基于线性内点法及正交变换的抗差参数估计

WLAV Robust Parameter Estimation Based on Linear Interior Point Algorithm and Orthogonal Transformation

参数估计目前主要采用加权最小二乘算法。由于包含参数估计,该方法在处理电网坏数据混杂及数值稳定性上都面临着困难。考虑到实际电网中有时会同时出现量测错误及参数错误,提出了使用基于线性内点法及正交变换的加权最小绝对值(WLAV)增广参数估计。基于L1范数的WLAV估计具有良好的抗差特性,用以应对混杂有坏数据、拓扑错误及参数错误的混杂估计。为提高增广参数估计的数值稳定性,将正交变换引入线性内点法修正方程的求解中。对算法的测试结果表明,该方法具有良好的应用前景。

The predominant approach to solving the parameter estimation is the weighted least squares method.However,it＇s difficult to process gross errors in the measurement incorporated with the network parameter error and prevent numerical ill-conditioning using this method.Considering the fact that both bad measurement data and false network parameters may simultaneously appear in power system state estimation,a weighted least absolute value（WLAV） robust estimation based on the linear interior point algorithm and orthogonal transformation is proposed for augmented parameter estimation.The WLAV estimators based on the L1 norm have good properties of bad data rejection and are exploited for mixed estimation of bad data /topology error and parameter error.The orthogonal transformation is introduced into the computation process of the linear interior point algorithm for improving the robust behavior of the numerical problem of the augmented parameter estimation.The test results show that the method has good application prospects.