Search

CN-122008252-A - Robot driving control system based on particle swarm algorithm

CN122008252ACN 122008252 ACN122008252 ACN 122008252ACN-122008252-A

Abstract

The invention relates to the technical field of robot control and intelligent optimization control, in particular to a robot driving control system based on a particle swarm algorithm, which comprises a capability observation module, a manifold construction module, a manifold optimizing module, a driving closed-loop module and a safety driving capability, wherein the capability observation module is used for receiving theoretical track instructions to output a system capability tensor, the manifold construction module is used for reading physical parameters of a robot body and initializing a knowledge manifold matrix containing initial damping coefficients based on the physical parameters of the robot body to generate a target manifold matrix, the manifold optimizing module is used for initializing a particle swarm on the surface of the target manifold matrix to output target control parameters of a current control period, the driving closed-loop module is used for converting the target control parameters into driving waveforms to control a driving motor, and carrying out numerical adjustment on the initial damping coefficients of the knowledge manifold matrix based on track residual errors.

Inventors

  • CAO FULU

Assignees

  • 陕西伏特安电气科技有限公司

Dates

Publication Date
20260512
Application Date
20260410

Claims (9)

  1. 1. A robot drive control system based on a particle swarm algorithm, the system comprising: The capacity observation module is used for receiving theoretical track instructions issued by the upper computer through the communication interface, collecting current data and position data of a driving motor of the robot in real time through the configured current sensor and the position encoder, and calculating mechanical impedance characteristics based on the current data and the position data so as to output a system capacity tensor comprising physical dimensions such as friction characteristics, load inertia characteristics and the like; the manifold construction module is used for reading physical parameters of a robot body, initializing a knowledge manifold matrix which comprises initial damping coefficients calibrated in advance based on rated torque of the driving motor and a maximum allowable overload current ratio and is used for representing potential energy distribution states in a physical parameter space based on the physical parameters of the robot body, and carrying out numerical update on local curvature of the knowledge manifold matrix by taking the system capacity tensor as update weight so as to generate a target manifold matrix; The manifold optimizing module is used for initializing a particle swarm on the surface of the target manifold matrix, deriving the target manifold matrix to extract a manifold gradient matrix, and carrying out single-step iterative optimization on the particle swarm in a current control period based on the manifold gradient matrix and the system capacity tensor to output target control parameters of the current control period; The driving closed loop module is used for converting the target control parameters into driving waveforms to control the driving motor, collecting the actual running track of the driving motor, extracting track residual errors of the actual running track and a theoretical track corresponding to the theoretical track instruction, and carrying out numerical adjustment on the initial damping coefficient of the knowledge manifold matrix based on the track residual errors.
  2. 2. The robot drive control system based on a particle swarm algorithm according to claim 1, wherein the capability observation module comprises: The data synchronization unit is used for synchronously collecting the current data and the position data at a preset frequency; The differential calculation unit is used for carrying out differential processing on the position data to obtain speed differential data, and carrying out on-line parameter identification based on the current data and the speed differential data so as to calculate friction force data and load inertia data; and the tensor generation unit is used for packaging the friction force data and the load inertia data into the system capacity tensor.
  3. 3. The robot drive control system based on a particle swarm algorithm according to claim 1, wherein said manifold construction module comprises: The parameter reading unit is used for reading the reduction ratio parameter and the rated torque parameter of the driving motor as the physical parameters of the robot body; the matrix initializing unit is used for constructing a three-dimensional manifold curved surface based on the reduction ratio parameter and the rated torque parameter and converting the three-dimensional manifold curved surface into the knowledge manifold matrix; And the target definition unit is used for mapping the control state representing that the deviation rate of the theoretical track and the actual track at the target position is towards zero into the minimum value point in the knowledge manifold matrix.
  4. 4. The robot drive control system based on the particle swarm algorithm according to claim 2, wherein the manifold construction module further comprises a curvature updating unit for: under the condition that the load inertia data is higher than a preset inertia threshold value, determining a target position area according to a theoretical track corresponding to the theoretical track instruction, and increasing the curvature value of matrix elements corresponding to the knowledge manifold matrix in the target position area according to a preset updating step length; Reducing the curvature value of the knowledge manifold matrix according to a preset updating step length under the condition that the load inertia data is lower than a preset inertia threshold value; Under the condition that the load inertia data is equal to a preset inertia threshold value, keeping the current curvature value of the knowledge manifold matrix unchanged; The preset inertia threshold value is determined based on superposition of a calibration inertia reference value of the robot in an idle state and a dynamic margin of a current working condition.
  5. 5. The robot drive control system based on a particle swarm algorithm according to claim 1, wherein the manifold optimization module comprises: The speed updating unit is used for initializing a current updating speed vector of the particle swarm, multiplying the current updating speed vector by the manifold gradient matrix in a matrix manner, and carrying out element-by-element weighted calculation on the multiplication result and the system capacity tensor so as to obtain a target updating speed; and the parameter output unit is used for determining the spatial position of the particle swarm in the current control period based on the target update speed and decoding the spatial position into target control parameters of the current control period, wherein the target control parameters comprise a proportional coefficient, an integral coefficient, a differential coefficient and a feedforward coefficient.
  6. 6. The robot drive control system based on a particle swarm algorithm according to claim 1, wherein said drive closed loop module comprises: The waveform conversion unit is used for writing the target control parameters into a controller register of the driving motor and converting the target control parameters into pulse width modulation waveforms; A motor driving unit for driving the driving motor by using the pulse width modulation waveform, and enabling the joints of the robot to execute actions through a transmission part of the driving motor; and the residual error extraction unit is used for collecting the actual running track corresponding to the joint execution action and comparing the actual running track with the theoretical track corresponding to the theoretical track instruction so as to output the track residual error.
  7. 7. The robot drive control system of claim 6, wherein the drive closed loop module further comprises a damping fine tuning unit for: Under the condition that the track residual is higher than a preset residual threshold, increasing the initial damping coefficient of the knowledge manifold matrix according to a preset adjustment step length; reducing the initial damping coefficient of the knowledge manifold matrix according to a preset adjustment step length under the condition that the track residual is lower than a preset residual threshold; And under the condition that the track residual is equal to a preset residual threshold, maintaining the initial damping coefficient of the knowledge manifold matrix unchanged.
  8. 8. The robot drive control system based on the particle swarm optimization according to claim 1, wherein the system is applied to an industrial robot arm control scene, the physical parameters of the robot body comprise an arm span length parameter and a joint quality parameter of the industrial robot arm, and the knowledge manifold matrix is used for restraining overshoot of an end effector of the industrial robot arm in a grabbing process.
  9. 9. The robot drive control system based on the particle swarm algorithm according to claim 1, wherein the system is applied to a pipeline robot driving scene, the system capacity tensor comprises a real-time friction coefficient variation of the inner wall of the pipeline, and the target manifold matrix is used for adapting to the abrupt change of the pipeline load and suppressing driving oscillation.

Description

Robot driving control system based on particle swarm algorithm Technical Field The invention relates to the technical field of robot control and intelligent optimization control, in particular to a robot driving control system based on a particle swarm algorithm. Background Along with the continuous improvement of the requirements of scenes such as refining maintenance, pipeline inspection, industrial assembly and the like on the motion precision and real-time performance of the robot, the stability control problem of a robot driving control system under the conditions of complex load, friction mutation and multi-working condition switching is increasingly outstanding; The traditional robot driving control is mainly dependent on a servo control mode based on fixed parameters, a parameter adjustment mode based on experience setting and an off-line parameter searching mode based on a conventional optimization algorithm at present; However, the fixed parameter control, the experience setting, the conventional offline optimization and other modes have certain defects, such as difficulty in adapting to real-time changes of load inertia and friction states, easy overshoot or tracking error increase near a target position, strong dependence on operators and preset working conditions in the experience setting mode, difficult adaptation to dynamic switching in a complex environment, difficulty in meeting the real-time requirement of millisecond control period due to large iteration quantity in the conventional optimization algorithm, insufficient constraint on physical boundary and actual execution residual error of a robot body, and easy influence on driving stability. Disclosure of Invention The invention aims to provide a robot driving control system based on a particle swarm algorithm, which solves the following technical problems: the problems that the traditional parameter optimization is separated from a physical boundary and the calculated amount is too large are avoided, so that the rapid positioning and low overshoot control of the robot under the requirements of complex load, complex friction and high real-time performance are realized, and the robot is easier to adapt to abrupt changes of a physical environment and inhibit driving oscillation. The aim of the invention can be achieved by the following technical scheme: a robot drive control system based on a particle swarm algorithm, the system comprising: The capacity observation module is used for receiving theoretical track instructions issued by the upper computer through the communication interface, collecting current data and position data of a driving motor of the robot in real time through the configured current sensor and the position encoder, and calculating mechanical impedance characteristics based on the current data and the position data so as to output a system capacity tensor comprising physical dimensions such as friction characteristics, load inertia characteristics and the like; the manifold construction module is used for reading physical parameters of a robot body, initializing a knowledge manifold matrix which comprises initial damping coefficients calibrated in advance based on rated torque of the driving motor and a maximum allowable overload current ratio and is used for representing potential energy distribution states in a physical parameter space based on the physical parameters of the robot body, and carrying out numerical update on local curvature of the knowledge manifold matrix by taking the system capacity tensor as update weight so as to generate a target manifold matrix; The manifold optimizing module is used for initializing a particle swarm on the surface of the target manifold matrix, deriving the target manifold matrix to extract a manifold gradient matrix, and carrying out single-step iterative optimization on the particle swarm in a current control period based on the manifold gradient matrix and the system capacity tensor to output target control parameters of the current control period; The driving closed loop module is used for converting the target control parameters into driving waveforms to control the driving motor, collecting the actual running track of the driving motor, extracting track residual errors of the actual running track and a theoretical track corresponding to the theoretical track instruction, and carrying out numerical adjustment on the initial damping coefficient of the knowledge manifold matrix based on the track residual errors. Preferably, the capability observation module includes: The data synchronization unit is used for synchronously collecting the current data and the position data at a preset frequency; The differential calculation unit is used for carrying out differential processing on the position data to obtain speed differential data, and carrying out on-line parameter identification based on the current data and the speed differential data so as to calculate friction force data