Tuning of Extended Kalman Filter for Speed Estimation of PMSM Drive using Particle Swarm Optimization

ABSTRACT This paper presents a speed estimation technique for Permanent Magnet Synchronous Motor (PMSM) based on Extended Kalman Filter (EKF). Particle Swarm Optimization (PSO) method is used to optimize the noise covariance matrices of EKF, thereby ensuring filter stability and accuracy in speed estimation. The proposed method will be performed in two steps; in the first the covariance matrices are optimized in off-line manner and then in the second step these error covariance matrices values are injected into EKF to estimate rotor speed. Simulation results shows that the covariance matrices improve the convergence of estimation process and quality of the system performance. Keywords: Permanent Magnet Synchronous Motor, Extended Kalman Filter, Particle swarm optimization, Objective function.