开关磁阻电机直接自适应神经网络控制  

Direct adaptive neural network control of switched reluctance motors

作　　者：李存贺 王国峰[1] 李岩 范云生[1] 许爱德[1] 

机构地区：[1]大连海事大学信息科学技术学院,辽宁大连116026

出　　处：《电机与控制学报》2018年第1期29-36,共8页Electric Machines and Control

基　　金：国家自然科学基金(51407021);辽宁省自然科学基金(2015020022);中央高校基本业务费(3132016312)

摘　　要：针对开关磁阻电机调速系统存在的未知参数波动和外部负载扰动问题,提出了直接瞬时转矩控制下的基于最小学习参数的直接自适应神经网络控制算法实现开关磁阻电机高品质调速控制。采用RBF神经网络对包含未知参数波动和外部负载扰动等不确定项的理想控制律进行整体逼近。将神经网络理想权值的范数作为在线估计参数,使在线学习参数由多个权值减少为一个,降低了控制器的计算负担。基于李雅普诺夫函数的稳定性分析保证了闭环调速系统半全局一致最终有界稳定。与PI控制的对比仿真试验表明,直接自适应神经网络控制器能够有效地提高开关磁阻电机调速系统对参数波动的自适应性和对外部负载扰动的鲁棒性。In order to solve parameter variations and external load disturbance of SRM drive, a novel direct adaptive neural network controller was presented to achieve the high quality speed control of SRM drive.The controller based on the technology of minimal learning parameters was designed wit!i direct instanta-neous tor〈que control. The RBF neural networl^ was adopted to approximate the ideal control law which in-cluded the parameter variations and external load of t!ie SRM drive. The ntor was used as an on-line estimation parameter, which reduces t!ie on-line learning parameters from multi-ple to one and reduces the computational burden of the controller. The stability analysis based on lya-punov stability theory guarantees the uniform ultimate boundedness stability of the cparative studies were carried out between the proposed controller and PI controller on a 6/4 pole SRM,and the results show that the proposed control scheme has a beter performance fand external load disturbance.

关 键 词：开关磁阻电机 速度控制 直接自适应神经网络 最少学习参数&P控制 直接瞬时转矩控制 

分 类 号：TM352[电气工程—电机]