In view of sensorless control of permanent magnetic linear synchronous motor（ PMLSM）

put forward a speed and position estimation method based on modified unscented Kalman filtering（ UKF） to estimate moving speed and position. Compared to conventional UKF algorithm

the modified UKF algorithm improves at obtaining of sampling point. It uses spherical sampling strategy for obtaining sample points

rather than traditional square root symmetric sampling strategy. It greatly reduces number of sampling points

and reduce calculation of state estimation process. When the estimation properties are equivalent

the modified spherical sampling strategy UKF algorithm is better than the traditional square root symmetric sampling UKF algorithm in permanent magnet linear synchronous motor sensorless real-time control system has obvious advantages and obtain good control effect.