基于高维随机矩阵大数据分析模型的输变电设备关键性能评估方法

The Key State Assessment Method of Power Transmission Equipment Using Big Data Analyzing Model Based on Large Dimensional Random Matrix

输变电设备状态参数众多,其变化与电网运行、气象环境等因素密切相关,亟需采用有效的大数据技术对大量的相关数据进行挖掘分析和信息提取,提高输变电设备状态评估的及时性和准确性。文中首先在大量故障和缺陷样本的基础上,通过关联规则的置信度挖掘出设备关键性能与状态量的对应关系,然后通过高维随机矩阵理论对设备状态量的时间序列进行大数据表征,研究了含有时间序列模型的高维矩阵的特征根谱分布与圆环率,分析状态量的历史和当前状态信息,实现设备关键性能的评估和异常检测。以500 k V变电站为例,对变压器的负荷、在线监测和环境气象数据融合构成关键性能的高维矩阵,采用高维随机矩阵理论对历史、当前各时段矩阵的谱性质进行比对,以实现变压器关键性能评估和异常检测。研究结果表明,高维随机矩阵理论对分析设备的运行状态是有效的,为大数据技术在电力设备状态评估中的应用提供了一种新的思路。

Parameters of power equipment are highly related to the operation and environmental situation so that big data processing technology is needed for promotion of the efficiency and accuracy of condition assessment. Based on abundant fault samples, the confidence rate of association rules was used to extract the correlation between key states and parameters. Then, the eigenvalue spectrum distribution and ring law were derived of high dimension matrices containing time series model. The circle ring of eigenvalues of state matrices of different time domains were constructed and compared in order to analyze the trend of each state and detect the abnormal key state. The data of load, online monitoring and environment of one 500 k V substation were used for case study and big data model of each key state was built. The evaluation of key state and anomaly detection were accomplished through the comparison of spectrum and circle ring. The result indicates the random matrices theory is effective, bringing a novel idea for condition assessment of power equipment.