节点文献

基于空间电力负荷动态分布辨识技术的配网规划研究

Reaserach on Distribution Network Palnning Based on Identification of Dynamic Distributed Spatial Load Technology

【作者】 陈晋

【导师】 李扬;

【作者基本信息】 东南大学 , 电气工程, 2016, 硕士

【摘要】 随着城市的发展,城市地区土地资源愈发稀缺,配电站建设费用不菲。而用户的用电需求不断增长,用户报装容量增长迅速。但是在实际情况中,配电站利用效率不高,部分间隔开放完毕的配电站长期处于轻载运行状态。传统的配电网规划以用户报装容量为基础,未能充分考虑报装容量与实际负荷之间的关系。因此研究报装容量与实际负荷之间的关系,考虑实际负荷的增长趋势,将其纳入到配电网规划优化,有助于提高配电站利用率,实现资源优化配置。本文提出了空间电力负荷动态分布辨识技术,研究了配电网规划模型以及求解算法,重点研究了基于空间电力负荷动态分布辨识技术的配电网规划模型,同时在该模型中考虑了行业间负荷同时率。首先,分析了目前配电网规划中负荷增长趋势分析研究现状,重点对配电网规划国内外研究现状作了分析,从模型分类、目标函数、约束条件和求解方法四个方面作了详细的介绍,总结前人的研究成果。其次,建立了实用系数和阶段系数这两个指标,为报装容量与实际负荷之间关系的研究奠定了基础,在阶段系数的分析中引入了聚类算法。提出了空间电力负荷动态分布辨识技术,实现了对区域新增用户负荷的阶段性预测。通过算例分析了实用系数和阶段系数对于提高配电站利用率的重要意义。然后,研究了配电网规划模型,基于负荷出力模型和配电网可靠性指标建立了配电网规划模型的目标函数和约束条件,重点研究了模型求解算法。采用遗传算法求解配电站选址问题,利用基于Voronoi图理论的分配算法求解负荷点分配问题,采用基于蒙特卡洛的区间估计方法计算线路电流限值和配电站的容量。通过算例验证了该模型的有效性。最后,在传统的配电网规划模型基础上,引入行业间负荷同时率的概念,提出了基于空间电力负荷动态分布辨识技术的配电网规划模型,该模型实现了配电网采用逐年建设的方法,代替了传统的一次性建设方法,从而节省了配电网的规划建设成本。通过算例与传统模型对比验证了该模型的有效性,并且通过算例分析研究了不同行业负荷类型和负荷同时率对配电网规划的影响。

【Abstract】 Because of land resources of urban areas are growing scarce with the development of the city, the costs of sub-station construction are expensive, while the electricity demand of users are growing and the new capacity of users are growing so quickly. But in the actual situation, the utilization of sub-station was not high and some sub-stations which bays were all opened were in the light load running state for a long period of time. Traditional distribution network planning based on the new capacity of users could not take full account of the relation between the new capacity and actual load. So studying the relation between the new capacity and actual load and considering the development characteristics of actual load, then bring them into distribution network optimal planning, that was help to improve the utilization of sub-station and realize the optimal allocation of resources.This paper proposed identification of dynamic distributed spatial load technology and studied the distribution network planning model and its solution algorithm. The important point was to study the distribution network planning model based on identification of dynamic distributed spatial load technology, and load coincidence factor was considered in this model.Firstly, research method of load characteristic and research status of load forecasting were analyzed, and research actuality of distribution network planning was the important point to analyze from four aspects, such as model classification, objective function, constraint condition and solution algorithm. Research achievement of forefathers was summarized.Secondly, through establishing practical coefficient and staged coefficient, this paper proposed identification of dynamic distributed spatial load technology to lay the foundation for the research between new capacity and actual load. Clustering algorithm was brought in the analysis of staged coefficient. Identification of dynamic distributed spatial load technology was proposed, and this technology realized periodic forecasting of new user load in some areas. The significance of practical coefficient and staged coefficient for improving the utilization of sub-station was analyzed by an example.Thirdly, distribution network planning model was studied, and objective function and constraint condition of this model were established based on load output model and distribution network reliability index. Solution algorithm of this model was focused on. Genetic algorithm was used to solve location problem of sub-station. Allocation algorithm based on Voronoi map theory was used to solve load point distribution problem. The interval estimation method based on Monte-Carlo simulation was used to calculate the line current limit and sub-station capacity. The validity of the model was verified by an example.Finally, distribution network planning model based on identification of dynamic distributed spatial load technology with load coincidence factor between different industries on the basis of traditional model. This model saved cost of distribution network planning and construction, because of construction year after year was used to take place of traditional one-step construction in distribution network. The validity of the model was verified by an example by an example compared to traditional model. And the influences of different industries load type and load coincidence factor between different industries on distribution network planning were analyzed by examples.

  • 【网络出版投稿人】 东南大学
  • 【网络出版年期】2017年 03期
节点文献中: 

本文链接的文献网络图示:

本文的引文网络