中文摘要：

提出基于Hadoop和HBase的电力设备状态高速采样数据的存储方案，基于MapReduce设计实现了设备状态高速采样数据的并行查询方法。创建了1个包含20个节点（每个节点配置4核CPU）Hadoop集群，并对集群进行了基准测试，测试结果表明所建集群适合进行大量数据的读写。以绝缘子泄漏电流数据为例，使用YCSB对所建存储系统进行了性能测试，测试结果表明，Hadoop和Hbase在存储容量、吞吐量以及查询延迟上提供了足够高的性能，能够满足智能电网状态监测数据可靠性及实时性要求。

英文摘要：

A storage scheme based on Hadoop and HBase is proposed for the power equipment state data sampled with high speed. An approach is designed and implemented based on MapReduce to query in parallel the created and built for the and Hbase requirements sampled data. A Hadoop cluster containing 20 nodes,each equipped with a 4-core CPU,is benchmarked. Results show it is suitable for massive data reading/writing. A storage system is insulator leakage current data as an example and tested by YCSB,results indicate that,Hadoop pro,ride sufficiently high performance in storage capacity,throughput and latency,meeting the of smart grid for the reliability and real-time performance of state monitoring data.