基于盲源分离算法的柴油机噪声激励源诊断  

Diesel Engine Noise Excitation Source Diagnosis Based on Blind Source Separation Algorithm

作　　者：胡溧[1] 贺献博 谭征宇 杨啟梁[1] 王佳 陈雷磊 HU Li;HE Xianbo;TAN Zhengyu;YANG Qiliang;WANG Jia;CHEN Leilei(Wuhan University of Science and Technology,Wuhan 430081,China;Dongfeng Commercial Vehicle Co.,Ltd.,Wuhan 430056,China)

机构地区：[1]武汉科技大学,武汉430081 [2]东风商用车有限公司,武汉430056

出　　处：《内燃机工程》2023年第1期27-33,44,共8页Chinese Internal Combustion Engine Engineering

基　　金：国家自然科学基金项目(51105283)。

摘　　要：基于负熵极大化算法的盲源分离技术,结合小波变换谱分析技术提出了自适应系数负熵极大化盲源分离算法及连续小波变换(adaption negentropy maximum–continuous wavelet transform,ANM–CWT)综合分析技术。并利用ANM–CWT分析技术,成功识别出了怠速工况下发动机由燃烧产生的燃烧噪声激励源、由机械运动产生的机械噪声激励源及由活塞敲击气缸壁而产生的异响噪声激励源。Based on the blind source separation technology of negative entropy maximization algorithm and wavelet transform spectrum analysis technology,the adaption negentropy maximum(ANM)–continuous wavelet transform(CWT)was proposed.With the ANM–CWT analysis technology,the combustion noise excitation source generated by combustion,the mechanical noise excitation source generated by mechanical motion and the abnormal noise excitation source generated by piston knocking on the cylinder wall were successfully identified.

关 键 词：负熵 盲源分离 振源诊断 小波分析 

分 类 号：TK42[动力工程及工程热物理—动力机械及工程]