一种新的变步长自适应噪声消除算法

摘要

本文针对电力线噪声的特点,提出了一种新的变步长自适应噪声消除算法.在自适应算法的步长与梯度之间建立了新的关系,弥补了基于误差的变步长算法在自适应噪声消除方面的不足,克服了标准LMS算法的收敛性对输入信号的敏感性,并能根据梯度调整步长大小从而实现算法的快速收敛.通过理论分析设计了新的变步长自适应噪声消除算法,并进行了仿真和实测数据验证,证明了算法相对于其他算法的优势.%Considering the features of carrier power line noise,a novel variable step-size adaptive interference cancellation algorithm is proposed in this paper.A new function relation formula between the varying step-size of adaptive algorithm and gradient is established.this algorithm not only make up for the deficiency of the variable step size algorithm based on the error in the adaptive interference cancellation,but also overcome the limitation of the standard LMS algorithm that its convergence is sensitive to the power of input signal.And the algorithm can adjust step size along with the gradient in order to converge faster.Also,theoretical analysis about the novel variable step-size adaptive interference cancellation algorithm is conducted in this paper,and the performance of the algorithm is verified by simulation and practical results.Simulation and practical results indicate the correctness of theoretical analysis and demonstrates the advantages of this algorithm.