Nonlinear Friction Compensator Design for Mechatronics Servo Systems Using Neural Network

A neural network compensator for stick-slip friction phenomena in meachatronics servo systems is practically proposed to supplement the traditionally available position and velocity control loops for precise motion control. The neural network compensa- tor plays the role of canceling the effect of nonlinear slipping friction force. It works robustly and effectively in a real control system. This enables the mechatronics servo systems to provide more precise control in the digital computer. It was confirmed that the con- trol accuracy is improved near zero velocity and the points of changing the moving direction through numerical simulation. Ho wever, asymptotic property of the steady state error of the normal operating points is guaranteed by the integral term of traditional velocity loop controller. Stick-slip friction is the natural resistance phenomena to relative motion between two contacting bodies in the mecha- tronics servo systems such as industrial robot arms and nu- merical control (NC) machine tools . It is commonly composed of Coulomb friction, static friction, and viscous frictions, etc., but it has highly nonlinear characteristics. In many motion control systems, this friction phenomenon becomes a dom i- nant factor near zero velocity that prevents them from high- precise control because of its nonlinear nature and difficulties in handling it effectively and in compensating it adequately with the linear feedback control system. In order to achieve high precision motion control, these frictions must be acc u- rately compensated to cancel the effects in the real time con- trol system.