基于小波偏相干分析的车内噪声源识别

Vehicle Interior Noise Source Identification based on wavelet partial-coherence analysis

为了对复杂多相关声场下的汽车车内噪声源进行准确、快速地识别,提出了一种基于小波偏相干分析的噪声源识别方法。该方法通过连续复小波变换与时频偏相干分析获取声源测点与接收点噪声的时频偏相干函数与瞬时相位关系,并通过模拟信号加以说明;利用该方法对车内异常噪声进行声源识别,并基于分析结果改善车内噪声水平。结果表明:车内异常噪声源于发动机激励前副车架,从而导致副车架与车身连接点的结构振动噪声所致,进而对连接点橡胶衬套进行优化设计;改进后,通过道路试验主观评价并结合声品质分析得知车内噪声水平改善效果明显,进一步验证了所提出的噪声源识别方法的有效性。

To quickly and accurately identify complicated and multi-correlated vehicle interior noise sources,a noise source identification method based on wavelet partial-coherence analysis was proposed.The approach was performed through the continuous complex wavelet transformation and the time-frequency partial-coherence analysis to get time-frequency partial coherence functions of noise source’s measured point and receiver measured point,and relations between them and transient phase.This was illustrated with simulated signals.The proposed method was utilized to identify an abnormal noise source in a vehicle,and then to improve the sound quality of interior noise with identified results.The results showed that the interior abnormal noise source is induced due to the engine exciting the front sub-frame to cause vibration and noise of the connecting point between the sub-frame and car body;through optimizing the connecting point rubber bushes,the subjective and objective evaluations of vehicle interior sound quality are improved evidently,the effectiveness of the proposed noise source identification method is verfied.