中文摘要：

文章梳理了智能电网大数据流实时分析技术的发展趋势,研究了具有流式计算与主动管理双重特性的复杂事件处理（complex event processing,CEP）技术特点。基于实时数据分发服务（data distribution service,DDS）中间件设计了自动需求响应（automated demand response,ADR）系统与CEP实时监测服务的集成架构,给出CEP服务节点的功能组件与分析流程,论述了应用架构的互操作优势。以用户需求响应性能实时监测分析为案例,说明应用CEP实现用户基线负荷与响应性能计算的关键技术,由ADR事件定义反映资源响应时序的上下文规则,设计用户侧监测事件模型与关键查询算法。特别针对实时与增强历史监测数据上下文的混合查询算法,给出CEP引擎内部扩展设计与实现流程,并实验验证算法性能。应用CEP内嵌R算法实现快速统计计算。最后,建立原型系统与CEP实时分析算法的可视化仿真实验,验证了CEP应用的可行性。

英文摘要：

This paper summarized development trends of smart grid big data stream real-time analytic technology, considering complex event processing（CEP） technology with dual characteristics of stream computing active management potentials. Automated demand response（ADR） system CEP real time monitoring service integration architecture was designed based on real time data distribution service（DDS） middleware with detailed CEP monitoring service node design of functional structure analysis workflow, and its application architecture interoperability advantages were explained. As a case study of real time monitoring analysis of customer demand response performance, CEP is applied to implement real time customer baseline load response performance computation using key technology. The paper presented CEP key technology, using ADR event model to reflect chronological steps of a demand response event with CEP contextual rule definition, and designed customer side event monitoring model and key query algorithm. Especially for real-time history-enhanced monitoring contextual hybrid query, an extended CEP engine internal design algorithm flow was given with experiments validating algorithm performance. To achieve fast statistical calculation, CEP used embedded R algorithm. Finally, prototype system was built with CEP real time analytic algorithm related programs to conduct visualized simulation to evaluate CEP application efficiency.