Multi-manipulator collaborative control method based on fish scale prediction algorithm and PID

Abstract

The present invention belongs to the field of PID control technology, and specifically relates to a multi-manipulator collaborative control method based on a fish scale prediction algorithm and PID, wherein the steps include introducing a fish scale prediction algorithm in a multi-manipulator collaborative operating system; in the multi-manipulator collaborative operating system, each manipulator has multiple degrees of freedom, and each degree of freedom needs to be controlled, so as to establish a state space of the multi-manipulator collaborative operating system; in the multi-manipulator collaborative operating system, PID control is introduced; a disturbance offset PID control theorem is established, and the disturbance offset PID control theorem is applied to the fish scale prediction algorithm to obtain a multi-manipulator parameter optimization controller; and the multi-manipulator parameter optimization controller is used to perform parameter setting of the multi-manipulator collaborative operating system, thereby realizing collaborative control of the multi-manipulator collaborative operating system. The present invention has strong anti-interference ability and fast response speed, and can better perform collaborative control of the multi-manipulator collaborative operating system.

Description

Technical Field

[0001] The present invention belongs to the technical field of PID control, and in particular relates to a multi-robot arm collaborative control method based on a fish scale prediction algorithm and PID. Background Art

[0002] Control systems play a key role in industrial production. The application of automated control technology makes the production process more efficient and accurate, and it can reduce the risk of human error. PID is a traditional method in control technology, and many new control methods have been derived based on this method. In the control technology of complex systems, many control parameters are involved. How to optimize and set the control parameters directly determines the implementation effect of the control system. Many researchers focus on the control algorithm, that is, to optimize the PID parameters online through the control algorithm to complete multi-machine collaborative control.

[0003] For example, in the existing literature, "An Online PID Parameter Optimization Control Method Based on Genetic Algorithm", this method optimizes the parameters of the PID controller by improving the genetic algorithm and genetic mechanism. Compared with the traditional static parameter optimization strategy, it can perform parameter optimization online in real time. Scholars have also tried to use the Chaos Atom Search Optimization Algorithm (CHASO) to optimize PID control parameters. This method uses chaotic sequences and logical mappings to avoid stagnation in local minima, thereby improving the convergence speed and result accuracy.

[0004] Although existing methods have achieved certain success in control parameter optimization, there are still some shortcomings when facing high-dimensional complex systems (such as multi-manipulator collaborative operation systems). Existing methods face problems such as low search efficiency, slow convergence speed, and high computational complexity when dealing with high-dimensional parameter spaces. Due to the complexity and diversity of high-dimensional parameter spaces, traditional optimization algorithms are affected by the curse of dimensionality, resulting in reduced algorithm performance. Therefore, a new research idea and solution is needed for the optimization of control parameters of high-dimensional complex systems such as multi-manipulator collaborative operation systems. Summary of the invention

[0005] In view of the above deficiencies in the prior art, the present invention provides a multi-robotic arm collaborative control method based on fish scale prediction algorithm and PID, which has strong anti-interference ability and fast response speed, and can better perform collaborative control of multi-robotic arm collaborative operation system.

[0006] To achieve the above objectives, the present invention provides a multi-manipulator collaborative control method based on fish scale prediction algorithm and PID, comprising the following steps:

[0007] S1. Introduce fish scale prediction algorithm in multi-robot collaborative operation system;

[0008] S2. In a multi-manipulator collaborative operation system, each manipulator has multiple degrees of freedom, and each degree of freedom needs to be controlled to establish the state space of the multi-manipulator collaborative operation system;

[0009] S3. Introduce PID control in the multi-robot collaborative operation system;

[0010] S4. Establish a disturbance offset PID control theorem, and apply the disturbance offset PID control theorem to the fish scale prediction algorithm to obtain a multi-manipulator parameter optimization controller;

[0011] S5. Parameters of the multi-robotic arm collaborative operating system are adjusted through the multi-robotic arm parameter optimization controller, thereby realizing collaborative control of the multi-robotic arm collaborative operating system.

[0012] As a preferred embodiment of the present invention, in S1, the process of constructing a fish scale prediction algorithm is as follows:

[0013] S11. In the fish scale prediction algorithm, the number of fish scales is N, which is the same as the number of robotic arms in the multi-robotic arm collaborative operating system. Each fish scale corresponds to a robotic arm, and the dimension of each fish scale is D. The initial data set of all fish scales is It is expressed as: , the initial data of each fish scale is the fusion of n data, that is, D=n;

[0014] S12. The expected range of each fish scale regarding the regulation object is , where p1 is the lowest standard in the expected range; p2 is the highest standard in the expected range;

[0015] S13, define the parameter difference of fish scale prediction algorithm as ,and ; is a steady-state parameter used to adjust In stability, The size of a length within the scope of the field, that is Therefore, when When , the parameter is abnormal;

[0016] S14, through the fish scale difference matrix To indicate the state of each parameter difference:

[0017] (1);

[0018] about The structure matrix It is expressed as:

[0019] (2);

[0020] S15. Settings As the disturbance factor, the disturbance balance equation is established:

[0021] (3);

[0022] In the formula, Represents the anti-interference output of the fish scale prediction algorithm; It is the steady-state input parameter of the fish scale prediction algorithm; Control input parameters for fish scale prediction algorithm; The current input for the fish scale prediction algorithm;

[0023] S16, fish scale parameters after iteration It is expressed as: ; Represents the learning rate of each iteration; parameter fitness function of the kth fish scale It is expressed as;

[0024] (4);

[0025] (5);

[0026] (6);

[0027] (7);

[0028] In the formula, is the parameter fitness function of the k-1th fish scale; is the allocation weight of the kth fish scale; is the stable error of the parameters corresponding to the kth fish scale; and express and The impact function on iterative updates; is the fish scale parameter after iterative optimization; Formula (5) indicates Affiliate ; Formula (6) and Formula (7) are constraints; Formula (6) means satisfy within the neighborhood of; Represents the iteration range of fish scale parameters; Represents the fish scale parameter vector composed after iteration (that is, the parameter vector composed of the parameters after iteration).

[0029] As a preferred solution of the present invention, in S2, the process of establishing the state space of the multi-robot collaborative operation system is:

[0030] S21. In a multi-robot collaborative operating system, each robot has multiple degrees of freedom, and each degree of freedom needs to be controlled. The number of robots is set to R1, and the degree of freedom of each robot is set to R2, which can be expressed as:

[0031] (8);

[0032] (9);

[0033] Among them, Rot is a set of manipulators, and the elements in Rot represent each manipulator; Rod is a set of degrees of freedom, and the elements in Rod represent each degree of freedom;

[0034] S22, the number of fish scales is N, that is, N=R1, the initial data of each fish scale is the fusion of n data, and then D=n=R2, the state space of the multi-manipulator collaborative operation system can be expressed as a size of The rotation matrix :

[0035] (10);

[0036] in, is the rotation state of the x-th robot arm in the y-th degree of freedom, , ;

[0037] S23, taking the rotation speed of the robot arm (in rpm) as the control object of the multi-robot collaborative operation system, As the state space about the rotation speed, the conventional expected range of the control object is:

[0038] (11).

[0039] As a preferred embodiment of the present invention, in S3, the method of introducing PID control is:

[0040] S31, establish a high-dimensional PID input module, and pass the high-dimensional PID input module through the high-dimensional space matrix H PID To express:

[0041] (12);

[0042] In the formula, represents the PID control input of the yth degree of freedom in the xth robot;

[0043] S32. Select the time integral of absolute error ITAE as the performance evaluation index of PID control. It is known that ITAE represents the integral of error over time, and the rotation speed of the robot arm is equivalent to one of the outputs of the multi-robot collaborative operation system. The change of the rotation speed reflects the response speed of the multi-robot collaborative operation system to the error. The difference between the expected rotation speed and the actual rotation speed is taken as the error function, which is set as , belong The expansion object, It is expressed as:

[0044] (13);

[0045] (14);

[0046] In the formula, is the expected rotation speed; is the actual rotation speed; then ITAE is defined as:

[0047] (15);

[0048] In the formula, t is time; T represents the time period;

[0049] S33, structural matrix of fish scale parameters Yes H PID adjust The result is now expressed as:

[0050] (16);

[0051] S34. In a multi-robot collaborative operating system, control , the control decision changes continuously with the operation mechanism of the robot arm, and the proportional gain K of the PID control parameter is optimized online p , integral gain K i and differential gain K d , to minimize ITAE; During the operation of the robot, the control output characteristics of each robot joint are expressed as:

[0052] (17);

[0053] In the formula, , and They represent the proportional gain, integral gain and differential gain of the yth degree of freedom in the xth robot arm respectively;

[0054] Among them, the control output characteristic of each robot arm joint refers to the control signal generated during PID control, which is , It is also the input signal for controlling the joints of the robot arm, which means PID control input.

[0055] As a preferred solution of the present invention, in S4, the process of establishing the disturbance offset PID control theorem is:

[0056] S41. When an interference signal appears during the control output of the robot arm joint, the control output characteristic becomes:

[0057] (18);

[0058] In the formula, is the disturbance signal; D xy is the disturbance cancellation gain;

[0059] S42, the disturbance cancellation PID control theorem is expressed as: when designing a feedback controller gain matrix The system state matrix When the characteristic polynomial of is the desired characteristic polynomial, that is, When , the interference signal can be offset, thus realizing the disturbance offset PID control; among them, is the structural matrix (that is, the structural matrix of PID control, which is combined with the fish scale prediction algorithm ),Right now ; H PID adjust The structure matrix after that; det() is the determinant function; is I over the field of complex numbers Q The product of I and s, Q is the identity matrix over the complex field, s is a complex variable; Represents the transfer function matrix of the multi-manipulator collaborative operation system in the frequency domain.

[0060] As a preferred embodiment of the present invention, in S4, in the multi-robot parameter optimization controller, the disturbance cancellation PID control theorem enables the multi-robot parameter optimization controller to respond to external interference and internal changes by real-time monitoring and compensation of disturbances of the multi-robot collaborative operating system.

[0061] As a preferred solution of the present invention, the process of the multi-manipulator parameter optimization controller responding to external interference and internal changes is:

[0062] System state equation of multi-robot cooperative operation system It is expressed as:

[0063] (19);

[0064] Where x(t) is the system input vector before the robot joint (the previous control cycle);

[0065] right Perform linearization and give the working point and , near the working point and Perform a linear approximation, expressed as:

[0066] (20);

[0067] (twenty one);

[0068] In the formula, g is about and The state function of , Respectively represent the and The result after linear approximation; the length near the working point is Neighborhood range;

[0069] Substituting equation (20) and equation (21) into The linearized state equation is obtained:

[0070] (twenty two);

[0071] when hour, , thereby achieving the cancellation of interference signals; at this time, according to the fish scale prediction algorithm, the parameters of the controlled object are continuously updated, and finally the parameters of the online optimization PID control are achieved, that is, the K p , K i and K d Online optimization.

[0072] As a preferred solution of the present invention, in S5, the parameter setting process of the multi-manipulator parameter optimization controller is expressed as:

[0073] (twenty three);

[0074] In the formula, w xy represents the distribution weight of the yth degree of freedom in the xth robot; g0 represents the state function of the previous cycle, represents the value of the state function g0 of the previous cycle; g1 is the state function of the current cycle; Represents the iteration range of the gain matrix; Represents the PID parameter vector formed after iteration.

[0075] The algorithm involved in the present invention can be executed by an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The algorithm is implemented by executing software through the processor.

[0076] The beneficial effects of the present invention are:

[0077] The present invention proposes a fish scale prediction algorithm with high robustness, which can adapt to nonlinear complex systems in high-dimensional space, and thus is better suitable for multi-manipulator collaborative operation systems. Through this algorithm, the system can more accurately predict and respond to the operating status in various complex environments, thereby achieving more stable and reliable control.

[0078] The present invention establishes the disturbance offset PID control theorem, which further improves the stability of the fish scale prediction algorithm in the PID parameter optimization process. The application of this theorem enables the multi-manipulator collaborative operating system to better cope with external disturbances and changes, ensure the normal operation of high-dimensional spatial states, and provide a guarantee for the long-term stability of the system.

[0079] The present invention designs a multi-machine parameter optimization controller based on the fish scale prediction algorithm. The controller can realize parameter setting of the controlled objects of the multi-robotic arm collaborative operation system in a high-dimensional nonlinear space, so that the system can adapt to different working environments and task requirements more flexibly, thereby improving the adaptability and efficiency of the system.

[0080] In summary, the present invention establishes a fish scale prediction algorithm, which monitors the operating status of the multi-robotic arm collaborative operating system in real time and dynamically adjusts the PID controller parameters to meet the performance indicators of the system. Compared with traditional algorithms, it is more suitable for high-dimensional nonlinear control systems, has strong anti-interference ability and fast response speed, shows more superior performance, and can be applied to other high-dimensional complex systems, providing an effective solution for the control parameter optimization of high-dimensional complex systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 It is a schematic diagram of the process of the present invention;

[0082] Figure 2 It is a schematic diagram of the principle of the present invention;

[0083] Figure 3 It is a control flow chart of the multi-robot arm parameter optimization controller of the present invention;

[0084] Figure 4 It is a diagram of the optimization process of the output states of each controller during the verification process of the embodiment of the present invention; Figure 4(a) is the optimization process diagram of GAPID; Figure 4 (b) in the figure is the optimization process diagram of CHASO; Figure 4 (c) in FIG. 1 is an optimization process diagram of the present invention. DETAILED DESCRIPTION

[0085] The embodiments of the present invention are further described below in conjunction with the accompanying drawings:

[0086] like Figure 1 As shown, the multi-manipulator collaborative control method based on fish scale prediction algorithm and PID includes the following steps:

[0087] S1. Introduce fish scale prediction algorithm in multi-robot collaborative operation system;

[0088] S2. In a multi-manipulator collaborative operation system, each manipulator has multiple degrees of freedom, and each degree of freedom needs to be controlled to establish the state space of the multi-manipulator collaborative operation system;

[0089] S3. Introduce PID control in the multi-robot collaborative operation system;

[0090] S4. Establish a disturbance offset PID control theorem, and apply the disturbance offset PID control theorem to the fish scale prediction algorithm to obtain a multi-manipulator parameter optimization controller;

[0091] S5. Parameters of the multi-robotic arm collaborative operating system are adjusted through the multi-robotic arm parameter optimization controller, thereby realizing collaborative control of the multi-robotic arm collaborative operating system.

[0092] The fish scale prediction algorithm draws on the biological response of fish to changes in environmental temperature. It is optimized and upgraded based on the idea of ​​the bionic fish scale algorithm, and prediction is achieved by simulating the arrangement of fish scales at different temperatures. The algorithm includes steps such as initializing the fish scale population, calculating parameter differences, updating the fish scale state, and selecting crossover mutations. The purpose of this algorithm is to achieve reasonable prediction of the parameters of the control object with a small number of iterations and control it within the expected range. In this way, the adjustment of parameters can be achieved and the occurrence of abnormal parameters can be avoided. The main definition of the fish scale prediction algorithm is detailed in the applicant's prior patent CN118092195B, and CN117784852B also has a detailed description of the fish scale bionic optimization algorithm, one of the variants of the fish scale prediction algorithm.

[0093] In S1, the process of constructing the fish scale prediction algorithm is:

[0094] S11. In the fish scale prediction algorithm, the number of fish scales is N, which is the same as the number of robotic arms in the multi-robotic arm collaborative operating system. Each fish scale corresponds to a robotic arm, and the dimension of each fish scale is D. The initial data set of all fish scales is It is expressed as: , the initial data of each fish scale is the fusion of n data, that is, D=n;

[0095] S12. The expected range of each fish scale regarding the regulation object is , where p1 is the lowest standard in the expected range; p2 is the highest standard in the expected range;

[0096] S13, define the parameter difference of fish scale prediction algorithm as ,and ; is a steady-state parameter used to adjust In stability, The size of a length within the scope of the field, that is Therefore, when When , the parameter is abnormal;

[0097] S14, through the fish scale difference matrix To indicate the state of each parameter difference:

[0098] (1);

[0099] about The structure matrix It is expressed as:

[0100] (2);

[0101] S15. Settings As the disturbance factor, the disturbance balance equation is established:

[0102] (3);

[0103] In the formula, Represents the anti-interference output of the fish scale prediction algorithm; It is the steady-state input parameter of the fish scale prediction algorithm; Control input parameters for fish scale prediction algorithm; The current input for the fish scale prediction algorithm;

[0104] S16, fish scale parameters after iteration It is expressed as: ; Represents the learning rate of each iteration; parameter fitness function of the kth fish scale It is expressed as;

[0105] (4);

[0106] (5);

[0107] (6);

[0108] (7);

[0109] In the formula, is the parameter fitness function of the k-1th fish scale; is the allocation weight of the kth fish scale; is the stable error of the parameters corresponding to the kth fish scale; and express and The impact function on iterative updates; is the fish scale parameter after iterative optimization; Formula (5) indicates Affiliate ; Formula (6) and Formula (7) are constraints; Formula (6) means satisfy within the neighborhood of; Represents the iteration range of fish scale parameters; Represents the fish scale parameter vector formed after iteration.

[0110] In S2, the process of establishing the state space of the multi-robot collaborative operation system is as follows:

[0111] S21. In a multi-robot collaborative operating system, each robot has multiple degrees of freedom, and each degree of freedom needs to be controlled. The number of robots is set to R1, and the degree of freedom of each robot is set to R2, which can be expressed as:

[0112] (8);

[0113] (9);

[0114] Among them, Rot is a set of manipulators, and the elements in Rot represent each manipulator; Rod is a set of degrees of freedom, and the elements in Rod represent each degree of freedom;

[0115] S22, the number of fish scales is N, that is, N=R1, the initial data of each fish scale is the fusion of n data, and then D=n=R2, the state space of the multi-manipulator collaborative operation system can be expressed as a size of The rotation matrix :

[0116] (10);

[0117] in, is the rotation state of the x-th robot arm in the y-th degree of freedom, , ;

[0118] S23, taking the rotation speed of the robot arm as the control object of the multi-robot arm collaborative operation system, As the state space about the rotation speed, the conventional expected range of the control object is:

[0119] (11).

[0120] In S3, the method of introducing PID control is:

[0121] S31, establish a high-dimensional PID input module, and pass the high-dimensional PID input module through the high-dimensional space matrix H PID To express:

[0122] (12);

[0123] In the formula, represents the PID control input of the yth degree of freedom in the xth robot;

[0124] In a multi-robot collaborative operating system, PID control is extremely complex. Therefore, it is necessary to optimize the PID control parameters online. In system performance evaluation, commonly used error performance indicators include absolute error integral, cumulative error, square error integral, time integral of absolute error ITAE and maximum deviation MD. Among them, ITAE is considered to be a more engineering practical and selective performance evaluation indicator. It can not only quantify the absolute size of the error, but also take into account the change of the error over time. Therefore, ITAE is selected as the performance evaluation indicator.

[0125] S32. Select the time integral of absolute error ITAE as the performance evaluation index of PID control. It is known that ITAE represents the integral of error over time, and the rotation speed of the robot arm is equivalent to one of the outputs of the multi-robot collaborative operation system. The change of the rotation speed reflects the response speed of the multi-robot collaborative operation system to the error. The difference between the expected rotation speed and the actual rotation speed is taken as the error function, which is set as , belong The expansion object, It is expressed as:

[0126] (13);

[0127] (14);

[0128] In the formula, is the expected rotation speed; is the actual rotation speed; then ITAE is defined as:

[0129] (15);

[0130] In the formula, t is time; T represents the time period;

[0131] S33, structural matrix of fish scale parameters Yes H PID adjust The result is now expressed as:

[0132] (16);

[0133] S34. In a multi-robot collaborative operating system, control , the control decision changes continuously with the operation mechanism of the robot arm, and the proportional gain K of the PID control parameter is optimized online p , integral gain K i and differential gain K d , in order to minimize ITAE, the optimization goal is to make the rotation speed of the robot reach the expected value as soon as possible and minimize the accumulation of errors over time; During the operation of the robot, the control output characteristics of each robot joint are expressed as:

[0134] (17);

[0135] In the formula, , and They represent the proportional gain, integral gain and differential gain of the yth degree of freedom in the xth robot arm respectively;

[0136] The operation process refers to the movement and operation process of the robot in the multi-robot collaborative operation system. In this process, the control strategy will constantly change according to the movement of the robot. By optimizing the PID control parameters (proportional gain K p , integral gain K i , differential gain K d ) to minimize the ITAE value.

[0137] Among them, the control output characteristic of each robot arm joint refers to the control signal generated during PID control, which is , It is also the input signal for controlling the joints of the robot arm, which means PID control input.

[0138] In S4, the process of establishing the disturbance offset PID control theorem is:

[0139] S41. When an interference signal appears during the control output of the robot arm joint, the control output characteristic becomes:

[0140] (18);

[0141] In the formula, is the disturbance signal; D xy is the disturbance cancellation gain;

[0142] S42, the disturbance cancellation PID control theorem is expressed as: when designing a feedback controller gain matrix The system state matrix When the characteristic polynomial of is the desired characteristic polynomial, that is, When , the interference signal can be offset, thus realizing the disturbance offset PID control; among them, is the structure matrix, that is ; H PID adjust The structure matrix after that; det() is the determinant function; is I over the field of complex numbers Q The product of I and s, Q is the identity matrix over the complex field, s is a complex variable; Represents the transfer function matrix of the multi-manipulator collaborative operation system in the frequency domain.

[0143] In S4, in the multi-manipulator parameter optimization controller, the disturbance offset PID control theorem enables the multi-manipulator parameter optimization controller to respond to external disturbances and internal changes by real-time monitoring and compensation of the disturbances of the multi-manipulator collaborative operation system.

[0144] The process of the multi-manipulator parameter optimization controller responding to external disturbances and internal changes is:

[0145] System state equation of multi-robot cooperative operation system It is expressed as:

[0146] (19);

[0147] Where x(t) is the system input vector before the robot joint;

[0148] right Perform linearization and give the working point and , near the working point and Perform a linear approximation, expressed as:

[0149] (20);

[0150] (twenty one);

[0151] In the formula, g is about and The state function of , Respectively represent the and The result after linear approximation; the length near the working point is Neighborhood range;

[0152] Substituting equation (20) and equation (21) into The linearized state equation is obtained:

[0153] (twenty two);

[0154] when hour, , thereby achieving the cancellation of interference signals; at this time, according to the fish scale prediction algorithm, the parameters of the controlled object are continuously updated, and finally the parameters of the online optimization PID control are achieved, that is, the K p , K i and K d Online optimization.

[0155] In S5, the parameter tuning process of the multi-manipulator parameter optimization controller is expressed as:

[0156] (twenty three);

[0157] In the formula, w xy represents the distribution weight of the yth degree of freedom in the xth robot; g0 represents the state function of the previous cycle, represents the value of the state function g0 of the previous cycle; g1 is the state function of the current cycle; Represents the iteration range of the gain matrix; Represents the PID parameter vector formed after iteration.

[0158] The principle of the present invention is as follows Figure 2 As shown in the figure, the multi-machine collaborative system is a multi-robot collaborative operating system. The control flow of the multi-robot parameter optimization controller is as follows: Figure 3 As shown, Figure 3 The high-dimensional space matrix element in is the high-dimensional space matrix H PID The elements in .

[0159] The verification process is:

[0160] The present invention is experimentally compared with two traditional methods, GAPID and CHASO, and the comparison indicators include tracking error, output state and prediction effect. The purpose of tracking error is to reduce the error. In the controller parameter selection of GAPID, the proportional coefficient is K p =50, the integral time is K i =30. Differential time is K d =10, which is used to ensure system stability and suppress overshoot, but the controller is mainly used in wind energy systems. The control scenario of the present invention is different from the above method. The controlled object of the controller of the present invention is a multi-robot collaborative operating system. In the MATLAB simulation, the parameters of the fish scale prediction algorithm need to be adjusted according to the system characteristics, and the prediction step size is set to 10 and the prediction time domain length is 20. The parameters of the PID controller (multi-robot parameter optimization controller) in the FSPOC method (the control method of the present invention is referred to as FSPOC) also need to be adjusted, and the parameters are set to K p =60, K i =25 and K d =8.

[0161] Matlab / Simulink is used to establish a simulation model of the coordinated rotation of multiple robotic arms, and the multi-robotic arm parameter optimization controller of the present invention is used for control.

[0162] For the simulation comparison of tracking gap, CHASO has a large tracking gap during the tracking process, and the tracking effect of FSPOC is better than that of CHASO, which verifies the superiority of FSPOC in tracking error.

[0163] For the simulation comparison of the output state, the actual output curve of FSPOC is close to the control target. There is no oversaturation or undersaturation phenomenon, which shows its superiority and applicability. The optimization process of FSPOC output state in the continuous iteration of fish scales is shown in the figure. Figure 4 As shown, Figure 4 In the figure, the ellipse is the critical saturation zone, and the simulation time is controlled between 0 and 0.2 seconds. As the optimization proceeds, it can be seen that the position of the critical saturation point drops from 0.17 seconds to around 0.16 seconds, and finally to around 0.15 seconds, and the time efficiency is improved by nearly 11.8%.

[0164] For the simulation comparison of prediction effects, 50 tests were conducted to compare the prediction effects of FSPOC, GAPID and CHASO. The results showed that FSPOC had a better fit than GAPID and CHASO in the 50 tests, with a fit improvement of about 21% and an accuracy improvement of about 15%.

Claims

1. A multi-manipulator collaborative control method based on fish scale prediction algorithm and PID, characterized by The following steps are involved: S1. Introduce fish scale prediction algorithm in multi-robot collaborative operation system; S2. In a multi-manipulator collaborative operation system, each manipulator has multiple degrees of freedom, and each degree of freedom needs to be controlled to establish the state space of the multi-manipulator collaborative operation system; S3. Introduce PID control in the multi-robot collaborative operation system; S4. Establish a disturbance offset PID control theorem, and apply the disturbance offset PID control theorem to the fish scale prediction algorithm to obtain a multi-manipulator parameter optimization controller; S5. Parameters of the multi-manipulator collaborative operating system are adjusted by the multi-manipulator parameter optimization controller, thereby realizing collaborative control of the multi-manipulator collaborative operating system; In the above S1, the process of constructing the fish scale prediction algorithm is as follows: S11. In the fish scale prediction algorithm, the number of fish scales is N, which is the same as the number of robotic arms in the multi-robotic arm collaborative operating system. Each fish scale corresponds to a robotic arm, and the dimension of each fish scale is D. The initial data set of all fish scales is It is expressed as: , the initial data of each fish scale is the fusion of n data, that is, D=n; S12. The expected range of each fish scale regarding the regulation object is , where p1 is the lowest standard in the expected range; p2 is the highest standard in the expected range; S13, define the parameter difference of fish scale prediction algorithm as ,and ; is a steady-state parameter used to adjust In stability, The size of a length within the scope of the field, that is Therefore, when When , the parameter is abnormal; S14, through the fish scale difference matrix To indicate the state of each parameter difference: （1）； about The structure matrix It is expressed as: （2）； S15. Settings As the disturbance factor, the disturbance balance equation is established: （3）； In the formula, Represents the anti-interference output of the fish scale prediction algorithm; It is the steady-state input parameter of the fish scale prediction algorithm; Control input parameters for fish scale prediction algorithm; The current input for the fish scale prediction algorithm; S16, fish scale parameters after iteration It is expressed as: ; Represents the learning rate of each iteration; parameter fitness function of the kth fish scale It is expressed as; （4）； （5）； （6）； （7）； In the formula, is the parameter fitness function of the k-1th fish scale; is the allocation weight of the kth fish scale; is the stable error of the parameters corresponding to the kth fish scale; and express and The impact function on iterative updates; is the fish scale parameter after iterative optimization; Formula (5) indicates Affiliate ; Formula (6) and Formula (7) are constraints; Formula (6) means satisfy within the neighborhood of; Represents the iteration range of fish scale parameters; Represents the fish scale parameter vector formed after iteration; In S2, the process of establishing the state space of the multi-robot collaborative operation system is as follows: S21. In a multi-robot collaborative operating system, each robot has multiple degrees of freedom, and each degree of freedom needs to be controlled. The number of robots is set to R1, and the degree of freedom of each robot is set to R2, which can be expressed as: （8）； （9）； Among them, Rot is a set of manipulators, and the elements in Rot represent each manipulator; Rod is a set of degrees of freedom, and the elements in Rod represent each degree of freedom; S22, the number of fish scales is N, that is, N=R1, the initial data of each fish scale is the fusion of n data, and then D=n=R2, the state space of the multi-manipulator collaborative operation system can be expressed as a size of The rotation matrix : （10）； in, is the rotation state of the x-th robot arm in the y-th degree of freedom, , ; S23, taking the rotation speed of the robot arm as the control object of the multi-robot arm collaborative operation system, As the state space about the rotation speed, the conventional expected range of the control object is: （11）； In the above-mentioned S3, the method of introducing PID control is: S31, establish a high-dimensional PID input module, and pass the high-dimensional PID input module through the high-dimensional space matrix H PID To express: （12）； In the formula, represents the PID control input of the yth degree of freedom in the xth robot; S32. Select the time integral of absolute error ITAE as the performance evaluation index of PID control. It is known that ITAE represents the integral of error over time, and the rotation speed of the robot arm is equivalent to one of the outputs of the multi-robot collaborative operation system. The change of the rotation speed reflects the response speed of the multi-robot collaborative operation system to the error. The difference between the expected rotation speed and the actual rotation speed is taken as the error function, which is set as , belong The expansion object, It is expressed as: （13）； （14）； In the formula, is the expected rotation speed; is the actual rotation speed; then ITAE is defined as: （15）； In the formula, t is time; T represents the time period; S33, structural matrix of fish scale parameters Yes H PID adjust The result is now expressed as: （16）； S34. In a multi-robot collaborative operating system, control , the control decision changes continuously with the operation mechanism of the robot arm, and the proportional gain K of the PID control parameter is optimized online p , integral gain K i and differential gain K d , to minimize ITAE; During the operation of the robot, the control output characteristics of each robot joint are expressed as: （17）； In the formula, , and They represent the proportional gain, integral gain and differential gain of the yth degree of freedom in the xth robot arm respectively; Among them, the control output characteristic of each robot arm joint refers to the control signal generated during PID control, which is , It is also the input signal for controlling the joints of the robot arm, which means PID control input; In S4, the process of establishing the disturbance offset PID control theorem is: S41. When an interference signal appears during the control output of the robot arm joint, the control output characteristic becomes: （18）； In the formula, is the disturbance signal; D xy is the disturbance cancellation gain; S42, the disturbance cancellation PID control theorem is expressed as: when designing a feedback controller gain matrix The system state matrix When the characteristic polynomial of is the desired characteristic polynomial, that is, When , the interference signal can be offset, thus realizing the disturbance offset PID control; among them, is the structure matrix, that is ; H PID adjust The structure matrix after that; det() is the determinant function; is I over the field of complex numbers Q The product of I and s, Q is the identity matrix over the complex field, s is a complex variable; Represents the transfer function matrix of the multi-manipulator collaborative operation system in the frequency domain; In the above S4, in the multi-manipulator parameter optimization controller, the disturbance offset PID control theorem enables the multi-manipulator parameter optimization controller to respond to external disturbances and internal changes by real-time monitoring and compensation of disturbances of the multi-manipulator cooperative operating system; The process of the multi-manipulator parameter optimization controller responding to external disturbances and internal changes is: System state equation of multi-robot cooperative operation system It is expressed as: （19）； Where x(t) is the system input vector before the robot joint; right Perform linearization and give the working point and , near the working point and Perform a linear approximation, expressed as: （20）； （21）； In the formula, g is about and The state function of , Respectively represent the and The result after linear approximation; the length near the working point is Neighborhood range; Substituting equation (20) and equation (21) into The linearized state equation is obtained: （22）； when hour, , thereby achieving the cancellation of interference signals; at this time, according to the fish scale prediction algorithm, the parameters of the controlled object are continuously updated, and finally the parameters of the online optimization PID control are achieved, that is, the K p , K i and K d Online optimization.

2. The multi-manipulator collaborative control method based on fish scale prediction algorithm and PID according to claim 1 is characterized in that: In the above S5, the parameter setting process of the multi-manipulator parameter optimization controller is expressed as: （23）； In the formula, w xy represents the distribution weight of the yth degree of freedom in the xth robot; g0 represents the state function of the previous cycle, represents the value of the state function g0 of the previous cycle; g1 is the state function of the current cycle; Represents the iteration range of the gain matrix; Represents the PID parameter vector formed after iteration.

Images