Transmission friction measurement and suppression of dual-inertia system based on RBF neural network and nonlinear disturbance observer

Abstract

With the high demand for the anti-interference ability and motion accuracy of robot systems, the friction caused by the vibration of gears and connecting parts affects the stability and accuracy of the motor torque transmission. Thus, the measurement and compensation of the friction interference of reducers become more important. In this paper, a friction model is modified based on the gear system, the excitation transmission relationship between the gear system and the dual-inertia system is analyzed, and an adaptive control strategy is proposed to measure and suppress friction phenomena in the dual-inertia system. Subsequently, the transmission friction observation of different motor position signals, the influence of the control method, and tooth surface friction coefficient on system response are studied. The results show that the LuGre model based on a gear system can characterize the friction torque affected by multiple factors, and the proposed method can effectively suppress the friction phenomenon.