基于分布式多层架构的计量生产管理平台设计与研究  

Design and Research of Measurement Scheduling Platform Based on Distributed Multi-layer Architecture

作　　者：欧振国 张永旺 刘海斌 舒晔 黄博伟 彭强 OU Zhenguo;ZHANG Yongwang;LIU Haibin;SHU Ye;HUANG Bowei;PENG Qiang(Measurement Center,Guangdong Power Grid Co.Ltd.,Guangzhou 510000,China)

机构地区：[1]广东电网有限责任公司,计量中心,广东广州510000

出　　处：《微型电脑应用》2022年第3期193-196,205,共5页Microcomputer Applications

摘　　要：针对常规技术中的计量生产管理平台不完善、数据处理滞后等问题,提出一种新型的计量生产管理平台。构建包括数据层、访问层、计算层和应用层的分布式多层架构,实现数据的分层处理。设计包括计量客户服务层、计量业务处理层、计量工作质量监控层和计量调度层的计量生产调度数据管理平台,实现不同计量业务的管理,满足用户的多种需要。还融入了大数据处理算法,应用KNN算法实现不同类型数据的分类,通过构建AdaBoost算法模型,实现电能计量故障数据的诊断。试验表明,提出的方法数据处理能力强,工作效率高,误差低。Aiming at the problems of imperfect measurement scheduling platform in conventional technology and lagging data processing, a new measurement scheduling platform is proposed. A platform with distributed multi-layer architecture, including data layer, access layer, computing layer and application layer, is constructed to achieve data layering. A measurement scheduling data management platform, including measurement customer service layer, measurement business processing layer, measurement work quality monitoring layer and measurement scheduling layer, which achieves the management of different measurement services to meet the needs of users. This research incorporates big data processing algorithms, uses the KNN algorithm to classify different types of data, and builds the AdaBoost algorithm model to diagnose electrical energy measurement fault data. The experiment shows that the method designed has strong data processing ability, high working efficiency and low error.

关 键 词：计量生产管理平台 分布式多层架构 KNN算法 ADABOOST算法 计量业务 

分 类 号：TP393[自动化与计算机技术—计算机应用技术] TP309[自动化与计算机技术—计算机科学与技术]