【摘要】：近年來,對于有源噪聲控制算法的性能越來越重視。與基于維納濾波原理的最小均方濾波(Filtered-x Least Mean Square,Fx LMS)、最小二乘濾波(Filtered-x Recursive Least Square,Fx RLS)算法相比較,基于卡爾曼濾波的有源控制算法(Filtered-x Kalman,Fx Kalman)具有較快的收斂速度和良好的跟蹤性能,且對帶寬噪聲有較好的降噪性能。設(shè)計、仿真運行了Fx Kalman算法的有源控制器,并針對單頻、窄帶和寬帶信號,在實驗室封閉空間對Fx Kalman算法、Fx LMS算法和Fx RLS算法進行有源控制器驗證性實驗比較,證實了Fx Kalman有源控制器具有上述優(yōu)點。而如果初級噪聲為單頻信號且對算法收斂速度要求不高,Fx LMS算法是最經(jīng)濟穩(wěn)妥的選擇。當需要控制帶寬噪聲或?qū)λ惴ㄊ諗克俣纫筝^高時,Fx Kalman算法則為最好的選擇。
[Abstract]:In recent years, more and more attention has been paid to the performance of active noise control algorithms. Compared with the least mean square filter (Filtered-x Recursive Least Square,Fx RLS) algorithm based on Wiener filter principle, the active control algorithm based on Kalman filter (Filtered-x Kalman,) is compared with the least square filter (Filtered-x Recursive Least Square,Fx RLS) algorithm. Fx Kalman) has fast convergence speed, good tracking performance and good noise reduction performance for bandwidth noise. The active controller of Fx Kalman algorithm is designed and simulated. For single frequency, narrow band and wideband signals, the verification experiments of Fx Kalman algorithm, Fx LMS algorithm and Fx RLS algorithm are carried out in laboratory closed space. It is proved that the Fx Kalman active controller has the above advantages. However, if the primary noise is a single frequency signal and the convergence speed of the algorithm is not high, the, Fx LMS algorithm is the most economical and safe choice. , Fx Kalman algorithm is the best choice when the bandwidth noise control is needed or the convergence speed of the algorithm is high.
【作者單位】： 杭州應(yīng)用聲學(xué)研究所;
【分類號】：TB535
，

本文編號：2407986