拉格朗日乘子法电力系统网络参数错误辨识研究

Power System Network Parameter Error Identification by Lagrange Multiplier Based Method

如何辨识出错误的网络参数是电力系统建模中的一个难题。拉格朗日乘子法是一种有效的网络参数错误辨识算法,然而由于计算效率和应用模式问题,这类方法尚难以实现实际应用。为此,基于现有的拉格朗日乘子网络参数错误辨识方法,给出了采用大权重法或修正牛顿法处理零注入约束时该方法的实现模式;针对传统方法的计算效率问题,提出了一种基于稀疏逆矩阵法的拉格朗日乘子协方差矩阵对角元的高效计算方法,以及直接基于状态估计结果,无需反复进行状态估计和参数估计计算的实现模式。该方法提升了拉格朗日乘子法的计算效率,为这类方法的实际应用提供了条件。

The identification of the network parameter errors is a puzzling problem in power system modeling.Lagrange multiplier method is an effective network parameter error identification approach.However,the practical application of this method is difficult due to its calculation burden and application procedure problems.Based on the existing researches on Lagrange multiplier method,the procedure using large weights or modified Newton methods to deal with zero injection constraints was presented.To improve calculation efficiency,an efficient algorithm to calculate the value of the diagonal elements of the covariance matrix of Lagrange multipliers based on sparse inverse matrix method was proposed;and the implementation method directly based on a single state estimation calculation was proposed,in which reiterative state estimation or parameter estimation is not necessary.This research improves the calculation efficiency of the Lagrange multiplier method and makes this method fitful for practical application.