面向电力调度控制系统的多源异构数据处理方法研究

发布时间：2018-03-13 14:55

本文选题：电力调度控制系统　切入点：多源异构数据　出处：《华北电力大学(北京)》2017年硕士论文　论文类型：学位论文

【摘要】：随着智能电网建设的深入实施和智能传感设备的广泛使用,电力系统的数据量呈爆炸性增长趋势,电力行业迎来了大数据时代。电力大数据虽然拥有广阔的应用前景,但是面临着数据量大、数据异构和数据分散等问题。在电力调度控制系统中,为了实现调控大数据的有效利用,需要通过多源异构数据处理方法将调控多源异构数据进行信息共享,并建立统一的数据模型和全景调控数据,其对调控多源异构数据处理和数据融合技术的研究具有重要的意义。首先,本文在分析电力调度控制系统数据特点、需求及多源异构数据预处理方法的基础上,提出了全景调控统一数据模型和调控多源异构数据ETL处理模型来满足建立调控全景数据的需求,并对模型的体系框架和工作流程进行了分析。其次,针对构建调控全景数据过程中出现的不完整数据,本文提出了一种面向统一数据模型的缺失数据填补算法。该算法采用改进的混沌遗传优化方法估计不完整数据的均值和协方差对应的最佳参数,再根据已知数据利用改进马尔可夫蒙特卡洛方法估计缺失数据,解决了调控数据中的缺失问题。结果表明,该算法能通过较少的迭代次数获得不完整调控数据的最佳参数,同时,缺失数据的估计值更加准确,有效的保证了数据的准确性和完整性。最后,针对调控多源异构数据融合中存在的字符串匹配问题,本文提出了一种面向电力调控系统数据的字符串匹配算法。为了准确快速的匹配字符串,该算法依据电力调度控制系统数据特点制定了匹配规则,同时提出了一种匹配度计算方法,该方法将字符串的相似程度合理量化,促进调控字符串数据匹配正确率的提高。通过实验仿真分析,验证了该算法有助于提高调控系统字符串数据匹配的正确率,促进了调控多源异构数据的融合。
[Abstract]:With the deep implementation of smart grid construction and the extensive use of intelligent sensing equipment, the data volume of power system is increasing explosively, and the power industry has ushered in big data's time. However, in order to realize the effective use of big data in the electric power dispatching control system, there are many problems, such as large amount of data, heterogeneous data and scattered data, etc. It is necessary to share the information of multi-source and heterogeneous data through multi-source and heterogeneous data processing method, and establish a unified data model and panoramic control data. It is of great significance to study the technology of regulating multi-source heterogeneous data processing and data fusion. Firstly, based on the analysis of data characteristics, requirements and preprocessing methods of multi-source heterogeneous data, this paper analyzes the data characteristics, requirements and methods of multi-source heterogeneous data processing in power dispatching control system. The unified data model of panoramic control and the ETL processing model of multi-source heterogeneous data are put forward to meet the requirements of establishing the panoramic data, and the architecture and workflow of the model are analyzed. For the incomplete data that appears in the process of building the control panoramic data, In this paper, a missing data filling algorithm for uniform data model is proposed, which uses an improved chaotic genetic optimization method to estimate the optimal parameters corresponding to the mean and covariance of incomplete data. Based on the known data, the missing data is estimated by the improved Markov Monte Carlo method, and the problem of missing data is solved. The results show that the optimal parameters of incomplete control data can be obtained by the algorithm with fewer iterations. At the same time, the estimation of missing data is more accurate, which effectively ensures the accuracy and integrity of the data. Finally, aiming at the string matching problem in multi-source heterogeneous data fusion, In this paper, a string matching algorithm for power regulation and control system data is proposed. In order to match string accurately and quickly, the algorithm formulates matching rules according to the data characteristics of power dispatching control system. At the same time, a method of calculating matching degree is proposed, which quantifies the similarity of string reasonably, and promotes the improvement of matching accuracy of string data. It is verified that this algorithm can improve the accuracy of string data matching and promote the fusion of multi-source heterogeneous data.
【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】：TM73

【参考文献】