中文摘要：

以智能电网中电力变压器故障诊断为例,给出了基于MapReduce的电力变压器并行故障诊断过程,其应用4个MapReduce过程执行故障诊断算法的训练阶段,并得出分类模型,应用1个MapReduce过程完成对电力设备状态信息数据的故障诊断。建立了电力设备状态信息并行故障诊断实验平台,基于海量变压器油中溶解气体分析数据进行并行故障诊断实验,实验结果表明并行故障诊断速度高于传统单机环境下的诊断速度,满足智能电网环境下对海量电力设备状态信息快速故障诊断的要求。

英文摘要：

As an example,a parallel fault diagnosis algorithm based on MapReduce is given for diagnosing the power transformer faults in smart grid,which includes four MapReduce procedures in its training stage to obtain the classification model and one MapReduce procedure to complete the fault diagnosis based on the status data of electric power equipment. An experimental platform is built and a parallel fault diagnosis experiment based on the massive DGA（Dissolved Gas Analysis） data of power transformer is carried out.Results show that,the diagnosis speed of parallel fault diagnosis is faster than that of mono-computer environment,meeting the requirement of rapid fault diagnosis based on the massive status data of electric power equipments in smart grid.