基于最小二乘法对时变参数测量的新方法

摘要

针对现有的递推最小二乘算法对突变信号的跟踪能力不够,提出一种新的基于可变遗忘因子递推最小二乘算法.它对电力系统时变信号的幅值、相位和频率进行测量时,调节好引入的可变遗忘因子λ初始值可使算法的收敛速度和跟踪能力得到改善.仿真结果表明:与现有的递推最小二乘法相比,本算法收敛速度更快,参数估算精度更高,在一定低信噪比环境中抗噪性更好.%In view of the tracking performance defect in measuring the time-varying parameters by using the recursive least squares (RLS) algorithm.A new RLS method based on alterable forgetting factor was put forward.It can be improved on the convergence rate and the trace ability of algorithm by regulating the value of forgetting factor when measuring the time-varying parameters of power system.The simulation results show that the proposed method is faster in convergence rate and higher in parameter estimation accuracy as well as the noise resistance is better in a certainly low SNR environment.