基于ReCorre特征优化的气阀故障诊断方法

Diagnosis Method for Air Valve Failure Based on ReCorre Characteristics Optimization

振动信号包含信息丰富,反应状态直接,振动法是压缩机气阀故障诊断常用方法。振动信号的特征参数繁多,特征向量选择是否合理对故障诊断结果准确性影响很大。研究ReCorre方法对气阀振动信号特征参数进行优化选择,再通过神经网络进行分类识别。实例表明,基于特征优化的模糊神经网络分类识别结果正确率高,识别结果受数据来源影响小,是一种较好的气阀故障诊断方法。

Vibration signal contains rich information and reactive state directly. Vibration method is commonly used to diagnoze air valve failure in compressor. There are many characteristic parameters of vibration signal, so correctly selecting charicteristic vector is greatly important for the accuracy of failure diagnosis results. ReCorre method is studied to carry out the optimized selection of characteristic parameters of air valve＇ s vibration signal. Then a neural network is applied to carry out the classification and identification. The practical example shows that the classification and identification results of fuzzy neural network based on characteristic optimization have high correct rate and the identification results have been affected a little by data resource ,which is a better diagnosis method for air valve failure.