Unscented particle filter-based state estimation for permanent magnet linear synchronous motor

This paper investigates the rotor angular velocity state estimation of permanent magnet linear synchronous motor (PMLSM) sensorless system. The velocity feedback is an indispensable link to realize the accurate speed control of PMLSM. Accurate rotor information acquisition is the essential problem of the whole control system, since it is impossible to obtain rotor information by traditional mechanical sensors in the non-position sensor control of PMLSM drive system. To work out this problem, this paper raises a non-position sensor control method based on unscented particle filter (UPF). Firstly, the nonlinear dynamic model is established for PMLSM. In the framework of particle filter (PF), the latest measurement information is to generate the predicted particles. Finally, a numerical example and simulated results show that UPF can accurately estimate the rotor angular velocity.