基于并行多维近似动态规划的综合能源系统动态经济调度  

Dynamic Economic Dispatch for Integrated Energy System Based on Parallel Multi-dimensional Approximate Dynamic Programming

作　　者：张玉敏 孙鹏凯 吉兴全 杨明[2] 于一潇 张旋 ZHANG Yumin;SUN Pengkai;JI Xingquan;YANG Ming;YU Yixiao;ZHANG Xuan(College of Electrical Engineering and Automation,Shandong University of Science and Technology,Qingdao 266590,China;Key Laboratory of Power System Intelligent Dispatch and Control of Ministry of Education(Shandong University),Jinan 250061,China;Key Laboratory of Smart Grid of Ministry of Education(Tianjin University),Tianjin 300072,China)

机构地区：[1]山东科技大学电气与自动化工程学院,山东省青岛市266590 [2]电网智能化调度与控制教育部重点实验室(山东大学),山东省济南市250061 [3]智能电网教育部重点实验室(天津大学),天津市300072

出　　处：《电力系统自动化》2023年第4期60-68,共9页Automation of Electric Power Systems

基　　金：国家自然科学基金资助项目(52107111);山东省自然科学基金资助项目(ZR2022ME219);山东省自然科学基金青年基金资助项目(ZR2021QE117)。

摘　　要：针对随机动态经济调度模型难以高效求解的问题,提出一种用于快速求解电-气-热综合能源系统(IES)随机动态经济调度问题的并行多维近似动态规划算法。该算法将原有的高维状态空间中的状态变量聚合到IES的电储能系统(ESS)的可用容量和蓄热罐(HST)的可用热量中,实现了状态空间的降维,进而解决了动态规划中“维数灾”的问题。此外,利用ESS和HST的运行约束,将无效状态剔除,降低了问题的求解规模。在风电、电价和负荷等不确定场景充分训练后的值表模型中,通过近似值函数的迭代寻优,将含有经验知识的值表应用于在线测试中,实现应对不确定性的IES最优调度决策。以改进的2个系统为例,验证了所提模型和算法的有效性。With respect to the problem that the stochastic dynamic economic dispatch model is difficult to solve efficiently, this paper proposes a parallel multi-dimensional approximate dynamic programming algorithm for solving the stochastic dynamic economic dispatch rapidly for electricity-gas-heat integrated energy system(IES). This algorithm can aggregate the state variables in the original high-dimensional state space into the available capacity of the electrical storage system(ESS), and the available heat of the heat storage tank(HST) of the IES to achieve dimensionality reduction of the state space and solve the problem of “dimensional disaster” in the dynamic dispatch. In addition, this paper uses the operation constraints of ESS and HST to eliminate the invalid states and reduce the solving scale of the problem. In the fully trained value table model for uncertain scenarios such as wind power, electricity price and load, through the iterative optimization of the approximate value function, the value table containing empirical knowledge is employed to the online test to make the IES optimal dispatch decision deal with uncertainty. The effectiveness of the proposed model and algorithm is verified by taking the improved two power systems as examples.

关 键 词：综合能源系统 经济调度 近似动态规划 随机优化 状态空间压缩 

分 类 号：TK01[动力工程及工程热物理] TM73[电气工程—电力系统及自动化]