基于模糊神经网络的高阶频段振动信号故障诊断研究

Fault Diagnosis by Vibration Signal at the Range of High Frequency Based on Fuzzy Neural Network

将模糊神经网络技术应用于机械故障高阶频率振动信号的研究 ,参考齿轮和转子故障模式并结合专家经验建立了螺杆压缩机转子故障诊断专家系统知识库 ,利用振动频谱特征就螺杆压缩机的几种故障模式结合一种模糊神经网络故障诊断模型进行了模糊神经网络识别。算例诊断结果为压缩机阴阳转子型线加工误差 ,与试验结果一致。

Variations in amplitudes of a vibration signal at the range of high frequency acquired from rotating machinery of the screw compressor kind usually show the appearance of the faults in the machine. In the paper, the technique of fuzzy neural network is applied to the fault diagnosis of a screw compressor rotor by its vibration signals at the range of high frequency. In the light of the fault patterns of the gear and the rotor combined with expert knowledge, an expert knowledge database for the fault diagosis of the screw compressor rotor is established. A diagnosis on an actual screw compressor using the method proposed in the paper indicates the shape mismachining of its male and female rotors. Measurement results prove the diagnosis.