微电网中激励型需求响应的分散协调凸优化

For a microgrid, which generally has uncertain distributed generation, relatively weak grid structure and high autonomy requirment, demand response presents a great potential to improve its reliability and operating economy and thus has important theoretical and practical significance.This project will focus on a new control strategy of incentive-based demand response for three- and single-phase hybrid microgrids.The basic idea is to formulate the design of demand response into a nonlinear constrained optimization problem in terms of the benefits of all participants and then to solve it with a decentralized and coordination convex optimization approach.The key scientific issues of the project include: the optimal power flow model of demand response and its "convexification" method, the conditions for conversion and exactness between the non-convex optimization problem and the convex optimization problem, the design of the decentralized coordinated optimization algorithm and the proof of its convergence and scalability.Compared with previous work, the striking advantages of the proposed new method are as follows: either the safety and economy of the microgrid or the comfortness of the customers is taken into account; globally optimized demand response strategy can be achieved with the convex approach; the decentralized and coordinated decision-making mechanism not only protects the data privacy of the various participants but also makes the method have excellent scalability.In the project, we will develop a wireless-communication based microgrid-DR test platform, on which the proposed control system is to be implemented and fully tested, thereby providing a demonstrative prototype system for further practical applications.

微电网面临分布式发电不确定性、电网结构相对薄弱性和较高自治性等挑战，使得采用需求响应(DR)来提高其供电可靠性和运行经济性具有重要理论价值和实际意义。本项目针对三、单相混合微电网提出一种新的激励型需求响应决策方法，将需求响应策略设计规范为一个兼顾各方利益的非线性约束优化潮流问题，并采用分散-协调凸优化方法来实现互动求解。其中关键性科学问题包括：微网需求响应的优化潮流模型及其“凸化”方法、非凸优化问题与凸优化问题之间转换和一致性的条件、分散-协调优化算法的设计及其收敛性与可扩展性证明。与以往方法相比，新方法的突出优点是：兼顾了微电网的安全经济性与用户舒适性，借助凸优化实现了需求响应策略的全局最优性，采用的分散-协调决策机制既保护了各方隐私权又具有良好的可扩展性。项目还将研发基于无线通信的微网-DR测试平台，在其上实施并测试理论研究成果，进而为成果实用化提供示范原型系统。

项目围绕微电网中需求响应（Demand Response，DR）优化控制和紧急需求响应提升短期电压和频率稳定性等开展研究。通过理论分析、模型构建、数值仿真以及工程实践，在系统建模、优化方法和控制策略等方面取得重要进展。具体包括：.1）在系统建模方面：建立了适用于微网DR分析与控制设计的并网和孤岛微电网动态模型，以及基于紧急DR的电压稳定预防控制模型、受端电网频率响应模型和短期频率稳定控制模型。.2）在优化方法方面：将非线性约束优化潮流模型进行松弛，构建凸优化问题，将凸优化问题进一步分解，提出了分散-协调优化方法。.3）在控制策略上：考虑分散-集中的协调控制，将需求响应的分布式控制与控制中心的在线集中优化控制相结合，更加经济有效地实现控制目标；提出了微电网能量管理优化策略、短期频率稳定性优化控制、分布式预防电压控制等控制策略。.4）发表学术论文13篇，其中SCI期刊论文7篇，EI检索论文6篇（不含SCI检索），获得中国石油工程建设协会科技进步一等奖1项、获得9项授权国家发明专利。.基于项目成果研发了微电网的DR优化控制与能量管理系统，并先后应用于广东某岛上微电网、渤海某海上油田群电网和某火电厂电池储能系统工程中，解决了实际工程问题，取得了显著的经济效益。