基于模型参考自适应的直线电机转速辨识  

Parameters identification on speed of linear motor on model reference adaptive control

作　　者：沈显庆[1] 周宝国[1] 

机构地区：[1]黑龙江科技学院电气与信息工程学院,哈尔滨150027

出　　处：《黑龙江科技学院学报》2007年第3期206-209,共4页Journal of Heilongjiang Institute of Science and Technology

基　　金：国家自然科学基金资助项目(50375102)

摘　　要：针对直线永磁同步电动机全闭环控制系统易受干扰、降低系统性能指标、存在不稳定因素等问题,提出了基于模型参考的自适应模糊神经网络在线辨识方法,用梯度法实时修正模糊控制器的输入和输出隶属度参数,建立了模型参考自适应模糊神经网络速度伺服系统模型,给出了模糊神经网络控制器的设计方法。仿真和实验结果表明:该方法提高了速度检测装置的分辨率和动态响应能力,系统具有很强的鲁棒性。Aimed at more likely disturbance affecting the closed-loop control system of linear permanent magnet synchronous motor, the resultant degradation of the performance of the control system and poor stability, the paper introduces an on-line identification method of model reference adaptive control based on fuzzy neural network. The inputs and membership parameters of the fuzzy controller are modified in real-time by a gradient method. It is proved that the method is capable of improving the resolution of velocity-detecting devices and dynamic response by simulation and practice. The system proves robust.

关 键 词：模型参考自适应 模糊神经网络 直线永磁同步电动机 梯度法 

分 类 号：TM341[电气工程—电机] TP273.2[自动化与计算机技术—检测技术与自动化装置]