A digital twin of multiple energy hub systems with peer-to-peer energy sharing

Faced with the intricate interplay and uncertainty of future energy markets, an extensive digital twin (EXDT) is proposed to perform predictive testing and evaluate the performance of MESs. This EXDT provides energy system operators with insights into the coordinated behavior of interconnected EHs under various future scenarios, thus contributing to smarter decision-making processes. Specifically, an array of scenarios including different decision-making strategies and peer-to-peer energy sharing strategies were considered. For each of these scenarios, “what-if” tests were conducted using a multi-agent reinforcement learning (MARL)-based method to model the stochastic decision-making process of EHs belonging to different stakeholders with access to local information. Uncertainties during operation can be mitigated using Markov Game (MG) by capturing knowledge from historical energy data. Subsequently, the economic and technical performance were evaluated using multidimensional evaluation indexes."Numerical results and figures.xlsx" contains the numerical data for Figures 9 through 15 of the paper. It includes seven sheets, each providing data corresponding to these figures.In the "Fig. 9" sheet, the x-axis represents time (unit: h), while the y-axis shows the purchasing price of electricity from the upstream grid (unit: Yuan/(kW·h)).In the "Fig. 10" sheet, the x-axis represents the energy hubs (EHs), and the y-axis displays the costs of EHs for both no peer-to-peer sharing (CO) and with peer-to-peer sharing (COP) on a typical summer day (unit: 10,000 Yuan), the costs of EHs for both no peer-to-peer sharing (CO) and with peer-to-peer sharing (COP) on a typical winter day (unit: 10,000 Yuan)In the "Fig. 11" sheet, the x-axis represents time (unit: h), and the y-axis shows the energy exchange between the 4-energy hub (4-EH) system and the grid on a typical summer day (unit: kW·h) as well as on a typical winter day (unit: kW·h).In the "Fig. 12" sheet, the x-axis represents time (unit: h), and the y-axis provides the multi-energy flow dispatch in Energy Hub 1, including: photovoltaic generation (unit: kW), wind turbine generation (unit: kW), combined heat and power generation (unit: kW), energy storage charging/discharging (unit: kW), peer-to-peer energy sharing (unit: kW), energy exchange with the upstream grid (unit: kW), electricity purchase ratio from the upstream grid (unit: kW), cooling demand met by absorption chiller (unit: kW), cooling demand met by electrical chiller (unit: kW), cooling demand met by heat pump (unit: kW), cooling demand met by energy storage (unit: kW), and total cooling demand (unit: kW).In the "Fig. 13" sheet, the x-axis represents time (unit: h), and the y-axis provides the multi-energy flow dispatch in Energy Hub 2, including: photovoltaic generation (unit: kW), wind turbine generation (unit: kW), combined heat and power generation (unit: kW), energy storage charging/discharging (unit: kW), peer-to-peer energy sharing (unit: kW), energy exchange with the upstream grid (unit: kW), electricity purchase ratio from the upstream grid (unit: kW), cooling demand met by absorption chiller (unit: kW), cooling demand met by electrical chiller (unit: kW), cooling demand met by heat pump (unit: kW), cooling demand met by energy storage (unit: kW), and total cooling demand (unit: kW).In the "Fig. 14" sheet, the x-axis represents time (unit: h), and the y-axis provides the multi-energy flow dispatch in Energy Hub 3, including: photovoltaic generation (unit: kW), wind turbine generation (unit: kW), combined heat and power generation (unit: kW), energy storage charging/discharging (unit: kW), peer-to-peer energy sharing (unit: kW), energy exchange with the upstream grid (unit: kW), electricity purchase ratio from the upstream grid (unit: kW), cooling demand met by absorption chiller (unit: kW), cooling demand met by electrical chiller (unit: kW), cooling demand met by heat pump (unit: kW), cooling demand met by energy storage (unit: kW), and total cooling demand (unit: kW).In the "Fig. 15" sheet, the x-axis represents time (unit: h), and the y-axis provides the multi-energy flow dispatch in Energy Hub 4, including: photovoltaic generation (unit: kW), wind turbine generation (unit: kW), combined heat and power generation (unit: kW), energy storage charging/discharging (unit: kW), peer-to-peer energy sharing (unit: kW), energy exchange with the upstream grid (unit: kW), electricity purchase ratio from the upstream grid (unit: kW), cooling demand met by absorption chiller (unit: kW), cooling demand met by electrical chiller (unit: kW), cooling demand met by heat pump (unit: kW), cooling demand met by energy storage (unit: kW), and total cooling demand (unit: kW).

面对未来能源市场的复杂相互作用与不确定性，提出了一种广泛的数字孪生体（EXDT），以执行预测测试并评估微能源系统（MESs）的性能。该EXDT为能源系统运营商提供了洞察，揭示了在不同未来情景下互联能源枢纽（EHs）的协同行为，从而有助于决策过程的智能化。具体而言，考虑了一系列场景，包括不同的决策策略和点对点能源共享策略。对于这些场景中的每一个，都通过基于多智能体强化学习（MARL）的方法进行了“如果发生”测试，以模拟不同利益相关者拥有本地信息的情况下，能源枢纽的随机决策过程。利用马尔可夫博弈（MG）通过捕捉历史能源数据中的知识，可以减轻运营过程中的不确定性。随后，使用多维评估指数对经济和技术性能进行了评估。“数值结果和图表.xlsx”包含了论文第9至15图的数值数据。它包括七个工作表，每个工作表都提供了对应这些图表的数据。在“图9”工作表中，x轴代表时间（单位：小时），y轴显示从上游电网购买的电价（单位：元/（千瓦时））。在“图10”工作表中，x轴代表能源枢纽（EHs），y轴显示典型夏季和冬季典型日无点对点共享（CO）和有点对点共享（COP）的能源枢纽成本（单位：10,000元）。在“图11”工作表中，x轴代表时间（单位：小时），y轴显示典型夏季和冬季典型日4-EH系统与电网之间的能源交换量（单位：千瓦时）。在“图12”至“图15”工作表中，x轴代表时间（单位：小时），y轴提供了能源枢纽1至4的多能源流调度，包括：光伏发电（单位：千瓦）、风力涡轮机发电（单位：千瓦）、热电联产发电（单位：千瓦）、储能充电/放电（单位：千瓦）、点对点能源共享（单位：千瓦）、与上游电网的能源交换（单位：千瓦）、从上游电网购买的电比例（单位：千瓦）、由吸收式冷水机组满足的冷却需求（单位：千瓦）、由电动冷水机组满足的冷却需求（单位：千瓦）、由热泵满足的冷却需求（单位：千瓦）、由储能满足的冷却需求（单位：千瓦）以及总冷却需求（单位：千瓦）。