【摘要】 基于Newton算法,本文导出了一种用于TLS估计的自学习算法,讨论了它在谐波检测中的应用。证明LMS算法和约束反Hebbian自学习算法(Constrained Anti-Hebbian Learning Algorithm)是本文提出的算法的两个特例。仿真结果表明,该算法的噪声剔出能力和收敛速度优于LMS算法和Constrained Anti-Hebbian自学习算法。更多还原

【Abstract】 Based on the Newton learning algorithm, a new algorithm for TLS estimation and its application in harmonics detection is presented in this paper. It is shown that the LMS algorithm and Constrained Anti-Hebbian algorithm are two examples of the proposed algorithm in this paper. The simulation result show that its noise rejection capability is supurior to those of the LMS and Constrained Anti-Hebbian algorithms and the new algorithm gives faster convergence and better precision than another two methods.更多还原