基于Smith预估和性能加权函数的永磁直线同步电机鲁棒迭代学习控制  

Robust Iterative Learning Control for PMLSM Based on Smith Predictor and Performance Weighting Function

作　　者：赵希梅[1] 马志军[1] 朱国昕[1] 

机构地区：[1]沈阳工业大学电气工程学院,沈阳110870

出　　处：《电工技术学报》2016年第19期141-146,共6页Transactions of China Electrotechnical Society

基　　金：国家自然科学基金(51175349);辽宁省教育厅科学技术研究(L2013060)资助项目

摘　　要：针对永磁直线同步电机(PMLSM)迭代学习控制(ILC)过程中,由于扰动及时间滞后引起的系统不稳定、误差难以收敛及跟踪精度下降等问题,提出一种基于Smith预估和性能加权函数的鲁棒ILC方案。Smith预估器与ILC相结合,可在不需要PMLSM精确数学模型的情况下,减少时间滞后对系统跟踪性能的影响,避免迭代过程中由于时间滞后的累积而引起的系统不稳定。由于系统存在外部扰动、参数变化、端部效应等不确定因素,充分利用性能加权函数的信息设计反馈控制器,在满足鲁棒收敛条件情况下,可使位置误差收敛到期望值。实验结果表明,所提出的控制方案可以提高PMLSM伺服系统的位置跟踪精度,增强系统的鲁棒性。In the process of iterative learning control（ILC） for permanent magnet linear synchronous motors （PMLSM）, due to the disturbances and time delay, the system might be unstable, the error is difficult to converge, and the tracking precision is reduced. A robust ILC scheme based on the Smith predictor and the performance weighting function is proposed. In the case that the precise mathematical model of PMLSM is unwanted, the combination of the Smith predictor and ILC decreases the influence of time delay on the system tracking performance. Then it can avoid the instability of the system in the iterative process caused by the accumulation of time delay. Due to these uncertain factors of the external disturbance, parameters change, and end effect etc. , the scheme makes full use of the information of the performance weighting function to design the feedback controller. In the case of fulfilling robust convergence condition, it can make the position error converge to the desired value. The experimental results indicate that the proposed control scheme improves the position tracking accuracy of the PMLSM servo system and enhances the robustness of the system.

关 键 词：永磁直线同步电机 迭代学习控制 性能加权函数 时间滞后 

分 类 号：TM359[电气工程—电机]