多源信息融合的微服务化电网事故追忆  

Microservice-based post disturbance review with multi-source information fusion

作　　者：韦洪波 曹伟 叶桂南 韦昌福 何伊妮 WEI Hongbo;CAO Wei;YE Guinan;WEI Changfu;HE Yini(Power dispatch and Control Center of Guangxi Power Grid Co.,Ltd.,Nanning 530024Guangxi,China)

机构地区：[1]广西电网电力调度控制中心,广西南宁530024

出　　处：《电力大数据》2020年第4期8-15,共8页Power Systems and Big Data

摘　　要：提出一种在云计算平台上构建大数据环境下支持多源信息融合的微服务化电网事故追忆系统方法,解决传统电网事故追忆系统耦合度高、灵活性差、不易扩展等问题,满足事故追忆系统对短时间内解析大规模故障数据的要求。根据功能需求,将系统重构为电网模型、事故记录、操作记录和人机交互四个细粒度微服务;在此基础上,将每一个数据源接入独立的微服务模块,避免服务间的耦合和堆叠。结合云平台监控系统采集的容器集群负载参数,提出了基于长短期记忆神经网络算法的资源预测模型,提前对容器资源进行预测调度,避免负载突变对系统效率的影响,提升容器资源调度水平。结果分析表明,采用所述方法实现的微服务化电网事故追忆系统可靠性达99.999980%,具有良好的响应效率。This paper proposes a method of constructing a microservice-based post disturbance review system that supports multi-source information fusion on cloud computing platform,which can solve the problems such as high system coupling degree,poor flexibility and difficult expansion of current power grid post disturbance review system,and meet the requirements of post disturbance review system for analyzing large-scale fault data within few minutes.According to the functional requirements,the system is reconstructed into four fine-grained microservices of grid model,accident record,operation record and human-computer interaction;on this basis,each data source is separately connected to a microservice module to avoid coupling between services and stacked.Combining the load parameters of the container cluster collected by the cloud platform monitoring system,a resource prediction model based on long short-term memory neural network is proposed to predict and schedule container resources in advance,avoiding the impact of sudden load changes on system efficiency and improve the container resource scheduling level.The result analysis shows that the reliability of the microservice-based post disturbance review system realized by the method reaches 99.999980%,which has pretty response efficiency.

关 键 词：事故追忆 微服务 容器 网络 云计算 

分 类 号：TM73[电气工程—电力系统及自动化]