基于CEP引擎的配电网运行监控信号大数据ETL方法

摘要

The scale of power system becomes larger and larger, and the number of electrical equipment in distribu-tion network increases sharply and becomes further precise. Massive and random operation monitoring and control-ling data cause various applications in active distribution network unable to extract useful information quickly and efficiently so as difficult to form decision support. The article uses CEP engine as the operational monitoring and controlling signal processing core, defines and perfects rules library with Apriori machine learning algorithm, and does standardized treatment to signal data stream with core algorithms library. On the whole, ETL ( Extract-Trans-form-Load) framework is used to integrate, clean and load the distributed and disordered signal data in active dis-tribution network into the data warehouse, and output data to different media by different data interface to satisfy different applications. The CEP engine based Big Data ETL solution can implement the fast, accurate and effective standardization processing, and multi-source data integration and output of operation monitoring and controlling sig-nal, and can provide accurate data preparation for fast simulation, fault analysis, state estimation and other impor-tant application in active distribution network.%电力系统规模在不断扩大,配电网电气设备数量急剧增长且趋于精细化.大量且散乱的运行监控数据使得主动配电网各应用无法快速有效地提取有用信息以形成决策支持.本文利用复杂事件处理(CEP)引擎作为运行监控信号处理核心,通过Apriori机器学习算法定义和完善规则库,通过核心算法库对信号数据流进行规范化处理.整体采用ETL(Extract-Transform-Load)框架,将主动配电网中分散、零乱、标准不统一的信号数据整合、清洗后加载到数据仓库,并以多种数据接口输出至不同介质,供不同应用调用.基于CEP引擎的大数据ETL方法,可对运行监控信号进行快速、精确、有效的规范化处理,实现多数据源集成与输出,为主动配电网设备仿真、故障分析、状态估计等多种重要应用提供数据准备.