配电变压器运行状态评估的大数据分析方法  

Big Data Analytical Method for Operating State Assessment of Distribution Transformer

作　　者：张友强[1] 寇凌峰[2] 盛万兴[2] 王金丽[2] 梁英[2] 宋祺鹏[2] 

机构地区：[1]国网重庆市电力公司电力科学研究院,重庆市渝北区401123 [2]配电变压器节能技术北京市重点实验室(中国电力科学研究院),北京市海淀区100192

出　　处：《电网技术》2016年第3期768-773,共6页Power System Technology

基　　金：国家电网公司科技项目(PDB17201400055)~~

摘　　要：为提高配电变压器的故障预测水平和使用寿命,提出了配电变压器运行状态的大数据分析评估方法,构建了常规评估方法与大数据方法结合的业务流程架构。结合配电网的多源数据,分析了配电变压器运行状态相关的数据类型、来源及预处理方法,通过数据流处理方法量化各属性的相关关系;基于信息熵理论构建了配电变压器运行状态评估模型,同时将大数据在线评估和常规综合评估方法进行有机结合,实现概率预测与特定指标评价的统一。经实例验证,结果可以有效反映配电变压器的整体运行状态。In order to raise fault prediction level and working life of distribution transformers, a big data analysis method for distribution transformeroperating state assessment is presented, a business process architecture which combined regular assessment method and big data method is constructed. Combined with multi-source data of distribution network, analyze the data type, source and pretreatment methods for distribution transformers operating state, then quantify the properties correlation between each factor through data stream processing method. The assessment model of distribution transformer operating statebased on information entropy theory is developed, it combines the big data online assessment and regular comprehensive assessment methods, achieves the unity of probability predict and specific indicator assessment. The instanceshows that the result can effectively reflect the overall operating state of distribution transformers.

关 键 词：大数据 信息熵 在线评估 综合评估 

分 类 号：TM73[电气工程—电力系统及自动化]