摘要
提出了一种基于小波神经网络的电动机转子故障诊断方法,即提取转子的振动信息,利用小波神经网络的非线性模型,实现电机转子故障诊断;同时对多输入输出问题带来的网络规模大、收敛速度慢等问题,提出一种网络优化算法,即采用改进的遗传算法寻找最优小波基元,从而简化小波网络并加快收敛,仿真实例证明该方法是有效的。
A method for rotor fault diagnosis of electrical machinery based on wavelet network is proposed,which collects dynamic vibration in- formation of rotor by acceleration meter,realizes the on-line state detection by use of the non-liner model of wavelet NN.Then aiming at the'dimension disaster'and the slow learning speed caused by the'MIMO',the wavelet notwork is improved by optimization genetic algorithm in order to find the optimum wavelet neurons.Finally,the simpler structure and quickly convergent velocity of the new algo- rithm is demonstrated by simulation results.
出处
《电气自动化》
北大核心
2005年第3期6-8,共3页
Electrical Automation
基金
国家自然科学基金资助(NO:60374020)
河北省自然科学基金资助(NO:F2004000180)
河北省教育厅自然科学研究资助(NO:2003240)
关键词
转子故障
故障诊断
小波网络
优化的遗传算法
rotor fault
fault diagnosis
wavelet network
opimum genetic algorithm