Dynamic friction modelling and parameter identification for electromagnetic valve actuator

Abstract

A new modified LuGre friction model is presented for electromagnetic valve actuator system. The modification to the traditional LuGre friction model is made by adding an acceleration-dependent part and a nonlinear continuous switch function. The proposed new friction model solves the implementation problems with the traditional LuGre model at high speeds. An improved artificial fish swarm algorithm (IAFSA) method which combines the chaotic search and Gauss mutation operator into traditional artificial fish swarm algorithm is used to identify the parameters in the proposed modified LuGre friction model. The steady state response experiments and dynamic friction experiments are implemented to validate the effectiveness of IAFSA algorithm. The comparisons between the measured dynamic friction forces and the ones simulated with the established mathematic friction model at different frequencies and magnitudes demonstrate that the proposed modified LuGre friction model can give accurate simulation about the dynamic friction characteristics existing in the electromagnetic valve actuator system. The presented modelling and parameter identification methods are applicable for many other high-speed mechanical systems with friction.