定子电阻的小波神经网络辨识  

Wavelet neural network online identification of stator resistance

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》2008年第4期295-298,共4页Journal of Heilongjiang Institute of Science and Technology

基　　金：黑龙江省自然科学基金项目(E2007-34)

摘　　要：为改善直接转矩控制系统(DTC)低速运行时的性能,采用小波神经网络设计了定子电阻辨识器,对直接转矩控制系统的参数加以辨识。该方法能够充分利用小波变换或分解所表现的良好时频局部化特性以及多尺度功能,实现低速运行时对电阻变化的精确控制。仿真结果表明:该辨识器能够精确辨识定子电阻值,极大地提高了直接转矩控制系统的低速性能。Given the poor performance of the direct torque control system which runs slowly, this paper gives the design of the stator resistance identification equipment based on wavelet neural network to identify the direct torque control system （DTC） parameter. According to the best time frequency lo- cal specialty of wavelet and its multi scale function, the method permits the precise control of the resistance variety on slow speed. The results show that identification equipment can distinguish the resistance value precisely and improve the slow speed performance of the direct torque control system.

关 键 词：直接转矩控制 小波神经网络 辨识器 直线永磁同步电动机 

分 类 号：TP271.4[自动化与计算机技术—检测技术与自动化装置]