基于随机森林算法的电力计量大数据分析平台研究  

Research on Big Data Analysis Platform for Electric Power Measurement Based on Random Forest Algorithm

作　　者：文耀宽 王献军 王峻 苏沛 WEN Yao-kuan;WANG Xian-jun;WANG Jun;SU Pei(Electric Power Research Institute,State Grid Henan Electric Power Company,Zhengzhou 450000,China)

机构地区：[1]国家电网河南省电力公司电力科学研究院,河南郑州450000

出　　处：《计算机技术与发展》2021年第6期216-220,共5页Computer Technology and Development

基　　金：国家电网有限公司科技项目(521304170028)。

摘　　要：针对电力计量大数据复杂、用户检索困难的问题,该研究利用随机森林算法模型,构建出基于云计算的智能电网大数据处理平台SP-DPP(smart power system big data processing platform in cloud environment),并且通过物联网通讯方式实现底层设备到用户层的数据信息传递。SP-DPP软件平台具有较好的吞吐量与加速比,可在较短的时间内接收大量的数据,提高了数据的容纳能力。该研究通过随机森林算法实现了大数据的训练、学习,将接收的电力数据按照用户设定的属性进行训练、学习、检索,提高了电能计量装置数据检索的准确度。试验结果表明,该技术方案提高了数据处理能力。Aiming at the problems of complex big data in power measurement and difficult user retrieval, we build SP-DPP(smart power system big data processing platform in cloud environment) based on cloud computing with a random forest algorithm model, and realize the transfer of data information from the underlying device to the user layer through the internet of things communication. The SP-DPP,which has a better throughput and acceleration ratio, can receive a large amount of data in a short time, so as to improve the data capacity. In this study, training and learning of big data is realized by random forest algorithm. The received power data is trained, learned, and retrieved according to the attributes set by the user, which improves the accuracy of data retrieval of the energy metering device. The test shows that the proposed scheme improves the data processing ability.

关 键 词：云计算 电能计量装置 SP-DPP软件平台 物联网 随机森林算法模型 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]