VS-MFxLMS算法及其在汽车车内噪声有源控制中的应用  

Application of Variable Step-size Modified FxLMS Algorithm to Active Interior Noise Control for Vehicles

作　　者：张帅[1] 王岩松[1] 张心光[1] 李文武 ZHANG Shuai;WANG Yansong;ZHANG Xinguang;LI Wenwu(Automotive Engineering College, Shanghai University of Engineering Science,Shanghai 201620, China)

机构地区：[1]上海工程技术大学汽车工程学院,上海201620

出　　处：《噪声与振动控制》2019年第2期64-69,共6页Noise and Vibration Control

基　　金：国家自然科学基金资助项目(51675324)

摘　　要：为提高经典VS-FxLMS算法的收敛性能以及规避MFxLMS算法不能同时兼顾收敛速度和稳态误差的缺陷,结合修正反正切函数和归一化的方法,提出了一种可用于汽车车内噪声有源控制的VS-MFxLMS算法。应用MFxLMS算法、VS-FxLMS算法和VS-MFxLMS算法分别进行汽车车内噪声有源控制仿真实验。噪声有源控制结果的比较表明,与MFxLMS算法相比,VS-MFxLMS算法的收敛速度提高1.5倍以上,稳态误差降低55%以上;与VSFxLMS算法相比,VS-MFxLMS算法的收敛速度提高25%以上,稳态误差降低28%以上,这为汽车车内噪声的有源控制提供一种新方法。Combining the arc tangent function and normalization method, a variable-step Modified FxLMS algorithm for active control of vehicle' s interior noise is proposed. This algorithm solves the defect that the MFxLMS algorithm is unlikely to improve the convergence speed and reduce the steady-state error simultaneously. And the convergence performance is improved compared with the classic variable-step FxLMS algorithm. The MFxLMS algorithm, variable-step FxLMS algorithm and variable-step MFxLMS algorithm are used respectively to simulate the active control of the vehicle interior noise. The comparison of the results of active noise control shows that the steady-state error of the variable-step MFxLMS algorithm is reduced by more than 55 % compared with the fixed-step MFxLMS algorithm and by more than 28 % compared with the variable-step FxLMS algorithm, and the convergence speed of the variable-step MFxLMS algorithm is increased by more than 1.5 times compared with the fixed-step MFxLMS algorithm and by more than 25 % compared with the variable-step FxLMS algorithm. This work provides a new method for active control of vehicle interior noise.

关 键 词：声学 车内噪声 有源噪声控制 变步长 FxLMS MFxLMS 

分 类 号：U467.4[机械工程—车辆工程]