首页|基于深度强化学习的综合能源系统低碳经济调度

基于深度强化学习的综合能源系统低碳经济调度

扫码查看
综合能源系统(IES)能够实现多种能源形式的供应,但同时排放的大量CO2 也影响着周边环境.针对综合能源系统的低碳经济调度问题,本文提出一种基于双延迟深度确定性策略梯度(TD3)算法的优化调度策略.首先,以调度运行成本最小为目标函数,建立考虑碳捕集技术和电转气技术的包含电、热、冷多能互补的综合能源系统模型;其次,引入碳交易机制,提高优化调度策略节能减排的积极性;然后,根据强化学习框架设计优化模型的状态空间、动作空间和奖励函数等,利用TD3 算法中的智能体与环境互动,学习综合能源系统的运行策略;最后,利用历史数据对TD3 算法的智能体进行训练,并对比线性规划和粒子群算法在不同场景下进行算例分析.结果表明,本文所提方法可以减少综合能源系统运行时的碳排放和运行成本,能够实现综合能源系统的低碳经济调度.
Low-carbon economic dispatch of integrated energy system based on deep reinforcement learning
Integrated energy system(IES)enables the supply of multiple forms of energy,but the large amount of carbon dioxide it emitted affects the surrounding environment.Here,an optimal scheduling approach based on Twin Delayed Deep Deterministic Policy Gradient(TD3)is proposed for low-carbon economic scheduling of IES.First,taking the minimum operation cost as the objective function,an IES model with multiple complementary energies of electricity,heat and cold is established considering carbon capture technology and power-to-gas technology.Second,a carbon trading mechanism is introduced to stimulate the enthusiasm of energy conservation and emission reduction under optimal scheduling.Then,according to the reinforcement learning framework,the state space,action space and reward function of the optimization model are designed,and the agents in the TD3 algorithm are used to interact with the environment to explore strategies and learn the IES operation strategies.Finally,the historical data are used to train the agents of TD3 algorithm,and the linear programming and particle swarm optimization are compared under different scenarios.The results show that the proposed approach can reduce the IES carbon emission and operating cost,thus realizing the low-carbon economic dispatch of the integrated energy system.

carbon capture system(CCS)power-to-gas(P2G)deep reinforcement learningtwin delayed deep deterministic policy gradient(TD3)integrated energy system(IES)carbon trading mechanism

崔在兴、应雨龙、李靖超、王新友

展开 >

上海电力大学 能源与机械工程学院,上海,200090

上海电机学院 电子信息学院,上海,201306

碳捕获系统 电转气 深度强化学习 双延迟深度确定性策略梯度 综合能源系统 碳交易机制

国家自然科学基金上海市自然科学基金上海市"科技创新行动计划"启明星项目

6207616021ZR142470023QA1403800

2024

南京信息工程大学学报
南京信息工程大学

南京信息工程大学学报

CSTPCD北大核心
影响因子:0.737
ISSN:1674-7070
年,卷(期):2024.16(5)