基于电能表需求预测的配送优化研究

发布时间：2018-05-30 08:52

本文选题：电能表 + 需求预测　；参考：《北京交通大学》2017年硕士论文

【摘要】：近年来,电力企业改革日益深入,电力营销进入发展新形势。电网公司为深入推进“三集五大”,进一步集约计量业务,在北京、天津等公司开展计量器具直配试点工作。以业务需求为导向,以数据分析为驱动,进行电能表需求预测工作;按照省公司直接配送到县公司、市计量中心直接配送到供电所的配送方式,开展电能表配送工作。为实现计量中心“整体式授权、自动化检定、智能化仓储、物流化配送”的建设目标,电能表需要进行集中检定、统一配送等流程,这也对智能电能表的备货、配送等工作的质量和效率提出了更高的要求。本文通过对电能表安装数目的时间序列预测及对地区内电能表的配送优化,合理的安排计量中心的备表、配送工作,提高了电能表管理的效率及效益。首先,通过地区内电能表安装历史数据的统计,分析电能表安装数量发展变化的规律,运用科学时间序列预测方法建立了三种需求预测模型,并对模型拟合结果做出了评价。通过误差对比选出最优模型对未来电能表安装数量做出了精准预测,为计量中心电能表的备货工作提供参考,为电能表配送优化工作提供了基础数据。其次,在基本配送优化模型的基础上,针对计量中心配送现状建立了可对不同品规电能表混合配送的数学模型。考虑到供电所电能表需求的紧急程度,在模型中加入了时间窗约束。运用遗传算法设计了适用于配送优化模型的编码解码方式、适应度函数等。最后,将本文使用的算法和模型应用于某市计量中心电能表配送算例进行仿真计算,证明了模型中加入时间窗约束的有效性。通过不同电能表单独配送与电能表混合配送情况的比较,体现了混合配送对于节约配送成本的优势。另外,通过对车辆载重不同时配送成本的比较,得出了车辆载重对于配送成本的影响。经过案例分析给出了计量中心减少配送成本的建议。
[Abstract]:In recent years, the electric power enterprise reform is deepening day by day, the electric power marketing enters the development new situation. In order to further promote the "three sets and five" and further intensive measurement business, power grid companies in Beijing, Tianjin and other companies to carry out direct distribution of measuring instruments pilot work. Guided by business demand and driven by data analysis, the forecasting work of watt-hour meter demand is carried out; according to the way that provincial company distributes directly to county company and city metering center distributes directly to power supply station, the distribution work of watt-hour meter is carried out. In order to realize the construction target of "integral authorization, automatic verification, intelligent warehousing, logistics distribution" in the measurement center, the electric energy meter needs centralized verification and unified distribution flow, which also supplies goods to the intelligent watt-hour meter. The quality and efficiency of distribution and other work put forward higher requirements. By forecasting the time series of the number of watt-hour meters installed and optimizing the distribution of the watt-hour meters in the area, the efficiency and benefit of the management of the watt-hour meters are improved by the reasonable arrangement of the meters and the distribution work of the metering center. Firstly, through the statistics of the historical data of the installation of the watt-hour meter in the region, the law of the development and change of the installation quantity of the watt-hour meter is analyzed, and three kinds of demand forecasting models are established by using the scientific time series forecasting method, and the fitting results of the model are evaluated. Through error comparison, the optimal model is selected to predict the installation quantity of watt-hour meters in the future, which provides a reference for the preparation of watt-hour meters in the measurement center, and provides the basic data for the optimization of the distribution of watt-hour meters. Secondly, on the basis of the basic distribution optimization model, the mathematical model of mixed distribution for different watt-hour meters is established according to the distribution situation of metering center. Considering the urgency of electricity meter demand in power supply station, time window constraint is added to the model. The genetic algorithm is used to design the coding and decoding method, fitness function and so on. Finally, the algorithm and model used in this paper are applied to the simulation calculation of the distribution of electric energy meter in a city metering center, which proves the validity of adding time window constraints to the model. By comparing the distribution of different watt-hour meters separately with the mixed distribution of watt-hour meters, the advantages of hybrid distribution for saving distribution cost are demonstrated. In addition, the effect of vehicle load on distribution cost is obtained by comparing the cost of vehicle load distribution at the same time. Through the case analysis, the paper gives the suggestion of reducing the cost of distribution in the metrology center.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U116.2

【参考文献】