Time-optimal Speed Planning Algorithm For Industrial Robots

Industrial robots are playing an increasingly important role in current industrial production.With the advancement of technology year by year,the requirements for control methods and trajectory planning methods of industrial robots have become higher and higher.However,most controllers do not consider the dynamic model of the robot,and only consider the kinematic constraints of the joint,which limits the accuracy and performance of the robot.Based on the laboratory's 6-axis serial robot,we mainly discussed the kinematics and dynamics modeling of the robot,the parameter identification of the robot,and the time-optimal planning of the robot.This thesis first summarizes the current development of industrial robots and the research status of serial robot parameter identification and robot time-optimal trajectory planning problem.Then,we construct the kinematic model of the industrial robot.The Newton-Euler method is selected to construct the dynamics model of the robot.We rewrite the dynamic equations to linear form for the identification of dynamic parameters.After getting the dynamic model of the robot,the full inertial parameters of the robot were converted to base inertial parameters for identification.The excitation trajectory was optimized by the condition number of observation matrix with the constrain of joint motion ability and collision of the mechanical structure.A physically feasible system identification method is used to convert the least-square problem into a semipositive definite optimization problem with constraints of linear matrix inequality under consideration of physically feasible constraints.After obtaining the dynamic model of the robot,this thesis chooses an algorithm based on reachability analysis to achieve the time-optimal trajectory planning of the robot's path.Then we prove the correctness and optimality of this planning algorithm.A segmented speed planning method is put forward to obtain an online approximate optimal speed planning to deal with the problem that the global optimal speed planning can not be calculated online in real-time.This method of piecewise calculation can plan the approximate optimal robot trajectory online in real-time on the premise of a reasonable selection of the number of forwarding segments.Finally,the joint position and torque data of the robot are collected from an identification experiment.A zero-phase Butterworth filter is performed to decrease the noiseof the joint position and the torque from sensors.Then the dynamic parameters of the robot were identified,and the accuracy of the dynamic model was verified by using another excitation trajectory.Then,the global time-optimal speed planning method and the segmented time-optimal speed planning method were used to plan the trajectory of the robot,and the real-time and approximate optimality of the segmented time-optimal speed planning method was verified.