基于支持向量回归改进的工况传递路径分析方法及其应用  

A support vector regression based operational transfer path analysis method and its application

作　　者：胡涵 高亮 孟言 郑辉[1] HU Han;GAO Liang;MENG Yan;ZHENG Hui(Institute of Vibration,Shock&Noise,Shanghai Jiao Tong University,Shanghai 200240,China;XCMG Construction Machinery Co.Ltd,XCMG,Xuzhou 221004,China)

机构地区：[1]上海交通大学振动冲击噪声研究所,上海200240 [2]徐工集团徐工机械道路机械分公司,江苏徐州221004

出　　处：《振动与冲击》2021年第20期285-295,共11页Journal of Vibration and Shock

基　　金：国家自然科学基金(51705341,51905357,51675353);国家重点研发计划(2017YFC0703903)。

摘　　要：提出了一种结合支持向量回归理论改进的工况传递路径分析方法。利用改进方法良好的回归特性和自适应性及对随机噪声的量化,能够有效克服现有分析方法由于信号中混有随机噪声和串扰而造成对路径贡献的错误预测。运用“源-传递路径-接收者”数值算例和相应实验验证了所提出方法的有效性和可靠性。最后将改进的工况传递路径方法应用于某型振动压路机驾驶室噪声分析,成功辨识出空气声和结构声主导传递路径,为驾驶室的降噪工作提供了理论依据。In this paper,an improved operational transfer path analysis(OTPA)method based on Support Vector Regression(SVR)theory was proposed.Due to its promising regression property and adaptability and besides quantifying the random noise in the input signals,the proposed method outperforms existing OTPA methods,which may conduct false predictions on the path contributions because of random noise and crosstalk effect.The effectiveness and stability of the proposed method were validated through an“emitter-transfer path-receiver”model,both numerically and experimentally.Finally,the proposed method was applied on a cabin noise analysis of a vibratory roller.The dominant air-borne and structure-borne transfer paths were identified successfully,thus proving a theoretical support for the noise reduction project of the cabin.

关 键 词：传递路径 支持向量回归 不确定性 路径贡献 振动压路机 

分 类 号：TB5[理学—物理]