摘要
针对电机振动信号的频谱特点,提出基于小波 神经网络技术的电机故障模式识别与诊断的新方法。利用小波包的多维多分辨率特性,对电机振动信号进行分解与重构,获得振动信号的突变信息,提取与电机故障相关的特征信息,将其作为特征向量输入ART2神经网络,对其进行训练。经过训练后的神经网络可对电机工作状态进行在线监测和实时故障诊断,并在转子实验台上进行了模拟故障仿真试验。通过对仿真结果的分析,证实这种诊断方法的可行性。
A novel method of pattern recognition and fault diagnosis in electrical machine based on the wavelet-neural network is proposed according to the frequency spectrum characteristics of vibration signal. Based on the advantage of multi-dimensional, multi-scaling decomposition of wavelet packets, the abrupt change information can be obtained and the features related to the fault of electrical machine is extracted through the decomposing and reconstruction of the vibration signals of the electrical machine. The extracted features are inputted into ART2 neural network to diagnose the type of the fault. The trained ANN can be used to the online state monitor and real-time fault diagnosis of electrical machine. All simulated failures are emulated on experiment rig. The feasibility of this novel method is proved by the simulation results.
出处
《控制工程》
CSCD
2004年第2期152-154,176,共4页
Control Engineering of China
基金
国家高技术研究发展计划资助项目(2001AA411230)
