基于EKF-LESO级联观测的永磁直线同步电机无传感器线性自抗扰控制

Sensorless LADRC System of PMLSM based on EKF-LESO cascade observation

为克服永磁直线同步电机(Permanent-Magnet Linear Synchronous Motor, PMLSM)敏感扰动、易受噪声影响诱发抖振的不足,摆脱位置传感器对电机伺服系统的束缚,构建了一种以线性自抗扰控制器(Linear Active Disturbance Rejection Control, LADRC)为核心、扩展卡尔曼滤波算法(Extended Kalman Filter, EKF)和线性扩张状态观测器(Linear Extended State Observer, LESO)级联观测的PMLSM无传感器矢量控制系统。首先从理论上分析了控制系统的误差来源,通过推导得出在输入量测信号存在慢时变偏差的情况下,LESO对除第1阶和第n+1阶外的"中阶"状态进行观测时观测精度不受影响的结论;进而以此为依据,提出了EKF+LESO的级联观测方式,搭建了无传感器位置环自抗扰控制系统。通过仿真对LESO的容错性、对控制系统的跟随能力和噪声抑制能力进行了验证,实验结果则进一步表明,所设计系统能够准确实现直线电机的速度观测,系统运行平稳,具有较好的实用性。

To overcome a permanent-magnet linear synchronous motor(PMLSM)’s shortcomings of being sensitive to disturbance and easy to chatter due to noise, and get rid of position sensor’s constraint to motor servo system, a PMLSM sensor-less vector control system with the linear active disturbance rejection controller(LADRC) taken as its core, and the extended Kalman filter(EKF) algorithm and the linear extended state observer(LESO) as its cascade observation was constructed. Firstly, the error source of the control system was analyzed theoretically, it was shown that the observation accuracy of LESO is not affected in cases of input measured signals existing slow time-varying bias and observing "middle order" states except the first order state and the(n+1) one. Then, the cascade observation mode of EKF+LESO was proposed, and a sensor-less position loop LADRC system was built. Finally, simulation was done to verify fault tolerance of LESO, following ability and noise suppression ability of the control system. The test results further showed that the designed system can be used to correctly realize speed observation of the linear motor;the system operates smoothly and has a better practicability.