基于时频分布图像的柴油机故障诊断

摘要

为研究柴油机故障诊断中图像分析法存在的问题,提出了一种基于时频分布图像分析的柴油机故障诊断方法.采集缸盖表面振动信号进行经验模态(Empirical Mode Decomposition,EMD)分解,得到各阶内禀模态函数(Intrinsic Mode Function,IMF),由前2阶IMF重构振动信号,并经Hibert-Huang变换得到解析信号.试验结果表明,选择合适的时频分布分析方法得到解析信号的时频分布图,可有效地区分柴油机故障工况.%In order to study the remaining problems of image processing method used in diesel engine fault diagnosis, a new method is presented based on time-frequency distribution image processing. The vibration signals of cylinder head surface are obtained and decomposed into a finite number of intrinsic mode function components by the EMD method. Using the reconstruct vibration signals by IMF1 and IMF2, the analytic signals are got by Hibert-Huang transform. Experimental results show that selecting the appropriate time-frequency distribution analysis and obtaining time-frequency distribution of analytic signal can distinguish the diesel engine fault conditions effectively.