Study on the sound quality of unsteady exhaust noise based on RNR-WVD calculation and GA-wavelet network
ZENG Falin1,SUN Sumin2
1.Automobile Engineering Research Institute, Jiangsu University, Zhenjiang 212013, China;
2.School of Automobile and Traffic Engineering, Jiangsu University, Zhenjiang 212013, China
摘要基于心里声学客观参量的GA-BP声品质预测模型能够准确的预测稳态排气噪声声品质。对于非稳态噪声研究，引入正则化非稳态回归技术（RNR）优化计算维格纳-威尔分布(WVD)的时频方法，建立新的声品质参量SQP-RW(Sound quality parameter base on RNR-WVD)，用此参量替换掉与满意度相关性较小的客观参量。同时，以Morlet小波基函数作为隐含层结点的传递函数构建小波神经网络(Wavelet Neural Network,WNN)，并用GA优化小波神经网络层间的权值和阈值，构造出GA-WNN并用于非稳态排气噪声声品质预测。结果表明：GA-WNN在非稳态排气噪声声品质预测上比GA-BP神经网络更加准确；引入SQP-RW参量，模型具有更高的精度，更能体现出非稳态信号特征及声品质特点。
Abstract：The GA-BP acoustic quality prediction model based on the psychoacoustical objective parameters can accurately predict the sound quality of the steady state exhaust noise.For the research on unsteady state noise, RNR (regularization non-stationary regression technique) was applied to compute the WVD distribution (RNR-WVD), and a new parameter SQP-RW was built to replace the objective parameters that have less correlation with the satisfaction degree.Meanwhile, a WNN (Wavelet Neural Network, WNN) was constructed by using the Morlet wave basis function which was used as the transfer function of the hidden layer.And GA was used to optimize the weights and thresholds of the WNN layers, the GA-WNN was constructed and used to predict the acoustic quality of the unsteady exhaust noise.The results show that the GA-WNN is more accurate in predicting the sound quality of unsteady exhaust noise than the GA-BP neural network.With the introduction of SQP-RW, the model has higher accuracy and it can reflect the characteristics of unsteady signals and the sound quality characteristics.
曾发林1，孙苏民2. 基于RNR-WVD与GA-小波的非稳态排气噪声声品质研究[J]. 振动与冲击, 2019, 38(12): 74-80.
ZENG Falin1,SUN Sumin2. Study on the sound quality of unsteady exhaust noise based on RNR-WVD calculation and GA-wavelet network. JOURNAL OF VIBRATION AND SHOCK, 2019, 38(12): 74-80.
[1] Liu Hai, Zhang Junhong. Sound quality prediction for engine-radiated noise[J]. Mechanical Systems and Signal Processing,2015, 56/57:277-287.
[2] 申秀敏,左曙光,李林,张世炜.车内噪声声品质的支持向量机预测[J].振动与冲击,2010,29(06):66-68+236.
Shen X M, Zuo S G, Lin L I, et al. Interiror sound quality forecast for veheicles based on support vector machine[J]. Journal of Vibration & Shock, ,29(06):66-68+236.
[3] 毕凤荣, 李琳, 张剑, 马腾. 基于EEMD-HT与LSSVM的柴油机辐射噪声品质预测技术[J]. 天津大学学报(自然科学与工程技术版), 2017,50(01):28-34.
Bi F, Li L, Zhang J, et al. Sound quality prediction for diesel engine radiated noise based on EEMD-HT and LSSVM[J]. Tianjin Daxue Xuebao, 2017:28-34.
[4] Sang-Kwon Lee, Ho-Wuk Kim,Eun-Woo Na. Improvement of impact noise in a passenger car utilizing sound metric based on wavelet transform[J]. Journal of Sound and Vibration, 2010,329(17).
[5] .刘程, 贺岩松, 于海兴. 汽车制动工况下时变噪声响度特征[J].汽车工程学报, 2013,（01）:40-46.
Liu C, He Y, Yu H, et al. Loudness Characteristics for Vehicle Interior Time-Varying Noise Under Braking Condition[J]. Chinese Journal of Automotive Engineering, 2013.
[6] Lee S K, Objective evaluation of interior sound quality in passenger cars during acceleration. Journal of Sound and Vibration, 2008,310(1):149-168.
[7] 陈剑, 杨雯, 李伟毅. 汽车声品质主观评价试验方法探究[J]. 汽车工程, 2009,31(04):389-392+339.
Chen J, Yang W A. An Investigation into the Test Methods for Subjective Evaluation of Vehicle Sound Quality[J]. Automotive Engineering, 2009, 31(4):389-388.
[8] Amman S, Otto N. Sound quality analysis of vehicle windshield wiper systems.Proceedings 1993 SAE NVH Conference, Traverse City, ML,Paper 931345,1993.
[9] Hussain M，Golles J，Ronacher A. Statistical evaluation of an annoyance index for engine noise recording[J]. SAE Transactions, 1991, 100(6):527-535.
[10] Wu X, Liu T. Spectral decomposition of seismic data with reassigned smoothed pseudo Wigner–Ville distribution[J]. Journal of Applied Geophysics, 2009, 68(3):386-393.
[11] 王宏禹, 邱天爽, 陈喆. 非平稳随机信号分析与处理[M]. 2版.北京：国防工业出版社, 2008:311-326.
Wang Hongyu, Qiu Tianshuang, Chen Zhe. Non stationary random signal analysis and processing [M]. Beijing: National Defense Industry Press, 2008(02):311-326.
[12] 汪璇, 曹万强. Hilbert变换及其基本性质分析[J]. 湖北大学学报(自然科学版), 2008(01):53-55.
Wang X, Cao W Q. The Hilbert transform and its characters[J]. Journal of Hubei University, 2008.
[13] 张志飞, 陈思, 徐中明, 贺岩松, 黎术.双面声阵列波束形成的正则化改进算法[J]. 声学学报, 2017,42(02):178-185.
Zhang Z, Chen S, Xu Z, et al. Iterative regularization method in generalized inverse beamforming[J]. Journal of Sound & Vibration, 2017, 396.
[14] Fomel S. Adaptive multiple subtraction using regularized nonstationary regression. Geophysics, 2009,74(1):V25-V33.
[15] Fomel S. Shaping regularition in geophysical-estimation problems[J]. Geophysics, 72(2):R29-R36.
[16] 吴胜利,邵毅敏,王利明,袁意林,叶维军.基于自适应小波的二十辊轧机轧辊局部缺陷识别研究[J].振动与冲击,2017,36(10):117-120.
Wu S L, Shao Y M, Wang L M, et al. identification of local defects on the roll surface of a twenty-high roll mill based on the adaptive wavelet[J]. Sock and Vibration, 2017,36(10):117-120.
