【作者】 杨雪；

【导师】 夏明超；

【作者基本信息】 北京交通大学 ， 电气工程， 2018， 硕士

【摘要】 在化石能源和环境污染双重危机下,大力开发和利用可再生能源是解决这一问题的有效途径。我国资源分布和负荷中心呈现非同步特性,远距离输送可再生发电资源难度高,损耗大。分布式电源由于距离负荷近,装机容量小,控制方便、灵活的特点,得到了大量的支持和发展。但可再生分布式能源间歇性、波动性的特点,使其大规模并网会对传统配电网的调度,运行产生巨大的影响。配电网正从单“源”化向多“源”化发展。电池储能技术的提高,电动汽车的大量普及以及智能家居的发展,负荷侧也能体现柔性特性参与电力市场调度。柔性负荷作为需求响应的重要资源,越来越成为促进可再生能源消纳、实现削峰填谷和提高配电网灵活调节的重要手段,也是主动配电网“源-网-荷”协调调度的重要部分。本文首先针对上述问题深入研究了主动配电网中柔性负荷响应特性。将全网柔性负荷分为四大类:电动汽车负荷、工业高载能负荷、商业大负荷及居民负荷。根据四类用户可调度潜力类型、参与响应的方式,建立了负荷模型,制定了不同负荷参与电网调度的机制,以集中式和基于负荷商控制形式,参与电网调度。基于改进的IEEE33节点模型,研究计及柔性负荷对配电网调度的经济和可再生能源消纳的影响。在33节点模型中,加入可转移负荷、可中断负荷、可平移负荷、可削减负荷等,在可再生能源消纳最大化的前提下,建立以经济调度为目标的模型。分四种情况:可再生能源低渗透率时,不计柔性负荷调度和计及柔性负荷调度的情况;可在生能源高渗透率高时,不计柔性负荷和计及柔性负荷调度的情况。研究仿真表明计及柔性负荷的调度可以减少配电网的运行成本和促进可再生能源的消纳。考虑风电、光伏分布式发电在不同时间尺度上的预测精度不同,时间越近,预测精度越高的特点,提出了“日前-日内-实时”多时间尺度调度策略。以多时间调度来平抑可再生能源波动,使电网消纳更多可再生能源。根据柔性负荷参与电网的响应速度、可调度时长,提前通知时间等因素将柔性负荷分为低质量柔性负荷、一般柔性负荷和高质量柔性负荷。并且建立以可再生能源消纳最大、经济效益最优、联络线功率波动和网损最小为目标,计及柔性负荷的主动配电网多时间尺度多目标优化调度模型。对于建立的高维度、多约束的混合整数非线性多目标优化模型。提出了多目标粒子群优化算法,利用动态步长和增加权重系数等方法增强算法的整体性能,使得求解的模型取得更优的Pareto前沿。利用改进的IEEE33节点模型,验证本文提出的多时间尺度调度模型。算法求解获得Pareto最优解集,证明了多时间尺度计及柔性负荷的主动配电网协调调度下,在经济成本增加时,等效负荷的波动减少,线路有功功率损耗下降,能减少扩建压力,提高资源利用率。同时也能保证高渗透率可再生能源发电情况下,平抑可再生能源的波动,跟踪电网调度计划安排,保证在计划可接受的波动内,实现最优调度。更多还原

【Abstract】 Under the dual crisis of fossil energy and environmental pollution,vigorously developing and utilizing renewable energy is an effective way to solve this problem.The distribution of resources and the center of load in China are not synchronous.It is difficult to transport renewable generation resources in a long distance,and the loss is great.Due to close to load,small installed capacity,convenient control and flexible characteristics,distributed power supplies have received a lot of support and development.However,the intermittent and fluctuating characteristics of distributed renewable energy will have an enormous impact on the dispatching and operation of traditional distribution networks,when large-scale access to the power grid.Distribution network is developing from single "source" to multi "source".With the improvement of battery energy storage technology,the popularity of electric vehicles and the development of smart home,the load side can also reflect the flexible characteristics to participate in the power market scheduling.The flexible load as a crucial resource of demand response,has played a vital role to promote renewable energy consumption,to clip peak and fill valley and to improve flexible adjustment of the distribution network,is also an important part of the active distribution network’s the source network load scheduling.The response characteristics of flexible load in active distribution network are studied in this paper.The flexible load of the whole network is divided into four categories:electric vehicle load,industrial load,commercial load and resident load.According to the four kinds of schedulable potential types and the way to participate in response,load models are established,and the mechanism of different load participating in the power grid dispatching is formulated.The centralized and based on load aggregator’s control forms are adopted to participate in the power grid dispatching.Based on the improved IEEE33 node model,the paper considers the impact of flexible loads on the distribution of economic and renewable energy for distribution network dispatch.With the addition of transferable load,interruptible load,translatable load,and load reduction added to IEEE33 node model,an economic dispatch model is established in the premise of renewable energy consumptive maximization.In consideration of four cases:renewable energy in low permeability without flexible load;renewable energy in low permeability with flexible load;renewable energy in high permeability without flexible load;renewable energy in high permeability with flexible load.Simulation results show that flexible load participating in grid scheduling can reduce the operation cost of distribution network and promote renewable energy consumption.Considering the forecasting errors of wind power distributed generation and photovoltaic distributed generation on different time scales:the closer the time is,the higher the prediction accuracy is,a“day-ahead-day-real-time”multi-time scale scheduling strategy is proposed in this paper.Multi-time scheduling will stabilize the fluctuation of renewable energy sources and allow the grid to absorb more renewable energy.According to factors such as the flexible load’s response speed,schedulable duration,and notice time when participating in the grid scheduling,the flexible load can be divided into low-quality flexible load,general flexible load and high-quality flexible load.In addition,a multi-objective optimization scheduling model for active distribution networks with flexible loads is proposed,which aims to maximize renewable energy consumption,optimize economic efficiency,minimize power line fluctuations,and minimize network losses.A multi-objective particle swarm optimization algorithm is used to solve the high-dimensional,multi-constraint mixed-integer nonlinear multi-objective optimization model,and the overall performance of the algorithm is enhanced by using dynamic step size and increasing weight coefficient,so that the solved model obtains a better Pareto frontier.The effectiveness of the proposed multi-time-scale scheduling model is verified using the improved IEEE33 node model in this paper.Getting the Pareto optimal solution set,and it proves that under the coordinated scheduling of active distribution network with multiple time scales and flexible load,the fluctuation of the equivalent load and the active power loss of the line decreases when the economic cost increases.So,the expansion pressure can be decreased,and resource utilization can be improved.At the same time,it can also guarantee the high-permeability renewable energy power generation situation,stabilize the fluctuation of renewable energy,track the scheduling of power grid dispatching plans,and ensure that the optimal scheduling is achieved within the acceptable fluctuations of the plan.更多还原