%0 Journal Article
%T 基于深度强化学习的电力网–信息网–路网耦合系统韧性分析<br>Resilience Analysis of Power-Information- Road Network Coupling System Based on Deep Reinforcement Learning
%A 罗浩
%A 张巍
%J Modeling and Simulation
%P 1460-1468
%@ 2324-870X
%D 2023
%I Hans Publishing
%R 10.12677/MOS.2023.122136
%X 韧性作为衡量系统对于极端灾害抵御能力的指标，对于电力网–信息网–路网三网耦合系统在极端灾害背景下的最大性能运行保证具有重要意义。本文基于耦合网络的交互机理，首先建立了网间关联的网络物理模型，结合极端灾害环境下各阶段的时间承接性，提出了对应于灾害发生时段的多时间尺度韧性提升策略。其次基于所求问题的非凸、非线性特性，建立了深度强化学习求解框架。最后基于IEEE 33节点系统搭建耦合系统算例，结果显示本文方法对于耦合系统的韧性提升显著。<br />
Resilience, as an index to measure the system’s ability to withstand extreme disasters, is of great significance to ensure the maximum performance of the coupling system of power network, infor-mation network and road network under the background of extreme disasters. In this paper, based on the interaction mechanism of coupled networks, a network physical model of inter-network cor-relation is first established. Combined with the time undertaking of each stage in extreme disaster environment, a multi-time scale toughness enhancement strategy corresponding to the disaster occurrence period is proposed. Secondly, based on the non-convex and nonlinear characteristics of the problem, a deep reinforcement learning solution framework is established. Finally, an example of coupling system is built based on IEEE 33 node system, and the results show that the proposed method can significantly improve the resilience of the coupling system.
%K 韧性，耦合系统，韧性提升，多时间尺度，深度强化学习；Resilience
%K Coupling Systems
%K Resilience Enhancement
%K Multiple Time Scales
%K Deep Reinforcement Learning
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=63038