中文摘要：

针对永磁直线同步电机无位置传感器控制时，单一的位置估计算法难以在宽范围内精确估计动子速度和位置信号的问题，为了进一步适应直线电机往复运动、速度变化范围大的特点，提出了高频信号注入法和增广扩展卡尔曼滤波算法复合的估计算法。在起动和低速时采用高频信号注入法，在中高速时采用增广扩展卡尔曼滤波法，在过渡区域采用高频信号注入法和增广扩展卡尔曼滤波算法融合的方法，实现从零速到高速全速范围内高精确度无位置传感器控制。仿真和实验结果表明，基于该复合算法的无位置传感器控制系统在全速范围内都具有良好的动态响应特性。

英文摘要：

In position sensorless control for permanent magnet linear synchronous motor （ PMLSM ）, a sin- gle estimation algorithm cannot accurately estimate speed and position of the mover iu wide speed range. In order to adapt to the PMLSM＇ s characteristics of reciprocating motion and big velocity change range, a composite estimation algorithm is put forward. A full speed sensorless control method can be realized from zero speed to high speed range, using the high-frequency voltage signal injection method at starting and low speed, the state augmented extended Kalman filtering （AEKF） method at high speed using, and fu- sing the high-frequency voltage signal injection method and the AEKF method by weighting method in the transition speed area. Simulation and experimental results show the proposed sensorless control system based on the complex algorithm has good dynamic response performance within full speed.