基于大数据技术的用电行为分析关键技术研究

发布时间：2018-10-17 21:22

【摘要】：随着营配一体化协同工作机制的建立和营配数据的打通,用电客户的用电数据能够与客户档案、缴费等数据关联起来。客户的用电数据中隐含着客户的用电行为特征,对这些用电数据进行深度挖掘并研究客户类型,可以帮助电网了解客户的个性化、差异化服务需求,从而使电网公司进一步拓展服务的深度和广度,为未来的电力需求侧响应政策的制定提供数据支撑。本论文首先是基于大数据技术通过深入分析梳理,确定了用电行为分析的外部数据和内部数据的数据来源,分析了海量数据存储技术和海量数据预处理技术,其次,研究了用电行为分析的关键技术,主要包括聚类算法、最优聚类评价算法、日期匹配算法、曲线相似性度量算法等用电行为分析算法。再次,构建用电分析模型构建方案,详细描述了用电分析模型的建模思路,研究了主网的用电特征和用电行为,然后结合主网的用电模式研究主网该模式下海量客户的用电行为,从而利用模式匹配技术匹配主网用电模式和该模式下海量客户的用电模式,建立历史日削峰填谷匹配关系,并通过实证研究对用电分析模型进行验证。最后,对用电行为分析管理软件进行系统架构设计、系统功能设计、系统数据库设计,并阐述系统的实现成果。用电行为分析管理软件可以比较准确的对用电客户的用电行为进行分析,有助于电力公司进一步指导用户个性化智能用电,对提高电网需求侧能效管理水平具有建设性意义。
[Abstract]:With the establishment of the cooperative working mechanism of operation and distribution integration and the opening of the distribution data, the electricity consumption data of the electric customers can be associated with the data of the customers' files, payment and so on. The customer's electricity consumption data is implicit in the customer's electricity behavior characteristic. It can help the power grid to understand the customer's individualized and differentiated service demand by digging these data deeply and studying the customer's type. So the grid company can further expand the depth and breadth of the service, and provide data support for the future DSM policy formulation. Firstly, based on big data technology, the external and internal data sources of electrical behavior analysis are determined, and the mass data storage technology and mass data preprocessing technology are analyzed. The key technologies of electrical behavior analysis are studied, including clustering algorithm, optimal clustering evaluation algorithm, date matching algorithm, curve similarity measurement algorithm and so on. Thirdly, the construction scheme of power consumption analysis model is built, the modeling idea of power consumption analysis model is described in detail, and the characteristics and behavior of power consumption of main network are studied. Then combining the power consumption mode of the main network, the paper studies the power consumption behavior of the massive customers under the main network mode, and then uses the pattern matching technology to match the power consumption mode of the main network and the massive customers under this mode, and establishes the matching relationship between the historical peak cutting and filling valley. And through the empirical study to verify the electricity analysis model. Finally, the system architecture design, system function design, system database design, and the achievement of the system are described. The analysis and management software of power consumption behavior can accurately analyze the power consumption behavior of customers, which is helpful for power companies to guide users to use electricity for personal intelligence, and to improve the management level of energy efficiency on the demand side of power grid.
【学位授予单位】：华北电力大学(北京)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.13;TM714;TM732

【参考文献】