当LMS算法用于时变系统时,其跟踪性能不佳,常常不能满足应用要求。其主要原因是步长的选取影响收敛速度与稳态失调。而固定步长往往使二者不可兼得。本文提出一种具有快速跟踪能力和允许大输入动态范围的归一化变步长自适应滤波的新算法(NVS).计算机仿真计算表明,新算法的性能明显优于LMS算法,而计算量增加甚少。 When the LMS algorithm is used in time-varying systems, its tracking performance is poor and often can not meet the application requirements. The main reason is that the choice of the step size affects the convergence rate and steady-state imbalance. The fixed step often make both can not have both. In this paper, we propose a new algorithm (NVS) with fast tracking capability and normalized variable step adaptive filtering that allows large input dynamic range. Computer simulation shows that the performance of the new algorithm is obviously better than LMS algorithm, Very few.