含分布式能源的交直流混合配电网规划

发布时间：2018-06-22 16:53

本文选题：分布式电源 + 不确定性　；参考：《华北电力大学》2017年硕士论文

【摘要】：配电网直接面向用户并承载电网的主要负荷,其配置的合理与否直接关系供电质量并影响居民用电和社会效益。随着经济的发展,社会对用电的量与质的要求都不断提高,配电网网架的升级和扩增也亟待进行;近年来分布式电源(distributed generation,DG)在配电网中的渗透率逐步提高并对配电网产生了较大的影响,其合理的规划将对配电网的优化规划和运行具有重要的意义;此外,随着直流负荷密度的不断攀升,交直流混合配电网的研究也逐渐火热。在上述背景下,本文针对含DG的配电网规划问题做了如下的工作:首先,针对DG出力和负荷的不确定性,提出了一种基于时序特性划分和混合聚类的多场景分析方法。通过分析DG出力和负荷所受影响因素的特点,将各单元的时序特性划分为多个典型时序情形;在每种情形内,基于大量历史数据采用混合聚类算法得到若干个具有特定概率的场景,并以期望、标准差、偏度和峰度四个统计指标验证了生成的场景集与原始数据集的接近性。该方法将不确定性因素量化成了场景集,为考虑不确定性因素的配电网规划奠定了基础。其次,针对DG与配电网网架在协调规划中的相互影响,提出了一种基于启发式智能算法的嵌套式改进双层规划方法,并基于时序多场景建立了交流配电网中DG与网架的多目标双层协调规划模型。上下层模型分别决策网架配置和DG配置,并分别从系统全局最优和DG效益最大化出发建立目标函数。该模型可通过上下层的嵌套求解保证种群中各网架与DG的最优组合,进而解决了两者的相互影响。通过算例分析,时序多场景方法和嵌套式双层方法的优越性得到了验证。再次,提出了一种考虑网架直流改造和直流网架扩建的规划方法,并基于时序多场景建立了含DG的交直流混合配电网规划模型。该规划模型将DG配置、待直流改造的网架、扩建网架的接入位置及是否以直流形式接入作为决策变量,目标函数中考虑了DG与网架投资、DG及负荷的换流器投资、网损和电压稳定性等。算例表明该规划方法可有效地将交流配电网改造成交直流混合配电网,为交直流混合配电网的发展提供了较好的参考。
[Abstract]:The distribution network is directly oriented to users and carries the main load of the power network. Whether the distribution network is reasonable or not is directly related to the quality of power supply and affects the power consumption and social benefits of the residents. With the development of economy, the demand of the quantity and quality of electricity is raised constantly, and the upgrade and expansion of the distribution network are urgently needed. In recent years, the permeability of distributed generation (DG) in the distribution network has been gradually increased and has a great impact on the distribution network. Its reasonable planning will be of great significance to the optimal planning and operation of the distribution network. With the increasing of DC load density, the research of AC / DC hybrid distribution network is becoming more and more hot. Under the above background, the following work has been done in this paper for the distribution network planning problem with DG: firstly, a multi-scene analysis method based on time-series characteristic partitioning and hybrid clustering is proposed for the uncertainty of DG output and load. By analyzing the characteristics of the factors affecting the output and load of DG, the timing characteristics of each unit are divided into several typical time series cases, and in each case, Based on a large number of historical data, a number of scenarios with specific probability are obtained by using mixed clustering algorithm, and the proximity of the generated scene set to the original data set is verified by four statistical indicators: expectation, standard deviation, bias and kurtosis. This method quantifies uncertainty factors into scene sets, which lays a foundation for distribution network planning considering uncertainty factors. Secondly, aiming at the interaction between DG and distribution network in coordination planning, a nested improved bilevel programming method based on heuristic intelligent algorithm is proposed. A multi-objective bilevel coordination programming model of DG and grid in AC distribution network is established based on time series and multi-scene. The upper and lower layer models respectively decide the network configuration and DG configuration, and set up the objective function from the view of the global optimization and the maximization of DG benefit respectively. The model can be solved by nesting the upper and lower layers to ensure the optimal combination of each grid and DG in the population, thus solving the interaction between the two. The superiority of the time series multi-scene method and the nested double-layer method is verified by example analysis. Thirdly, a planning method considering DC transformation of grid and expansion of DC grid is proposed, and a DG hybrid distribution network planning model with DG is established based on time series multi-scene. In this model, the DG configuration, the network structure to be transformed by DC, the access position of the extended grid and whether the direct current access is taken as the decision variable are considered in the objective function. The investment of DG and the investment of the DG and the load converter are considered in the objective function. Network loss and voltage stability. The example shows that this planning method can effectively transform the AC distribution network into the AC / DC hybrid distribution network and provide a good reference for the development of the AC / DC hybrid distribution network.
【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM715

【参考文献】