基于窗口滑动总体最小二乘法的输电线路参数辨识

输电线路参数的准确度与电力系统稳定运行密切相关,而广域测量系统的发展为利用PMU实测数据获取线路参数提供了新的途径.提出基于滑动窗口总体最小二乘的线路参数辨识算法,该方法对PMU数据采取窗口滑动处理,将目标函数变为窗口误差平方和最小,因而抗噪声能力较强;同时,对于大量辨识结果的可信度分别进行核密度估计和点估计,以获取参数辨识值的统计结果.仿真分析表明,该方法与单时刻辨识相比在含噪声环境下仍然能够保持较高的准确性与稳定性.某电网500kV线路实测PMU数据计算结果也验证了该方法的有效性和准确性.

Transmission line parameters identification based on moving-window TLS

The accuracy of transmission line parameters is related to power system normal operation.The development of wide area measurement system supplies a new way to obtain the real value of line parameters.This paper proposes the moving-window total least squares(TLS) algorithm for transmission line parameters from PMU data.With the moving-window technique and TLS model,the objective function of proposed method is changed into the least squares error by windows.Thus,the algorithm has stronger adaptability to noise.Meanwhile,Kernel density estimation and point estimation are used to identify the credibility of identification results.PSCAD based simulation results demonstrate that the presented algorithm can maintain accuracy and stability under noisy conditions.Lastly,real PMU data identification of a 500 kV transmission line verifies the validity and accuracy of moving-window TLS.