【摘要】 为了识别空压机气门潜在和早期的微弱故障,提出了一种基于KNN和Wigner-Hough振动谱时频图像识别的气门故障诊断方法.利用Wigner-Hough算法对某船用空压机缸盖表面振动信号进行了时频分析,将振动谱时频图像作为二维输入信号进一步分析和挖掘,提取一阶时间矩、一阶频率矩作为气门故障诊断参数,采用模式识别KNN算法实现其故障的识别.整合上述研究成果,基于Labview、Matlab混合编程开发了气门故障诊断平台.试验结果表明:此方法能较好地识别空压机气门磨损故障.更多还原

【Abstract】 In order to identify potential and early weak faults of air valves of air compressors, a fault diagnosis method of air valves based on KNN and Wigner-Hough vibration spectrum time-frequency image recognition was proposed. The Wigner-Hough algorithm was used to analyze the vibration signal of the cylinder head surface of a marine air compressor in time and frequency. The time-frequency image of the vibration spectrum was further analyzed and mined as a two-dimensional input signal, and the first-order time moment and the first-order frequency moment were extracted as the valve fault diagnosis parameters. Pattern recognition KNN algorithm was used to realize fault recognition. Integrating the above research results, a valve fault diagnosis platform was developed based on Labview and Matlab mixed programming. The test results show that this method can better identify the air valve wear fault of air compressor.更多还原