车内路噪主动控制系统参考传感器布放方法  

Methodology of Reference Sensor Placement for Active Control of Road Noise Inside Vehicle

作　　者：贺岩松[1] 刘亚琪 夏子恒 张志飞[1] 周桃 HE Yansong;LIU Yaqi;XIA Ziheng;ZHANG Zhifei;ZHOU Tao(College of Mechanical and Vehicle Engineering,Chongqing University Chongqing,400030,China;Chongqing Acoutec Techndology Co.,Ltd.Chongqing,400000,China)

机构地区：[1]重庆大学机械与运载工程学院,重庆400030 [2]重庆市埃库特科技责任有限公司,重庆400000

出　　处：《振动．测试与诊断》2023年第4期669-676,826,共9页Journal of Vibration,Measurement & Diagnosis

基　　金：国家自然科学基金资助项目(51275540)。

摘　　要：由于合理布放参考传感器是车内路噪主动控制(road noise cancellation,简称RNC)系统的关键,提出了基于多重相干分析(multiple coherence analysis,简称MCA)法的参考传感器布置方法,结合多通道归一化滤波x最小均方(normalized filter-x least mean square,简称NFxLMS)算法,建立基于MCA的参考信号选取数学模型,选取具有最大多重相干系数的布放位置,采用截断奇异值分解法提高多重相干系数的计算精度,并使用遗传算法进行多重相干法布放结果寻优。针对某电动汽车进行了车内路噪主动控制的仿真分析和实车道路试验。结果表明,所提出的参考信号布放方法有利于提升车内路噪主动控制系统的降噪效果。Due to the reasonable placement of reference sensors being crucial for road noise cancellation(RNC)systems,a reference sensor placement method is proposed based on multiple coherence analysis(MCA).The model of reference sensor placement is established according to MCA combined with multi-channel normalized filter-x least mean square(NFxLMS)algorithm,and the location with the maximum multiple coherence coefficient is selected.In order to improve the accuracy of multiple coherence calculation,truncated singular value decomposition is used,and genetic algorithm is employed to optimize the multiple coherence analysis results.The simulation and experiment of RNC are carried out on an electric vehicle,and the results demonstrate that the proposed reference sensor placement method can improve the noise reduction of RNC system.

关 键 词：路噪主动控制 参考信号 多重相干法 遗传算法 

分 类 号：TB535[理学—物理] TH771.5[理学—声学]