TWI656940B - Tuning system, simulation unit and tuning and simulation method thereof - Google Patents

Abstract

A tuning method is to provide a tuning system configured with a virtual machine, wherein the virtual machine is modeled on a target machine, and the processing conditions of the target machine are set on the tuning system, and An analog command including an analog parameter is input to the virtual machine, and then the virtual machine performs a simulation operation of the target machine according to the simulation parameter to calculate a required control parameter, and then the control parameter is used as an adjustment. The basic parameters of the target machine.

Description

Tuning method and its tuning system and simulation unit and simulation method

This case relates to a method for automatically adjusting parameters, especially a method for adjusting the parameters of the user and adjusting the system.

With the rapid development of machine tool automation, the use of input related parameters for processing operations has become the mainstream today. Usually, the machine tool will go through the adjustment process before leaving the factory, and adjust the relevant parameters to achieve certain performance indicators, and then meet the processing requirements specified by the customer.

However, after the machine is shipped from the factory, it often affects the performance of the original organic platform due to on-site installation, site construction and environmental issues. In addition, mechanical components undergo long-term operation, which are subject to vibration, friction, dust, or structural deformation, which may cause the characteristics of the machine to change, resulting in deterioration of performance. It is quite difficult to implement the practice in order to disassemble the machine to repair the mechanical components or structures and improve the performance of the machine.

Furthermore, regarding the adjustment of the parameters of the controller of the machine, the user needs to be quite familiar with the function of the machine, and understand the performance of the machine, and at the same time have certain expertise in the design and processing of the machine. In other words, the user must be a senior operator, otherwise the machine on the factory line cannot be adjusted immediately.

Therefore, how to enable the general operator to instantly adjust various machines on the factory production line is a technical problem that is urgently solved by those skilled in the art.

In view of the above-mentioned various deficiencies of the prior art, the present disclosure discloses a tuning method, which includes: providing a tuning system configured with a virtual machine, wherein the virtual machine is modeled on a target machine; Setting a processing condition of the target machine; inputting an analog command including a simulation parameter into the virtual machine; and performing a simulation operation of the target machine response by the virtual machine according to the simulation parameter, and calculating the current situation immediately The control parameters are required to make the control parameter as a basic parameter for adjusting the target machine.

In the foregoing tuning method, the tuning system is further configured with an authentication module, and before the processing condition is set, the tuning method further comprises: inputting another simulation command including another simulation parameter into the target machine, Having the target machine generate response information; transmitting the response information to the authentication module; and the authentication module compares the other analog command with the response information, if the error of the other analog parameter is within a reasonable range, Then, the other simulation parameter and the response information are used as system parameters. Further, the tuning system is further configured with an analog unit for transmitting the another analog command to the target machine and the authentication module. For example, the target machine has a controller and the analog unit is used to simulate the controller.

In the foregoing tuning method, the tuning system is further configured with a reference model, which is constructed by using the virtual machine and has a control parameter, and the tuning method includes: selecting the reference according to the processing condition of the target machine. The model is compared to the control parameter to calculate another control parameter.

The present invention discloses a tuning system, which includes: a host, configured with a virtual machine, wherein the virtual machine is modeled on a target machine; and an authentication module is used to identify the target machine to construct the virtual machine. And an analog unit for simulating a controller of the target machine for identification by the authentication module.

In the foregoing tuning system, the analog unit is further configured to transmit an analog command to the target machine and the authentication module for identification by the authentication module.

The present invention discloses a simulation method for simulating a controller, the controller is for controlling the motion of a target machine, the simulation method includes: setting a parameter of a simulated motion trajectory; calculating an analog command according to the parameter; The simulation command is input to the target machine for response comparison to determine whether the analog command corresponds to a target command used by the controller to perform the manipulation of the target machine.

The present disclosure discloses a simulation unit, which includes: a data collection unit for setting parameters of a simulated motion trajectory; a calculation unit for calculating an analog command according to the parameter; and an analysis unit for using the simulation command A target machine is input for response comparison to determine whether the analog command corresponds to a target command for executing a controller that controls the target machine.

In the foregoing analog unit and its simulation method, the analog unit is another controller.

In the foregoing simulation unit and the simulation method thereof, the data collection unit sets parameters of the simulated motion trajectory by collecting motion parameter information of the controller.

In the foregoing simulation unit and its simulation method, the analysis unit is Simulate the correlation between the motion trajectory and the action of the target machine to determine whether the simulation command corresponds to the target command.

It can be seen from the above that the tuning method and the tuning system of the present invention mainly enable the virtual machine to acquire the relevant parameters of the target machine by simulating the virtual machine to simulate the response of the target machine, and Calculate the required control parameters, and then provide the control parameters as the initial value of the parameters of the target machine. Therefore, compared with the prior art, the tuning method of the present case is adopted, regardless of the user's work experience. The control parameters can be quickly input to the target machine, thereby greatly reducing the time for the user to adjust the parameters of the target machine.

1‧‧‧simulation unit

10‧‧‧Information Collection Department

11‧‧‧ Calculation Department

12‧‧‧Analysis Department

1a‧‧‧ trajectory

1b‧‧‧Controller type

1c‧‧‧ model formula

1d‧‧‧simulation time

2‧‧‧Target machine

20‧‧‧ Controller

3‧‧‧ Identification module

3a‧‧‧System parameter values

3b‧‧‧System type

3c, 30‧‧‧ model of system parameters

31‧‧‧ Dynamic simulation of system parameters

40‧‧‧Dynamic simulation

41‧‧‧Execution options

9‧‧‧ Tuning system

9a‧‧‧Virtual Machine

90‧‧‧Host

91‧‧‧ screen

a, b‧‧‧ values

t, t’‧‧‧ error value

d‧‧‧Value of target system parameters

D’‧‧‧ Estimated system parameter values

E1‧‧‧Amplitude chart

E2‧‧‧ phase chart

L, L’‧‧‧ actual movement path

S, S’‧‧‧ simulated motion path

R1, R3‧‧‧ solid line

R2, R4‧‧‧ dotted line

T1, T2, T3, Ta, Tb, Tc‧‧ ‧ time

U1-U4‧‧‧ Curve

V1-V4‧‧‧ speed curve

P1-P4‧‧‧ acceleration curve

X, Y’, Z‧‧‧ arrow direction

S21~S27‧‧‧Steps

S31~S37‧‧‧Steps

S41~S47‧‧‧Steps

S51~S55‧‧‧Steps

S60~S65‧‧‧ steps

1A is a schematic diagram of the application configuration of the tuning system of the present case; FIG. 1B is a schematic diagram of the functional architecture of the tuning method of the present invention; FIG. 2A is a schematic diagram of the simulation operation of the simulation unit of the present embodiment; 2B to 2C The diagram is a schematic diagram of the simulation parameters generated by the data collection unit of the simulation unit of the present case; the 2D-1 to 2D-4 diagrams are schematic diagrams for generating different simulation trajectories of the simulation unit of the present case; the 3A diagram is the virtual machine platform cooperation of the case. Schematic diagram of the construction process of the authentication module; the 3B-1 and 3B-2 diagrams are schematic diagrams of the interface of the authentication module of the present invention; and the 3C diagram is a flow chart for system identification of the tuning system of the present invention; Figure 4A is a schematic diagram of the process of the tuning method of the present case; Figure 4B is a schematic diagram of the processing requirements of the tuning method of the present case; the 4C to 4D drawings are the automatic adjustment parameters of the virtual machine of the tuning system of the present case. Schematic diagram of the interface of the process; Figure 4E is a schematic diagram of another interface of the virtual machine of the tuning system of the present case; and Figures 4F-1 to 4F-3 are the inspection state of the processing state after the adjustment of the 4B chart. Schematic diagram; Figure 5 is a schematic diagram of the process of establishing a reference model for the tuning system of the present invention; and Figure 6 is a schematic diagram of the flow of the fifth embodiment of the tuning method of the present application.

The embodiments of the present invention are described below by way of specific embodiments, and those skilled in the art can readily appreciate the other advantages and functions of the present disclosure.

It is to be understood that the structure, the proportions, the size and the like of the drawings are only used in conjunction with the disclosure of the specification for the understanding and reading of those skilled in the art, and are not intended to limit the practice of the present invention. The qualifications are not technically meaningful. Any modification of the structure, change of the proportional relationship or adjustment of the size should not be affected by the effects of the case and the objectives that can be achieved. The technical content can be covered. At the same time, the terms "一" and the like quoted in this specification are for convenience only, and are not intended to limit the case. The scope of the application, the change or adjustment of its relative relationship, is also considered to be within the scope of the case under the technical content of no substantive change.

Please refer to Figures 1A to 1B for a schematic diagram of the tuning system and the method of adjusting the same.

As shown in FIG. 1A, the tuning system 9 is used to adjust the parameters of a target machine 2, and the tuning system 9 is a computer or other device having a human-machine interface, and includes a host 90 and a screen. 91 and a simulation unit 1.

In the present embodiment, the target machine 2 is an automated control machine tool (such as a CNC lathe), which is provided with a controller 20 to control the operation of the target machine 2 (such as a machining action).

The host 90 is used to simulate the operation of the target machine 2, and the simulated condition is presented on the screen 91 in the form of a virtual machine.

In this embodiment, the virtual machine is used to simulate the motion state of the target machine 2 before and after the parameter adjustment of the controller 20. For example, the virtual machine has a plurality of interfaces for presenting parameter operations (eg, One of the interfaces shown in Figure 4D to 4F). It should be understood that the virtual machine can also selectively present the appearance of the target machine 2.

Therefore, the simulation accuracy of the virtual machine is extremely important. The construction of the virtual machine will be detailed below.

Please refer to Figures 2A to 2D-4 for a schematic diagram of the simulation unit 1 and its simulation work in this case.

Since the application of the virtual machine involves various parameters of the controller 20, it is necessary to configure a device capable of simulating the controller 20, that is, the analog unit 1. Specifically, the analog unit 1 is used to simulate the controller 20, The configuration can be a special controller, for example, consisting of a control chip with associated electronic components, and can be selectively integrated into the host 90 or separately disposed outside the host 90.

In this embodiment, as shown in FIG. 2A, the functional architecture of the simulation unit 1 may include a data collection unit 10, a calculation unit 11, and an analysis unit 12. The data collecting unit 10 is configured to collect motion parameter information of the controller 20 to set a simulation parameter simulating a motion trajectory of the controller 20. The calculating unit 11 is configured to calculate at least one simulation command according to the simulation parameter. The analyzing unit 12 is configured to input the analog command to the target machine 2 for response comparison to determine whether the analog command corresponds to the controller 20 for executing the target command for manipulating the target machine 2.

Furthermore, the specific steps of the analog unit 1 for simulating the controller 20 are as follows:

First, in step S21, the data collecting unit 10 first captures the motion parameter information of the controller 20 by using a communication transmission mode, and the user arbitrarily sets a motion track (such as a processing path), such as a straight line, a turning line, an arc, and the like. There is no particular limitation, and the motion trajectory is input to the data collecting unit 10, and according to step S22, the user sets parameters of the simulated motion trajectory according to the motion trajectory, such as a spatial parameter (simulation parameter), wherein the space can be It is a two-dimensional plane space or a three-dimensional space. Next, in step S23, the user sets an acceleration/deceleration parameter (ie, an analog parameter) according to the spatial parameter, wherein the acceleration/deceleration parameter may be a pre-interference time constant (Tb), a post-complement time constant (Ta), Any one or a combination of the fulcrum front axle image acceleration (Ap) or the axial maximum speed difference limit (Vc). User set acceleration and deceleration parameters The setting state can be presented via a graph. As shown in FIG. 2B, it is a graph of the pre-compensation acceleration/deceleration parameters set by the user. The solid line R1 in the figure represents the speed curve, and the broken line R2 is the auto-converted acceleration curve. T1 is the time when the acceleration rises, T2 is the time when the acceleration rises and is flat, and T3 is the time when the acceleration rises, is flat and falls, and after tween, the curve of the acceleration and deceleration parameter before and after the tween as shown in Fig. 2C can be obtained. In the figure, the solid line R3 represents the acceleration/deceleration curve before the tween, the dotted line R4 represents the acceleration/deceleration curve after the tween, Tb is the time of the acceleration and deceleration before the tween, and T2 is the time of the acceleration and deceleration before the tween and the flat, Tc is the tween before adding When decelerating and descending, Ta is the time of acceleration and deceleration after tween, so the total acceleration and deceleration time before tween is the sum of T2 and Tc, and the total acceleration and deceleration time after tween is the sum of T2, Tc and Ta. It should be noted that if the acceleration/deceleration parameter exceeds an allowable value, the spatial parameter is reset (as in step S23').

Next, in step S24, the calculating unit 11 automatically calculates a speed command and an acceleration command according to the acceleration/deceleration parameter. For example, the speed command and the acceleration command can be calculated by using a built-in mathematical integration method, wherein There are many ways to integrate points, and there are no special restrictions. Next, in step S25, the calculation unit 11 automatically calculates a position command (simulation command) based on the speed command and the acceleration command. Specifically, since the target machine 2 cannot directly act according to the speed command and the acceleration command, the speed command and the acceleration command calculate a position command (simulation command) to make the target machine use the relative relationship of the positions. Table 2 produces an action.

Then, in step S26, the simulation command (position command) is automatically input to the target machine 2, and the analysis unit 12 performs response comparison, that is, comparison The position command and historical data (such as the controller 20 is used to execute the target command to manipulate the target machine 2) to obtain an analog error value, thereby determining whether the analog command corresponds to the target command. Specifically, if the analog error value is less than a preset value, it indicates that the analog command can correspond to the target command of the controller 20, thereby ending the simulation job (such as step S27); if the analog error value is greater than a preset value , it means that the simulation command cannot correspond to the target command of the controller 20, so the calculation unit 11 needs to recalculate the speed command and the acceleration command (step S26').

By the above simulation mode, the simulation unit 1 can effectively simulate the trajectory generation of the controller 20 of the target machine 2 (tool machine) to predict the influence of the parameter adjustment on the motion trajectory, and further analyze whether the adjustment of the control parameter is It meets the actual demand, that is, whether the trajectories of the two are close. Specifically, as shown in the 2D-1 to 2D-4 diagrams, the trajectory curve of the position of the analog unit 1 and the trajectory curve of the position of the controller 20 almost overlap, and the speed, acceleration, and acceleration of the two are The trajectory curves of the accelerations also overlap almost, so that the simulation steps of FIG. 2A can effectively obtain the control parameters of the controller 20.

It should be understood that, in steps S21-S23, if the condition set by the user does not match the condition of the controller 20, the response comparison cannot be performed in step S26.

Please refer to Figures 3A to 3C for a schematic diagram of the system identification method of this case.

When the tuning system 9 first simulates the target machine 2, the simulation unit 1 can be used to provide the function of simulating the controller 20 to construct the required The virtual machine 9a, but still needs to identify the servo system of the target machine 2, so that the operation of the virtual machine 9a is the same as the operation of the target machine 2, wherein the servo system includes a position, speed, current control loop, and The dynamic characteristics of the electrical part to understand the effects of different motor and servo parameters on the servo system.

As shown in FIG. 3A, in the present embodiment, the host 90 of the tuning system 9 is provided with an authentication module 3 for identifying the servo system of the target machine 2. When performing the system identification operation, the simulation unit 1 inputs an analog command including the analog parameters into the target machine 2 and the authentication module 3, so that the target machine 2 generates response information, and then transmits the response information to the The authentication module 3, at this time, the authentication module 3 compares the simulation command with the response information, and if the error of the simulation parameter is within a reasonable range, the simulation parameter and the response information are used as the system of the target machine 2. The parameter (such as the value 3a of the system parameter shown in FIG. 4C) is used to make the virtual machine table 9a, wherein the system parameter is, for example, a mass coefficient (M), a damping coefficient (B) or a spring coefficient (K). The combination. In other words, after the combination of the quality coefficient (M), the damping coefficient (B) or the spring coefficient (K) of the system parameter is obtained, the virtual machine table 9a can be made to simulate the real motion behavior of the target machine 2.

The calculation soft system of the identification module 3 uses the Method of Least Square (LS) for comparison calculation, as follows:

Where i = a positive integer from 1 to t ; y(i) : the actual system output; : Estimating system output; : system input matrix; : transposed matrix of system input matrix; θ : system coefficient matrix; : system estimation coefficient matrix; Φ: Transposed matrix; Y(t) : system output matrix; e(i) : estimated error; E(t) : estimated error matrix.

Next, matrix multiplication is used, as follows, to estimate an error value.

, where V( θ , t) is the output value, its equal sign right Enter the value, and E(t) ^T is the transposed matrix of the estimated error matrix, and since θ is the system coefficient matrix, The coefficient matrix is estimated for the system, so the θ value can be calculated to obtain the mass coefficient (M), the damping coefficient (B) or the spring coefficient (K).

In addition, in order to minimize the error value, another multiplication matrix can also be used, as follows: And multiply Φ ^T ( t ) at the same time, so that the right formula can find the inverse matrix, as follows: System estimation coefficient matrix For the output value, the equal sign right (Φ and Φ ^T ) is the input value, and Y is a constant. Since the above formula is a mathematical formula, the condition represented by it can be arbitrarily defined. For example, the input value can be an analog command of the simulation unit 1 including the simulation parameter, and the output value can be a mass coefficient (M), a damping coefficient. (B) or spring coefficient (K).

In this embodiment, the inspection interface of the authentication module 3 as shown in Figures 3B-1 and 3B-2 may include the system parameters (mass coefficient (M), damping coefficient (B) or spring coefficient (K). )), the model 30 of the system parameters and the dynamic simulation of the system parameters. Specifically, when the accuracy of the authentication module 3 is checked to determine the validity of the calculation method, the value d of a target system parameter can be arbitrarily input, for example, the spring coefficient (K) is adjusted from 5 to 1 for testing. Calculate the value d' of the estimated system parameter by the LS calculation method described above. (ie, 4.997976 becomes 0.99974), which is approximately equal to the value d of the target system parameter (ie, 4.99776 approaches 5, and 0.99974 approaches 1). Therefore, it can be seen that the authentication module 3 can be effectively obtained by the above LS calculation method. The system parameters of the target machine 2.

Furthermore, the system parameters such as the calculated mass coefficient (M), damping coefficient (B) and spring coefficient (K) can be modeled as follows: ms ² X(s)+bsX(s)+ksX( s)=F(s), s is a constant to be converted into the model 30 of the system parameter, and the model formula is built in the host 90, and the appropriate mode can be selected according to the obtained system parameters, so The model 30 of the system parameters is for the user to understand the physical and mechanical relationship, and thus may or may not be displayed.

Moreover, the dynamic simulation diagram 31 of the system parameter is a state curve of the system parameters such as the calculated quality coefficient (M), the damping coefficient (B) and the spring coefficient (K) when the system of the target machine 2 operates. Therefore, the dynamic simulation of the system parameters is shown in Figure 31 for the user to understand the actual state of the model 30 of the system parameters when the system of the target machine 2 is operating, as shown in Figures 3B-1 and 3B-2. The curve is thus selectable for display or not.

Therefore, the motion trajectory of the analog unit 1 (special controller) can be input to the authentication module 3 of the tuning system 9 to simulate the acceleration and deceleration of the system response in the target machine 2 (real machine tool). Specifically, as shown in FIG. 3C, first, step S31: the user first selects an input signal type (such as speed, position, or other options), and then according to steps S32 to S32: the analog unit 1 is sequentially set and processed. Path and acceleration/deceleration parameters (as in steps S21 to S23); note that if the acceleration/deceleration parameter exceeds an allowable value, then This spatial parameter is reset (as in step S33').

Then, in step S34: as in steps S24 to S25, the simulation unit 1 converts the acceleration/deceleration parameter into an analog command (speed command, acceleration command or position command), and then proceeds to step S35 by using the simulation command to obtain the target machine. The output of station 2 (ie response information).

Then, in step S36, the authentication module 3 performs system identification of the target machine 2 by using the response information of the LS algorithm and the simulation command (as input) by the calculus software of the LS algorithm to obtain a The error value is discriminated, and the relevant system parameters of the target machine 2 are obtained, thereby completing the estimation of the servo system of the target machine 2 (step S37). It should be noted that if the discrimination error is too large, the simulation unit 1 will regenerate the simulation command, as by step S36'.

It should be understood that, in steps S31-S33, if the condition set by the user does not match the condition of the controller 20, the response comparison cannot be performed in step S35.

Therefore, in order to prevent the motion track selected by the simulation unit 1 from meeting the system operation of the target machine 2, the system identification of the target machine 2 is performed by the authentication module 3 (for example, the simulation unit 1) The arc motion is required, but the system of the target machine 2 can only perform linear motion, so the authentication module 3 can be modified so that the virtual machine 9a can only perform linear motion), so when the user utilizes the tuning system 9 When constructing the virtual machine 9a, the identification module 3 is first used to perform the identification of the system of the target machine 2 to obtain the required system parameters.

Please refer to Figures 4A to 4F for the flow of the method of tuning for this case. It is intended that the tuning system 9 has completed the construction of the virtual machine 9a.

As shown in FIG. 4A, first, according to steps S41 to S42, the user sets a processing condition, such as precision requirement, speed requirement or roughness requirement, in the built-in interface (not shown) of the tuning system 9. Any one or combination thereof (as shown in Figure 4B). Next, in step S43, the user sets an error limit according to the processing condition in the built-in interface (not shown) of the tuning system 9, for example, the error limit may be a path error limit, a processing time limit, or an axial error. Any one or combination of limitations, wherein the accuracy requirement corresponds to a path error limit, the speed requirement corresponds to a processing time limit and the roughness requirement corresponds to an axial error limit. Then, in step S44, the user switches to one of the interfaces of the virtual machine 9a (as shown in FIG. 4C or 4D) to set the required tuning parameters (such as the value a of the input control parameter), and then Step S45: The user sets a required movement (machining) path (such as the motion track 1a shown in FIG. 4C) on the interface, and then causes the virtual machine 9a to automatically perform parameter calculation according to the above related setting ( Press Execute option 41). Then, in step S46, the existing parameters of the controller 20 (such as the value a of the control parameter of the PID) are input (manually or automatically) into the virtual machine 9a for calculation and measurement, thereby obtaining the Controlling the state of the parameter under the motion trajectory 1a, such as the simulated motion path S, S' presented in the dynamic simulation map 40, and comparing it to the actual motion path L, L' of the target machine 2, and obtaining a Error value t, t'. Finally, if the error value t' is less than the preset value (according to the setting of step S43), then according to step S47: the user adopts the control parameter and stores the value b of the control parameter; if the error value t is greater than the preset value , according to step S46': The user resets the tuning parameters.

In this embodiment, one of the interfaces of the virtual machine 9a is configured according to the information acquired by the analog unit 1 and the authentication module 3, and the interface of the 4C or 4D is time response or 4E. The interface shown in the figure is the frequency response of the virtual machine 9a, which is presented with the controller type 1b, the motion track 1a, the model formula 1c of the controller 20, the simulation time 1d, the execution option 41, and the authentication mode. The value 3a of the system parameter defined by the group 3, the model 3c of the system type 3b and the system parameter, the values a, b of the control parameter, and the dynamic simulation map 40.

The control system according to the type of the parameter controller 20 may be, such as PID type control element (proportional - integral - derivative controller) parameter of a proportional gain (K _p), integral time (T _i), the differential time (T _d ), natural frequency (N) or other equivalent values a, b. It should be understood that the control parameter can also be a parameter of the PI type control element. Therefore, the parameter patterns captured by the simulation unit 1 in steps S21 to S27 will be presented on the interface of the 4C or 4E diagram.

The motion track 1a is a tuning path selected when the parameters are adjusted, and includes a plurality of processing paths for the user to select and set. Therefore, when the user inputs the value a, b of the control parameter, and then selects the motion track 1a, the execution option 41 can be pressed (the value 3a of the system parameter is based on the system identification operation of the 3A to 3C figure) Decide).

The controller type 1b is simulated by the simulation unit 1, so that the single tuning system 9 can store a plurality of controller types 1b, such as the controller 20 being of a PID type or a PI type, for the user to select and set up.

The model formula 1c is a mathematical formula built in accordance with a well-known model in the industry, as follows: G _PID (S) = K _p [(1/T _i S) + 1 + T _d / (1 + T _d S/N)], and S is a constant, wherein the model formula 1c can select an appropriate state according to the control parameter of the controller 20 obtained by the simulation unit 1, so the model formula 1c is for the user to understand the physical mechanics. Relationships, so you can choose to display or not. Therefore, when the user selects the controller type 1b, the interface of the 4C or 4E diagram automatically presents the model formula 1c.

The simulation time 1d is the time (unit: second) at which the virtual machine 9a performs parameter calculation. Therefore, after the user presses the execution option 41, after the calculation of the virtual machine 9a is completed, the interface of the 4C or 4E diagram presents the number of seconds of the simulation time 1d.

The dynamic simulation map 40 displays various analog alignment states as required. Specifically, in one of the interfaces, as shown in FIGS. 4C and 4D, the operation of inputting the values a, b into the target machine 2 presents the actual motion path L, L', and the value a, b The simulated motion path S, S' is obtained after the calculation by the virtual machine 9a, so that the user can know the actual motion path L, L' and the simulated motion path through the curve comparison of the dynamic simulation map 40. The error value of S, S' is t, t'. Alternatively, in another interface, as shown in FIG. 4E, the dynamic simulation map 40 presents an amplitude chart E1 and a phase chart E2.

The value 3a of the system parameter is determined at the time of the system identification operation of Figures 3A to 3C.

The system type 3b is the servo system of the target machine 2 identified by the authentication module 3, so the single tuning system 9 can be based on the factory area. The model of the machine stores a variety of system types 3b for user selection and setting.

The model 3c of the system parameter is determined in the system identification operation of the 3A to 3C diagrams, that is, according to the selected system parameter selection manner, it can be the ms ² X(s)+ shown in the 3B-1 diagram. bsX(s)+ksX(s)=F(s), or the equations shown in Figures 4C and 4D: Therefore, the aspect of the model 3c of the system parameter is not particularly limited. Therefore, when the user selects the system type 3b, the interface of the 4C or 4E diagram automatically presents the value 3a of the system parameter and the model 3c of the system parameter.

However, any interface regarding the virtual machine 9a is presented in a variety of ways, and can be changed as needed, and is not limited to the above.

Furthermore, in step S45, when the virtual machine 9a calculates the control parameter at the time, the control parameter (such as the value of any one of K _p , T _i , T _d , N) is adjusted by a loop type to The control parameters that produce the smallest error value are obtained. For example, the virtual machine 9a is used for real-time calculation of the built-in software of the control parameter, and the mathematical calculation method used is a classical Runge-Kutta fourth-order method, as follows: given system initial value: y '( x )= f ( x , y ), and y ( x ₀ )= y ₀ , then the y value can be approximated as: y _{( i +1)} = y _i +( k ₁ +2 k ₂ +2 k ₃ + k ₄ )/ 6, where k ₁ = hf ( x _i , y _i ); k ₂ = hf ( x _i + h /2, y _i + k ₁ /2); k ₃ = hf ( x _i + h /2, y _i + k ₂ /2); k ₄ = hf ( x _i + h , y _i + k ₃ ); x _i = x ₀ + ih , where y _{(i + 1)} is the output value, its equal sign y _i +( k ₁ +2 k ₂ +2 k ₃ + k ₄ )/6 ( y _i and k ₁ , k ₂ , k ₃ , k ₄ ) are input values, and h is a constant. Since the above formula is a mathematical formula, the condition represented by it can be arbitrarily defined. For example, the input value can be the current parameter of the controller 20 (such as K _p , T _i , T _d , N shown in FIG. 4D). One of the outputs), and the output value may be the adjusted control parameter (such as the value b of one of K _p , T _i , T _d , N shown in FIG. 4E ). Specifically, in the operation process, the obtained values a, b will present curves of different simulated motion paths S, S′ and actual motion paths L, L′ in the dynamic simulation map 40, thereby facilitating user judgment. The error value t, t' changes.

Moreover, when checking the validity of the process shown in FIG. 4A, it can be compared by means of comparison. Specifically, when the processing condition is the roughness requirement aspect, as shown in FIG. 4F-1, the curves U1-U4 in the calculation process of the virtual machine 9a are integrated into a graph, and it is known that in the same speed interval, The acceleration limit becomes larger, and the error of the axial dynamic becomes smaller. For example, the arrow direction Z indicates that the error from the curve U1 to the curve U4 gradually becomes smaller, wherein the speed of the 4F-1 chart is in millimeters per second (mm/ s), and the unit of error is 0.001 mm.

When the machining condition is the speed demand aspect, as shown in FIG. 4F-2, P1-P4 means the first to fourth acceleration curves, V1-V4 means the first to fourth speed curves, and the first acceleration curve P1 Corresponding to the first speed curve V1, the second acceleration curve P2 corresponds to the second speed curve V2, the third acceleration curve P3 corresponds to the third speed curve V3, and the fourth acceleration curve P4 corresponds to the fourth speed curve V4, so the same speed is reached. First to fourth speed Under the demand of the degree curve V1-V4 having the same maximum value, the larger the acceleration (such as the arrow direction Y'), the shorter the acceleration time (such as the arrow direction X).

When the processing conditions are in terms of accuracy requirements, as shown in Fig. 4F-3, it is known that the corner error is small and the geometric error is small.

Therefore, when the tuning system 9 performs the adjustment parameter, the simulation unit 1 and the authentication module 3 respectively complete the simulation operation and the system identification operation (that is, construct a virtual machine 9a) to present the information to at least one. The interface (such as the interface shown in FIG. 4D or 4E), for the user to set the processing conditions of the target machine 2 to be simulated and the current control parameters of the controller 20 to be adjusted (ie, the simulation command including the analog parameters) ) is input to the virtual machine 9a as in steps S41-S45.

Then, in steps S45-S46, after the setting is completed, the user presses the execution option 41, so that the virtual machine 9a can be built by the built-in software (for example, according to the classical Runge-Kutta fourth-order method described above. Software) real-time automatic calculation of required control parameters (or optimization parameters), and the actual response of the target machine 2 (such as the actual motion path L, L') and the virtual machine 9a simulate the target machine 2 The simulated response (eg, the simulated motion path S, S') is presented in the dynamic simulation map 40 (ie, the virtual machine 9a performs a simulated job of responding to the target machine 2 based on the simulation parameters). In actual operation, the virtual machine 9a calculates the value a of the control parameter of FIG. 4C (such as the parameter before or during the calculation, or the current parameter of the controller 20) into the control parameter of the 4D map. The value b (the error value t' is smaller).

After that, in step S47, the user can calculate the virtual machine 9a. The value b of the control parameter is used as an initial value for adjusting the basic parameters of the target machine 2, such as the control parameters of the controller 20. Finally, the user can quickly adjust the parameters on the target machine 2 by manual mode or the automatic transmission mode of the tuning system 9, thereby greatly reducing the time for adjusting the parameters of the target machine 2.

In addition, the motion track 1a provided by the simulation unit 1 can be input into the virtual machine table 9a to simulate the response of the acceleration/deceleration to the mechanical structure, and can also simulate the compensation effect of the servo system on the friction and backlash caused by the mechanism. , or the effectiveness of structural vibration suppression.

Please refer to Figures 5 and 6, which are schematic diagrams of another embodiment of the method of tuning of the present invention.

In another embodiment, the tuning system 9 can use the authentication module 3 to measure the control parameters obtained by the virtual machine 9a after each parameter adjustment or measure the target machine 2 (the same machine tool) The control parameters obtained by operating under normal conditions are archived as control parameters to establish a reference model.

As shown in FIG. 5, the reference model is formed by first inputting a motion command (simulation command) into the virtual machine 9a and the target machine 2 via the simulation unit 1 according to steps S51 to S53. The target machine 2 generates the first response information, and the virtual machine 9a generates the second response information, and then according to step S54: comparing the first response information with the second response information, when the error values of the two are less than the preset value. When the error is within the allowable range, the second response information is stored in step S55 to establish the reference model. If the error value of the two is greater than the preset value, then according to step S54': correct the analog command Analog parameters.

Therefore, in this case, the reference model of the target machine 2 can be obtained through the identification module 3, so when performing the adjustment parameter operation, as shown in FIGS. 1B and 6 , the user first sets the processing condition, and then follows the steps. S60 to S63: input the control parameter (motion command) extracted from the controller 20 into the target machine 2 and the authentication module 3, so that the tuning system 9 simulates the response of the target machine 2. And obtaining a system model (such as the first virtual machine), and the tuning system 9 provides a reference model (such as a second virtual machine) that meets the processing conditions according to step S64, so that the system model utilizes mathematical calculations. The software (based on the classical Runge-Kutta fourth-order method described above) instantly calculates the difference between the reference model and the system model (as in the dynamic simulation map 40, the motion path of the reference model replaces the actual motion path L) , L', to obtain the error value between the motion path of the system model and the motion path of the reference model), and to derive an optimization parameter (such as step S65), this parameter (such as the control parameter in the 4D figure) The number in the field) can be used as Automatically adjusts a flow of the initial value of the basic parameters, which can reduce the user to adjust the parameters of the time 2 of the target machine.

In summary, the tuning system 9 and the tuning method of the present invention are constructed by the simulation unit 1 and the authentication module 3 to construct a virtual machine 9a capable of effectively simulating the target machine 2. The user's control experience, the user only needs to set the relevant conditions on the interface of the tuning system 9, and the virtual machine 9a can quickly and automatically calculate the parameters of the required controller 20 (step S45). Up to S46), so that the user can directly use the calculated parameter (such as the value b) or only need to fine-tune the value of the parameter, thereby The user can easily adjust the parameters of the controller 20 of the target machine 2 without constantly trying a large number of parameter values.

Furthermore, the tuning system 9 of the present invention can perform the tuning operation at a remote end (the current parameters of the controller 20 still need to be known to be input into the field of the value a of the interface of FIG. 4C), and can simultaneously target Multiple target machines 2 (tool machines) perform adjustment parameters.

Moreover, the system identification method of the authentication module 3 uses the analog command of the analog unit 1 as an input signal.

In addition, the simulation unit 1 can be used alone to simulate the controller 20 without limiting the need to cooperate with the tuning system 9.

The above embodiments are intended to illustrate the principles of the present invention and its effects, and are not intended to limit the present invention. Anyone who is familiar with the art can modify the above embodiments without violating the spirit and scope of the case. Therefore, the scope of protection of the rights disclosed in this case should be listed in the scope of the patent application described later.

Claims (14)

A tuning method includes: providing a tuning system configured with a virtual machine, an authentication module, and an analog unit, wherein the virtual machine is constructed by imitating a target machine by using the simulation unit; The group performs a system identification operation, so that the simulation unit transmits a first simulation command including the first simulation parameter to the target machine and the authentication module, so that the target machine generates response information, and then transmits the response information to the The authentication module is configured to compare the first simulation parameter with the response information by using the first simulation command and the response information, and if the error of the first simulation parameter is within a reasonable range, the first simulation parameter and the response information are used as a system a parameter; setting a processing condition of the target machine in the tuning system; inputting a second simulation command including a second simulation parameter into the virtual machine; and performing, by the virtual machine, the target according to the second simulation parameter The simulated operation of the machine's response, and the required control parameters are calculated in real time, so that the control parameter is used as a basic parameter for adjusting the target machine. The method of tuning according to claim 1, wherein the target machine has a controller, and the analog unit is used to simulate the controller. The tuning method of claim 1, wherein the tuning system is further configured with a reference model constructed using the virtual machine and having control parameters. For example, the method of adjusting the machine mentioned in item 3 of the patent application includes The processing condition of the target machine is selected, and the reference model is compared and compared with the control parameter to calculate another control parameter. A tuning system includes: a host, configured with a virtual machine, wherein the virtual machine is modeled on a target machine; and an authentication module is used to identify the target machine to construct the virtual machine; And an analog unit for simulating a controller of the target machine, wherein an analog command can be input to the target machine for response comparison to determine whether the analog command corresponds to a target command for executing the controller For the authentication module to use the analog command of the analog unit as an input signal for authentication. The tuning system of claim 5, wherein the simulation unit is configured to transmit the analog command to the target machine and the authentication module for identification by the authentication module. A simulation method for simulating a controller for controlling a motion of a target machine, the simulation method comprising: setting a parameter of a simulated motion trajectory; calculating an analog command according to the parameter; The analog command inputs the target machine for response comparison to determine whether the analog command corresponds to a target command used by the controller to execute the target machine. The simulation method of claim 7, wherein the simulation unit is another controller. The simulation method of claim 7, wherein the simulation unit sets the parameters of the simulated motion trajectory by collecting motion parameter information of the controller. The simulation method of claim 7, wherein the simulation unit determines whether the simulation command corresponds to the target command by correlating the simulated motion trajectory with the target machine actuation. An analog unit includes: a data collecting unit for setting a parameter of a simulated motion trajectory; a calculating unit for calculating an analog command according to the parameter; and an analyzing unit for inputting the analog command into the The target machine performs a response comparison to determine whether the analog command corresponds to a target command for executing a controller that controls the target machine. The simulation unit of claim 11, wherein the simulation unit is another controller. The simulation unit of claim 11, wherein the data collection unit sets the parameters of the simulated motion trajectory by collecting motion parameter information of the controller. The simulation unit of claim 11, wherein the analysis unit determines whether the simulation command corresponds to the target command by correlating the simulated motion trajectory with the target machine.