基于运行风险分析和协调优化调度的极端气象灾害电力系统防灾调度问题研究

项目名称： 基于运行风险分析和协调优化调度的极端气象灾害电力系统防灾调度问题研究

项目编号： No.51277160

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 电工技术

项目作者： 汪震

作者单位： 浙江大学

项目金额： 76万元

中文摘要： 近年来，台风、冻雨等极端气象灾害在国内的频繁发生使基于定量分析的防灾调度研究变得非常紧迫。本项目致力于从运行风险控制的角度实现多种调控手段的协调优化，力求建立灾害条件下的多时间尺度协同优化调度机制。主要研究内容为：提出基于逻辑-物理互补模型的输电设备运行可靠性综合评估方法，综合考虑多种影响因素，同时探索ELM快速在线学习技术用于气象-设备逻辑建模；分析考虑安全风险约束和负荷预测误差的随机机组组合优化模型和随机最优输电网切换模型，研究基于场景树的分解算法和安全风险约束降维策略；研究风险共享的预防控制和紧急控制协调双层优化模型及其分层分解算法，研究考虑场景概率的电力系统低频减载主/辅控制协调设计方法，探索异步分布式技术在其中的应用。此项目的研究成果有助于提高极端气象灾害条件下的电力系统应急管理水平，降低气象灾害给电力系统安全运行带来的风险，健全电力系统停电防御体系。

中文关键词： 气象灾害；运行风险分析；协调优化调度；分布鲁棒优化；随机规划

英文摘要： In recent years, extreme adverse weathers such as typhoon and ice rain occur frequently in China, which have activated the imperative study on disaster-relief dispatch in power systems based on quantitative analysis methods. The project devotes to the coordinated optimization of multiple dispatch and control measures for the operational risk control purpose and construction of a collaborative dispatch mechanism with different time scale under extreme adverse weather conditions. The project will cover the following aspects: 1) A synthesis operational reliability evaluation method for transmission equipments based on a logical and physical complementary model is proposed and the ELM on-line study technique is introduced for the logical modeling; 2) Two types of stochastic programming problems are studied, the stochastic unit commitment (SUC) and stochastic optimal transmission switching (SOTS), both of which will consider the security risk constraints and load forecast errors (uncertainty). The scenario-tree decomposition algorithm for the SUC and SOTS problems and dimension reduction strategy of the security risk constraints will be elaborately explored; 3) A risk-share coordinated bi-level optimization model of preventive and emergency control is proposed and a hierarchical decomposition algorithms is developed

英文关键词： extreme adverse weather conditions；operational risk analysis；coordinated optimal dispatch；distributionally robust optimization；stochastic programming