一种变步长VLMP算法及其在发动机降噪中的应用  

A Variable Step Size Volterra LMP Algorithm and Its Application for Reducing Noise of Engine

作　　者：崔婷玉 张俊[1,3] 黄薛龙 CUI Ting-yu;ZHANG Jun;HUANG Xue-long(Wuhu Institute of Technology,Wuhu 241003,China;Anhui Huadong Polytechnic Institute Residents,Wuhu 241002,China;Wuhu Big Data and Artificial Intelligence Engineering Technology Research Center,Wuhu 241000,China)

机构地区：[1]芜湖职业技术学院,安徽芜湖241003 [2]安徽华东光电技术研究所,安徽芜湖241002 [3]芜湖市大数据与人工智能工程技术研究中心,安徽芜湖241000

出　　处：《长春师范大学学报》2022年第2期57-63,共7页Journal of Changchun Normal University

基　　金：2020年度安徽省高校自然科学研究重点项目“电梯安全监管大数据平台研究与应用”(KJ2020A0914)。

摘　　要：针对发动机振动引起的噪声控制问题,提出一种变步长Volterra滤波器最小平均p范数(VSS-VLMP)算法,依据Volterra算法建立了发动机噪声的非线性滤波器模型。采用sigmoid函数更新步长因子,对一阶和二阶权系数采取迭代更新,从而达到滤除发动机噪声中所含的高斯噪声和脉冲噪声的目的。通过与Volterra LMS滤波、Volterra LMP滤波以及wiener滤波的仿真实验结果对比分析可知,改进后算法对发动机振动引起的噪声的降噪效果要优于进行对比的其它滤波算法。According to the problems in Control of Noise due to Engine Vibration,A variable step size adaptive Volterra least mean th power(VSS-VLMP)filtering algorithm is proposed in this paper.In the proposed algorithm,the filter model of engine noise is established based on the Volterra algorithm.Firstly,the step factor updated by the sigmoid function,Additionally,the first-order and second-order weight coefficients were iteratively updated,so as to achieve the purpose of cancellation of gaussian noise and impulse noise contained in the engine noise.The proposed algorithm filtering is compared with Volterra LMS filtering,Volterra LMP filtering and wiener filtering cancellation of noise caused by engine.The simulation results show that the proposed algorithm outperforms the other filtering algorithm.

关 键 词：VOLTERRA滤波器 LMS算法 VLMP算法 发动机噪声降噪 

分 类 号：TN912[电子电信—通信与信息系统]