JP7530815B2 - Wearable Robot - Google Patents

Description

The present invention relates to a wearable robot.

In recent years, with the declining birthrate and aging population, efforts are being made to reduce the burden on humans at construction sites by introducing robots. For example, Patent Document 1 discloses a screw-driving robot and a ceiling panel transport robot that install ceiling panels. Patent Document 2 discloses a robot device that is composed of a cart-type mobile robot with a multi-joint arm attached, and that performs interior work such as driving screws into ceiling panels.

JP 2012-11653 A Japanese Patent Application Publication No. 5-321454

However, because construction sites involve a variety of environments, there are tasks that are difficult for the conventional robotic devices disclosed in Patent Documents 1 and 2 to perform. Furthermore, at construction sites, it is effective in utilizing robotic devices to enable people and robots to coexist and work together, such as by working together or sharing tasks. To enable people and robots to work together, it is also effective to have the person wear a robot. However, because robots are relatively heavy, it can be difficult to work with them.

The present invention aims to solve the above-mentioned problems by providing a wearable robot that can easily collaborate with humans and robots.

In order to achieve the above-mentioned objective, the wearable robot of the present invention comprises a multi-joint arm, a mounting unit to which the base end of the multi-joint arm is attached and which is configured to be attached and detached to the human body, a support leg that supports the multi-joint arm by contacting a support surface , an attitude measurement unit that measures the attitude of the support leg, and a control device that controls the operation of the multi-joint arm and corrects the operation of the multi-joint arm based on the amount of movement of the multi-joint arm calculated using the measurement results of the attitude measurement unit .

The wearable robot of the present invention makes it easy for humans and robots to work together.

1 is a schematic diagram of a wearable robot according to a first embodiment of the present invention; FIG. 1B is a rear view of the wearable robot shown in FIG. Diagram showing the support legs in the retracted position Diagram showing the support legs tilted Block diagram of wearable robot Flowchart of a program executed by the control device Flowchart of a program executed by the control device Diagram showing the operation of the wearable robot Diagram showing the operation of the wearable robot Diagram showing the operation of the wearable robot Schematic diagram of a wearable robot according to a second embodiment of the present invention. FIG. 9 is an enlarged partial cross-sectional view of the support leg shown in FIG. FIG. 13 is an enlarged partial cross-sectional view of a support leg according to a modified example of the second embodiment of the present invention. Block diagram of a wearable robot according to a third embodiment of the present invention. FIG. 13 is a diagram showing the operation of the wearable robot according to the third embodiment of the present invention. FIG. 1 is a schematic diagram of a wearable robot according to a modified example of the first embodiment of the present invention;

First Embodiment
A wearable robot according to a first embodiment of the present invention will be described below with reference to the drawings. In the following description, the upper and lower sides in Fig. 1A are respectively above and below the wearable robot 1.

The wearable robot 1 is a robot that works in collaboration with humans, for example, at a construction site to perform tasks such as driving screws into ceiling panels. As shown in Figs. 1A and 1B, the wearable robot 1 includes a mounting unit 10, support legs 20, an articulated arm 30, a posture measurement unit 40, a camera 50, and a control device 60.

The mounting unit 10 is for fixing the multi-joint arm 30 to the human body H. The mounting unit 10 is configured to be detachable from the human body H. The mounting unit 10 includes a frame 11 and multiple belts 12, 13, and 14. The frame 11 is formed in a frame shape. The frame 11 is attached to the human body H by having the arm passed through it and hanging it on the shoulder. The multiple belts 12, 13, and 14 are for fixing the frame 11 to the human body H. In this embodiment, the number of belts is three, but it goes without saying that this is not limited to this. The base end (upper end) of the support leg 20 is attached to the lower side of the mounting unit 10. The multi-joint arm 30 is attached to the upper side of the mounting unit 10.

The support legs 20 support the mounting unit 10 and the articulated arm 30 by contacting the support surface F. The support surface F is the surface on which the worker wearing the wearable robot 1 is located. The support surface F is, for example, the floor or the ground. The support legs 20 are formed in a rod shape and are configured to be extendable and retractable.

When the wearable robot 1 performs a target task, the length of the support leg 20 is adjusted so that the tip (lower end) of the support leg 20 comes into contact with the support surface F (FIGS. 1A and 1B). When the tip of the support leg 20 comes into contact with the support surface F, the support leg 20 supports the weight of the articulated arm 30 and the mounting unit 10. Thus, the support leg 20 can reduce the load on the human body H.

On the other hand, when the worker finishes work and moves, the articulated arm 30 can be easily transported by retracting the support legs 20 (Figure 2).

The multi-joint arm 30 is a manipulator. In this embodiment, the multi-joint arm 30 has five degrees of freedom, but it goes without saying that this is not limited to this. The multi-joint arm 30 includes a base 31, a shoulder block 32, a number of upper arm links 33, an elbow block 34, a forearm link 35, and a hand 36.

The base 31 is attached to the upper surface of the mounting part 10. A first end of a shoulder block 32 is connected to the base 31 so as to be rotatable around a first joint axis J1. The first joint axis J1 is coaxial with the axis 20a of the support leg 20. An elbow block 34 is connected to a second end of the shoulder block 32 via multiple upper arm links 33.

The shoulder block 32, the upper arm links 33, and the elbow block 34 form a parallel link mechanism. The rotation axis that is the center of rotation between the shoulder block 32 and the upper arm links 33, and the rotation axis that is the center of rotation between the upper arm links 33 and the elbow block 34 are collectively referred to as the second joint axis J2. The angle between the second joint axis J2 and the first joint axis J1 is a right angle. In this embodiment, the number of upper arm links 33 and the number of second joint axes J2 are four, but it goes without saying that this is not limited to four.

The first end of the forearm link 35 is connected to the elbow block 34 so as to be rotatable around the third joint axis J3. The third joint axis J3 is parallel to the first joint axis J1. As described above, the shoulder block 32, the upper arm links 33, and the elbow block 34 form a parallel link mechanism, so that the elbow block 34 moves relative to the shoulder block 32 so that the third joint axis J3 and the first joint axis J1 are parallel to each other.

The hand 36 is connected to the second end of the forearm link 35 so as to be rotatable around the fourth joint axis J4 and the fifth joint axis J5. The fourth joint axis J4 and the fifth joint axis J5 are perpendicular to each other. The angles between the fourth joint axis J4 and the fifth joint axis J5 and the first joint axis J1 are right angles.

The hand 36 holds a tool D, such as an electric screwdriver.

The articulated arm 30 also includes a plurality of drive units M and a plurality of angle detection units E. The drive unit M is, for example, a motor. The drive unit M rotates the shoulder block 32, the elbow block 34, the forearm link 35, and the hand 36 around the corresponding joint axes. The first drive unit M1 rotates the shoulder block 32 around the first joint axis J1 relative to the base 31. The second drive unit M2 rotates the elbow block 34 around the second joint axis J2 relative to the shoulder block 32. The third drive unit M3 rotates the forearm link 35 around the third joint axis J3 relative to the elbow block 34. The fourth drive unit M4 rotates the hand 36 around the fourth joint axis J4 relative to the forearm link 35. The fifth drive unit M5 rotates the hand 36 around the fifth joint axis J5 relative to the forearm link 35. The multiple drive units M are controlled by the control device 60.

The multiple angle detection units E detect the rotation angle of the drive unit M. The multiple angle detection units E are, for example, encoders. The first to fifth angle detection units E1 to E5 are arranged to correspond to the first to fifth drive units M1 to M5, respectively. The detection results of the angle detection units E are transmitted to the control device 60.

Further, for the articulated arm 30, a reference coordinate system Σ ₀ and a hand coordinate system Σ _e are defined.

The reference coordinate system _Σ0 is a coordinate system disposed on the base 31. The origin of the reference coordinate system _Σ0 is disposed at the intersection of the bottom surface of the base 31 and the axis 20a of the support leg 20, and the reference coordinate system Σ0 includes a reference Z-axis _Z0 that is coaxial with the axis 20a of the support leg ₂₀ and the first joint axis J1, a reference X-axis _X0 that is perpendicular to the reference Z-axis Z0, and a reference Y-axis _Y0 that is perpendicular to the reference Z-axis _Z0 and the reference X-axis _X0 . The intersection of the bottom surface of the base 31, where the origin of the reference coordinate system _Σ0 is disposed, and the axis 20a of the support leg 20 corresponds to the base end of the articulated arm 30.

The hand coordinate system _Σe is a coordinate system arranged in the hand 36. It includes a hand X-axis _Xe that is coaxial with the fourth joint axis J4, a hand Y-axis _Ye that is coaxial with the fifth joint axis J5, and a hand Z-axis _Ze that is perpendicular to the hand X-axis _Xe and the hand Y-axis _Ye .

The coordinates (x, y, z) of the origin position of the hand end coordinate system _Σe in the reference coordinate system _Σ0 are defined as the position of the hand 36. x is the value of the reference X-axis _X0 . y is the value of the reference Y-axis _Y0 . z is the value of the reference Z-axis _Z0 . In addition, the rotation angle (θ1, ψ1) of the hand end coordinate system _Σe relative to the reference coordinate system _Σ0 is defined as the posture of the hand 36. θ1 is the rotation angle around the reference Y-axis _Y0 . ψ1 is the rotation angle around the reference X-axis _X0 .

Here, in the reference coordinate system _Σ0 , the hand end vector r representing the position and posture of the hand 36 is defined as r = [x, y, z, θ1, ψ1] ^T. Strictly speaking, the hand end vector r is time series information (i.e., a time function) and can be expressed as r(t) = [x(t), y(t), z(t), θ1(t), ψ1(t)] ^T. The velocity of the hand 36 can be obtained by first-order differentiation of the hand end vector r, and the acceleration of the hand 36 can be obtained by second-order differentiation.

The posture measuring unit 40 is disposed in the mounting unit 10 and measures the posture of the support leg 20. Specifically, the posture of the support leg 20 is the inclination of the support leg 20 with respect to the vertical direction and the rotation angle of the support leg 20 around the axis 20a. As shown in FIG. 3, the inclination of the support leg 20 with respect to the vertical direction is represented by the leg pitch angle θ2 and the leg yaw angle ψ2. The leg pitch angle θ2 and the leg yaw angle ψ2 represent the rotation angles around two axes that are perpendicular to each other on the horizontal plane on which the tip of the support leg 20 is located. The rotation angle of the support leg 20 around the axis 20a is represented by the leg roll angle φ2. The leg pitch angle θ2, the leg yaw angle ψ2, and the leg roll angle φ2 change as the human body H swings. The swing of the human body H is a movement in which the human body H swings or tilts.

The attitude measurement unit 40 includes a gravity sensor 41 and a gyro sensor 42. The gravity sensor 41 is an acceleration sensor that detects the angle of the support leg 20 relative to the vertical direction, and measures the leg pitch angle θ2 and the leg yaw angle ψ2. The gyro sensor 42 is a sensor for detecting the angle of the support leg 20. Specifically, the gyro sensor 42 measures the leg roll angle φ2, which is the rotation angle of the support leg 20 around the axis 20a. The measurement results of the attitude measurement unit 40 are transmitted to the control device 60. The attitude measurement unit 40 can easily measure the attitude of the support leg 20 using the gravity sensor 41 and the gyro sensor 42.

The camera 50 captures images of the work area where the desired work is to be performed. The images captured by the camera 50 are sent to the control device 60.

The control device 60 is a computer that controls the wearable robot 1. As shown in FIG. 4, the control device 60 includes an input unit 61, an arm trajectory control unit 62, a work processing unit 63, and an output unit 64. The input unit 61, the arm trajectory control unit 62, the work processing unit 63, and the output unit 64 are configured by dedicated hardware or a processor that executes software. In this embodiment, the arm trajectory control unit 62 and the work processing unit 63 are configured by a processor that executes software. In other words, the control device 60 functions as the input unit 61, the arm trajectory control unit 62, the work processing unit 63, and the output unit 64 by executing a program described below.

The control device 60 may be separate from the wearable robot 1. In this case, the control device 60 is connected to the posture measurement unit 40, the angle detection unit E, and the drive unit M in a wired or wireless manner so as to be able to communicate with them. In other words, the multi-joint arm 30 is remotely controlled by the control device 60. In this case, the control device 60 may be, for example, a smartphone.

The input unit 61 acquires the measurement results of the orientation measurement unit 40, the detection results of the multiple angle detection units E, and the images captured by the camera 50.

The measurement results of the posture measurement unit 40 are the changes in the leg pitch angle θ2, leg yaw angle ψ2, and leg roll angle φ2 (hereinafter referred to as posture change amount ΔΦ (= {θ2, ψ2, φ2})). The input unit 61 transmits the posture change amount ΔΦ to the arm trajectory control unit 62.

The detection results of the multiple angle detection units E are represented by _q1 , _q2 , _q3 , _q4 , and _q5 . _q1 is the rotation angle of the first drive unit M1. _q2 is the rotation angle of the second drive unit M2. _q3 is the rotation angle of the third drive unit M3. _q4 is the rotation angle of the fourth drive unit M4. _q5 is the rotation angle of the fifth drive unit M5. Hereinafter, the rotation angle of the drive unit M will be referred to as the joint angle. The input unit 61 transmits the current joint angle q (={ _q1 , _q2 , _q3 , _q4 , _q5 }), which is the current joint angle of each drive unit M1 to M5, to the arm trajectory control unit 62.

The input unit 61 transmits the images captured by the camera 50 to the work processing unit 63.

The arm trajectory control unit 62 controls the operation of the articulated arm 30 by executing a program (described later) that is pre-stored in a memory area (not shown) of the control device 60. Specifically, the arm trajectory control unit 62 acquires a target position and a movement time, and controls the operation of the articulated arm 30 so as to move the position of the hand 36 to the acquired target position in the acquired movement time. Specifically, the arm trajectory control unit 62 calculates the operation amount (specifically, angular velocity) of each of the drive units M1 to M5. The arm trajectory control unit 62 acquires the target position and the movement time from the work processing unit 63 (details will be described later).

The arm trajectory control unit 62 includes a trajectory generation unit 62a, a swing correction unit 62b, an inverse kinematics calculation unit 62c, an angle error calculation unit 62d, and an angle error compensation unit 62e. The arm trajectory control unit 62 functions as the trajectory generation unit 62a, the swing correction unit 62b, the inverse kinematics calculation unit 62c, the angle error calculation unit 62d, and the angle error compensation unit 62e by the processor executing a program described below.

The trajectory generating unit 62a generates time series data of the hand tip vector r, which corresponds to the trajectory of the position of the hand 36 as it moves from the current position to the target position over a movement time. The current position of the hand 36 corresponds to the previous target position acquired by the arm trajectory control unit 62. The previous target position is stored in the memory area.

The trajectory generating unit 62a outputs the hand vector r in order along the generated time series data from the hand vector r indicating the previous target position to the hand vector r indicating the acquired target position (i.e., the current target position). The hand vector r output by the trajectory generating unit 62a is referred to as the trajectory hand vector r _d .

The swing correction unit 62b corrects the trajectory hand vector r _d output from the trajectory generation unit 62a in accordance with the swing of the human body H. The swing of the human body H is transmitted from the posture measurement unit 40 via the input unit 61 as a posture change amount ΔΦ. The swing correction unit 62b calculates a corrected hand vector r _dm by correcting the trajectory hand vector r _d based on the formula (1).

(Equation 1)
r _dm = R _ψ R _θ R _φ {r _d + [0,0,L] ^T }...(1)

R _ψ , R _θ , and R _φ are rotation matrices of the leg roll angle φ2, the leg yaw angle ψ2, and the leg pitch angle θ2, respectively. L is the distance from the tip of the supporting leg 20 to the origin of the reference coordinate system Σ _0. Equation (1) converts the posture change amount ΔΦ of the supporting leg 20 caused by the swinging of the human body H into the movement amount of the multi-joint arm 30, and corrects the trajectory hand vector r _d . The movement amount of the multi-joint arm 30 is specifically the displacement amount seen from the tip of the supporting leg 20 of the origin of the reference coordinate system Σ ₀ corresponding to the base end of the multi-joint arm 30, and the rotation amount of the supporting leg 20 around the axis 20a.

It should be noted that formula (1) is based on the premise that the support leg 20 inclines with respect to the horizontal plane and rotates around the axis 20a when the tip of the support leg 20 is in contact with the support surface F and does not move with respect to the support surface F. Formula (1) is based on the premise that the origin of the reference coordinate system Σ ₀ is located on the axis 20a of the support leg 20 and the axis 20a of the support leg 20 and the reference Z-axis Z ₀ are coaxial. Since formula (1) is thus constructed, the operation of the multi-joint arm 30 and the position of the hand 36 can be easily corrected based on the amount of movement of the multi-joint arm 30 calculated using the inclination of the support leg 20 with respect to the horizontal plane caused by the swing of the human body H and the rotation angle of the support leg 20 around the axis 20a.

The inverse kinematics calculation unit 62c calculates the target joint angle _qdm for each of the actuators M1 to M5 using the calculated corrected hand vector _rdm based on the inverse kinematics. The target joint angle _qdm indicates the target joint angle for each of the actuators M1 to M5.

The angle error calculation unit 62d calculates the difference _qe between the calculated target joint angle _qdm and the current joint angle q transmitted from the input unit 61 for each of the drive units M1 to M5.

The angular error compensating unit 62e calculates a target angular velocity _ωd , which is a target angular velocity, for each of the driving units M1 to M5 based on the calculated difference _qe , and transmits the calculated target angular velocity ωd to the output unit 64. The angular error compensating unit 62e is, for example, a PID compensator. Specifically, the angular error compensating unit 62e calculates the target angular velocity _ωd so as to converge the difference _qe to zero by appropriately adjusting three gains, proportional, differential, and integral, which are diagonal matrices of constants. The angular error compensating unit 62e transmits the calculated target angular velocity _ωd to the output unit 64.

When the program described below is executed, the work processing unit 63 sets a target position and a movement time so that the articulated arm 30 performs a predetermined operation. The predetermined operation is a home position operation and a work operation.

The home position operation is the operation of the articulated arm 30 that moves the position of the hand 36 from the current position to a predetermined reference position P1 (Figure 7A). The work operation is the operation of the articulated arm 30 that moves the position of the hand 36 from the standby position P2 (Figure 7B) to the work position P3 (Figure 7C), keeps the hand 36 at the work position P3 for a predetermined work time, and then returns to the standby position P2. If the target work is driving screws, the work position P3 is the position where the screws are driven. The standby position P2 is a position that is a predetermined distance below the work position P3. The predetermined work time is the time required to drive the screws.

The work processing unit 63 also controls the camera 50 to acquire an image from the camera 50. The work processing unit 63 then sets the target position by executing image processing, described below, on the acquired image. The work processing unit 63 also sets a movement time based on the movement distance from the previous target position to the current target position. The movement time is set to become longer as the movement distance increases. The work processing unit 63 transmits the set target position and movement time to the arm trajectory control unit 62.

The output section 64 converts the target angular velocity ω _d transmitted from the arm trajectory control section 62 and the control amount transmitted from the work processing section 63 into control command values, and outputs them to the respective drive sections M1 to M5.

Next, the operation of the wearable robot 1 and the program executed by the control device 60 when a worker wearing the wearable robot 1 causes the wearable robot 1 to perform a target task will be described with reference to the flowcharts in FIGS. 5 and 6. In this embodiment, the target task is driving screws into a ceiling panel.

The program represented by the flowchart in FIG. 5 is executed by the work processing unit 63. The program represented by the flowchart in FIG. 6 is executed by the arm trajectory control unit 62. In addition, the arm trajectory control unit 62 constantly executes the program when the power supply of the wearable robot 1 is on.

Before making the wearable robot 1 perform the desired task, the worker brings the tip of the support leg 20 into contact with the support surface F and adjusts the worker's own posture so that the axis 20a of the support leg 20 is roughly aligned vertically. When the worker performs a predetermined operation (e.g., turns on a start switch (not shown) provided on the mounting unit 10), the task processing unit 63 starts the program, and the desired task begins.

In S10, the work processing unit 63 causes the articulated arm 30 to perform a home position operation. Specifically, the work processing unit 63 sets the target position to the reference position P1, and sets a movement time based on the distance between the current position of the hand 36 (i.e., the previous target position) and the set target position (i.e., the reference position P1, which is the current target position). The work processing unit 63 transmits the set target position and movement time to the arm trajectory control unit 62.

In response to acquiring the target position and the movement time, the arm trajectory control unit 62 executes S30 shown in Fig. 6. In S30, the trajectory generation unit 62a generates time series data of the hand vector r based on the previous target position stored in the storage area, the acquired current target position, and the acquired movement time. Furthermore, the arm trajectory control unit 62 outputs a trajectory hand vector r _d from the generated time series data of the hand vector r.

Next, the swing correction unit 62b acquires the posture change amount ΔΦ in S32, and calculates the corrected hand vector _{r_dm} from the trajectory hand vector _{r_d} output in S34.

In S36, the inverse kinematics calculation unit 62c calculates the target joint angle _qdm from the calculated corrected hand vector _rdm . Then, in S38, the angle error calculation unit 62d obtains the current joint angle q, and in S40, calculates the difference _qe between the calculated target joint angle _qdm and the current joint angle q.

Furthermore, the angular error compensation unit 62e calculates a target angular velocity _ωd based on the calculated difference _qe in S42. Then, the output unit 64 outputs the calculated target angular velocity _ωd as a control command value in S44.

In response to the control command values being sent to each of the drive units M1 to M5, each of the drive units M1 to M5 and thus the articulated arm 30 operate, and the hand 36 moves from its current position toward the reference position P1, which is set as the target position.

Even if the human body H sways or tilts while the hand 36 is moving, the trajectory hand vector r _d is corrected based on the change in posture of the support leg 20 caused by the swaying or tilting of the human body H, so that the position reached by the hand 36 is prevented from shifting from the target position.

As the articulated arm 30 operates over the set movement time, the hand 36 moves from the previous target position to the reference position P1 (FIG. 7A). When the set movement time has elapsed, the home position operation of S10 ends. Next, the work processing unit 63 calibrates the orientation measurement unit 40 in S12.

Calibration is a process in which, when the gravity sensor 41 detects that the axis 20a of the support leg 20 has remained approximately vertical to the support surface F for a predetermined period of time, the leg pitch angle θ2, leg yaw angle ψ2, and leg roll angle φ2 at the end of the predetermined period of time are set to reference values (zero). The operator maintains his or her own posture so that the axis 20a of the support leg 20 is approximately vertical until the calibration is completed. When the calibration is completed, the operator is notified that the calibration has been completed by, for example, a buzzer sound generated by the control device 60.

Furthermore, in S14, the work processing unit 63 detects the standby position P2 and the work position P3. Specifically, the work processing unit 63 controls the camera 50 to obtain an image of, for example, a ceiling panel placed on the ceiling, and detects the contour of the ceiling panel from the obtained image by image processing (specifically, edge detection).

Next, the work processing unit 63 detects the work position P3 where the screws are to be driven in based on the detected contour of the ceiling panel. Furthermore, the work processing unit 63 detects the waiting position P2 that is a predetermined distance below the work position P3. In this embodiment, multiple work positions P3 and multiple waiting positions P2 (e.g., three) are detected.

Then, in S16, the work processing unit 63 causes the multi-joint arm 30 to perform a work operation. Specifically, the work processing unit 63 sets the standby position P2 closest to the hand 36 among the detected standby positions P2 as a new target position for the hand 36. Furthermore, the work processing unit 63 sets a new movement time based on the distance between the current position (reference position P1) of the hand 36 and the newly set target position (standby position P2). The work processing unit 63 transmits the newly set target position and movement time to the arm trajectory control unit 62.

In response to acquiring the newly set target position (standby position P2) and movement time, the arm trajectory control unit 62 executes the program represented by the flowchart in FIG. 6 as described above. As a result, the articulated arm 30 operates for the newly set movement time, and the hand 36 moves from the reference position P1 to the standby position P2 (FIG. 7B).

Even if the human body H sways or tilts while the hand 36 is moving, the trajectory hand vector r _d is corrected based on the change in posture of the support leg 20 caused by the swaying or tilting of the human body H, so that the position reached by the hand 36 is prevented from shifting from the target position.

Next, the work processing unit 63 sets a new target position for the hand 36 to a work position P3 corresponding to the current standby position P2 as the result of the elapse of the set movement time. Furthermore, the work processing unit 63 sets a new movement time based on the distance between the current position (standby position P2) of the hand 36 and the newly set target position (work position P3). The work processing unit 63 transmits the newly set target position and movement time to the arm trajectory control unit 62.

In response to acquiring the newly set target position (working position P3) and movement time, the arm trajectory control unit 62 executes the program represented by the flowchart in FIG. 6 as described above. As a result, the articulated arm 30 operates for the newly set movement time, and the hand 36 moves from the standby position P2 to the working position P3 (FIG. 7C).

When the hand 36 is positioned at the working position P3, the tool D held by the hand 36, which is an electric screwdriver, starts to operate, for example by the tip of the electric screwdriver coming into contact with the ceiling panel and being pushed in, to drive the screws. The screw driving is completed when a predetermined operating time has elapsed.

Next, the work processing unit 63 sets a new target position for the hand 36 to a standby position P2 corresponding to the current work position P3 in response to the passage of a predetermined operating time. Furthermore, the work processing unit 63 sets a new movement time based on the distance between the current position of the hand 36 (work position P3) and the newly set target position (standby position P2). The work processing unit 63 transmits the newly set target position and movement time to the arm trajectory control unit 62.

In response to acquiring the newly set target position (standby position P2) and movement time, the arm trajectory control unit 62 executes the program represented by the flowchart in FIG. 6 as described above. As a result, the articulated arm 30 operates for the newly set movement time, and the hand 36 moves from the working position P3 to the waiting position P2 (FIG. 7B). When the hand 36 is moved from the working position P3 to the waiting position P2, the work operation of S16 ends when the newly set movement time has elapsed. At this time, the target work is completed for one of the multiple working positions P3.

Furthermore, in S18, the work processing unit 63 determines whether all work has been completed. If work has not been completed for all of the multiple work positions P3, not all work has been completed. In this case (NO in S18), the work processing unit 63 returns the program to S16 and causes the multi-joint arm 30 to perform the work operation again. Specifically, the work processing unit 63 sets the standby position P2 corresponding to the work position P3 closest to the hand 36 among the work positions P3 for which work has not been completed as the target position, and causes the multi-joint arm 30 to start the work operation.

On the other hand, if work has been completed for all of the multiple work positions P3 (YES in S18), the work processing unit 63 causes the multi-joint arm 30 to execute a home position operation in S20, and moves the hand 36 to the reference position P1. This causes the multi-joint arm 30 to operate, and the hand 36 is positioned at the reference position P1 (FIG. 7A).

In this way, the wearable robot 1 can perform the intended task with high precision even if the human body H sways. In addition, the load on the human body H of the wearable robot 1 is reduced by the support legs 20. Then, while the wearable robot 1 is performing the intended task, the worker can perform a task other than that of the wearable robot 1. This makes it easy for humans and robots to work together.

Second Embodiment
Next, the wearable robot 1 according to a second embodiment of the present invention will be described with reference to Figures 8 and 9, mainly focusing on the differences from the first embodiment. In the wearable robot 1 according to the second embodiment, the support leg 120 includes a leg portion 121, a leg tip portion 122, a joint portion 123 that rotatably connects the leg portion 121 and the leg tip portion 122, and a movement amount measuring unit 124.

The leg 121 is configured to be extendable and contractible, similar to the support leg 20 of the first embodiment described above.

The leg tip 122 is connected to the tip (lower end) of the leg 121 via a joint 123. The bottom surface of the leg tip 122 has a larger area in contact with the support surface F than the tip of the support leg 20 of the first embodiment. Therefore, the frictional force of the leg tip 122 against the support surface F is larger than that of the tip of the support leg 20 of the first embodiment, and therefore movement relative to the support surface F is suppressed when the wearable robot 1 is performing the intended task.

When the bottom surface of the leg tip 122 is in contact with the support surface F, the leg 121 rotates relative to the leg tip 122. Specifically, the leg 121 rotates relative to the leg tip 122 around the sixth joint axis J6 and the seventh joint axis J7 of the joint portion 123. The sixth joint axis J6 and the seventh joint axis J7 are perpendicular to each other and are parallel to the bottom surface of the leg tip 122. The axis 20a of the leg 121 passes through the intersection of the sixth joint axis J6 and the seventh joint axis J7. The leg 121 also rotates relative to the leg tip 122 around the axis 121a of the leg 121.

The movement amount measuring unit 124 measures the amount of movement of the tip of the support leg 120 relative to the support surface F. The amount of movement of the tip of the support leg 120 relative to the support surface F is specifically the amount of movement of the leg tip portion 122 along the horizontal direction relative to the support surface F. The movement amount measuring unit 124 includes a laser light source 124a and an image sensor 124b.

The laser light source 124a irradiates the support surface F with laser light. The image sensor 124b is a sensor that acquires an image of the support surface F irradiated with the laser light and detects the amount of movement of the leg tip 122 along the horizontal direction. The amount of movement ΔS of the leg tip 122 along the horizontal direction is represented by the amount of movement (ΔSx and ΔSy) on each of two axes that are parallel to the horizontal plane and perpendicular to each other. The amount of movement ΔS (= {ΔSx, ΔSy}) detected by the image sensor 124b is transmitted to the control device 60.

Moreover, in the second embodiment, the swing correction unit 62b calculates a corrected hand vector _{r_dm} by correcting the trajectory hand vector _{r_d} based on the formula (2).

(Equation 2)
r _dm = R _ψ R _θ R _φ {r _d + [0,0,L] ^T }+ΔS...(2)

Here, ΔS=[ΔSx, ΔSy, 0] ^T . In this way, the correction of the trajectory hand vector r _d is performed not only for the posture change amount ΔΦ of the support leg 120 but also for the movement amount ΔS of the leg tip 122 along the horizontal direction. Therefore, the position of the hand 36 can be corrected with high accuracy. In addition, since the movement amount measuring unit 124 measures the movement amount ΔS in a non-contact state with the support surface F, it can stably measure the movement amount ΔS even when the support surface F is dirty.

<Modification of the second embodiment>
Next, a wearable robot 1 according to a modification of the second embodiment described above will be described with reference to Fig. 10. The wearable robot 1 according to this modification may include first and second angle sensors A26 and A27 instead of the gravity sensor 41 of the attitude measurement unit 40 according to the first embodiment described above.

The first and second angle sensors A26 and A27 measure the angle of the support leg 220 with respect to the support surface F. The first and second angle sensors A26 and A27 are configured with encoders. Specifically, the first angle sensor A26 detects the rotation angle (θ3) of the leg 221 around the sixth joint axis J6 relative to the leg tip 222. The second angle sensor A27 detects the rotation angle (ψ3) of the leg 221 around the seventh joint axis J7 relative to the leg tip 222. In this case, instead of the leg pitch angle θ2 and leg yaw angle ψ2 output from the gravity sensor 41 in the first embodiment described above, the rotation angles detected by the first and second angle sensors A26 and A27 are used as elements of the posture change amount ΔΦ.

In addition, the wearable robot 1 in this modified example may be provided with a third angle sensor A28 that detects the rotation angle of the leg 221 relative to the leg tip 222 around the axis 221a of the leg 221, instead of the gyro sensor 42 of the posture measurement unit 40 in the first embodiment described above. The third angle sensor A28 is, for example, an encoder. In this case, the rotation angle (φ3) detected by the third angle sensor A28 is used as an element of the posture change amount ΔΦ, instead of the leg roll angle φ2 output from the gyro sensor 42 in the first embodiment described above.

The above-mentioned movement amount measuring unit 124 is composed of a laser light source 124a and an image sensor 124b, but the movement amount measuring unit 224 of this modified example includes a rolling unit 224a and a rolling amount measuring unit 224b. The rolling unit 224a is formed in a spherical shape and contacts the support surface F and rolls on the support surface F. The rolling amount measuring unit 224b measures the rolling amount of the rolling unit 224a, and is composed of, for example, an encoder (not shown) that detects the amount of rotation of a shaft member (not shown) that contacts the rolling unit 224a and rotates in response to the rolling of the rolling unit 224a. The movement amount measuring unit 224 calculates the movement amount ΔS based on the measurement result of the rolling amount measuring unit 224b. When the movement amount measuring unit 224 uses the rolling unit 224a and the rolling amount measuring unit 224b, the movement amount ΔS can be continuously measured with high accuracy even when the pattern on the support surface F is difficult to distinguish using the image sensor 124b described above.

Third Embodiment
Next, the wearable robot 1 according to a third embodiment of the present invention will be described with reference to Figures 11 and 12, focusing mainly on the differences from the first embodiment. The control device 60 according to the third embodiment selects one of the motion correction mode and the target position change mode to control the motion of the articulated arm 30.

The motion correction mode is a control mode in which the control device 60 corrects the motion of the multi-joint arm 30 based on the amount of movement of the multi-joint arm 30 calculated using the measurement results of the posture measurement unit 40, as in the first embodiment described above.

The target position change mode is a mode in which the control device 60 changes the target position using the measurement results of the posture measurement unit 40. In the target position change mode, the worker can change the posture of the support leg 20 to move the articulated arm 30 and therefore the hand 36 a desired distance (details will be described later).

The wearable robot 1 of the third embodiment further includes a swing operation unit 362f, a trajectory switching unit 362g, and a changeover switch 370. The swing operation unit 362f and the trajectory switching unit 362g are provided in the control device 60.

The swing operation unit 362f acquires the trajectory hand vector r _d from the trajectory generation unit 62a, and acquires the posture change amount ΔΦ from the posture measurement unit 40 via the input unit 61. The swing operation unit 362f calculates the changed hand vector r _dc from the acquired trajectory hand vector r _d based on the formula (3).

(Equation 3)
r _dc = r _d +KcΔΦ...(3)

Here, Kc is a constant gain matrix. Also, ΔΦ=[θ2, ψ2, φ2] ^T. In formula (3), the modified hand vector r _dc is calculated by adding the posture change amount ΔΦ multiplied by the constant gain matrix to the trajectory hand vector r _d .

The trajectory switching unit 362g acquires the corrected hand vector r _dm from the swing correction unit 62b, and acquires the changed hand vector r _dc from the swing operation unit 362f. The trajectory switching unit 362g outputs the hand vector r corresponding to the control mode. Specifically, the trajectory switching unit 362g outputs the corrected hand vector r _dm in the motion correction mode, as in the first embodiment described above. Also, the trajectory switching unit 362g outputs the changed hand vector r _dc in the target position change mode. When the trajectory switching unit 362g acquires an operation signal described later, it switches the hand vector r to be output.

The changeover switch 370 is a switch for switching the control mode. The changeover switch 370 is arranged, for example, on the mounting unit 10. When the operator operates the changeover switch 370, an operation signal is output to the control device 60. The operation signal is acquired by the track switching unit 362g via the input unit 61.

Next, the operation of the wearable robot 1 when the control mode is the target position change mode will be described. The wearable robot 1 executes the program shown in the flowchart of Fig. 6 described above, and the control mode will be described from the state where the control mode is the motion correction mode. In the motion correction mode, the control device 60 controls the motion of the multi-joint arm 30 so as to position the hand 36 at the target position set in S16 of Fig. 5 while correcting the motion of the multi-joint arm 30 in response to the swinging of the human body H. As described above, the trajectory switching unit 362g outputs the correction hand vector r _dm in the motion correction mode.

In this state, when the operator operates the changeover switch 370, an operation signal is transmitted and the control mode is switched from the motion correction mode to the target position change mode. In response to receiving the operation signal, the trajectory switching unit 362g switches the output hand vector r from the corrected hand vector r _dm to the changed hand vector r _dc . Furthermore, the operator changes the posture of the support leg 20.

As described above, the modified hand vector r _dc is calculated by adding the posture change amount ΔΦ multiplied by the constant gain matrix to the trajectory hand vector r _d . Therefore, when the posture of the support leg 20 changes, the articulated arm 30 and the hand 36 move a distance proportional to the amount of change. For example, when the leg pitch angle θ2 changes by Δθ (FIG. 12), the position of the hand 36 moves by Δx in the direction of the reference X-axis _X0 .

When the hand 36 has moved the distance desired by the operator, the operator operates the changeover switch 370 to return the control mode from the target position change mode to the motion correction mode. In response to this, the trajectory switching unit 362g switches the hand vector r to be output from the changed hand vector r _dc to the corrected hand vector r _dm . As a result, the arm trajectory control unit 62 controls the hand 36 to be positioned at a position moved by the distance desired by the operator from the target position set in S16 of FIG. 5.

In this way, the target position change mode allows the operator to adjust the target position of the hand 36 by changing the posture of the support leg 20. This allows the operator to improve the work efficiency of the wearable robot 1. In addition, the support leg 20 can be used as an interface to change the target position of the hand 36.

<Modification>
The present invention is not limited to the embodiments described above. As long as they do not deviate from the gist of the present invention, various modifications of the present embodiments and combinations of components in different embodiments are also included within the scope of the present invention.

For example, the base end of the support leg 20 is attached to the mounting part 10, but instead, it may be attached to, for example, the base 31 of the articulated arm 30.

The gyro sensor 42 of the posture measurement unit 40 is used to measure the leg roll angle φ2, but may also be used to measure at least one of the leg pitch angle θ2 and the leg yaw angle ψ2. When the gyro sensor 42 is used to measure both the leg pitch angle θ2 and the leg yaw angle ψ2, the posture measurement unit 40 does not include a gravity sensor 41 for measuring the leg pitch angle θ2 and the leg yaw angle ψ2. In this way, the gyro sensor 42 may measure the angle of the support leg 20 with respect to the vertical direction instead of the gravity sensor 41.

Furthermore, the origin of the reference coordinate system _Σ0 is located on the axis 20a of the support leg 20, but may be located at a position other than on the axis 20a of the support leg 20. Furthermore, the reference coordinate system _Σ0 is disposed on the base 31, but may be disposed on a component of the multi-joint arm 30 other than the base 31. Note that if the position of the origin of the reference coordinate system _Σ0 is changed to a position other than the base end of the multi-joint arm 30, the position of the origin of the reference coordinate system _Σ0 relative to the tip of the support leg 20 changes, and therefore the above-mentioned formula (1) is changed.

In addition, the movement amount measuring units 124, 224 are arranged at the leg tip portions 122, 222 in the second embodiment, but may be arranged at the lower end portions of the support legs 20 in the first embodiment.

In addition, in the second embodiment, the leg tip portion 122 is equipped with a movement amount measuring unit 124, but alternatively, it may not be equipped with the movement amount measuring unit 124.

The wearable robot 1 also has one support leg 20, but as shown in FIG. 13, it may have two support legs 20 to further reduce the load on the human body H.

As described above, the control device 60 may be separate from the wearable robot 1. Also, a camera or the like installed at a work site or the like that is separate from the wearable robot 1 and capable of measuring the posture of the wearable robot 1 may replace the posture measuring unit 40. Also, a camera or the like installed at a work site or the like that is separate from the wearable robot 1 and capable of measuring the amount of movement of the tip of the support leg 20 relative to the support surface F may replace the movement amount measuring unit 124.

Additionally, the drive unit M is a motor, but it may be a pneumatic actuator, a solenoid, or a linear motor, or may be a combination of these.

In addition, multiple drive units M are arranged on the multi-joint arm 30, but at least one drive unit M may be configured to drive multiple joints.

This invention can be widely used in wearable robots.

REFERENCE SIGNS LIST 1 wearable robot 10 wearable unit 20 support leg 20a axis 30 multi-joint arm 40 attitude measurement unit 41 gravity sensor 42 gyro sensor 60 control device 121 leg 122 leg tip 124 movement amount measurement unit 124b image sensor 224 movement amount measurement unit 224a rolling unit 224b rolling amount measurement unit A26 first angle sensor A27 second angle sensor A28 third angle sensor F support surface H human body

Claims (11)

A multi-joint arm;
a mounting part to which a base end of the articulated arm is attached and which is configured to be detachable from a human body;
a support leg that supports the articulated arm by contacting a support surface;
a posture measuring unit that measures the posture of the support leg;
a control device that controls the operation of the multi-joint arm and corrects the operation of the multi-joint arm based on a movement amount of the multi-joint arm calculated using the measurement result of the posture measurement unit,
Wearable robot. the control device corrects the movement of the multi-joint arm by using a movement amount of the base end of the multi-joint arm calculated based on the measurement result of the posture measurement unit.
The wearable robot according to claim 1 . The posture measurement unit includes a gyro sensor for detecting an angle of the support leg.
The wearable robot according to claim 1 or 2 . The posture measurement unit further includes a gravity sensor for detecting an angle of the support leg with respect to a vertical direction.
The wearable robot according to any one of claims 1 to 3 . The posture measurement unit includes an angle sensor that measures an angle of the support leg relative to the support surface.
The wearable robot according to any one of claims 1 to 3 . A movement amount measuring unit that measures a movement amount of the tip of the support leg relative to the support surface,
the control device corrects the operation of the multi-joint arm based on a movement amount of the multi-joint arm calculated using the measurement result of the attitude measurement unit and the measurement result of the movement amount measurement unit.
The wearable robot according to any one of claims 1 to 5 . The movement amount measuring unit is
an image sensor for acquiring an image of the support surface;
measuring the amount of movement of the tip of the support leg based on the image of the support surface acquired by the image sensor;
The wearable robot according to claim 6 . The movement amount measuring unit is
A rolling part that rolls on the support surface;
A rolling amount measuring unit that measures the rolling amount of the rolling part,
measuring a movement amount of the tip of the support leg based on the rolling amount measured by the rolling amount measuring unit;
The wearable robot according to claim 6 . the control device selects one of a motion correction mode in which the motion of the multi-joint arm is corrected based on a movement amount of the multi-joint arm calculated using the measurement result of the posture measurement unit, and a target position change mode in which a target position of the tip of the multi-joint arm is changed based on the measurement result of the posture measurement unit, to control the motion of the multi-joint arm.
The wearable robot according to any one of claims 1 to 8 . The support leg may be one or two.
The wearable robot according to any one of claims 1 to 9 . The support legs include:
The legs,
The leg portion is rotatably disposed at the tip of the leg portion and is provided with a leg tip portion that contacts the support surface.
The wearable robot according to any one of claims 1 to 10 .

Images