Servo Motor Control Based On Friction Compensation

Servo control technology is an important control technology in industrial automation control.Servo control system has high performance requirements,especially the tracking performance of position reference signal.The research shows that the main factor affecting the performance of the servo system is the nonlinear error caused by friction,so the control performance of the servo system can be improved by eliminating the nonlinear error.This thesis takes the three-phase permanent magnet synchronous motor as the main body,uses vector control and three closed-loop control of position,speed and torque to establish the AC servo control system,analyzes the characteristics of friction and establishes the friction model,uses the friction model-based method to design the friction compensation controller,designs the neural network to identify nonlinear variables,and the robust adaptive controller to estimate other unknown parameters and eliminate errors.The main research contents are as follows:(1)The three-phase permanent magnet synchronous motor servo control system is established by using vector control and three closed-loop control algorithm.The established servo control system can quickly recover and track the given reference signal when the load disturbance is applied.(2)The characteristics of nonlinear friction in servo control system are studied and the friction model is analyzed.Lu Gre friction model is selected to describe the friction.Experiments show that the friction described by Lu Gre friction in servo control will lead to overshoot of tracking given signal during steady-state operation and "crawling" phenomenon during slow-speed operation.Aiming at the nonlinear friction and the nonlinear function in Lu Gre friction model,a neural network is designed to identify the nonlinear variables.(3)The friction compensation controller is designed by the methods of not based on friction model and based on friction model to solve the influence of friction on the performance of servo control system.According to the designed parameter identification neural network,the robust adaptive algorithm is used to eliminate the identification error of neural network and estimate other unknown parameters in the system.The performance of the designed friction compensation controller is analyzed.Simulation experiments show that the designed global approximation neural network robust adaptive controller based on friction model can achieve the expected performance index of the system.