基于负荷曲线和多代理技术的含微电网配电网动态供电恢复

发布时间：2018-03-05 23:37

本文选题：供电恢复　切入点：微电网　出处：《重庆大学》2014年硕士论文　论文类型：学位论文

【摘要】：可靠持续地供电是电力系统运行的基本要求之一，，但量大面广的配电网在实际运行中难以避免故障发生。故障发生后，应准确定位故障，快速隔离故障，及时供电恢复。其中，供电恢复的主要目标是指在不违背操作约束和安全约束的条件下，通过网络重构技术将非故障停电区的负荷尽可能多地转移到正常馈线。有效的供电恢复能减少因故障造成的停电区域和缩短用户的停电时间，减少用户的停电损失和增加供电公司的售电量。因此，研究有效的配电网供电恢复方法，对提高配电网供电运行的安全性和可靠性具有十分重要的意义。在研究当前国内外配电网供电恢复模型和方法的基础上，本文主要从以下几个方面开展研究工作： ①从负荷模型的角度看，本文针对现有负荷模型在制定配电网供电恢复方案时所存在的不足，提出了一种基于负荷曲线的配电网动态供电恢复方法。该方法在制定恢复方案时，基于负荷曲线按小时考虑负荷大小的变化，以失负荷量最少为目标，支路电流,节点电压以及辐射状运行为安全约束为条件，优选配电网故障修复期间各小时的网架结构形成候选网架集合，进而得到故障修复期间能随负荷变化而改变，同时使开关动作次数最少的动态供电恢复方案。该方法具有时间空间动态特性，能充分利用负荷曲线，进一步有效减少因故障造成的停电区域、缩短用户停电时间、减小用户停电损失和增加电力公司售电量。 ②结合配电网发展趋势，探讨了配电网动态供电恢复方法在含微电网配电网中的应用，针对故障发生后备用馈线含微电网的情况，提出了一种考虑开关动作次数的含微电网配电网动态供电恢复方法。该方法在时间层面上按小时进行分割，在结合考虑负荷变化特性的同时，为故障修复期间各小时建立兼顾开关动作次数和失负荷量的多目标多约束数学模型，通过求解故障修复期间各小时对应的数学模型，得到各小时的网架结构，进而形成候选网架集合，再根据失负荷量和开关动作次数的重要程度量化关系，得到排序后的多个动态供电恢复方案，以便工作人员根据实际情况进行选择。同时，该方法通过调整微电网内可调度分布式电源出力使其满发，增大备用馈线的余量，从而扩大供电恢复面积，缩短用户停电时间。 ③未来的智能配电网将电力网和通信网紧密相连，在电力传输的同时具有信息交互的能力。本文针对现有配电自动化集中控制在供电恢复问题中的不足，建立了一种分布式分层结构的多代理系统，该多代理系统由馈线代理、开关代理、微电网代理以及分布式电源代理组成，通过代理间的相互通信及时更新用于供电恢复的网络数据，确保了制定的供电恢复方案安全最优；同时，通过代理间的相互合作共同制定并实现了动态供电恢复过程。本文建立的多代理系统可以充分利用通信网络和配电网的分布式特性减少人工参与，智能化实施安全可靠的最优动态供电恢复方案，提高了配电自动化智能水平。 ④采用70节点的网络作为算例验证了上述方法的可行性和有效性。其中①和②在Matlab编程环境实现，③在JADE平台上采用Java-Matlab混合编程方法实现。
[Abstract]:Reliable power supply continuously is the basic requirement of power system operation of distribution network, but a large amount of wide in the actual operation is difficult to avoid failure. After the failure, should be accurate fault location, fast fault isolation, timely recovery of power supply. The main goal of restoration is in does not violate the operation and security constraints under the condition, through the network reconfiguration technology will be the non fault zone load as much as possible transfer to normal feeders. Effective restoration can reduce the outage area caused by faults and shorten the user outage time, reduce user outage loss and increase the power company electricity sales. Therefore, the study of effective power distribution net recovery method, has very important significance to improve the safety and reliability of power supply operation of distribution network. Based on the research at home and abroad, the current power supply distribution network restoration model and method on In this paper, the main research work is to be carried out in the following aspects:
From the angle of load model, aiming at the existing problem formulated in the load model of distribution network power supply restoration scheme exists, proposed a dynamic recovery method based on power load curve. The method in the formulation of restoration scheme, when considering the variation of load curve according to the small load based on the loss of load minimum volume as the target, the branch current, node voltage and radial transport behavior safety constraint condition, optimization of distribution network fault restoration during each hour of the formation of network structure and network candidate set, during the repair of the fault can vary with the change of load and the dynamic power supply minimum switch recovery scheme. This method is dynamic the characteristics of time and space, can make full use of the load curve, and further reduce the outage area caused by failure, shorten the outage time, reduce the loss of power outages And increase the sale of electricity power company.
Based on the development trend of distribution network, discusses the dynamic power recovery method in the application of micro grid in the distribution network, aiming at the failure of backup feeders containing micro grid, this paper presents a micro dynamic power grid distribution network restoration method with switch times are considered. The method in time according to the level of hour segmentation, in combination with the load change characteristics at the same time, the establishment of both switch and failure load of multi-objective and multi constraint mathematical model for fault repair during the hour, every hour during the repair of solving mathematical model, obtain the grid structure of each hour, and then form the candidate frame according to the set. The importance of quantitative load and switching loss frequency relationship, get a number of dynamic power supply after sorting recovery program, so that workers choose according to the actual situation. This method can make full use of the dispatch of distributed power in microgrid, increase the allowance of spare feeder, expand the area of power supply recovery and shorten the time of user power failure.
The future of the intelligent distribution network of power network and communication network is closely linked with the ability of information interaction in power transmission and distribution automation. Aiming at the existing problems in the centralized control of power supply restoration in a multi-agent system, a distributed hierarchical structure, the multi agent system is composed of feeder switch agent, agent the micro grid and distributed power supply agency, through mutual communication between agents to update the network for data restoration, to ensure that the formulation of the restoration scheme is safe and optimal; at the same time, through mutual cooperation between agents to develop and realize the dynamic power recovery process. Multi agent system is established in this paper can fully utilize the characteristics of distributed communication network and distribution network to reduce the artificial participation, the implementation of intelligent recovery scheme of optimal dynamic safe and reliable power supply, improve the distribution of self The level of dynamic intelligence.
Fourth, a 70 node network is used as an example to verify the feasibility and effectiveness of the above methods. 1. And II are implemented in the Matlab programming environment. On the JADE platform, Java-Matlab mixed programming is applied.

【学位授予单位】：重庆大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM732

【参考文献】