Compensation control method for robot flexible joints based on lstm hysteresis model

Abstract

The invention discloses a robot flexible joint compensation control method based on the LSTM hysteresis model, which uses the improved LSTM-based industrial robot flexible joint hysteresis characteristic model to online predict the robot joint output angle, and calculates the output torque angle by comparing with the ideal joint output. Corresponding to the angle compensation amount of the input end of the joint, the set input angle of the input end of the joint is compensated. From the motor drive end of the joint, the complex hysteresis characteristic of the joint is offset, and the joint conversion accuracy of the industrial robot is effectively improved. The hysteresis model has online learning ability, which not only compensates the transmission nonlinear error caused by the structure of the industrial robot joint itself, but also compensates the conversion error caused by the slow drift characteristic of the industrial robot joint under long-term operation, which improves the long-term operation of the joint. Stability under precision.

Description

technical field [0001] The invention relates to the technical field of industrial robots, in particular to a compensation control method for flexible joints of a robot based on an LSTM hysteresis model. Background technique [0002] Industrial robots have the advantages of high repeatability, automation, safety and strong applicability. In the context of intelligent manufacturing, the functions of industrial robots are becoming more and more powerful, and at the same time, higher requirements are placed on the execution accuracy of industrial robots. In recent years, robots such as lightweight robots and collaborative robots that have emerged use harmonic drives to drive their joint motions. The complex nonlinear hysteresis characteristics of the flexible joints of industrial robots composed of motors and harmonic reducers often affect the conversion accuracy of joints and directly affect the positioning accuracy of industrial robots. Modeling the nonlinear hysteresis char...

AI Technical Summary

Problems solved by technology

[0003] What the present invention aims to solve is the problem that the complex hysteresis characteristics of the joints in the existing industrial robots are not accurately compensated, resulting in a decrease in the joint conversion accuracy, and provides a compensation control method for the robot's flexible joints based on the LSTM hysteresis model

Images