Abstract: Friction-induced self-excited vibration is harmful to engineering systems. Understanding this phenomenon and developing some strategies to effectively control the vibration, has theoretical and practical significances. The reasons of friction self-excited vibration are explored through analyzing some numerical characteristics of two kinds of friction models. A fuzzy modeling technique based adaptive active control strategy is proposed to eliminate or reduce the amplitude of limit cycle and steady-state error. Analytic solutions of friction self-excited vibration are discussed and numerical simulations of limit cycles caused by Coulomb friction and Stribeck friction are carried out. In view of the defects and lack of self-learning ability and environmental adaptability of conventional friction models, it is proposed that the adaptive fuzzy modeling technique is adopted for on-line approximating the friction force, and the identification result is then employed in the active controller design. Simulation results demonstrate the effectiveness of the PD active control technique based on the adaptive fuzzy system compensation, which can effectively restrain the amplitude of limit cycle and eliminate the system steady-state error.

Key words: Active control, Adaptive, Friction model, Fuzzy system, Self-excited vibration