Position and Speed Estimation of Permanent Magnet Linear Synchronous Motor Based on UKF

Peiqiong Yu,Zichen Chen

DOI: https://doi.org/10.3321/j.issn:1000-6753.2007.09.011

2007-01-01

Abstract:Unscented Kalman filter (UKF) has been successfully applied in many nonlinear estimation and control problems. For the position sensorless control of permanent magnet linear synchronous motor (PMLSM) direct drive system, UKF is applied in the position and speed estimation of PMLSM. An UKF-based position and speed estimator is applied in PMLSM drive system. The position sensorless PMLSM drive system using UKF is presented. To obtain the best estimated values of the states, the UKF has to be tuned by iterative searching for covariance matrices which yield the best estimates. Both the PMLSM drive system with position sensor and the PMLSM drive system without position sensor are studied by simulations. The simulation results show, the UKF-based position and speed estimator has good dynamic and steady state estimation performance, the UKF-based PMLSM drive system without position sensor has good speed response performance over a wide speed range, even for large initial position errors. The experiments show that UKF-based estimator is feasible and efficient for estimating position and speed of PMLSM.