考虑多时段量测数据随机误差的变压器正序参数辨识与评估方法

变压器离线参数已难以满足日益增长的电网安全、稳定分析计算需求。而在线参数的准确获取依赖于对量测数据随机误差的有效处理和变压器参数辨识结果的合理评估。基于变压器多时段同步相量测量单元(phasormeasurement unit,PMU)量测数据,首先通过在线辨识和查表修正相结合的方式求取准确变比参数,同时根据功率匹配分析法和箱形图数据筛选策略对PMU数据进行质量分析和异常值排除。在此基础上采用最小二乘法对电抗参数进行辨识,最后依据矩阵范数的相容性定理推导出相对误差率因子对结果准确度进行定量评估、优选。对某区域电网潮流模拟和实际PMU数据的算例测试验证了文中算法的有效性和实用性。

Identification and Evaluation Method of Transformer Positive Sequence Parameters Considering Random Error of Multi-period Measurement Data

Off-line parameters of transformers are difficult to meet growing demand for grid safety and stability analysis. Accurate acquisition of online parameters depends on effective processing of random error of measurement data and reasonable evaluation of transformer parameter identification results. Based on the multi-period PMU measurement data of transformers, accurate ratio parameters are obtained through combination of on-line identification and look-up table correction. At the same time, the PMU data quality is analyzed and the outliers are excluded according to power matching analysis method and boxplot method. Based on this, the least square method is used to identify reactance parameters. Finally, based on the compatibility theorem of matrix norm, the relative error rate factor is derived to evaluate and optimize accuracy of the results quantitatively. A numerical case study of power flow simulation data and actual PMU data of a regional power grid verifies effectiveness and practicability of the proposed algorithm.