基于相量测量单元实测数据的变压器参数在线估计方法

摘要

给出了基于相量测量单元(PMU)实测数据的变压器变比(分接头位置)和电抗参数的计算方法.通过对变压器参数和变比的多时段计算结果进行统计分析发现,多时段参数计算结果符合正态分布.利用正态分布参数估计理论,给出了变压器参数和变比辨识结果的置信区间和点估计值,最大限度地减少了量测随机误差的影响.算法考虑了电力系统的生产实际情况,在三绕组变压器低压侧无量测的情况下,能够准确给出变压器分接头实际位置的计算结果.算例表明,该方法准确可靠,能够应用于实践.%Based on phasor measurement unit (PMU) measurements, an online algorithm is presented to calculate power transformer's winding reactances and transformation ratio. It is found that due to the random error of PMU measurements, the multi-period parameter calculation results comply with a normal distribution. Furthermore, based on the statistics theory of normal distribution, a robust parameter identification algorithm is proposed to give confidence interval and point estimation of transformer's parameters and transformation ratio. Considering that low voltage side of three-winding transformer is usually connected with reactive power compensation devices, the algorithm can precisely identify the tap changer position without low voltage side PMU measurements. Examples from a wide area measurement system (WAMS) in North China power grid show that the methods proposed is feasible and reliable.