【摘要】 本文就参数呈快时变(包括阶跃变化)时系统辨识问题提出了两种在线自适应辨识算法,它们是通过参数时变检验对最小二乘递推算法进行适时校正得来的算法,并通过仿真与其它方法比较,验证了它们的有效性。更多还原

【Abstract】 This paper is about two simple adaptive identification algorithms for estimating time-varying parameters ( including abrupt and smooth changes). A parameter variation detector and two kinds of adjustment to the estimate covariance matrix P(k) are proposed to embed in the Recursive Least Squares algorithm(RLS). A number of simulations are carried out to prove their validity and superiority in comparison with some other methods.更多还原