【摘要】 在分析传统定步长LMS(Least Mean Square)算法和变步长LMS算法的基础上,提出了一种改进的变步长LMS算法。新算法利用瞬时误差绝对值三次方的指数形式和遗忘因子同时调整步长,更好地解决了收敛速度和稳态误差的矛盾。将三种算法均用到噪声对消中进行比较,仿真结果表明:新算法收敛速率优于传统定步长LMS算法和变步长LMS算法。更多还原

【Abstract】 Based on the analysis of traditional LMS algorithms with fixed step size and variable step size,this paper presents an improved variable step size LMS algorithm.By utilizing the exponential form of the cube of instantaneous error’s absolute value and the forgetting factor simultaneously,the new algorithm adjusts the step size so that it better solves the contradictions of convergence speed and steady error.Applying three algorithms to noise cancellation and comparing them,the simulation results show that,the new algorithm performs better than conventional fixed step size LMS algorithm and variable step size LMS algorithm in convergence rate.更多还原