Research on Parameter Identification Method for LuGre Friction Model in Permanent Magnet Synchronous Motor Servo Systems

To address the insufficient identification accuracy of LuGre friction parameters in permanent-magnet synchronous servo systems, an Adaptive Weight-Inertia Particle Swarm Optimization (AWI-PSO) is proposed. Nonlinear inertial-weight decay balances global search and local convergence; a fitness-feedback adjustment enhances self-adaptive search; and a staged static-dynamic identification framework reduces parameter coupling. Under steady-state and dynamic conditions, AWI-PSO confines the relative errors of Stribeck velocity and maximum static friction within 1 %, cuts the mean error and root-mean-square error by 38.6 % and 30.3 % compared with basic PSO, and lifts the goodness-of-fit R² to 0.9932. Convergence curves show phased optimization from static to dynamic identification. The study verifies AWI-PSO’s marked superiority in LuGre model parameter identification, providing an efficient and reliable optimization strategy for friction modeling in complex servo systems.