Multi-joint torque control method of mechanical arm for cleaning foreign matters on power distribution line

Abstract

The invention discloses a multi-joint torque control method of a mechanical arm for cleaning foreign matters on a power distribution line. The multi-joint torque control method comprises the following steps: S1, setting an initial value and an expected track of a joint in the mechanical arm; s2, a mechanical arm multi-joint torque control model is constructed based on a fuzzy neural network; s3, performing initialization processing on the fuzzy rule, and determining an initial value of a fuzzy neural network parameter; s4, constructing an error function to calculate an error, and performing adjustment and iterative correction on each parameter; s5, the error and the error change rate are calculated to serve as input of the fuzzy neural network, and the control torque parameter of the mechanical arm serves as output; s6, performing iterative optimization processing on the output parameters of the fuzzy neural network based on a particle swarm and grey wolf hybrid optimization algorithm to obtain a global optimal solution of the output parameters of the fuzzy neural network, and obtaining the joint torque of the mechanical arm; according to the method, the joint output torque can be accurately controlled, the joint position tracking error is reduced, and the overall robustness and stability of a mechanical arm control system are improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of mechanical arm control, and particularly relates to a multi-joint torque control method for a mechanical arm for cleaning foreign matters in a power distribution line. BACKGROUND

[0002] As a core equipment in the process of power transmission, the safe and stable operation of the power distribution line is of great significance. In the daily operation and maintenance of the power distribution line, wind kite lines, plastic bags, advertising cloth strips and various foreign matters blown up by the wind or thrown by people are often entangled on the power distribution line. The foreign matters can shorten the distance between phases and reduce the distance between the conductor and the ground. Especially in rainy and snowy weather, short circuit and grounding accidents can be caused. Therefore, in order to improve the reliability of power supply, the foreign matters must be removed in time.

[0003] In actual operation, the live working or power-off manual approach is usually adopted to remove the foreign matters, which is relatively high in cost. With the use of the mechanical arm, live working can be realized without manual approach, which is higher in safety and lower in cost.

[0004] The multi-joint mechanical arm is a motion structure formed by a series of rigid body chains in series, which has the characteristics of nonlinearity and time variation. The traditional mechanical ratio has the problems of modeling error and friction between joint degrees of freedom. These uncertain factors can cause the control performance of the mechanical arm to deteriorate. The chattering of the joint itself has an impact on the overall stability of the mechanical arm system. Therefore, it is very important to study the multi-joint torque control of the mechanical arm.

[0005] Therefore, it is necessary to develop a multi-joint torque control method for a mechanical arm for cleaning foreign matters in a power distribution line to solve the above problems. SUMMARY

[0006] The purpose of the application is to overcome the shortcomings of the prior art, and to provide a multi-joint torque control method for a mechanical arm for cleaning foreign matters in a power distribution line, which can accurately control the joint output torque, reduce the joint position tracking error, and improve the overall robustness and stability of the mechanical arm control system.

[0007] The purpose of the application is achieved by a multi-joint torque control method for a mechanical arm for cleaning foreign matters in a power distribution line, comprising the following steps:

[0008] S1, setting the initial value and the expected trajectory of the joint in the mechanical arm;

[0009] S2, constructing a multi-joint torque control model of the mechanical arm based on a fuzzy neural network;

[0010] S3, initializing the fuzzy rule to determine the initial value of the fuzzy neural network parameter;

[0011] S4. Construct an error function to calculate the error corresponding to each layer in the fuzzy neural network, and adjust and iteratively correct the parameters of each fuzzy neural network based on the error calculation results;

[0012] S5. Calculate the error and error rate of change corresponding to the sample, and use the error and error rate of change as the input of the fuzzy neural network, and the control torque parameter of the robotic arm as the output.

[0013] S6. Based on the particle swarm optimization and gray wolf hybrid optimization algorithm, the output parameters of the fuzzy neural network are iteratively optimized to obtain the global optimal solution of the output parameters of the fuzzy neural network and obtain the joint torque of the robotic arm.

[0014] Further, the fuzzy neural network used to construct the multi-joint torque control model of the robotic arm in step S2 includes an input layer, a fuzzification layer, a rule layer, a normalization layer, and an output layer. The input layer is used to pass input variables to the next layer. The fuzzification layer is used to convert the input variables into fuzzy vectors after fuzzification processing, and a Gaussian function is used as the membership function in the fuzzification layer. The rule layer is used to use the fuzzy subsets of the deviation and the rate of change of deviation in the fuzzification layer as conditional rules for fuzzy inference output. The normalization layer is used to normalize the output of the rule layer. The output layer is used for declarative calculation to generate the output control quantity.

[0015] Further, the input and output values ​​of the input layer nodes in step S2 are represented as follows: In the formula: Indicates the first One input variable, Indicates the first The input variable in the th... Membership degrees on fuzzy subsets; the input and output values ​​of the fuzzification layer nodes are represented as: In the formula: and They represent the first The input variable in the th... Membership function center and basis width on a fuzzy subset Indicates the first The input variable in the th... The corresponding linguistic variable values ​​on each fuzzy subset; the input and output values ​​of the rule layer nodes are represented as follows: In the formula: Representing regular points, , These represent the first and second inputs from the upper layer, respectively. , The linguistic variable values ​​corresponding to each fuzzy subset The input and output values ​​of the normalization layer nodes are represented as follows: The input and output values ​​of the output layer node are represented as follows: In the formula: This indicates the output connection weights. This indicates the number of nodes in the output layer.

[0016] Furthermore, the fuzzy neural network parameters in step S3 include the membership function basis width. The central value of the membership function and output connection weights The initial value.

[0017] Furthermore, the error function in step S4 is expressed as: In the formula: Indicates the given output value, Representing the actual output value; then the modified central and basis width of the membership function of the fuzzy neural network based on the error function are formally expressed as: The modified representation of the output connection weights is as follows: , , , These represent the center and basis width of the membership function, as well as the correction values ​​for the output connection weights. For learning rate, The inertia coefficient, This represents the number of iterations.

[0018] Furthermore, the joint rotation angle error of the robotic arm during its movement at the current moment in step S5 is expressed as: The error in the joint rotation angle during the previous moment of motion is expressed as: ,in, and These represent the joint rotation angles set in the robotic arm control system at the current and previous moments, respectively. and Let represent the joint rotation angles at the current and previous moments during the actual movement of the robotic arm, respectively; then the rate of change of the joint rotation angle at the current moment is expressed as: .

[0019] Furthermore, the control torque parameters of the robotic arm in step S5 include the change values ​​of various control factors of the controller, specifically including the change value of the controller proportional factor. Controller integral factor change value and the change value of the differential factor of the controller .

[0020] Furthermore, the parameter optimization process in step S6 specifically includes the following steps: ① initializing particle parameters; ② calculating the fitness value of the particles to obtain the positions of all particles; ③ sorting the fitness values ​​of the particle swarm and assigning the top three particles to corresponding positions according to their fitness. Wolf, wolves and Wolves; ④ Update via the gray wolf algorithm The wolf's position is found to be optimal, completing the entire optimization process.

[0021] Furthermore, the particle position and velocity update formulas in the particle swarm optimization algorithm in step S6 are expressed as follows: , In the formula, and Represents the velocity of the particle. and Indicates the particle position. This represents the optimal position of an individual particle. Indicates the optimal position of the population. Indicates the weighting coefficient. and Represents the learning coefficient. and It is a random number.

[0022] Furthermore, the position update formula in the gray wolf optimization algorithm in step S6 is expressed as: , In the formula: and Indicates the location of the gray wolf. and This indicates the location of the prey, which is also the location of the optimal solution. and Indicates the update coefficients. Indicates distance control parameters, The maximum number of iterations, , It is a random number. Indicates the first three levels of gray wolves Wolf, wolves and individual wolves The position at this iteration.

[0023] Due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0024] (1) Based on the fuzzy neural network, a multi-joint torque control model for the robotic arm is constructed to achieve precise control of the joint output torque, reach the desired torque, reduce joint position tracking error, reduce control time, improve control performance, and reduce pressure loss;

[0025] (2) The output parameters of the fuzzy neural network are iteratively optimized based on the particle swarm optimization and gray wolf hybrid optimization algorithm to obtain the global optimal solution of the output parameters of the fuzzy neural network, thereby optimizing the controller parameters, improving the overall robustness and stability of the robotic arm control system, and enhancing its anti-interference ability. Attached Figure Description

[0026] Figure 1 This is a flowchart of the present invention.

[0027] Figure 2 This is a flowchart of the parameter optimization process in step S6 of the present invention. Detailed Implementation

[0028] The technical solution of the present invention will be further described in detail below through embodiments and in conjunction with the accompanying drawings.

[0029] like Figure 1 , Figure 2 As shown, a method for controlling the torque of a multi-joint robotic arm used for removing foreign objects from power distribution lines includes the following steps:

[0030] S1. Set the initial values ​​and desired trajectory of the joints in the robotic arm.

[0031] S2. Construct a multi-joint torque control model for a robotic arm based on a fuzzy neural network.

[0032] Preferably, the fuzzy neural network used to construct the multi-joint torque control model of the robotic arm in step S2 includes an input layer, a fuzzification layer, a rule layer, a normalization layer, and an output layer; the input layer is used to pass input variables to the next layer; the fuzzification layer is used to convert the input variables into fuzzy vectors after fuzzification processing, and a Gaussian function is used as the membership function in the fuzzification layer; the rule layer is used to use the fuzzy subset of the deviation and the rate of change of deviation in the fuzzification layer as conditional rules for fuzzy inference output; the normalization layer is used to normalize the output of the rule layer; and the output layer is used for declarative calculation to generate the output control quantity.

[0033] Preferably, in this application, the input layer in step S2 includes two nodes, and the input and output values ​​of the input layer nodes are represented as follows: In the formula: Indicates the first One input variable, Indicates the first The input variable in the th... Membership degrees on fuzzy subsets; In this application, the fuzzification layer contains 2 inputs, each corresponding to 7 linguistic variable values, for a total of 17 nodes, and the input and output values ​​of its nodes are represented as follows: In the formula: and They represent the first The input variable in the th... Membership function center and basis width on a fuzzy subset Indicates the first The input variable in the th... The corresponding linguistic variable values ​​on a fuzzy subset; in this application, the rule layer contains 49 nodes, i.e., rule points, each of which corresponds to a fuzzy rule, and the input and output values ​​of its nodes are represented as follows: In the formula: Representing regular points, , These represent the first and second inputs from the upper layer, respectively. , The linguistic variable values ​​corresponding to each fuzzy subset In this application, the normalization layer contains 49 nodes, and the input and output values ​​of its nodes are represented as follows: In this application, the output layer contains 3 nodes, and the input and output values ​​of its nodes are represented as follows: In the formula: This indicates the output connection weights. This indicates the number of nodes in the output layer.

[0034] S3. Initialize the fuzzy rules and determine the initial values ​​of the fuzzy neural network parameters.

[0035] Preferably, the fuzzy neural network parameters in step S3 include the membership function basis width. The central value of the membership function and output connection weights The initial value.

[0036] S4. Construct an error function to calculate the error corresponding to each layer in the fuzzy neural network, and adjust and iteratively correct the parameters of each fuzzy neural network based on the error calculation results.

[0037] Preferably, the error function in step S4 is expressed as: In the formula: Indicates the given output value, Representing the actual output value; then the modified central and basis width of the membership function of the fuzzy neural network based on the error function are formally expressed as: The modified representation of the output connection weights is as follows: , , , These represent the center and basis width of the membership function, as well as the correction values ​​for the output connection weights. For learning rate, The inertia coefficient, This represents the number of iterations.

[0038] S5. Calculate the error and error rate of change corresponding to the sample, and use the error and error rate of change as the input of the fuzzy neural network, and the control torque parameter of the robotic arm as the output.

[0039] Preferably, the joint rotation angle error of the robotic arm during its movement at the current moment in step S5 is expressed as: The error in the joint rotation angle during the previous moment of motion is expressed as: ,in, and These represent the joint rotation angles set in the robotic arm control system at the current and previous moments, respectively. and Let represent the joint rotation angles at the current and previous moments during the actual movement of the robotic arm, respectively; then the rate of change of the joint rotation angle at the current moment is expressed as: .

[0040] Preferably, the control torque parameters of the robotic arm in step S5 include the change values ​​of various control factors of the controller, specifically including the change values ​​of the controller proportional factor. Controller integral factor change value and the change value of the differential factor of the controller .

[0041] S6. Based on the particle swarm optimization and gray wolf hybrid optimization algorithm, the output parameters of the fuzzy neural network are iteratively optimized to obtain the global optimal solution of the output parameters of the fuzzy neural network and obtain the joint torque of the robotic arm.

[0042] Preferably, the parameter optimization process in step S6 specifically includes the following steps: ① initializing particle parameters; ② calculating the fitness values ​​of the particles to obtain the positions of all particles; ③ sorting the fitness values ​​of the particle swarm and assigning the top three particles to corresponding positions according to their fitness. Wolf, wolves and Wolves; ④ Update via the gray wolf algorithm The wolf's position is found to be optimal, completing the entire optimization process.

[0043] Preferably, the particle position and velocity update formulas in the particle swarm optimization algorithm in step S6 are expressed as follows: , In the formula, and Represents the velocity of the particle. and Indicates the particle position. This represents the optimal position of an individual particle. Indicates the optimal position of the population. Indicates the weighting coefficient. and Represents the learning coefficient. and It is a random number.

[0044] Preferably, the position update formula in the gray wolf optimization algorithm in step S6 is expressed as: , In the formula: and Indicates the location of the gray wolf. and This indicates the location of the prey, which is also the location of the optimal solution. and Indicates the update coefficients. Indicates distance control parameters, The maximum number of iterations, , It is a random number. Indicates the first three levels of gray wolves Wolf, wolves and individual wolves The position at this iteration.

[0045] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of the claims of the present invention.

Claims

1. A multi-joint torque control method for a mechanical arm for foreign matter cleaning of a power distribution line, characterized by: The method comprises the following steps: S1, setting initial values and expected trajectories of joints in a mechanical arm; S2, constructing a multi-joint torque control model of the mechanical arm based on a fuzzy neural network; S3, initializing fuzzy rules to determine initial values of parameters of the fuzzy neural network; S4, constructing an error function to calculate errors corresponding to each layer in the fuzzy neural network, and adjusting and iteratively correcting each parameter of the fuzzy neural network according to the error calculation result; S5, calculating errors and error change rates corresponding to samples, taking the errors and error change rates as inputs of the fuzzy neural network, and taking control torque parameters of the mechanical arm as outputs; S6, iteratively optimizing the output parameters of the fuzzy neural network based on a particle swarm and grey wolf hybrid optimization algorithm to obtain a global optimal solution of the output parameters of the fuzzy neural network and obtain joint torque of the mechanical arm.

2. The multi-joint torque control method of a mechanical arm for cleaning foreign matters of a power distribution line according to claim 1, characterized in that: The fuzzy neural network used for constructing the multi-joint torque control model of the mechanical arm in the step S2 comprises an input layer, a fuzzification layer, a rule layer, a normalization layer and an output layer; the input layer is used for transmitting input variables to the next layer; the fuzzification layer is used for converting the input variables into fuzzy vectors after fuzzy processing; the fuzzification layer adopts a Gaussian function as a membership function; the rule layer is used for taking fuzzy subsets of the deviation and the deviation change rate in the fuzzification layer as conditional rules to perform fuzzy reasoning and output; the normalization layer is used for performing normalization calculation on the output of the rule layer; and the output layer is used for clearization operation to generate an output control quantity.

3. The multi-joint torque control method of a mechanical arm for cleaning foreign matters of a power distribution line according to claim 2, characterized in that: The input and output values ​​of the input layer nodes in step S2 are represented as follows: In the formula: Indicates the first One input variable, Indicates the first The input variable in the th... Membership degrees on fuzzy subsets; the input and output values ​​of the fuzzification layer nodes are represented as: In the formula: and They represent the first The input variable in the th... Membership function center and basis width on a fuzzy subset Indicates the first The input variable in the th... The corresponding linguistic variable values ​​on each fuzzy subset; the input and output values ​​of the rule layer nodes are represented as follows: In the formula: Representing regular points, , These represent the first and second inputs from the upper layer, respectively. , The linguistic variable values ​​corresponding to each fuzzy subset The input and output values ​​of the normalization layer nodes are represented as follows: The input and output values ​​of the output layer node are represented as follows: In the formula: This indicates the output connection weights. This indicates the number of nodes in the output layer.

4. The multi-joint torque control method of a mechanical arm for cleaning foreign matters of a power distribution line according to claim 1, characterized in that: The fuzzy neural network parameters in the step S3 include initial values of membership function base width , membership function center value , and output connection weight .

5. The multi-joint torque control method of a mechanical arm for cleaning foreign matters of a power distribution line according to claim 1, characterized in that: The error function in step S4 is expressed as follows: In the formula: Indicates the given output value, Representing the actual output value; then the modified central and basis width of the membership function of the fuzzy neural network based on the error function are formally expressed as: The modified representation of the output connection weights is as follows: , , , These represent the center and basis width of the membership function, as well as the correction values ​​for the output connection weights. For learning rate, The inertia coefficient, This represents the number of iterations.

6. The multi-joint torque control method of a mechanical arm for cleaning foreign matters of a power distribution line according to claim 1, characterized in that: The joint rotation angle error in the motion process of the robot arm at the current time is expressed as: The joint rotation angle error in the motion process at the previous time is expressed as: wherein, and respectively represent the joint rotation angle set in the robot arm control system at the current time and the previous time, and respectively represent the joint rotation angle in the actual motion process of the robot arm at the current time and the previous time; and the error change rate of the joint rotation angle at the current time is expressed as .

7. The multi-joint torque control method of a mechanical arm for cleaning foreign matters of a power distribution line according to claim 1, characterized in that: The control torque parameters of the mechanical arm in the step S5 include change values of various regulating factors of the controller, specifically including a change value of a proportional factor of the controller , a change value of an integral factor of the controller , and a change value of a differential factor of the controller .

8. The multi-joint torque control method of a mechanical arm for cleaning foreign matters of a power distribution line according to claim 1, characterized in that: The parameter optimization process in step S6 specifically includes the following steps: ① Initialize particle parameters; ② Calculate the fitness value of the particles to obtain the positions of all particles; ③ Sort the fitness values ​​of the particle swarm and assign the top three particles to corresponding positions according to their fitness. Wolf, wolves and Wolves; ④ Update using the gray wolf algorithm The wolf's position is found to be optimal, completing the entire optimization process.

9. The multi-joint torque control method of a mechanical arm for cleaning foreign matters of a power distribution line according to claim 1, characterized in that: The particle position and velocity updating formula in the particle swarm optimization algorithm in the step S6 is expressed as: , , wherein, and denote the velocity of the particle, and denote the particle position, denote the individual optimal position in the particle, denote the population optimal position, denote the weight coefficient, and denote the learning coefficient, and are random numbers.

10. The multi-joint torque control method of a mechanical arm for cleaning foreign matters of a power distribution line according to claim 1, characterized in that: The position update formula in the gray wolf optimization algorithm in step S6 is expressed as follows: , In the formula: and Indicates the location of the gray wolf. and This indicates the location of the prey, which is also the location of the optimal solution. and Indicates the update coefficients. Indicates distance control parameters, The maximum number of iterations, , It is a random number. Indicates the first three levels of gray wolves Wolf, wolves and individual wolves The position at this iteration.

Images