汽车车内噪声主动控制归一化LMS算法

Active Noise Control for Vehicle Interior Noise Using Normalized LMS Algorithm

通过对汽车车内噪声采集试验数据进行分析,运用最小均方(LMS)算法和归一化LMS算法,分别对自适应滤波器中的权向量按照最速下降算法进行更新,并利用建立的自适应滤波器进行汽车车内噪声主动控制。通过比较汽车车内噪声主动控制结果,表明:与LMS算法相比,归一化LMS算法具有更快的算法收敛速度和较小的稳态误差,其收敛速度提高58.3%,其稳态误差降低62.5%。

By analyzing the test data of vehicle interior noise, LMS algorithm and normalized LMS algorithm are respectively applied to update the weight vectors in the adaptive filter based on the steepest descent algorithm. Comparison of active noise control results obtained by using the established adaptive filter shows that the normalized LMS algorithm has a faster convergence rate and a smaller steady state error, i.e., 58. 3 percentage point faster and 62.5 percentage point lower than LMS algorithm.