of the reducer 3 when driving the arm of the robot device. In the following, the terms "torque", "load torque", and "rotational force" may be used interchangeably.

The motor 2 drives the input shaft 31 to rotate. As an example, the motor 2 is an electric motor. As an example, the motor 2 is a brushless motor. The motor 2 includes a magnet as a rotor, a stator as a fixed part, and a power supply unit. The stator includes a coil and generates a magnetic field by power supplied from the outside via the power supply unit. The rotor, which is a magnet, rotates by the force it receives from the magnetic field generated by the coil included in the stator. Note that the motor 2 may be an electric motor of a type other than a brushless motor.
The input shaft 31 is connected to the rotor of the motor 2 and rotates together with the rotor. The input shaft 31 is hollow and cylindrical, and the output shaft 32 passes through the inside of the input shaft 31. One end (first end) of the input shaft 31 is connected to a connection portion 301 on the input side of the reduction gear 3. One end (first end) of the output shaft 32 is connected to a connection portion 302 on the output side of the reduction gear 3.

The reducer 3 is a power transmission device that reduces the rotation of the input shaft 31 and transmits it to the output shaft 32. In other words, the reducer 3 reduces (changes) the rotation of the rotor of the motor 2 transmitted from the connection part 301 on the input side, and outputs the power from the connection part 302 on the output side. In other words, the reducer 3 reduces the power of the rotor of the motor 2 transmitted from the input shaft 31 at a predetermined reduction ratio, and outputs the reduced power from the output shaft 32. One end (first end) of the output shaft 32 is connected to a driven part (load). The driven part is driven by the power output from the reducer 3. The reducer 3 includes at least one of one or more gears, a belt device, a chain device, and a drive shaft device. The reducer 3 may be a transmission. A transmission is a device that changes the ratio of rotational speeds. A transmission is capable of changing speed.

As shown in FIG. 2, as an example, the input shaft encoder 4 is disposed on the opposite side of the motor 2 from the reducer 3. The input shaft encoder 4 detects position information of the input shaft 31 (which may be rotation information, rotation position information, angle information, etc.). The position information of the input shaft 31 is transition information including the rotation angle, angular velocity, and/or multi-rotation information indicating how many times the input shaft 31 has rotated.

The input shaft encoder 4 is, for example, an optical encoder. The input shaft encoder 4 includes a first rotating portion (which may be described as a disk, a scale plate, etc.) 41 and a first detection portion 42. The first rotating portion 41 is provided at the other end (second end) of the input shaft 31. The other end (second end) of the input shaft 31 is opposite to the one end (first end) connected to the connection portion 301 on the input side of the reducer 3. The first rotating portion 41 is provided with a first scale (which may be described as a pattern). The first scale is formed with a pattern that changes in the circumferential direction of the input shaft 31. A specific example of the first scale is at least an absolute type, and may be other types. For example, the first scale may be a reflective type that reflects light from the first detection portion 42, or a transmissive type that transmits light from the first detection portion 42.

The first detection unit 42 optically detects the pattern formed on the first scale. The input shaft encoder 4 detects position information of the input shaft 31 from the pattern on the first scale detected by the first detection unit 42. The input shaft encoder 4 can communicate with the calculation unit 8 via wire or wirelessly, and supplies position information of the input shaft 31 to the calculation unit 8.

As shown in FIG. 2, as an example, the output shaft encoder 5 is disposed on the opposite side of the motor 2 from the reducer 3. The output shaft encoder 5 detects position information (which may be rotation information, rotation position information, angle information, etc.) of the output shaft 32. The position information of the output shaft 32 is transition information including the rotation angle, angular velocity, and/or multi-rotation information indicating how many times the output shaft 32 has rotated.

The output shaft encoder 5 is, for example, an optical encoder. The output shaft encoder 5 includes a second rotating section (which may be described as a disk, a scale plate, etc.) 51 and a second detection section 52. The second rotating section 51 is provided at the other end (second end) of the output shaft 32. The second end is the other end of the output shaft 32 on the side different from the first end described above. The second rotating section 51 is provided with a second scale (which may be described as a pattern). The second scale is formed with a pattern that changes in the circumferential direction of the output shaft 32. Specific examples of the second scale include at least an absolute type, and may be other types. For example, the second scale may be a reflective type that reflects light from the second detection section 52, or a transmissive type that transmits light from the second detection section 52.

The second detection unit 52 optically detects the pattern formed on the second scale. The output shaft encoder 5 detects position information of the output shaft 32 from the pattern on the second scale detected by the second detection unit 52. The output shaft encoder 5 can communicate with the calculation unit 8 via a wired or wireless connection, and supplies position information of the output shaft 32 to the calculation unit 8.

In the present embodiment, as an example, a case will be described in which the input shaft encoder 4 and the output shaft encoder 5 communicate with the calculation unit 8 via wires. In the present embodiment, as an example, the input shaft encoder 4, the output shaft encoder 5, and the calculation unit 8 can communicate with each other via a bus connection.
As described above, when the input shaft encoder 4 and the calculation unit 8 communicate with each other via wireless communication, or when the output shaft encoder 5 and the calculation unit 8 communicate with each other via wireless communication, short-range wireless communication is preferably used for the wireless communication. Examples of short-range wireless communication include short-range wireless communication using BLE (Bluetooth Low Energy (registered trademark)), short-range wireless communication using NFC (Near Field Communication), and short-range wireless communication using infrared communication.

In this embodiment, the position information supplied from the input shaft encoder 4 and the output shaft encoder 5 is referred to as rotation angle information B1. The rotation angle information B1 is information indicating the input shaft rotation angle θ1 (rotation angle of the input shaft 31) detected by the input shaft encoder 4, and the output shaft rotation angle θ2 (rotation angle of the output shaft 32) detected by the output shaft encoder 5.

The actuator 1 includes a temperature sensor 10. The temperature sensor 10 measures the temperature of the actuator 1. In this embodiment, the temperature of the actuator 1 is, for example, the temperature of the motor 2. That is, the temperature sensor 10 measures the temperature of the motor 2. The temperature of the actuator 1 may be the temperature of any of the parts constituting the actuator 1, or may be the average temperature of those parts. The temperature sensor 10 of the actuator 1 according to this embodiment outputs, for example, information indicating the temperature of the coil wound around the stator of the motor 2 (motor coil temperature information C1). Here, the temperature sensor 10 is, for example, installed near the coil. The motor coil temperature information C1 is generated by detecting the temperature of the coil by the temperature sensor 10. The generated motor coil temperature information C1 is supplied from the temperature sensor 10 to the motor control unit 7 and the calculation unit 8.

The driver 6 controls the entire actuator 1. The driver 6 includes a motor control unit 7, a calculation unit 8, and a memory unit 9. For example, the driver 6 is one or more electronic components (such as an integrated circuit) in which the motor control unit 7, the calculation unit 8, and the memory unit 9 are provided on a substrate. A detailed description of the driver 6 will be given later.

The motor control unit 7 controls the motor 2 based on control (commands, upper control commands) from an external control device (upper control device). The motor control unit 7 includes a control signal generation unit 71 and a motor driver 72.

The control signal generating unit 71 generates a control signal based on the higher-level control command, the control power supply, the rotation angle information B1, and the motor coil temperature information C1. The control power supply supplies the motor control unit 7 and the calculation unit 8 with power for their respective operations.

The motor driver 72 generates a motor drive current from power supplied from the power source based on a control signal for controlling the motor 2 supplied from the control signal generating unit 71. The power source supplies power for driving the motor 2 to the motor driver 72. The control signal is a signal for controlling at least one of the angular position, angular velocity, and angular acceleration of the rotor of the motor 2. The motor driver 72 controls (drives/rotates/rotationally drives) the motor 2 by supplying the generated motor drive current to the motor 2. In other words, the motor driver 72 generates motor drive current information A1 indicating the value (effective current value/current value) of the motor drive current to be supplied to the motor 2, and supplies the generated motor drive current information A1 to the calculation unit 8.

The calculation unit 8 performs various calculations (for example, calculation of the torque applied to the actuator 1). More specifically, the calculation unit 8 generates torque information E1 and a spring constant. The detailed operation of the calculation unit 8 will be described later. The torque information E1 is generated by the calculation unit 8 based on the rotation angle information B1 and the like. The calculation unit 8 includes a calculation circuit that performs calculations in a hardware manner, such as an ASIC (Application Specific Integrated Circuit). The calculation unit 8 may be realized by a computer including a CPU (Central Processing Unit) and memory executing processing according to a program.

The memory unit 9 stores various types of information. The memory unit 9 stores a spring constant K1 and a torque constant Kt. The spring constant K1 is a value related to the torsional rigidity (also written as torsional rigidity) in the rotational direction of the reducer 3. The torsional rigidity is a value that represents the stiffness against twisting. The torque constant Kt is used to calculate the second load torque T2 from the motor drive current. The memory unit 9 includes, for example, a rewritable non-volatile memory.

Here, the reducer 3 has spring characteristics. Due to the spring characteristics of the reducer 3, when the input side is fixed and torque (also referred to as load or load torque) is applied to the output side, a twist proportional to the load torque is generated on the output side. In this embodiment, an input shaft 31 is connected to the input side of the reducer 3, and an output shaft 32 is connected to the output side of the reducer 3.
Specifically, when no load torque is applied to the output side of the reducer 3 (no load), no torsion occurs in the reducer 3. When a load torque is applied to the output side of the reducer 3, torsion occurs on the output side of the reducer 3.

In other words, in the reducer 3, torsion occurs due to the load torque applied to the output side. Therefore, the load torque can be calculated from the magnitude of torsion (amount of torsion) of the reducer 3. More specifically, in this application, encoders are attached to the input side (input shaft 31) and output side (output shaft 32) of the reducer 3. Therefore, the amount of torsion of the reducer can be obtained from the difference value of the position information obtained from the input shaft encoder 4 and the output shaft encoder 5. The magnitude of torsion is indicated by the torsion angle α1 in the rotational direction of the reducer 3.

The input shaft rotation angle θ1, which is the rotation angle of the input shaft 31, is detected by the input shaft encoder 4. The output shaft rotation angle θ2, which is the rotation angle of the output shaft 32, is detected by the output shaft encoder 5. The calculation unit 8 calculates the torsion angle α1 in the rotation direction of the reducer 3 based on the difference between the input shaft rotation angle θ1 and the output shaft rotation angle θ2. The calculation unit 8 calculates the load torque by multiplying the calculated torsion angle α1 by the spring constant K1 of the reducer 3. The load torque is the load torque in the rotation direction of the reducer 3 applied between the input shaft 31 and the output shaft 32. The load torque may also be considered as the torque loaded on the actuator 1.

The spring constant K1 is a value related to the torsional rigidity (torsional stiffness) in the rotational direction of the reducer 3. The spring constant K1 is stored in the memory unit 9. Here, the torsional rigidity in the rotational direction of the reducer 3 represents the stiffness (resistance) of the output side against the load torque.

The torsional rigidity of the reducer 3 in the rotational direction may change depending on various conditions under which the actuator 1 is used. For example, the torsional rigidity of the reducer 3 in the rotational direction may change as the actuator 1 is used. In other words, the torsional rigidity of the reducer 3 in the rotational direction may change due to changes over time.

Fig. 3 is a diagram showing an example of the relationship between the rotation speed of the motor 2 according to this embodiment and the amount of change in the amount of torsion and the effective current value, and Fig. 4 is a diagram showing an example of the change in the amount of torsion and the effective current value over time after the motor 2 according to this embodiment is driven.
The effective current value is the value of the current (motor drive current) for driving the motor 2. As shown in Figures 3 and 4, the effective current value changes in a short period of time depending on the rotation speed, but does not change with the elapsed time.
3 and 4, the amount of torsion (torsion angle α1) does not change with the rotation speed, but changes with the elapsed time. This is believed to be due to a change in the torsional rigidity of the reducer 3 caused by the actuator 1 being continuously operated for a long period of time.

As a result, the value of the spring constant K1 changes according to changes in the torsional rigidity of the reducer 3 in the rotational direction. The change in the value of the spring constant K1 results in an error when calculating the load torque. In other words, if the spring constant K1 is not updated, the error in the load torque calculated by the actuator 1 will increase over time. Therefore, the actuator 1 according to this embodiment has a function for self-calibrating the spring constant K1. Details of the self-calibration function will be described later. Self-calibration refers to calibrating (also described as correcting, amending, correcting, changing, updating, etc.) the spring constant K1 inside the actuator 1.

Here, the load torque is calculated based on the torsion angle α1 in the rotational direction of the reducer 3, but the method is not limited to this. The actuator 1 can calculate the load torque based on the current for driving the motor 2 (motor drive current) as a method for calculating the load torque. The method for calculating the load torque based on the motor drive current will be described in detail later.

The actuator 1 detects a change in the torsional rigidity in the rotational direction of the reducer 3 by comparing the load torque calculated based on the motor drive current with the load torque calculated based on the torsion angle α1. Hereinafter, the load torque calculated based on the torsion angle α1 is referred to as a first load torque T1, and the load torque calculated based on the motor drive current is referred to as a second load torque T2.
The configuration of the actuator 1 according to this embodiment will be described in more detail below, focusing on the calculation of the load torque.

FIG. 5 is a diagram showing an example of the functional configuration of the calculation unit 8 according to this embodiment. The calculation unit 8 includes a first load torque calculation unit 81, a second load torque calculation unit 82, a comparison unit 83, and a corrected spring constant calculation unit 84.

The first load torque calculation unit 81 calculates the first load torque T1 based on the rotation angle information B1 and the spring constant K1. The first load torque calculation unit 81 acquires the rotation angle information B1 supplied from the input shaft encoder 4 and the output shaft encoder 5. The first load torque calculation unit 81 calculates the torsion angle α1 as the difference between the input shaft rotation angle θ1 and the output shaft rotation angle θ2 based on the input shaft rotation angle θ1 and the output shaft rotation angle θ2 indicated by the rotation angle information B1. The first load torque calculation unit 81 reads out the spring constant K1 from the storage unit 9. The first load torque calculation unit 81 generates torque information E1 indicating the value of the calculated first load torque T1. The first load torque calculation unit 81 outputs the generated torque information E1.

The first load torque calculation unit 81 also generates first torque information F1 indicating the value of the calculated first load torque T1. The first load torque calculation unit 81 supplies the generated first torque information F1 to the comparison unit 83.

The second load torque calculation unit 82 calculates the second load torque T2 based on the motor drive current information A1 and the torque constant Kt. The second load torque calculation unit 82 acquires the motor drive current information A1 supplied from the motor driver 72. The second load torque calculation unit 82 reads out the torque constant Kt from the memory unit 9. The second load torque calculation unit 82 generates second torque information F2 indicating the value of the calculated second load torque T2. The second load torque calculation unit 82 supplies the generated second torque information F2 to the comparison unit 83.

The motor driver 72 (Figure 1) constantly monitors the motor drive current (at clock cycle intervals) regardless of the conditions (operating conditions and rotation speed). The second load torque calculation unit 82 constantly (at clock cycle intervals) monitors the motor drive current and also generates the second load torque T2. The torque constant Kt changes in response to changes in operating conditions and rotation speed, but does not change over time. Therefore, when a constant torque is applied, it is considered that the motor drive current information A1 and the value of the second load torque T2 under specific operating conditions and at a specific rotation speed will not change over a long period of time. By utilizing the second load torque T2 at this specific timing, the self-calibration process described below can be performed.

The comparison unit 83 compares the second load torque T2 and the first load torque T1 based on the second torque information F2 and the first torque information F1. The second torque information F2 indicates the value of the second load torque T2 calculated by the second load torque calculation unit 82. The first torque information F1 indicates the value of the first load torque T1 calculated by the first load torque calculation unit 81. The comparison unit 83 calculates the change in the torsional rigidity in the rotational direction of the reducer 3 by comparing the second load torque T2 and the first torque information F1.

The comparison unit 83 compares, for example, the second load torque T2 generated at a specific time with the first load torque T1 based on the second torque information F2 and the first torque information F1. Time information G1 indicating the time at which the second load torque T2 was generated is generated, for example, by the motor control unit 7. The comparison unit 83 obtains the time information G1 supplied from the motor control unit 7.

The comparison unit 83 determines that the motor drive current is generated at a specific time if the time indicated by the acquired time information G1 is included in a specific period (timing).

The modified spring constant calculation unit 84 calculates a modified spring constant according to the change in torsional rigidity in the rotational direction of the reducer 3 detected by the comparison unit 83. Here, the modified spring constant calculation unit 84 calculates the modified spring constant Kc from the second torque information F2 and the rotation angle information B1. The modified spring constant calculation unit 84 stores the calculated modified spring constant Kc in the memory unit 9 as the spring constant K1.

In other words, the calculation unit 8 self-calibrates the spring constant K1 based on the change in the torsional stiffness in the rotational direction of the reducer 3 detected by comparing the first load torque T1 and the second load torque T2. The calculation unit 8 detects the change in the torsional stiffness in the rotational direction of the reducer 3. The calculation unit 8 calculates a corrected spring constant Kc according to the detected change in torsional stiffness. The calculation unit 8 stores the calculated corrected spring constant Kc in the memory unit 9 as the spring constant K1.

In the process in which the calculation unit 8 detects the change in the torsional rigidity in the rotational direction of the reducer 3, there are several modes depending on the conditions (called self-calibration conditions) for using the second load torque T2 as a reference for self-calibration. Examples of the self-calibration conditions include the temperature conditions of the coil of the motor 2, the temperature conditions of the encoder, the temperature conditions of the reducer 3, and the temperature conditions of the driver 6.

When the self-calibration condition is the temperature of the coil of the motor 2, the calculation unit 8 determines the condition based on the motor coil temperature information C1 supplied from the temperature sensor 10. When the self-calibration condition is the temperature of the encoder, the calculation unit 8 determines the condition based on the encoder temperature information D1 supplied from the input shaft encoder 4. The encoder temperature information D1 indicates one or more of the temperature of the input shaft encoder 4 and the temperature of the output shaft encoder 5. The part of the encoder temperature information D1 indicating the temperature of the input shaft encoder 4 is supplied from the input shaft encoder 4 to the calculation unit 8. The part of the encoder temperature information D1 indicating the temperature of the output shaft encoder 5 is supplied from the output shaft encoder 5 to the calculation unit 8. Therefore, the encoder temperature information D1 is supplied to the calculation unit 8 from one or more of the input shaft encoder 4 and the output shaft encoder 5.

Now, with reference to Figures 6 to 8, other examples of self-calibration conditions used to calculate the second load torque T2 when comparing the second load torque T2 with the first load torque T1 will be described. Note that the same components as those in the calculation unit 8 (Figure 5) described above are given the same reference numerals, and descriptions of the same components and operations may be omitted.

6 is a diagram showing an example of the functional configuration of a calculation unit 8a according to a modification of this embodiment. Comparing the calculation unit 8a (FIG. 6) with the calculation unit 8 (FIG. 5), a comparison unit 83a is different.
The comparison unit 83a compares the first load torque T1 with a second load torque T2 calculated based on the motor drive current at a time when the rotation shaft angle of the input shaft 31 and/or the output shaft 32 is a specific angle. The comparison unit 83a determines whether the rotation shaft angle of the actuator 1 is a specific angle based on the rotation angle information B1 supplied from the input shaft encoder 4 and the output shaft encoder 5.

7 is a diagram showing an example of the functional configuration of a calculation unit 8b according to a modification of this embodiment. Comparing the calculation unit 8b (FIG. 7) with the calculation unit 8 (FIG. 5), a comparison unit 83b is different.
The comparison unit 83b compares the second load torque T2 calculated based on the motor drive current at a time when the temperature of the coil of the motor 2 is at a specific temperature with the first load torque T1, based on the second torque information F2 and the first torque information F1. The comparison unit 83a determines whether the temperature of the coil of the motor 2 is at a specific temperature based on the motor coil temperature information C1 supplied from the temperature sensor 10.

8 is a diagram showing an example of the functional configuration of a calculation unit 8c according to a modification of this embodiment. Comparing the calculation unit 8c (FIG. 8) with the calculation unit 8 (FIG. 5), a comparison unit 83c is different.
The comparison unit 83c compares the first load torque T1 with a second load torque T2 calculated based on the motor drive current at a time when the temperature of the input shaft encoder 4 is a specific temperature, based on the second torque information F2 and the first torque information F1. The comparison unit 83c determines whether the temperature of the input shaft encoder 4 or the temperature of the output shaft encoder 5 is a specific temperature, based on encoder temperature information D1 supplied from one or more of the input shaft encoder 4 and the output shaft encoder 5. Here, the comparison unit 83c may determine whether both the temperature of the input shaft encoder 4 and the temperature of the output shaft encoder 5 are a specific temperature, or may determine whether either one of the temperature of the input shaft encoder 4 and the temperature of the output shaft encoder 5 is a specific temperature.

Next, referring to FIG. 9, the self-calibration process in which the actuator 1 corrects the value of the spring constant K1 will be described. FIG. 9 is a diagram showing an example of the self-calibration process according to this embodiment. The self-calibration process is executed by the calculation unit 8.

The self-calibration process is executed, for example, when the actuator 1 is in operation. When the self-calibration process is executed when the actuator 1 is in operation, the spring constant K1 is calibrated in real time. The calculation unit 8 repeatedly executes the self-calibration process while the actuator 1 is powered on.
Furthermore, the timing for executing the self-calibration process may be the timing for calibration before driving the actuator 1. Calibration refers to an operation for calibration that is periodically performed by a user after assembling the actuator 1 in an apparatus.

Step S10: The first load torque calculation unit 81 calculates the first load torque T1 based on the rotation angle information B1 and the spring constant K1. The first load torque calculation unit 81 calculates the first load torque T1 by multiplying the torsion angle α1 indicated by the rotation angle information B1 by the spring constant K1. The first load torque calculation unit 81 supplies the comparison unit 83 with first torque information F1 indicating the calculated first load torque T1.

Step S20: The second load torque calculation unit 82 calculates the second load torque T2 based on the motor drive current information A1 and the torque constant Kt. The second load torque calculation unit 82 calculates the second load torque T2 by multiplying the value of the motor drive current indicated by the motor drive current information A1 by the torque constant Kt. The second load torque calculation unit 82 supplies the comparison unit 83 with second torque information F2 indicating the calculated second load torque T2.

Step S30: The comparison unit 83 determines whether the self-calibration conditions are met. If the second load torque T2 is generated at a specific time, the comparison unit 83 determines that the self-calibration conditions are met. The comparison unit 83 acquires time information G1 from the motor control unit 7. If the time indicated by the acquired time information G1 is included in a specific period, the comparison unit 83 determines that the motor drive current is generated at a specific time.

The specific time is, for example, a specific time when the motor drive current stabilizes during the process in which the actuator 1 performs a repetitive operation. Another example of a specific time is a time when the actuator 1 is performing a specific movement.

If the comparison unit 83 determines that the self-calibration conditions are met (step S30; YES), it executes the process of step S40. On the other hand, if the comparison unit 83 determines that the self-calibration conditions are not met (step S30; NO), the calculation unit 8 ends the self-calibration process.

Step S40: The comparison unit 83 calculates the change in the torsional rigidity in the rotational direction of the reducer 3 by comparing the second load torque T2 calculated by the second load torque calculation unit 82 with the first load torque T1 calculated by the first load torque calculation unit 81 based on the second torque information F2 and the first torque information F1. The comparison unit 83 calculates the difference value of the first load torque T1 relative to the second load torque T2 as the magnitude of the change in the torsional rigidity in the rotational direction of the reducer 3.

Step S50: The corrected spring constant calculation unit 84 calculates the corrected spring constant Kc according to the change in the torsional rigidity of the reducer 3 in the rotational direction detected by the comparison unit 83. Here, in this embodiment, the first load torque T1 is approximated by a cubic expression of the variable x when the variable x is the amount of torsion (torsion angle α1), as in equation (1).

Each of the parameters a, b, and c indicates the coefficient of each order. Therefore, the spring constant K1 is given by a set of three parameters a, b, and c. In this embodiment, each of these parameters a, b, and c is a function of the variable x, which is the amount of twist. This function is calculated in advance by actual measurement.

In this embodiment, the true value of the load torque is determined by adding the magnitude of the change in torsional rigidity calculated in step S40 to the first load torque T1. In other words, the true value of the load torque is determined to be the value of the second load torque T2. The modified spring constant calculation unit 84 calculates a, b, and c such that the amount on the right side of equation (1) is equal to the value of the second load torque T2. The comparison unit 83 determines the set of the three calculated parameters a, b, and c as the modified spring constant Kc.

Step S60: The modified spring constant calculation unit 84 stores the calculated modified spring constant Kc as the spring constant K1 in the memory unit 9. That is, the modified spring constant calculation unit 84 stores the set of three parameters a, b, and c calculated in step S50 as the spring constant K1.
With the above, the calculation unit 8 ends the self-calibration process.

If the specific time determined in step S30 is a specific time when the motor drive current is stable while the actuator 1 is performing a repetitive operation, the second load torque calculation unit 82 calculates the second load torque T2 based on the motor drive current at the specific time when the motor drive current is stable while the actuator 1 is performing a repetitive operation, and the comparison unit 83 compares the second load torque T2 calculated by the second load torque calculation unit 82 with the first load torque T1 based on the second torque information F2 and the first torque information F1, thereby calculating the change in the torsional rigidity in the rotational direction of the reducer 3.

In addition, if the specific time determined in step S30 is a time when the actuator 1 is performing a specific motion, the second load torque calculation unit 82 calculates a second load torque T2 based on the motor drive current during the time when the actuator 1 is performing the specific motion, and the comparison unit 83 compares the second load torque T2 calculated by the second load torque calculation unit 82 with the first load torque T1 based on the second torque information F2 and the first torque information F1, thereby calculating the change in the torsional rigidity in the rotational direction of the reducer 3.

Furthermore, the processes in steps S30, S40, and S50 may be performed based on the modified examples described above with reference to FIGS.
For example, if the specific time determined in step S30 is a time when the rotation shaft angle of actuator 1 is a specific angle, the second load torque calculation unit 82 calculates a second load torque T2 based on the motor drive current at the time when the rotation shaft angle of actuator 1 is the specific angle, and the comparison unit 83 compares the second load torque T2 calculated by the second load torque calculation unit 82 with the first load torque T1 based on the second torque information F2 and the first torque information F1, thereby calculating a change in the torsional rigidity in the rotational direction of the reducer 3.

Furthermore, for example, if the specific time determined in step S30 is a time when the rotational speed of actuator 1 is a specific speed, second load torque calculation unit 82 calculates second load torque T2 based on the motor drive current at the time when the rotational speed of actuator 1 is the specific speed, and comparison unit 83 compares the second load torque T2 calculated by second load torque calculation unit 82 with the first load torque T1 based on second torque information F2 and first torque information F1, thereby calculating the change in torsional rigidity in the rotational direction of reducer 3.

Also, for example, if the specific time determined in step S30 is a time after a certain time has elapsed, the second load torque calculation unit 82 calculates the second load torque T2 based on the motor drive current at the time after the certain time has elapsed, and the comparison unit 83 compares the second load torque T2 calculated by the second load torque calculation unit 82 with the first load torque T1 based on the second torque information F2 and the first torque information F1, thereby calculating the change in the torsional rigidity in the rotational direction of the reducer 3.

Furthermore, for example, if the specific time determined in step S30 is a time when the temperature of actuator 1 is a specific temperature, the second load torque calculation unit 82 calculates the second load torque T2 based on the motor drive current at the time when the temperature of actuator 1 is the specific temperature, and the comparison unit 83 compares the second load torque T2 calculated by the second load torque calculation unit 82 with the first load torque T1 based on the second torque information F2 and the first torque information F1, thereby calculating the change in the torsional rigidity in the rotational direction of the reducer 3.

The calculation unit 8 may include a notification unit that notifies the outside that the change in the torsional rigidity in the rotational direction of the reducer 3 exceeds a predetermined amount. When the change in the torsional rigidity in the rotational direction of the reducer 3 exceeds a predetermined amount, the notification unit notifies the outside that the change in the torsional rigidity in the rotational direction of the reducer 3 exceeds the predetermined amount.
The notification unit may notify by, for example, blinking an LED lamp. The notification unit may also notify by sound or a signal.

The calculation unit 8 may also include a stopping unit that stops the operation of the actuator 1. The stopping unit stops the operation of the actuator 1 when the change in the torsional rigidity of the reducer 3 in the rotational direction exceeds a predetermined amount. The stopping unit may stop the actuator 1 after causing the actuator 1 to perform a predetermined operation when the change in the torsional rigidity of the reducer 3 in the rotational direction exceeds a predetermined amount.

As described above, the actuator 1 according to this embodiment includes the motor 2, the reducer 3, a first encoder (in this embodiment, the input shaft encoder 4), a second encoder (in this embodiment, the output shaft encoder 5), a motor control unit 7, a memory unit 9, and a calculation unit 8.
The first encoder (in this embodiment, the input shaft encoder 4 ) detects the rotation angle of the input shaft 31 .
The second encoder (in this embodiment, the output shaft encoder 5 ) detects the rotation angle of the output shaft 32 .
The storage unit 9 stores the spring constant K1 of the reduction gear 3 .
The calculation unit 8 calculates a torsion angle α1 in the rotational direction of the reducer 3 based on the rotational angle (input shaft rotation angle θ1 in this embodiment) of the input shaft 31 detected by the first encoder (input shaft encoder 4 in this embodiment) and the rotational angle (output shaft rotation angle θ2 in this embodiment) of the output shaft 32 detected by the second encoder (output shaft encoder 5 in this embodiment), and calculates a load torque (first load torque T1 in this embodiment) in the rotational direction of the reducer 3 applied between the input shaft 31 and the output shaft 32 by multiplying the calculated torsion angle α1 by the spring constant K1 of the reducer 3 stored in the memory unit 9.
The calculation unit 8 detects changes in the torsional rigidity in the rotational direction of the reducer 3 (in this embodiment, the comparison unit 83), calculates a modified spring constant Kc corresponding to the detected change in torsional rigidity (in this embodiment, the modified spring constant calculation unit 84), and stores the calculated modified spring constant Kc as a spring constant K1 in the memory unit 9 (in this embodiment, the modified spring constant calculation unit 84).

With this configuration, the actuator 1 according to this embodiment can calculate the load torque using the corrected spring constant Kc as the spring constant K1, thereby reducing errors in the load torque due to changes in the torsional rigidity of the reducer 3 in the rotational direction.

Second Embodiment
The second embodiment of the present invention will now be described in detail with reference to the drawings.
In this embodiment, a case will be described in which the actuator outputs, as the load torque, one of the first load torque T1 calculated based on the amount of torsion and the second load torque T2 calculated based on the motor drive current, whichever is determined to be more likely.
The actuator according to this embodiment is referred to as actuator 1d. Note that the same components as those in the first embodiment described above are denoted by the same reference numerals, and descriptions of the same components and operations may be omitted.

As explained in the first embodiment, the first load torque T1 is calculated based on the amount of torsion (torsion angle α1) of the reducer 3 and the spring constant K1. The amount of torsion has hysteresis, and in the region with hysteresis, the error between the calculated first load torque T1 and the actual load torque tends to be large.

Figure 10 shows the relationship between the actual load torque and the first load torque T1 ("calculated torque" in Figure 10) calculated based on the amount of torsion. As Figure 10 shows, in the range where the load torque value is small (low torque region), the error is larger than in the range where the load torque value is large (high torque region). This is thought to be because the hysteresis in the characteristics of the spring of the reducer 3 is larger in the low torque region than in the high torque region.

Figure 11 shows the relationship between the actual load torque and the second load torque T2 calculated based on the motor drive current ("calculated torque" in Figure 11). As shown in Figure 11, the error is larger in the high torque region or in the medium range (medium torque region) compared to the low torque region. In the high torque region or medium torque region, the value of the second load torque T2 calculated based on the motor drive current tends to vary depending on cogging and temperature.

Actuator 1d according to this embodiment reduces the error in the load torque by selecting a load torque calculated based on the motor drive current in the low torque region and selecting a load torque calculated based on the amount of torsion outside the low torque region. The configuration of actuator 1d differs from that of actuator 1 in the configuration of the calculation unit. Actuator 1d includes calculation unit 8d as the calculation unit.

FIG. 12 is a diagram showing an example of the functional configuration of the calculation unit 8d according to this embodiment. Comparing the configuration of the calculation unit 8d (FIG. 12) with the configuration of the calculation unit 8 (FIG. 5), the calculation unit 8d differs in that it includes a comparison unit 83d and does not include a corrected spring constant calculation unit 84.

The comparison unit 83d calculates the load torque from the first load torque T1 and the second load torque T2 based on the second torque information F2 and the first torque information F1. In this embodiment, the comparison unit 83d selects the first load torque T1 calculated based on the amount of torsion as the load torque outside the low torque region, and selects the second load torque T2 as the load torque in the low torque region. In other words, the comparison unit 83d selects the first load torque T1 as the load torque when the load torque value is within a predetermined range, and selects the second load torque T2 as the load torque when the load torque value is not within the predetermined range.

Next, a torque calculation process in which the actuator 1d calculates the load torque will be described with reference to Fig. 13. Fig. 13 is a diagram showing an example of the torque calculation process according to this embodiment. The torque calculation process is executed by the calculator 8d. The torque calculation process is repeatedly executed while the actuator 1d is being driven.
The processes in steps S110 and S120 are similar to those in steps S10 and S20 in FIG. 9, and therefore will not be described.

Step S130: The comparison unit 83d determines whether the calculated load torque is in the low torque region. Based on the second torque information F2 and the first torque information F1, the comparison unit 83d uses either the value of the first load torque T1 or the value of the second load torque T2 to determine whether the load torque is in the low torque region.

For example, the comparison unit 83d uses the smaller of the first load torque T1 and the second load torque T2 to determine whether or not the torque range is low, based on the second torque information F2 and the first torque information F1. The larger of the first load torque T1 and the second load torque T2 may be used to determine whether or not the torque range is low. The average value of the first load torque T1 and the second load torque T2 may be used to determine whether or not the torque range is low. In addition, whether the first load torque T1 or the second load torque T2 is used as the load torque to be used to determine whether or not the torque range is low may be determined in advance, regardless of the value.

The range of the low torque region is determined in advance based on actual measured values. For example, the range of the low torque region is determined in advance based on the relationship between the actual load torque shown in FIG. 10 and the first load torque T1 calculated based on the amount of torsion, and the relationship between the actual load torque shown in FIG. 11 and the second load torque T2 calculated based on the motor drive current.

If the comparison unit 83d determines that the calculated load torque is in the low torque region (step S130; YES), it executes the process of step S150. On the other hand, if the comparison unit 83d determines that the calculated load torque is not in the low torque region (step S130; NO), it executes the process of step S140.

Step S140: The comparison unit 83d selects the load torque (first load torque T1) calculated based on the amount of torsion as the load torque.

Step S150: The comparison unit 83d selects the load torque (second load torque T2) calculated based on the motor drive current as the load torque.

The following equation (2) shows the load torque calculated as a result of the processing in steps S130, S140, and S150.

In equation (2), T is the load torque, the variable x is the amount of torsion (torsion angle α1), the parameters a, b, and c are coefficients of each order, I is the motor drive current, and Kt is the torque constant Kt. Tlow is the threshold value for the low torque region.

Step S160: The comparison unit 83d outputs the selected load torque as the torque information E1.
With this, the calculation unit 8d ends the torque calculation process.

In this embodiment, an example has been described in which the comparison unit 83d selects the first load torque T1 as the load torque when the value of the load torque is within a predetermined range based on the second torque information F2 and the first torque information F1, and selects the second load torque T2 as the load torque when the value of the load torque is not within the predetermined range, but this is not limited to this.
The comparison unit 83d may select the first load torque T1 or the second load torque T2, whichever is more likely to be the load torque, based on a predetermined parameter, the second torque information F2, and the first torque information F1. The predetermined parameter is the temperature of the actuator 1d. The temperature of the actuator 1d is, for example, the temperature of the input shaft encoder 4. The temperature of the actuator 1d may be the temperature of any of the parts constituting the actuator 1d, or may be the average value of the temperatures of those parts. For example, when the temperature of the input shaft encoder 4 is equal to or lower than a predetermined temperature, the comparison unit 83d selects the first load torque T1 as the load torque, and when the temperature of the input shaft encoder 4 is higher than the predetermined temperature, the comparison unit 83d selects the second load torque T2 as the load torque. The comparison unit 83d determines the temperature of the input shaft encoder 4 or the temperature of the output shaft encoder 5 based on the encoder temperature information D1 supplied from at least one of the input shaft encoder 4 and the output shaft encoder 5. Here, the comparison unit 83d may determine whether both the temperature of the input shaft encoder 4 and the temperature of the output shaft encoder 5 are specific temperatures, or may determine whether either one of the temperature of the input shaft encoder 4 and the temperature of the output shaft encoder 5 is a specific temperature.

The specified parameters may be the operating time of the actuator 1d, the rotation speed of the motor 2, etc.

The comparison unit 83d may also calculate the load torque as an average value calculated by multiplying each of the first load torque T1 and the second load torque T2 by a predetermined weight. Equation (3) shows the load torque as an average value calculated by multiplying each of the first load torque T1 and the second load torque T2 by a predetermined weight.

In formula (3), T1 and T2 respectively represent the first load torque T1 calculated based on the amount of twist, and the second load torque T2 calculated based on the motor drive current. In formula (3), σ represents a predetermined weight.

The predetermined weight is determined in advance, for example, based on the following formula (4):

In equation (4), Q represents the temperature of actuator 1d, V represents the rotational speed of motor 2, and T represents the load torque. As shown in equation (4), the predetermined weight is determined, for example, based on the range of values of the load torque, or the temperature of actuator 1d, the rotational speed of motor 2, etc. Note that equation (4) is only one example for determining the predetermined weight, and any one of the temperature of actuator 1d, the rotational speed of motor 2, and the load torque may be omitted from the right-hand side of equation (4). The predetermined weight may also be determined based on parameters other than the temperature of actuator 1d, the rotational speed of motor 2, and the load torque.

The calculation unit 8d may include a notification unit that notifies of a failure of the actuator 1d. The notification unit notifies of a failure of the actuator 1d when a difference between a load torque value calculated in advance by a simulation and the calculated load torque value is equal to or greater than a threshold value. In the simulation, for example, a value of the load torque acting on the actuator 1d is calculated based on the amount of friction, mass, and the like from the operating conditions of the actuator 1d.
The notification unit may notify by, for example, blinking an LED lamp. The notification unit may notify by an alarm sound or a signal.

As described above, the actuator 1d according to this embodiment includes the motor 2, the reducer 3, a first encoder (in this embodiment, the input shaft encoder 4), a second encoder (in this embodiment, the output shaft encoder 5), the motor control unit 7, an external sensor (in this embodiment, the motor control unit 7), and a calculation unit 8d.
The first encoder (input shaft encoder 4 in this embodiment) detects the rotation angle of the input shaft.
The second encoder (in this embodiment, the output shaft encoder 5) detects the rotation angle of the output shaft.
The calculation unit 8d calculates a torsion angle α1 in the rotational direction of the reducer 3 based on the rotational angle (input shaft rotation angle θ1 in this embodiment) of the input shaft 31 detected by the first encoder (input shaft encoder 4 in this embodiment) and the rotational angle (output shaft rotation angle θ2 in this embodiment) of the output shaft 32 detected by the second encoder (output shaft encoder 5 in this embodiment), and calculates the load torque in the rotational direction of the reducer 3 applied between the input shaft 31 and the output shaft 32 as a first load torque (first load torque T1 in this embodiment) based on the calculated torsion angle α1 (first load torque calculation unit 81 in this embodiment).
An external sensor (in this embodiment, the motor control unit 7) measures the state of the actuator 1d.
The calculation unit 8d calculates a second load torque (in this embodiment, the second load torque T2) based on the measurement results (in this embodiment, the value of the motor drive current) by an external sensor (in this embodiment, the motor control unit 7) (in this embodiment, the second load torque calculation unit 82), and calculates a load torque based on the first load torque (in this embodiment, the first load torque T1) and the second load torque (in this embodiment, the second load torque T2) (in this embodiment, the comparison unit 83d).

With this configuration, the actuator 1d according to this embodiment can calculate the load torque based on the first load torque and the second load torque, and therefore the accuracy of the calculated load torque can be improved. Here, improving the accuracy of the load torque means improving the accuracy of the load torque compared to the case where the first load torque is calculated as the load torque as is.

Third Embodiment
The third embodiment of the present invention will now be described in detail with reference to the drawings.
In the above second embodiment, the actuator outputs, as the load torque, one of the first load torque T1 calculated based on the amount of torsion and the second load torque T2 calculated based on the motor drive current, whichever is determined to be more likely. In the present embodiment, the actuator corrects an error due to hysteresis in the first load torque T1 calculated based on the amount of torsion.
The actuator according to this embodiment is referred to as actuator 1e. Note that the same components as those in the second embodiment are denoted by the same reference numerals, and descriptions of the same components and operations may be omitted.

FIG. 14 is a diagram showing an example of the functional configuration of the calculation unit 8e according to this embodiment. Comparing the configuration of the calculation unit 8e (FIG. 14) with the configuration of the calculation unit 8d (FIG. 12), the difference is that the calculation unit 8e includes a correction unit 84e.

The correction unit 84e corrects the value of the first load torque T1 calculated by the first load torque calculation unit 81 based on the correction value information H1 and the first torque information F1. The correction value information H1 indicates the correction value of the first load torque T1. The correction value information H1 is stored in the memory unit 9.

The comparison unit 83d calculates the load torque based on the first load torque T1 corrected by the correction unit 84e based on the correction value information H1 and the first torque information F1, and the second load torque T2 based on the second torque information F2. In this embodiment, the comparison unit 83d selects the first load torque T1 corrected by the correction unit 84e based on the correction value information H1 and the first torque information F1 as the load torque outside the low torque region, and selects the second load torque T2 based on the second torque information F2 as the load torque in the low torque region.

Now, referring to Figures 15 and 16, the error in the first load torque T1 due to hysteresis will be described. Figure 15 is a diagram showing an example of the relationship between the amount of twist and the load torque according to this embodiment. Two graphs are shown in Figure 15: a graph of "actual measured values" and a graph of "approximation formula".

The "Actual Values" graph shows the actual load torque versus torsion amount. The "Actual Values" graph shows the actual load torque versus torsion amount when the pressure is reduced and increased in a direction that reduces the torsion amount, and then reduced and increased in a direction that increases the torsion amount. As the "Actual Values" graph shows, the load torque versus torsion amount shows that hysteresis exists between the direction in which the torsion amount decreases and the direction in which the torsion amount increases.
On the other hand, the "approximation formula" graph is a graph showing the value of the load torque versus the amount of torsion when the spring constant K1 is approximated by a third-order polynomial.

FIG. 16 is a diagram showing an example of the error of the approximate equation for the load torque in this embodiment with respect to the actual measured value. The graph shown in FIG. 16 is a graph (sometimes referred to as an error correction graph in the following explanation) that shows the error (difference) of the graph of the "approximate equation" shown in FIG. 15 with respect to the graph of the "actual measured value" for each value of the load torque.

In the actuator 1e according to this embodiment, the error as shown in FIG. 16 is stored in advance in the storage unit as correction value information H1. In this case, the correction value information H1 is, as an example, two-dimensional tabular data consisting of rows and columns in which the load torque correction value is stored for each load torque value during pressurization and depressurization. The correction value information H1 may be information indicating an approximation formula that approximates a curve showing the relationship between the load torque value and the correction value.

Next, a torque calculation process in which the actuator 1e calculates the load torque will be described with reference to Fig. 17. Fig. 17 is a diagram showing an example of the torque calculation process according to this embodiment. The torque calculation process is executed by the calculator 8e. The torque calculation process is repeatedly executed while the actuator 1e is being driven.
The processes in steps S210 and S230 are similar to those in steps S110 and S120 in FIG. 13, and therefore will not be described.

Step S220: The correction unit 84e corrects the value of the first load torque T1 calculated by the first load torque calculation unit 81 based on the correction value information H1 and the first torque information F1. Here, the correction unit 84e acquires the first torque information F1 supplied from the first load torque calculation unit 81. The correction unit 84e reads out the correction value information H1 from the memory unit 9. The correction unit 84e corrects the value of the first load torque T1 indicated by the first torque information F1 with the correction value indicated by the correction value information H1. The correction unit 84e supplies the first load torque T1 with the corrected value to the comparison unit 83d.

The processes after step S240 (steps S240 to S270) are similar to the processes after step S130 in FIG. 13 (steps S130 to S160) except that the value of the first load torque T1 corrected by the correction unit 84e is used as the value of the first load torque T1. Therefore, the description thereof will be omitted.
With this, the calculation unit 8e ends the torque calculation process.

18 to 33, the correction process (step S220) in which the correction unit 84e corrects the value of the first load torque T1 based on the correction value information H1 will be described in detail along with modified examples.
FIG. 18 is a diagram for explaining an example of the correction process according to the present embodiment. The error correction graph shown in FIG. 18 is the same as the error correction graph shown in FIG. 16. The correction unit 84e compares the current value of the first load torque T1 (shown as the torque value Tn in FIG. 18) with the value of the first load torque T1 calculated immediately before, and determines whether the load torque is being increased or decreased. The torque value Tn, which is the current value of the first load torque, is the first load torque T1 calculated by the first load torque calculation unit 81 in step S210 of the current torque calculation process. The first load torque T1 calculated immediately before is the first load torque T1 calculated by the first load torque calculation unit 81 in step S210 of the previous torque calculation process.
In addition, the correction unit 84e may compare the amount of torsion (torsion angle α1) between the current time and the immediately previous time instead of the first load torque T1 to determine whether pressure is being increased or decreased.

When the correction unit 84e determines that pressure is being increased, it corrects the value of the first load torque T1 by subtracting a correction value h1 from the value of the first load torque T1. When the correction unit 84e determines that pressure is being decreased, it corrects the value of the first load torque T1 by adding a correction value h2 to the value of the first load torque T1.

In other words, the correction unit 84e calculates a correction value for correcting the first load torque T1 based on the relationship between the difference between the approximate expression for the spring constant K1 and the actual measured value, and the value of the load torque, depending on whether pressure is being increased or decreased, and the value of the first load torque T1. The correction unit 84e corrects the first load torque T1 based on the calculated correction value.

According to the error correction graph shown in FIG. 18 (or FIG. 16), when the load torque is increased from a value of 0 to a torque value Tn, the correction unit 84e reads out a correction value h1 from the correction value information H1. Immediately after the load torque value is increased to the torque value Tn, at the moment when the load torque is switched to reduced pressure, the correction unit 84e reads out a correction value h2 from the correction value information H1. Therefore, at the time when the load torque is switched from increased pressure to reduced pressure, the detected load torque value changes suddenly from a value obtained by subtracting the correction value h1 from the torque value Tn to a value obtained by adding the correction value h2 to the torque value Tn. Therefore, as a result of the correction process, the change in the value of the first load torque T1 after correction becomes discontinuous.

In order to make the change in the value of the first load torque T1 after correction continuous, the correction unit 84e may calculate a correction value corresponding to the value of the first load torque T1 based on a straight line with a specific slope during pressurization and depressurization. FIG. 19 is a diagram for explaining an example of the correction process according to this embodiment. Line segment L1 and line segment L2 shown in FIG. 19 are examples of line segments that are part of a straight line with a specific slope that is used in the correction process during pressurization. Line segment L4 is an example of a line segment that is part of a straight line with a specific slope that is used in the correction process during depressurization. In the example shown in FIG. 19, the slope of the straight line with a specific slope is the same during pressurization and depressurization.

When pressure is being applied and the current value of the first load torque T1 is torque value Tn1, the correction unit 84e uses the correction value h11, which is the value of the error on line segment L1 for the torque value Tn1, as the correction value. Similarly, when pressure is being applied and the current value of the first load torque T1 is torque value Tn2, which is greater than torque value Tn1, the correction unit 84e uses the correction value h12, which is the value of the error on line segment L2 for the torque value Tn2, as the correction value.

When pressure is being applied and the current value of the first load torque T1 is between the torque values Tn1 and Tn2, the correction unit 84e uses the error value on the line segment L2 for that value as the correction value. The same applies when pressure is being reduced. In other words, when the current value of the first load torque T1 is between the torque values Tn1 and Tn2 both during pressure application and pressure reduction, the correction value for the current value of the first load torque T1 is calculated so that the pair of the value of the first load torque T1 and the correction value for the value of the first load torque T1 moves on a straight line with a predetermined slope in accordance with the change in the value of the first load torque T1.

Here, at the point where the straight line with a predetermined slope intersects with the error correction graph, a correction value for the current value of the first load torque T1 is calculated based on the error correction graph. Curve L3 shown in FIG. 19 is the portion of the error correction graph from intersection Q1 of the straight line with a predetermined slope and the error correction graph to intersection Q2 of the straight line with a predetermined slope and the error correction graph.

If the current value of the first load torque T1 is between torque value Tn2 and torque value Tn3, the correction value for the current value of the first load torque T1 is calculated so that the pair of the value of the first load torque T1 and the correction value for the value of the first load torque T1 moves on curve L3 in response to changes in the value of the first load torque T1.

When pressure is being applied and the current value of the first load torque T1 is the torque value Tn3, the correction unit 84e uses the correction value h13, which is the value of the error on the curve L3 for the torque value Tn3, as the correction value.
When pressure is being reduced and the current value of the first load torque T1 is the torque value Tn4, the correction unit 84e uses a correction value h14, which is the value of the error on the line segment L4 relative to the torque value Tn4, as the correction value.

In the example shown in Fig. 19, the slope of the straight line having a specific slope is the same during pressurization and depressurization, but the present invention is not limited to this. The slope of the straight line having a specific slope may be different during pressurization and depressurization.
Also, the slope of a straight line having a specific slope may be changed according to the range of torque values, for example, by making the slope steeper in a low torque region and shallower in a high torque region.
Also, instead of a straight line with a particular slope, a curved line may be used.

In the present embodiment, an example has been described in which a table storing pairs of load torque values and correction values showing an error correction graph, or an approximation equation showing an error correction graph, is stored in the memory unit 9 as correction value information H1, but this is not limiting. Below, a description will be given of a case in which two types of spring constants based on hysteresis are stored in the memory unit 9. In that case, the correction unit 84e may be omitted from the configuration of the calculation unit 8e (FIG. 14).

For example, two spring constants for pressurization and depressurization are stored in the storage unit 9. Fig. 20 is a diagram showing an example of the relationship between the torsion amount and the load torque for two spring constants for pressurization and depressurization according to this embodiment. In Fig. 20, the arrow P1 indicates the direction in which the torsion amount (torsion angle α1) is reduced from a negative value and the torsion amount is increased until pressurization occurs. On the other hand, the arrow P2 indicates the direction in which the torsion amount (torsion angle α1) is reduced from a positive value and the torsion amount is decreased until pressurization occurs.
The spring constant in this embodiment is a set of coefficients of a third-order polynomial, as in the above-described embodiments. That is, sets of coefficients of a third-order polynomial for pressurization and depressurization are stored in the storage unit 9.

The first load torque calculation unit 81 determines whether pressure is being increased or decreased. For example, based on the rotation angle information B1 supplied from the input shaft encoder 4, the first load torque calculation unit 81 compares the torsion angle α1 in the previous torque calculation process with the current torque calculation process, and determines whether pressure is being increased or decreased from the amount of change in the torsion angle α1.

The first load torque calculation unit 81 reads out either the spring constant for pressurization or the spring constant for depressurization from the memory unit 9 depending on the result of the determination. The first load torque calculation unit 81 calculates the first load torque T1 from the torsion angle α1 based on the read spring constant.

If two spring constants are used, one for pressurization and one for depressurization, at the time of switching from pressurization to depressurization, the value of the first load torque T1 calculated due to the difference between the two spring constants becomes discontinuous with respect to the change in the amount of torsion.
Therefore, in the following, a case will be described in which the spring constant (coefficient of a cubic polynomial) is generated in real time based on the torsion angle (or load torque) while the actuator 1 is in operation.

FIG. 21 is a diagram showing an example of the measured load torque value with respect to the amount of twist according to this embodiment. In FIG. 21, the spring constant is measured by measuring the load torque when the load torque is reduced from a load torque value lower than the upper limit of the measured load torque. The upper limit of the measured load torque is sometimes called the peak torque. FIG. 21 shows the measured peak torque values of ±XNm, ±YNm, ±ZNm, and ±WNm. Here, the letters "X", "Y", "Z", and "W" each represent a predetermined numerical value. The numerical value indicated by the letter "X", the numerical value indicated by the letter "Y", the numerical value indicated by the letter "Z", and the numerical value indicated by the letter "W" are larger in that order. The way in which hysteresis occurs between pressurization and depressurization changes for each peak torque.

Furthermore, in the case of any peak torque, the point indicating the pair of the measured load torque value and the amount of torsion is on the curve during decompression, out of the curve during pressurization and the curve during decompression. In other words, the point does not move discontinuously from the curve during pressurization to the curve during decompression.

The correction unit 84e generates coefficients of a third-order polynomial corresponding to the spring constant during pressurization for each peak torque shown in FIG. 21. During depressurization, the last generated coefficient is used as the coefficient of the third-order polynomial. FIG. 22 shows the value of the first load torque T1 calculated from the amount of torsion using a third-order polynomial based on the coefficients generated by the correction unit 84e.

The method for generating the coefficients of the third order polynomial corresponding to the spring constant is described in detail below.
The relationship between the coefficients of the cubic polynomial and the peak torque is expressed by an approximation equation for each peak torque. Here, the peak torque corresponds to the maximum torsion amount, which is the maximum value of the torsion amount. Based on the approximation equation that expresses the relationship between the coefficients of the cubic polynomial and the peak torque, the coefficients of the cubic polynomial are calculated from the peak torque (peak torsion amount). The coefficients of the cubic polynomial that take into account the current hysteresis are generated from the actually generated torque (torsion amount). In other words, a spring constant that takes into account the current hysteresis is generated.

23 to 31, the relationship between the coefficients and the amount of twist for each of the coefficients of the third degree to the zeroth degree of the third degree polynomial will be described.
FIG. 23 is a diagram showing an example of the relationship between the amount of twist and the third-order coefficient of a third-order polynomial according to this embodiment. The third-order coefficient is almost equal when the amount of twist is positive and when it is negative. Therefore, as shown in FIG. 24, linear approximation was performed collectively without considering the positive and negative amounts of twist. Depending on the individual actuator, the third-order coefficient may not be equal when the amount of twist is positive and when it is negative. In that case, approximation may be performed using different approximation expressions when the amount of twist is positive and when it is negative. Furthermore, when the third-order coefficient is not equal within the range of positive and negative amounts of twist, approximation may be performed using a function other than a linear function.

FIG. 25 is a diagram showing an example of the relationship between the amount of twist and the quadratic coefficient of a cubic polynomial according to this embodiment. The absolute values of the quadratic coefficients are almost equal when the amount of twist is positive and when it is negative, and the signs are reversed. Therefore, as shown in FIG. 26, the amount of twist and the quadratic coefficient are approximated by a power function without considering whether they are positive or negative. Depending on the individual actuator, the absolute values of the quadratic coefficients may not be equal when the amount of twist is positive and when it is negative. In that case, approximation may be performed by different approximation expressions when the amount of twist is positive and when it is negative. In addition, when the absolute values of the quadratic coefficients are not equal when the amount of twist is positive and when it is negative, approximation may be performed by a function other than a power function.
In the process of generating the coefficients, the absolute value of the amount of twist is substituted into the approximation formula, and the sign of the value obtained as a result of the substitution is set to the opposite sign of the amount of twist.

FIG. 27 is a diagram showing an example of the relationship between the amount of twist and the first-order coefficient of a third-order polynomial according to this embodiment. The first-order coefficient is almost equal when the amount of twist is positive and when it is negative. Therefore, as shown in FIG. 28, approximation is performed by a power function without considering whether the amount of twist is positive or negative. Depending on the individual actuator, the first-order coefficient may not be equal when the amount of twist is positive and when it is negative. In that case, approximation may be performed by different approximation expressions when the amount of twist is positive and when it is negative. Furthermore, when the first-order coefficient is not equal within the range of the positive and negative amounts of twist, approximation may be performed by a function other than a power function.
In the process of generating the coefficients, the absolute value of the amount of twist is substituted into the approximation formula, and a negative sign is set to the value obtained as a result of the substitution.

FIG. 29 is a diagram showing an example of the relationship between the amount of twist and the zeroth-order coefficient of a third-order polynomial according to this embodiment. As shown in FIGS. 30 and 31, the zeroth-order coefficient was approximated by a second-order polynomial for both the positive and negative amounts of twist.

The approximation equations for the coefficients of each degree from 3rd to 0th order of the above-mentioned cubic polynomial are stored in the memory unit 9 as correction value information H1. The correction unit 84e reads out the correction value information H1 from the memory unit 9, and calculates the coefficients of each degree of the cubic polynomial as the current spring constant K1 from the torsion angle α1 based on the approximation equations for each degree of the cubic polynomial.

FIG. 32 shows an example of a spring constant reproduced by calculating the coefficients of a third-order polynomial using the approximation formula according to this embodiment. In FIG. 32, a graph showing a third-order polynomial having coefficients corresponding to the spring constant for each peak torque is shown.

FIG. 33 is a schematic diagram showing an enlarged fourth quadrant of the graph shown in FIG. 32. Line M1 corresponds to the spring constant generated when the amount of torsion is negative. Line M2 corresponds to the spring constant generated in the torque calculation process following the torque calculation process in which line M1 was generated. Line M3 corresponds to the spring constant generated in the torque calculation process following the torque calculation process in which line M2 was generated. Line M3 corresponds to the spring constant generated last in the torque calculation process during pressurization. During depressurization, the spring constant generated last during pressurization is used.

(Fourth embodiment)
The fourth embodiment of the present invention will now be described in detail with reference to the drawings.
In the above first embodiment, a case has been described in which the actuator corrects the spring constant K1 based on a change in the torsional rigidity in the rotational direction of the reducer 3. In the above second embodiment, a case has been described in which the actuator outputs, as the load torque, one of the first load torque T1 calculated based on the amount of torsion and the second load torque T2 calculated based on the motor drive current, whichever is determined to be more likely. In the present embodiment, a case will be described in which the actuator corrects the spring constant K1 based on a change in the torsional rigidity in the rotational direction of the reducer 3, and outputs, as the load torque, one of the first load torque T1 calculated based on the amount of torsion and the second load torque T2 calculated based on the motor drive current, whichever is determined to be more likely.

The actuator according to this embodiment is referred to as actuator 1f. The configuration of actuator 1f differs from that of actuator 1 in the configuration of the calculation unit. Actuator 1f includes calculation unit 8f as the calculation unit. Note that the same reference numerals are used for the same configurations as in the first and second embodiments described above, and descriptions of the same configurations and operations may be omitted.

Fig. 34 is a diagram showing an example of the functional configuration of a calculation unit 8f according to this embodiment. When the configuration of the calculation unit 8f (Fig. 34) is compared with the configuration of the calculation unit 8 (Fig. 5), a comparison unit 83f is different.
The comparison unit 83f has both the function of the comparison unit 83 (FIG. 5) and the function of the comparison unit 83d (FIG. 12). That is, the comparison unit 83f calculates a change in the torsional rigidity in the rotational direction of the speed reducer 3 by comparing the second load torque T2 calculated by the second load torque calculation unit 82 with the first load torque T1 calculated by the first load torque calculation unit 81 based on the second torque information F2 and the first torque information F1. The comparison unit 83f also calculates the load torque from the first load torque T1 and the second load torque T2 based on the first torque information F1 and the second torque information F2.

Next, a self-calibration and torque calculation process in which the actuator 1f corrects the value of the spring constant K1 and calculates the load torque will be described with reference to Fig. 35. Fig. 35 is a diagram showing an example of the self-calibration and torque calculation process according to this embodiment.
Note that the processes from step S310 to step S360 are similar to the processes from step S10 to step S60 in FIG. 9, and the processes from step S370 to step S3100 are similar to the processes from step S130 to step S160 in FIG. 13, so their explanations are omitted.

It should be noted that this embodiment may be further combined with the third embodiment. In other words, the actuator 1f may further correct an error due to hysteresis in the first load torque T1 calculated based on the amount of torsion. In this case, the calculation unit 8f (Fig. 34) further includes a correction unit 84e (Fig. 14).

In the above embodiment, an example of a case where the second load torque T2 is calculated based on the motor drive current has been described. Here, the motor control unit 7 is an example of an external sensor for measuring a physical quantity for calculating the load torque. The external sensor is not limited to the motor control unit 7. The second load torque T2 may be calculated based on a measurement value of a predetermined physical quantity by an external sensor other than the motor control unit 7.

The actuators according to the above-described embodiments (actuators 1, 1a, 1b, 1c, 1d, 1e, and 1f) are preferably used in a robot device.

In addition, a part of the actuators (actuators 1, 1a, 1b, 1c, 1d, 1e, 1f) in the above-mentioned embodiment, for example, the calculation unit (calculation unit 8, 8a, 8b, 8c, 8d, 8e, 8f) may be realized by a computer. In that case, a program for realizing this control function may be recorded in a computer-readable recording medium, and the program recorded in the recording medium may be read into a computer system and executed to realize the control function. Note that the "computer system" here refers to a computer system built into the actuators (actuators 1, 1a, 1b, 1c, 1d, 1e, 1f) and includes hardware such as an OS and peripheral devices. In addition, the "computer-readable recording medium" refers to a portable medium such as a flexible disk, a magneto-optical disk, a ROM, a CD-ROM, and a storage device such as a hard disk built into a computer system. Furthermore, the term "computer-readable recording medium" may include a medium that dynamically stores a program for a short period of time, such as a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line, and a medium that stores a program for a certain period of time, such as a volatile memory inside a computer system that serves as a server or client in such a case. The above program may be one that realizes part of the above-mentioned functions, or may be one that can realize the above-mentioned functions in combination with a program already recorded in the computer system.
In addition, a part or all of the arithmetic units (arithmetic units 8, 8a, 8b, 8c, 8d, 8e, 8f) in the above-mentioned embodiments may be realized as an integrated circuit such as an LSI (Large Scale Integration). Each functional block of the arithmetic units (arithmetic units 8, 8a, 8b, 8c, 8d, 8e, 8f) may be individually processed, or may be integrated into a processor in part or in whole. The integrated circuit method is not limited to LSI, and may be realized by a dedicated circuit or a general-purpose processor. In addition, when an integrated circuit technology that replaces LSI appears due to the advancement of semiconductor technology, an integrated circuit based on that technology may be used.

Although one embodiment of the present invention has been described in detail above with reference to the drawings, the specific configuration is not limited to that described above, and various design changes can be made without departing from the spirit of the present invention.

1, 1a, 1b, 1c, 1d, 1e, 1f... actuator, 2... motor, 3... reducer, 4... input shaft encoder, 5... output shaft encoder, 7... motor control unit, 9... memory unit, 8, 8a, 8b, 8c, 8d, 8e, 8f... calculation unit, T1... first load torque, T2... second load torque, 31... input shaft, 32... output shaft, θ1... input shaft rotation angle, θ2... output shaft rotation angle, α1... torsion angle, K1... spring constant, Kc... corrected spring constant

Claims (25)

A load torque calculation device for calculating a load torque value of an actuator including an input shaft, a motor that rotates the input shaft, a transmission to which power of the motor is input from the input shaft, an output shaft that rotates by power output from the transmission, a first encoder that detects first rotation information of the input shaft, and a second encoder that detects second rotation information of the output shaft,
a calculation unit that calculates a first load torque value of the actuator based on the first rotation information, the second rotation information, and a first spring constant stored in a storage unit. The load torque calculation device according to claim 1 , wherein the calculation unit calculates a second load torque value of the actuator based on a motor drive current that drives the motor and a torque constant stored in the storage unit. the calculation unit calculates a second spring constant based on the first load torque value and the second load torque value;
The load torque calculation device according to claim 2 , wherein the second spring constant is stored in the storage unit as the first spring constant. The load torque calculation device according to claim 3 , wherein the calculation unit outputs the first load torque value to an outside source. The load torque calculation device according to claim 2 , wherein the calculation unit outputs the second load torque value to an outside. 3. The load torque calculation device according to claim 2, wherein the calculation unit selects one of the first load torque value and the second load torque value, whichever is more likely, as the load torque value to be output to the outside, based on a predetermined parameter. The calculation unit is
calculating the second load torque value based on the motor drive current at a specific time when the motor drive current is stable during a process in which the actuator performs a repetitive operation, or at a time when the actuator performs a specific motion;
The load torque calculation device according to claim 3 , wherein the second spring constant is calculated by comparing the calculated second load torque value with the first load torque value. The calculation unit is
calculating the second load torque value based on the motor drive current at a time when the rotation shaft angle of the actuator is at a specific angle;
The load torque calculation device according to claim 3 , wherein the second spring constant is calculated by comparing the calculated second load torque value with the first load torque value. The calculation unit is
calculating the second load torque value based on the motor drive current at a time when the rotation speed of the actuator is a specific speed;
The load torque calculation device according to claim 3 , wherein the second spring constant is calculated by comparing the calculated second load torque value with the first load torque value. The calculation unit is
calculating the second load torque value based on the motor drive current at a time point after a certain time has elapsed;
The load torque calculation device according to claim 3 , wherein the second spring constant is calculated by comparing the calculated second load torque value with the first load torque value. The calculation unit is
calculating the second load torque value based on the motor drive current at a time when the temperature of the actuator is at a specific temperature;
The load torque calculation device according to claim 3 , wherein the second spring constant is calculated by comparing the calculated second load torque value with the first load torque value. The load torque calculation device according to claim 3 , wherein the calculation unit notifies an external device that a change in the torsional rigidity in the rotational direction of the transmission has exceeded a predetermined amount when the second spring constant exceeds a predetermined amount. The load torque calculation device according to claim 3 , wherein the calculation unit stops the operation of the actuator when the second spring constant exceeds a predetermined amount. The load torque calculation device according to claim 3 , wherein the calculation unit, when the second spring constant exceeds a predetermined amount, stops the actuator after causing the actuator to perform a predetermined operation. The load torque calculation device according to claim 3, wherein the calculation unit selects the first load torque value as the load torque value when the first load torque value is included in a predetermined range, and selects the second load torque value as the load torque value when the load torque value is not included in the predetermined range. The load torque calculation device according to claim 3 , wherein the calculation unit multiplies the first load torque value and the second load torque value by a predetermined weight and calculates an average value as the load torque value. 4. The load torque calculation device according to claim 3, wherein the calculation unit corrects the first load torque value based on correction value information indicating a correction value of the first load torque value, and calculates the load torque value based on the first load torque value corrected based on the correction value information and the second load torque value. The load torque calculation device according to claim 17 , wherein the correction value indicated by the correction value information is a correction value based on an error caused by hysteresis regarding a torsion angle in a rotational direction of the transmission. The load torque calculation device according to claim 3 , wherein the calculation unit notifies a failure of the actuator when a difference between the load torque value calculated in advance by a simulation and the calculated load torque value is equal to or greater than a threshold value. A load torque calculation device for calculating a load torque value of an actuator including an input shaft, a motor that rotates the input shaft, a transmission to which power of the motor is input from the input shaft, an output shaft that rotates by power output from the transmission, a first encoder that detects first rotation information of the input shaft, and a second encoder that detects second rotation information of the output shaft,
a load torque calculation device comprising: a calculation unit that calculates a second load torque value of the actuator based on a motor drive current that drives the motor and a torque constant stored in a storage unit. A load torque calculation device according to any one of claims 1 to 20,
The first encoder;
The second encoder;
An encoder system comprising: The encoder system of claim 21;
The input shaft;
The motor;
The transmission;
The output shaft;
An actuator comprising: A robotic device comprising the actuator of claim 22. 1. A load torque calculation method for calculating a load torque value of an actuator including an input shaft, a motor that rotates the input shaft, a transmission to which power of the motor is input from the input shaft, an output shaft that rotates by power output from the transmission, a first encoder that detects first rotation information of the input shaft, and a second encoder that detects second rotation information of the output shaft,
calculating a first load torque value of the actuator based on the first rotation information, the second rotation information, and a first spring constant stored in a storage unit;
calculating a second load torque value of the actuator based on a motor drive current that drives the motor and a torque constant stored in the storage unit;
A load torque calculation method comprising: A program for causing a computer to calculate a load torque value of an actuator including an input shaft, a motor that rotates the input shaft, a transmission to which power of the motor is input from the input shaft, an output shaft that rotates by power output from the transmission, a first encoder that detects first rotation information of the input shaft, and a second encoder that detects second rotation information of the output shaft,
On the computer,
calculating a first load torque value of the actuator based on the first rotation information, the second rotation information, and a first spring constant stored in a storage unit;
calculating a second load torque value of the actuator based on a motor drive current that drives the motor and a torque constant stored in the storage unit;
A program for executing.

Images