神经网络优化二阶滑模观测器的PMSM无感控制  

Sensorless Control of PMSM Based on Second-order Sliding Mode Observer Optimized by Neural Network

作　　者：张会林 王国强 杨海马[2] ZHANG Hui-lin;WANG Guo-qiang;YANGHai-ma(School of Mechanical Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China;School of Optical-electrical and Computer Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)

机构地区：[1]上海理工大学机械工程学院,上海200093 [2]上海理工大学光电信息与计算机工程学院,上海200093

出　　处：《控制工程》2022年第11期2075-2081,共7页Control Engineering of China

基　　金：国家自然科学基金资助项目(61701296,U1831133);上海市自然科学基金资助项目(17ZR1443500)。

摘　　要：针对表贴式永磁同步电机的速度和位置观测性能不佳的问题,提出一种结合反向传播(BP)神经网络的二阶滑模观测器,用于优化基于超螺旋算法的滑模观测器(STA-SMO)的滑模增益及改进的锁相环(PLL)。采用BP神经网络算法优化二阶滑模观测器的滑模增益,实现了增益在线调节,提高了系统的动态响应。采用改进的PLL,提高了转子位置和转速的观测精度。仿真结果表明,与变增益STA-SMO相比,所提出的二阶滑模观测器的位置误差和抖振减少了25%和40%以上,有效地抑制了滑模抖振,具有更强的鲁棒性。In order to solve the problem of poor speed and position observation performance of surface-mounted permanent magnet synchronous motor, a second-order sliding mode observer combined with back propagation(BP) neural network is proposed to optimize the sliding mode gain of sliding mode observer based on super-twisting algorithm(STA-SMO) and the improved phase-locked loop(PLL). BP neural network algorithm is used to optimize the sliding mode gain of the second-order sliding mode observer, which realizes the on-line adjustment of the gain and improves the dynamic response of the system. The improved PLL is used to improve the observation accuracy of rotor position and speed. The simulation results show that the position error and chattering of the proposed second-order sliding mode observer are reduced by more than 25% and 40% compared with the STA-SMO with variable gain. The proposed controller can effectively suppress sliding mode chattering and improve the robustness.

关 键 词：表贴式永磁同步电机 二阶滑模观测器 BP神经网络 增益在线调节 改进的PLL 

分 类 号：TM341[电气工程—电机]