数据驱动的复杂电网故障区域定位方法研究

Research on Fault Area Location Method of Complex Power Grid Based on Data-driven

针对电网结构复杂,采集数据量过大且难以准确定位故障等难题,本文应用数据驱动技术,提出一种基于随机矩阵理论的复杂电网故障区域定位方法。利用复杂电网故障数据建立增广矩阵,并通过单环定理对故障数据进行分析,建立平均谱半径差积分指标,得出各个区域内的数据分析结果。所提方法是大数据分析方法,不需要物理模型和拓扑信息,准确性高。此外,还可以减少数据量,提高确定电压源潜在影响范围的分析效率,速度快。最后,通过实例表明,该方法能够准确定位复杂电网故障区域。

Aiming at the problems of complex power grid structure such as large amount of collected data and difficult to accurately locate faults,this paper applies data-driven technology to propose a complex power grid fault area location method based on random matrix theory.The augmented matrix is established by using the complex power grid fault data,the fault data is analyzed by the single loop theorem,the average spectral radius difference integral index is established,and the data analysis results in each region are obtained.The proposed method is a big data analysis method,does not need physical model and topology information,and has high accuracy.Finally,this method can accurately locate the fault area of complex power grid was verified by simulation.