基于UKF和IGG改进聚类算法的特高压交流输电线路参数辨识

Identification method of UHV AC transmission line parameters based on UKF and IGG improved clustering

为有效辨识特/超高压交流输电线路的实时运行参数,尽可能剔除相量测量单元(PMU)量测数据存在的随机噪声,提出一种改进算法。该算法首先采用无迹卡尔曼滤波(UKF)对原始量测数据进行初始滤波,然后通过分布参数等值电路模型辨识线路参数,最后利用三段式IGG(Institute of Geodesy&Geophysics)原理设置权函数的改进聚类方法对辨识的参数进行聚类处理,将多时间断面的参数聚类中心作为线路参数最终辨识结果。对1000k V特高压交流示范工程晋东南-南阳-荆门线路进行仿真分析,算例仿真结果表明了新型混合算法的准确性与有效性。

In order to identify on-line parameters of UHVAC/EHVAC transmission line effectively, as well as to re- move random measurement noises existed in PMU （phasor measurement unit） data, a new method is proposed which can be used to provide accurate basic data in power system state estimation, power flow calculation, fault lo- cation and protection setting and so on. In the presented method, unscented Kalman filters （UKF） is firstly used for original data. Then the transmission line resistance, reactance and capacitance are calculated with long-trans- mission-line distribution parameter model. Finally, the resistance, reactance and capacitance are classified by im- proved clustering program and the multi-time section clustering centers are taken as results of the transmission line parameters. In the improved clustering method, the IGG （ Institute of Geodesy ＆ Geophysics） principle is utilized to set weight value and the robust identification is realized. The effect and precision of the new method are verified by a numerical simulation on lO00kV UHV AC demonstration project - Southeast Shanxi, Nanyang, Jingmen trans- mission line.