基于声强试验与EMD-ICA模型的柴油机噪声源识别研究  

Source Identification of Diesel Engine Noise Based on Sound Intensity Experiment and EMD-ICA Model

作　　者：杜宪峰[1,2,3] 范文强 孙福强 张磊[3] 

机构地区：[1]天津大学内燃机燃烧学国家重点实验室,天津300072 [2]东风朝阳朝柴动力有限公司 [3]辽宁工业大学省汽车振动与噪声工程技术研究中心

出　　处：《小型内燃机与车辆技术》2015年第5期79-84,共6页Small Internal Combustion Engine and Vehicle Technique

基　　金：辽宁省博士启动基金项目(20141200)

摘　　要：噪声源识别是实现柴油机噪声控制的关键因素,合理有效的试验测试与时频分析技术是实现噪声源分离识别的重要手段。基于经验模态分解(EMD)与独立分量分析(ICA)算法建立EMD-ICA分离识别模型,并结合声强测量法与EMD-ICA模型对噪声源进行分离及特征提取研究,即充分利用了声强测量法准确、辨识精度高的优势,也实现了分析模型强大的数据处理能力。同时,采用相关性与相干函数方法有效探讨了EMD分解分量与ICA分离分量信号特征的一致性。研究结果表明,EMD-ICA模型能够实现单一通道测量信号的信号源分离识别,声强测量法也验证了该模型在实际应用中的有效性和可行性,对于柴油机噪声源控制具有一定的指导价值。Noise source identification is one key factor in achieving the diesel noise control, reasonable experiment and time-frequency analysis technology are important methods to achieve separation and identification of diesel noise. This paper established EMD-ICA separation model based on empirical mode decomposition （EMD） and independent component analysis （ICA） algorithm, and then studied the noise source separation and feature extraction based on sound intensity measurement and EMD-ICA model. The task takes full advantage of sound intensity measurement accuracy and high identification, but also realizes the powerful data processing capabilities of model. Meanwhile, the paper discussed the signal feature consistency of EMD decomposition and ICA component in use of the relevance and coherence function method. The results show that, EMD-ICA model can achieve source separation for single channel signal, and the sound intensity measurement method also proved the practical effectiveness and feasibility of this model, and has a guiding value for diesel noise source control.

关 键 词：EMD-ICA模型 声强测量法 特征提取 噪声源 柴油机 

分 类 号：TK421.6[动力工程及工程热物理—动力机械及工程]