基于随机矩阵理论的状态监测数据关联分析方法  

Correlation Analysis Method of Condition Monitoring Data Based on Random Matrix Theory

作　　者：王劭菁 任茂鑫 曹培 季怡萍 盛戈皞[2] 罗林根[2] WANG Shaojing;REN Maoxin;CAO Pei;JI Yiping;SHENG Gehao;LUO Lingen(Electric Power Research Institute,State Grid Shanghai Municipal Electrical Power Company,Shanghai 200437,China;School of Electric Information and Electrical Engineering,Shanghai Jiao Tong University,Shanghai 200240,China)

机构地区：[1]国网上海市电力公司电力科学研究院,上海200437 [2]上海交通大学电子信息与电气工程学院,上海200240

出　　处：《高压电器》2023年第11期155-161,171,共8页High Voltage Apparatus

基　　金：国家自然科学基金资助项目(51777122)。

摘　　要：随着大数据技术在电力系统中应用研究的广泛开展,特别是针对设备多参量监测数据的关联分析,亟需新的数学工具对所获取的数据进行高效的数据表征及信息提取。文中采用高维随机矩阵对状态监测数据进行数学表征,并运用随机矩阵理论的特征根谱分析、MP定律、圆环率等理论,对设备状态监测数据进行数据挖掘。在理论方法基础上,文中以输电线路状态监测数据为研究对象,将输电线路电压及其谐波通过高维随机矩阵统一建模,并对其关联信息进行分析和提取。研究结果表明高维随机矩阵表征下的数据分析技术具备对设备状态信息的高度敏感性,适用于对多源数据关联关系的分析,且在设备状态分析应用中具有传统方法所缺乏的直观性与高效性。With the extensive development of the application of big data technology in power system,especially for the correlation analysis of equipment multi-parameter monitoring data,new mathematical tools are urgently needed for efficient data characterization and information extraction of acquired data.In this paper,high dimensional random matrix is adopted for mathematical characterization of the condition monitoring data.Moreover,the such theory as eigenvalue analysis of the random matrix theory and the MP law as well as the single ring law are used for data mining of the equipment’s condition monitoring data.On the basis of theoretical method,the condition monitoring data of the transmission line is taken as the research object,the unified modeling of the voltage of transmission line and its harmonics are set up through the high dimensional random matrix.Its correlated information is analyzed and extracted.The research results show that the data analysis technology under the high dimensional random matrix possesses the highly sensitivity to the condition information of equipment,is applicable for the analysis of multi-source data correlation and,in the condition analysis and application of equipment,has intuition and efficiency which is lacked by the conventional method.

关 键 词：状态监测 数据挖掘 关联分析 随机矩阵理论 大数据 

分 类 号：TM73[电气工程—电力系统及自动化]