基于不确定理论的含DG的配电网网架规划

发布时间：2018-10-19 13:39

【摘要】：摘要:随着新一轮能源革命的兴起,电网的发展已经进入到能源互联、智能的阶段。含分布式电源(distributiongeneration,DG)的配电网作为未来电网的一种重要形态,其规划发展对节约能源、推进地区城市建设、解决新兴技术并网等一系列问题都起着至关重要的作用。然而,DG本身带有的随机性、间歇性等不确定性特点却给配电网规划带来了一定的冲击和挑战。为此,如何做好这种形态下的配电网规划,解决各种不确定性因素对其经济性、安全可靠性带来的影响,提高配电网对可再生能源的接纳能力,获得配电网建立的最佳性能等一系列问题,变得尤其重要。本文基于不确定理论,对含DG的配电网网架进行规划,主要完成工作如下:考虑到含DG的配电网中存在的不确定因素,对不确定规划理论进行概述。在模糊规划中给出模糊集、模糊期望值的概念并进行模糊模拟;在随机规划中给出随机变量、随机期望值的概念和随机机会约束的数学原理。对不确定潮流计算和优化方法进行研究。首先考虑分布式电源出力和负荷功率的特性,并对应模糊规划和随机规划分别进行模糊潮流和随机潮流计算。模糊潮流以牛-拉法为基础,利用泰勒级数展开求解状态变量的模糊增量;随机潮流采用基于半不变量的方法,求出待求变量各节点电压和各支路功率的概率分布情况。其次重点介绍自适应遗传算法、单亲遗传算法和树形结构编码遗传算法这三种改进的遗传算法,比较其各自的应用特点。建立含DG的配电网网架规划模型。对风力发电、光伏发电进行建模,并建立了相应的模糊期望规划模型和机会约束规划模型。在模糊期望规划模型中,采用梯形模糊数对DG和负荷的不确定进行模糊模拟,以投资建设费用模糊期望值最小为目标函数;在机会约束规划中,考虑到节点电压、支路功率约束条件的置信水平。为说明以上模型的合理性,采用18节点系统进行分析验证。分别对是否含有DG接入系统和是否采用不确定规划理论方法验证,并采用基于树形结构编码的单亲遗传算法(TSE-PGA)进行优化求解。由建设投资费用最小和网损费用最小得出,含有DG接入、采用不确定规划理论,可以有效解决不确定因素对配电网规划的影响,降低网损,减少网络规划费用,从而获得可观的经济效益。通过以上研究,基于不确定理论,提供了解决含DG的配电网网架规划问题的解决思路,并采用案例计算说明了方法的可行性。
[Abstract]:Absrtact: with the rise of new energy revolution, the development of power grid has entered the stage of energy interconnection and intelligence. Distribution network with distributed generation (distributiongeneration,DG) is an important form of power grid in the future. Its planning and development plays an important role in saving energy, promoting the construction of regional cities, solving a series of problems such as connecting with new technologies and so on. However, the uncertain characteristics of DG itself, such as randomness and intermittence, have brought some challenges and challenges to distribution network planning. Therefore, how to do well the distribution network planning in this form, solve the impact of various uncertain factors on its economy, safety and reliability, improve the distribution network's ability to accept renewable energy, A series of problems, such as obtaining the best performance of the distribution network, have become particularly important. Based on the uncertainty theory, the main work of this paper is as follows: considering the uncertain factors in the distribution network with DG, the uncertain planning theory is summarized. In fuzzy programming, the concepts of fuzzy set and fuzzy expected value are given and fuzzy simulation is carried out. In random programming, the concepts of random variables, random expected values and the mathematical principles of stochastic chance constraints are given. The calculation and optimization methods of uncertain power flow are studied. Firstly, the characteristics of distributed power generation and load power are considered, and fuzzy power flow and stochastic power flow are calculated for fuzzy programming and random programming, respectively. The fuzzy power flow is based on the Niu-pull method and the Taylor series expansion is used to solve the fuzzy increment of the state variable, and the random power flow is based on the semi-invariant method to obtain the probability distribution of the voltage and branch power of each node. Secondly, three improved genetic algorithms, namely adaptive genetic algorithm, parthenogenetic genetic algorithm and tree structure coding genetic algorithm, are introduced, and their respective application characteristics are compared. The distribution network planning model with DG is established. The wind power generation and photovoltaic power generation are modeled, and the corresponding fuzzy expectation programming model and chance constrained programming model are established. In the fuzzy expectation programming model, the trapezoidal fuzzy number is used to simulate the uncertainty of DG and load, and the minimum expected value of investment and construction cost is taken as the objective function, and the node voltage is considered in the chance constrained programming. The confidence level of branch power constraints. In order to explain the rationality of the above model, the 18-bus system is used to analyze and verify the model. Whether the DG access system is included or not and whether the uncertain programming theory is adopted are verified, and the algorithm based on tree structure coding (TSE-PGA) is used to optimize the solution. From the minimum construction investment cost and the minimum network loss cost, including DG access, the uncertain programming theory can effectively solve the influence of uncertain factors on distribution network planning, reduce network loss, and reduce network planning costs. Thus, considerable economic benefits are obtained. Based on the theory of uncertainty, this paper provides a solution to the distribution network frame planning problem with DG, and the feasibility of the method is illustrated by a case study.
【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】：TM715

【参考文献】