随机矩阵理论在电力系统认知中的应用初探

发布时间：2019-02-12 10:00

【摘要】：数据是智能电网建设的战略资源乃至主要驱动力。如何处理智能电网中呈现海量、多样、实时、真实等4个特征的4 Vs数据,并从中提取信息,是电力系统大数据建设所面临的核心问题。描述了大数据的特征和引入了随机矩阵理论作为基础,以及提出电力系统大数据的应用思路和架构。具体电力应用方面,介绍了所开发的早期事件发现、事件诊断和定位、相关性分析、可视化3D Power Map辅助展示等一系列功能。在此基础上,建立起以随机矩阵为理论基础,以数据为主要驱动力的电力系统认知体系框架,并探讨其与传统经典认知方案的区别。进一步设计案例考查了其对坏数据的鲁棒能力,其结果表明,随机矩阵理论这种工具可以有效地处理电网中的复杂数据,具有很好的学术研究意义和工程应用价值。另通过仿真算例验证了随机矩阵方案对数据异步的鲁棒性。
[Abstract]:Data is the strategic resource and even the main driving force of smart grid construction. How to deal with 4 Vs data with 4 features, such as magnanimity, diversity, real time and reality in smart grid, and how to extract information from it are the core problems faced by big data in power system construction. This paper describes the characteristics of big data and introduces the stochastic matrix theory as the basis, and puts forward the applied train of thought and framework of power system big data. In the aspect of electric power application, a series of functions such as early event discovery, event diagnosis and location, correlation analysis, visual 3D Power Map display and so on are introduced. On this basis, a framework of power system cognitive system based on stochastic matrix and data as the main driving force is established, and the difference between it and traditional classical cognitive schemes is discussed. Further design cases are designed to examine its robustness to bad data. The results show that stochastic matrix theory is an effective tool to deal with complex data in power grid, and has good academic significance and engineering application value. The robustness of the random matrix scheme to data asynchrony is verified by a simulation example.
【作者单位】：上海交通大学电子信息与电气工程学院;国网上海市电力公司市区供电公司;
【基金】：国家自然科学基金项目(51577115,61571296)~~
【分类号】：TM76

【相似文献】