【作者】 孔涛；

【导师】 程浩忠；

【作者基本信息】 上海交通大学 ， 电力系统及其自动化， 2008， 博士

【摘要】 城市电力网(简称城网)通常由多个不同电压等级的网络组成。文中根据城市输电网、高压配电网和中压配电网网架的特点和规划中的要求,提出与之相适应的先进的多目标规划方法。其中,中压配电网是提高最终用户供电可靠性的关键,也是当前城网规划研究的重点,因此文中进一步研究了提高用户供电可靠性的新思路对中压配电网网架规划的影响。主要工作有:(1)提出城市输电网网架结构与分区方式的两层多目标联合规划方法。分析联合规划所涉及各方面问题的相互关系,在此基础上构建两层多目标联合规划模型。根据联合规划的物理意义设定问题求解的前提条件,从而将两层多目标规划问题的求解转化为单层多目标优化问题的求解。设计多目标遗传算法(Multiobjective GA,MOGA)的关键环节,使其在求解转化后的单层多目标优化问题的同时,逐渐形成接近原有两层多目标规划问题有效解集的一组解。(2)提出考虑负荷模糊不确定性的“手拉手”高压配电网网架多阶段多目标规划方法。使用等三角模糊数表示负荷预测值,并建立高压配电网网架多阶段多目标模糊规划模型。根据“手拉手”网架在分阶段规划中的特点及某一规划阶段各规划目标值的计算,得到各阶段优化目标的加权和,作为规划方案的评价指标。在MOGA的相关环节使用等三角模糊数的运算法则和基于满意度的模糊数大小比较与排序方法,使MOGA能够适用于模糊规划模型的求解。(3)提出基于组件式地理信息系统(Component GIS,ComGIS)网络分析功能与MOGA的城市中压配电网网架规划方法。通过ComGIS平台建立规划区域的网络数据集,在此基础上使用MOGA进行优化。利用ComGIS平台的网络分析功能优化单环最优路径,解决了遗传算法与地理信息结合时出现的网架编码复杂、进化算子效率低下的问题。利用基于Pareto秩的个体适应度函数引导MOGA的进化方向,最终得到接近多目标优化问题Pareto最优解集的一组两端供电网架方案。(4)研究基于绩效的(Performance Based Rate,PBR)可靠性管理模式对城市中压配电网网架多目标规划的影响,并对两种模式的作用进行比较,以找到其各自的适用场合。设定统一的可靠性电价-奖惩机制和可靠性指标参考值,在此基础上制定PBR可靠性管理模式的标准。根据规划网架供电可靠性指标的概率分布特性得到供电公司面临的奖惩风险指标,进而建立PBR可靠性管理模式下的中压配电网网架多目标规划模型。(5)研究分布式供电系统(Distributed Power System,DPS)对城市中压配电网网架多目标规划的影响。将DPS作为提高重要用户供电可靠性的措施,纳入中压配电网网架的统一规划中:提出DPS的运行方式及其在中压配电网网架规划建模中的处理方式;提出含DPS的中压配电网规划方案缺电成本计算方法,在此基础上建立含DPS的中压配电网网架多目标规划模型,并分析当前DPS建设中的不确定性问题及其对规划模型的影响。各章通过实际电力网规划算例来验证不同电压等级电力网网架多目标规划方法的有效性;并使用实际配电网规划结果分析了PBR可靠性管理模式与DPS对中压配电网网架规划的影响。更多还原

【Abstract】 Urban electric network usually includes several networks with different voltage grades. In this paper, according to their individual characteristics and different requirements in planning, the multiobjective planning methods for the structure of transmission network, High-Voltage (HV) distribution network, and Medium-Voltage (MV) distribution network are proposed. Among these, MV distribution network is the key of improving power supply reliability of the final customers, also is the emphasis of urban electric network planning research presently; therefore, the effects of new ideas that improving power supply reliability on MV distribution network structure planning are explored further. The following researches have been developped:(1) A united planning method for the network structure and partition mode of urban transmission network is proposed. The bilevel multiobjective united planning model is formulated, which is solved by a sepecial designed Multiobjective Genetic Algorithm (MOGA). Precondition of the solution is assumed, based on which the solution of the bilevel multiobjective model could be transferred to the solution of a corresponding single level multiobjective one. Key steps of the MOGA are designed, and the efficient solution set of the bilevel multiobjective problem could be formed with the solution of the corresponding single level problem.(2) A planning method for“hand in hand”urban HV distribution network structure is proposed. Power load forecasting values are presented as the equilateral triangular fuzzy numbers, and the multistage multiobjective fuzzy planning model is formulated. The characteristics of“hand in hand”network structure in multistage planning are explored, and then the optimization objectives in multistages planning are obtained, serving as the evaluation indices of the planning schemes. The arithmetic, comparison, and order criterions of triangle fuzzy numbers are explored, which make the MOGA applicable for solving the fuzzy planning model.(3) A method based on ComGIS network analysis and MOGA is proposed for the open-loop MV distribution network structure planning. The multiobjective planning model of open-loop MV distribution network structure is constructed. And a special MOGA is designed based on the network dataset built by ComGIS platform. The network analysis function of the ComGIS is embedded in the MOGA to find the single-loop optimal paths, which improves the efficiency of evolutionary coding and operators obviously. Finally, the Pareto Optimal Set (POS) of the multiobjective planning problem is obtained, including several candidate planning schemes.(4) The effects of Performance Based Regulation (PBR) reliability government modes on the multiobjective planning of urban MV distribution network structure are researched, and two modes are compared to find their individual applicable environment. The uniform reliability power price-compensation mechanism and reliability indices reference values are designed, and then the standards of the two PBR modes are established. The financial risk of the electric utility is obtained according to the probabilistic characteristic of power supply reliability indices. Then the MV distribution network structure multiobjective planning model under PBR reliability government is constructed.(5) The effects of Distributed Power System (DPS) on the multiobjective planning of urban MV distribution network structure are researched. Being treated as the measure of improving power supply reliability of the important customers, DPS are included in the uniform planning of MV distribution network structure: two operation modes are proposed for DPS connected in the MV distribution network, and the processing mode for DPS in the modeling of distribution network planning is presented; the outage cost calculation method for distribution network including DPS is proposed, based on which the multiobjective planning model is constructed. And the uncertain problems in DPS constructing and their effects on the planning model are analysed.Practical networks are selected to test the effectiveness of the structures planning methods for the different voltage grades networks. And the effects of PBR reliability government modes and DPS on MV distribution network structure planning are analysed through the planning results of the practical distribution network.更多还原