基于自适应线性神经元滤波的内置式永磁电机转子位置观测器  

Adaptive Linear Element Filtering Based Rotor Position Observer for Interior Permanent Magnet Synchronous Motors

作　　者：张国强[1] 王高林[1] 倪荣刚 徐进[1] 曲立志 徐殿国[1] 

机构地区：[1]哈尔滨工业大学电气工程及自动化学院,哈尔滨150001

出　　处：《电工技术学报》2016年第6期47-54,共8页Transactions of China Electrotechnical Society

基　　金：伺服驱动及电机测试规范;标准研究与测试平台国家科技重大专项(2012ZX04001051);国家自然科学基金(51207030)资助项目

摘　　要：为提高无位置传感器内置式永磁同步电动机(IPMSM)控制性能,提出一种改进的基于反电动势模型的自适应滤波转子位置观测器。针对逆变器非线性和磁场空间谐波引起扩展反电动势6k±1次谐波,进而产生6k次转子位置脉动观测误差问题,提出一种基于递归最小二乘算法自适应线性神经元滤波器的转子位置观测方法,从而实现IPMSM矢量控制系统准确解耦控制。该方案能够在线连续调整权重分量,保证了转子位置观测器的快速收敛性。通过自适应滤波器滤除指定的反电动势观测值谐波分量,从而提高正交软件锁相环获得转子位置信息的准确度。最后通过模型仿真和2.2k W IPMSM无传感器矢量控制系统验证了控制策略的有效性。To promote the performance for the position sensorless interior permanent magnet synchronous motor （IPMSM） drives, an improved back electromotive force model based adaptive filtering rotor position observer is presented. The （6k±l）th harmonics exist in the extended electromotive force estimates due to the influence of the inverter nonlinearities and the flux spatial harmonics, which give rise to the （6k）th harmonic ripple in the rotor position estimate. The recursive least square （RLS） algorithm based adaptive linear element （ADALINE） filter is proposed, whereby the accurate decoupling control can be accomplished for the vector controlled system. RLS algorithm can online update the weights of ADALINE filter continuously, which guarantees the fast convergence rate of the rotor position observer. The selective ripple harmonic components can be eliminated through the proposed adaptive filter. Therefore the accuracy of the rotor position obtained from the software quadrature phase-locked loop （PLL） can be improved. Model simulation and experiments on a 2.2kW IPMSM sensorless vector controlled drive have been carried out to verify the proposed scheme.

关 键 词：内置式永磁同步电机 无位置传感器控制 位置观测误差 自适应线性神经元滤波器 权重在线更新 

分 类 号：TM351[电气工程—电机]