摘要
根据汽轮机转子振动信号特点,提出了小波包分析和概率神经网络相结合的故障诊断方法。利用小波包对故障信号进行分解,然后将归一化后的数据用于概率神经网络,从而对信号特征及其故障类型建立非线性映射,实现故障诊断。MATLAB的实验仿真表明,小波包分析和概率神经网络的结合在汽轮机转子常见故障的诊断中是很有效的。
According to the signal features of turbine rotor vibration faults,a diagnosis method in combination of wavelet packet and probabilistic neural network is being proposed.Fault signals are decomposed using wavelet packet,and then the unified data will be processed by probabilistic neural network,whereafter a nonlinear mapping relationship between signal features and fault type can be obtained,and thus a fault diagnosis competed.Experimental results of MATLAB simulation show that the combined method by wavele...
出处
《发电设备》
2009年第6期397-399,共3页
Power Equipment
关键词
汽轮机
转子
故障诊断
小波包
神经网络
steam turbine
rotor
fault diagnosis
wavelet packet
neural network
