考虑随机过程的虚拟发电厂序贯优化调度研究

Virtual Power Plant (VPP), viewed as a controllable stand-alone power plant, organically integrates a large number of distributed units, such as distributed generations, energy storage devices and loads via advanced communication technologies and control strategies. The effectively coordinated control between distribution units would be the accomplished fundament of VPP optimal dispatch. This project adopts the structure advance successively by distributed unit modeling, coordinated control strategy and optimal dispatch, including:①establishing the stochastic process mathematical models of uncertainty distributed units, which have finite complexities and applicable for coordinated control and optimal dispatch;②studying the coordinated control patterns between distributed units based on stochastic process, and deducing the VPP coordinated control domain and the controllable state differential equations in multiple random environment;③establishing the VPP sequential optimal dispatch mathematical models with dynamic controllable parameters and differential - algebraic equations in continuous time domain, and considering VPP multiple external characteristic objectives of controllable generation plants (such as costs/capacity/rate), further proposing the solution structures based on quasi-sequential decomposition and interior point method (IPM) blended hybrid with intelligent algorithm. The feasibility of proposed mathematical models of distributed units、coordinated control strategies and dispatch optimization can be tested and verified by the VPP cyber-physical simulation platform. Research achievements of this project would lay the significant theoretical foundation and provide technical support for VPP coordinated and optimal operation.

虚拟发电厂将多个分布式电源、储能装置和负荷等分布式单元通过先进的通信技术和控制策略有机地集成为可控的独立发电厂。分布式单元的有效协调控制是虚拟发电厂优化调度的实现基础。本项目从分布式单元建模、协调控制策略和优化调度三个递进层次，研究：①建立有限复杂度、适用于协调控制和发电优化调度的不确定性分布式单元随机过程模型；②研究考虑随机过程的分布式单元协调控制模式，推导虚拟发电厂多重随机环境下的协调控制域以及可控状态微分方程；③建立连续时间域、动态可控参数、含微分-代数方程的虚拟发电厂序贯优化调度模型，考虑成本、容量、速率等可控发电厂外特性多维目标，构建拟序贯分解、结合内点法与混合智能计算的优化求解算法结构。依托虚拟发电厂信息物理融合仿真平台对分布式单元建模、分布式协调控制、优化调度策略进行试验验证。本项目研究成果将为虚拟发电厂协调控制与优化运行提供重要的理论基础和技术支持。

虚拟发电厂为分布式电源、储能和负荷等分布式单元提供了一个调度和管理平台，基于先进的通信技术和分布式控制技术，虚拟发电厂可以提高分布式电源的消纳，并为电网提供可控的电能和辅助服务。本项目从分布式单元建模、协调控制策略、优化调度、分布式调频四个递进层次，研究了基于分布式理论的虚拟发电厂优化调度：（1）分析了虚拟发电厂内部单元的发电特性和外特性；采用Copula理论研究了虚拟发电厂出力的随机性；（2）基于分布式理论及随机性，提出了虚拟发电厂分布式分层随机优化调度策略。考虑虚拟发电厂出力不确定性，结合虚拟发电厂分布式分层控制特点，建立了虚拟发电厂的随机优化模型；利用虚拟发电厂分布式分层优化算法对模型进行求解，并结合算例进行仿真与分析，结果表明了该模型与方法的有效性；（3）基于分布式理论及随机性，提出了一种虚拟发电厂分布式调频策略。分析了虚拟发电厂的调频性能，并基于通信技术建立了虚拟发电厂的调频架构；结合一致性协议和分布式牵引控制方法，提出了虚拟发电厂分布式调频策略，结合算例验证了其有效性；（4）设计了面向虚拟发电厂的电力-通信联合仿真平台。分析了通信环节在虚拟发电厂电力业务中的突出地位，提出了联合仿真平台的总体架构，并通过对MATLAB与OPNET的接口设计以及二者之间数据交互的协调控制，完成了电力-通信联合仿真平台设计。本项目研究成果为虚拟发电厂协调控制与优化运行提供了重要的理论基础和技术支持。