Abstract: The classical Filter-x least mean square (FxLMS) algorithm has been widely used in active noise control, but its performance would degrade dramatically in the presence of impulsive noise. In this paper, a new adaptive algorithm with nonlinear devices based on the concept of robust statistics has been presented, whose objective function is the M-estimate function instead of the mean square error(MSE). In this paper, four different M-estimate functions such as Huber function and Hampel’s three parts redescending M-estimate function are used as objective function and computer simulations are carried out to verify the efficiency of the presented algorithm for active impulsive noise control. Compared with the adding windows algorithm modified by Japanese scholar Akhtar, the simulation results also show that the performance of the presented algorithm has a better convergence.

摘要： 经典的滤波―X最小均方算法(Fx LMS)已经被广泛应用于有源噪声控制(ANC)领域。 但是当存在脉冲噪声时, 它的性能就会严重退化。基于鲁棒统计的概念介绍了一种新型自适应算法，采用的目标函数为M－估计函数, 而不是传统的最小均方误差。该算法分别采用了Huber 函数、Hampel 三段下降M估计函数等四种不同的M－估计函数作为目标函数，仿真结果表明所采用的算法能有效地消除脉冲噪声，并且与日本学者Akhtar 改进的加窗算法相比表现了更好的收敛性。

关键词: 声学, 有源噪声控制, 脉冲噪声, M―估计