【作者】 龚锦霞；

【导师】 蒋传文；

【作者基本信息】 上海交通大学 ， 电力系统及其自动化， 2014， 博士

【摘要】 在我国大力促进分布式能源发展和智能电网信息化、互动化建设的背景下，考虑分布式能源与我国当前电力调度模式的协调，解决分布式能源发电(用户)与电网之间的矛盾，使得分布式能源主动参与电网调度是值得研究的问题。主要研究内容和成果如下：结合虚拟发电厂运行特点，提出一种带优先级的多目标优化方法，将虚拟发电厂的技术层和经济层两层管理理念具体化为带有优先级的目标函数，并采用基于满意度表示的两步分解方法求解。建立虚拟发电厂和配电网的联合优化控制模型并采用近似动态规划方法求解。在与配电网的联合优化时虚拟发电厂的快速控制功能可以改善配电网的运行指标，如电网的频率偏差、电压偏差、功率控制等。基于充放储一体化电站的能量特性，研究了多个充放储电站入网的电网协调调度。利用电动汽车和电池的充放储过程的功率转移功能，改进SCOPF方法实现充放储节点功率调度的时间和空间解耦。首先采用虚拟发电厂的思想将区域电网中所有的分布式能源和充放储电站看作为一个整体，优化得到24个时段的调度结果，实现时间上的解耦；然后考虑电网拓扑结构，优化充放储功率实现空间解耦。提出一种内点法结合随机调节因子更新的方法来求解基于机会约束规划的随机最优调度模型，采用半不变量和Cornish-Fisher级数结合的方法来计算机会约束的等效条件。引入与机会约束概率相关的调节因子，将随机最优调度问题转化为可采用内点法求解的最优潮流模型，并提出调节因子的迭代更新方法实现随机环境下的功率解最优。最后以IEEE14和IEEE118节点系统为例验证所提方法并分析了所提方法的精度分析、灵敏度分析和计算量。提出一种多优化控制中心互动协调调度模式并应用于含分布式能源控制中心的电网优化调度。以优化控制中心为单位来分析分布式能源控制中心主动优化的特点。将传统的电网调度模型也转化为一个优化控制中心。各优化控制中心可按照自己的资源组成特点和多个优化目标并行独立优化以制定出力计划。基于互动协调调度模式，提出了并行优化和优先优化两种优化方式和正常和异常两种运行情况，并设计了两种优化方式和两种运行情况的协调优化方法。针对多优化控制中心并行优化中的冲突问题，采用协同优化来协调优化控制中心间的耦合变量不一致，并通过协调满意度设定值来解决多优化控制中心并行优化的目标冲突问题。最后采用不同运行情况下的算例仿真来说明和验证所提的模型和方法。更多还原

【Abstract】 Recently, Chinese policies greatly promote the development of the grid-connected Distributed Energy Resource (DER) and the informational andinteractive smart grid. In this context, the cooperation dispatching of thepower grid including lots of DERs is a meaningful issue to eliminate theconflicts between the main grid and DERs (or consumers).It firstly aims on the optimization model and its solving method ofaggregating DERs based on the operation characteristics of virtual powerplant(VPP). A multi-objective optimization model with priority is proposed tospecify the technical and economic two-layer management mode of thevirtual power plant as the target functions with priority. The proposed modelis solved by decomposing it into two step optimization models based on thesatisfaction function. Then a joint optimization model of distribution networkincluding virtual power plants is established and solved by the approximatedynamic programming. With the joint optimization model, the networkperformance, such as frequency deviation, voltage deviation, power controlcan be improved because of the rapid control ability of virtual power plants.The second concerned point focuses on the cooperative dispatch of DERand the charging-discharging-storage integrative station(CDSIS) containingelectric vehicles (EV) and the second-use of electric vehicle batteries. Animproved dispatching model is proposed based on security constrainedoptimal power flow (SCOPF) to realize the temporal and spatial decoupling,with the power-transferring function of EVs and batteries. The improvedmodel firstly obtains the24-hour power outputs of traditional generators andthe total output of DERs and the CDSIS by treating them as a whole(represented by a VPP). Then it optimizes the DERs’ outputs considering thetopological structure of power grid to decouple the spatial correlation.Thirdly, an algorithm combining the interior point method and the update of the regulatory variables is proposed for the chance constrained stochasticoptimal power flow (S-OPF) model of power system including DERs. Theequivalent conditions of chance constraints are calculated based on cumulantsand Cornish-Fisher series. The regulatory variables are proposed to changethe chance constrained stochastic optimal power flow into a deterministicoptimal power flow (D-OPF) model. The update of the regulatory variablesis proposed to abtain iteratively the optimal solution of the S-OPF model. Theproposed method are tested on the IEEE14-bus and118-bus systems. Itsprecision, sensitivity and calculation are also analyzed in the case study.Lastly, a multi-controller interactive coordination scheduling model ispresented and applied to the dispatch problem of the power system includingmultiple renewable controllers. The optimal controller models the activeoptimization of the DERs’ controlling center. The traditional power griddispatch model is also considered as a controlling center. In the dispatchingmodel, a controlling center can plan its power output on its own according toits resources characteristics and multiple optimization objectives. Based onthe interactive coordination dispatching model, two optimal models(peer-to-peer and with priority) and two operating model (normal andabnormal) are also proposed. The detailed coordination rules are alsodesigned accordingly. The collaborative optimization method based onsatisfaction is applied to solve the proposed models. The coupling variable iscoordinated using collaborative optimization. The coordination of satisfactionis proposed to resolve the objective conflict between multiple optimal controlcenters in parallel optimization. The model and the method are validated inthe simulation example.更多还原