基于声强和小波包分析的汽车发动机燃爆信息特征提取研究

Research on Extracting Engine Combustion Information Based on Sound Intensity and Wavelet Packets

针对发动机燃爆信息的非平稳性,利用声强近场测量和小波包多分辨率分析的特点,提出了一种基于小波包分析的声强算法,本文利用该方法对某型号发动机燃爆信息的特征提取进行了实验研究。实验结果表明,该方法可以从复杂的发动机缸盖辐射噪声中提取出燃爆信息,所提取的燃爆信息可作为发动机单缸失火的故障诊断指标。本文所提出的基于小波包分析的声强算法为汽车发动机运行状态的在线监测和故障诊断提供了一种有效的手段。

Aimed at nonstationary characteristic of the combustion information of engine this paper presented a sound intensity computing method based wavelet packets analysis. The method took advantage of near field sound intensity measurement and multiresolution analysis of wavelet packets. Using the method the shock characteristic of an engine combustion is extracted from complicated background noise by experiments. The extracted combustion information can be used as fault diagnosis index of single cylinder misfire. The research of the paper provides a tool for online monitoring and fault diagnosis of engine.