Sustainable Energy Grids & Networks最新文献_第2页

Enhancing resilience and cost efficiency in multi-microgrids through peer-to-peer energy trading and decentralized energy management systems 通过点对点能源交易和分散式能源管理系统，提高多微电网的弹性和成本效率

IF 5.6 2区工程技术 Q2 ENERGY & FUELS

Sustainable Energy Grids & Networks

Pub Date : 2026-01-02 DOI: 10.1016/j.segan.2025.102116

Wei Yu, Wenjian Wang

The study proposes a two-stage decentralized energy management system for multi-microgrids, includes peer-to-peer trading and resilient network reconfiguration. Physical constraints such as radiality, voltage stability, and line capacity are implemented in first stage to ensure safe operation in both normal and faulty conditions. In second stage, market-clearing mechanism enables supply-demand bidding and zonal cost allocation for multi-bilateral trades while maintains grid and marginal pricing. Renewable uncertainty is modeled using an upper-quantile approach from historical data to balance robustness and economic efficiency without probability distributions. Financial incentives are used in incentive-based demand response following faults to shift or curtailment loads. The model is implemented on a modified IEEE-33 bus system with three microgrids and 36 residential households equipped with photovoltaic panels, wind turbines, battery energy storage, and electric vehicles. Simulation results show that under normal conditions, grid reliance for residential loads is reduced by 43.66 %. During grid and line outages, demand falls by 16.33 % while grid usage increases modestly to 53.16 %. When distributed generators fail, peer to peer energy sharing within the faulty zone rises by a factor of 2.5, supported by battery and electrical vehicle discharges. The proposed framework thus enhances resilience, lowers operating costs, and strengthens local energy self-sufficiency through coordinated P2P trading, flexible storage, and fault-tolerant scheduling.

该研究提出了一种针对多微电网的两阶段分散式能源管理系统，包括点对点交易和弹性网络重构。在第一阶段实施了诸如径向、电压稳定性和线路容量等物理约束，以确保在正常和故障条件下的安全运行。第二阶段，市场出清机制在维持电网和边际定价的同时，实现了多双边贸易的供需竞价和区域成本分摊。可再生不确定性采用历史数据的上分位数方法建模，以平衡鲁棒性和经济效率，而不需要概率分布。财政激励用于基于激励的需求响应，在故障后转移或削减负荷。该模型在改进的IEEE-33总线系统上实施，该系统有三个微电网和36个配备光伏板、风力涡轮机、电池储能和电动汽车的住宅家庭。仿真结果表明，在正常情况下，住宅负荷的电网依赖降低了43.66%。在电网和线路中断期间，需求下降16.33%，而电网使用率小幅上升至53.16%。当分布式发电机发生故障时，故障区域内的点对点能量共享增加了2.5倍，由电池和电动汽车放电支持。因此，该框架通过协调P2P交易、灵活存储和容错调度，增强了弹性，降低了运营成本，并增强了本地能源自给自足。

{"title":"Enhancing resilience and cost efficiency in multi-microgrids through peer-to-peer energy trading and decentralized energy management systems","authors":"Wei Yu, Wenjian Wang","doi":"10.1016/j.segan.2025.102116","DOIUrl":"10.1016/j.segan.2025.102116","url":null,"abstract":"<div><div>The study proposes a two-stage decentralized energy management system for multi-microgrids, includes peer-to-peer trading and resilient network reconfiguration. Physical constraints such as radiality, voltage stability, and line capacity are implemented in first stage to ensure safe operation in both normal and faulty conditions. In second stage, market-clearing mechanism enables supply-demand bidding and zonal cost allocation for multi-bilateral trades while maintains grid and marginal pricing. Renewable uncertainty is modeled using an upper-quantile approach from historical data to balance robustness and economic efficiency without probability distributions. Financial incentives are used in incentive-based demand response following faults to shift or curtailment loads. The model is implemented on a modified IEEE-33 bus system with three microgrids and 36 residential households equipped with photovoltaic panels, wind turbines, battery energy storage, and electric vehicles. Simulation results show that under normal conditions, grid reliance for residential loads is reduced by 43.66 %. During grid and line outages, demand falls by 16.33 % while grid usage increases modestly to 53.16 %. When distributed generators fail, peer to peer energy sharing within the faulty zone rises by a factor of 2.5, supported by battery and electrical vehicle discharges. The proposed framework thus enhances resilience, lowers operating costs, and strengthens local energy self-sufficiency through coordinated P2P trading, flexible storage, and fault-tolerant scheduling.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102116"},"PeriodicalIF":5.6,"publicationDate":"2026-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145926049","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An optimal peer-to-peer market in energy communities: A game-theoretic approach with replicator dynamics 能源社区中最优点对点市场：具有复制因子动力学的博弈论方法

IF 5.6 2区工程技术 Q2 ENERGY & FUELS

Sustainable Energy Grids & Networks

Pub Date : 2025-12-31 DOI: 10.1016/j.segan.2025.102114

Sofía Chacón , Katerine Guerrero , Germán Obando , Andrés Pantoja

Energy communities (ECs) enable prosumers, consumers, and distributed energy resources (DERs) to jointly manage energy in a coordinated and economically efficient manner. In this work, we propose an energy management system (EMS) for ECs that integrates a demand response (DR) program with a peer-to-peer (P2P) market based on sealed-bid auctions and continuous Stackelberg dynamics. The buyers determine prices according to their energy demand and risk aversion, and generators decide on the amount of energy to sell based on the rewards received and their associated costs. Methodologically, we develop three algorithms to maximize the welfare of the community. The first algorithm incorporates a DR program and generation constraints to keep the EC competitive with grid prices over time. The second and third algorithms use replicator dynamics (RD) to find equilibria that optimize the system’s welfare, using Lagrangian relaxation (LR) to handle the model constraints. We integrate the models for sellers and buyers via a system of differential equations that simulate a Stackelberg game. Additionally, a filtering mechanism is employed to improve convergence and reduce computation time. We validate the EMS in a case study, showing that the proposed approach achieves greater self-sufficiency compared to a system without demand response and enables better resource management, enhanced fairness, and a more equitable distribution of benefits compared to a non-hierarchical and decoupled model.

能源共同体（ec）使产消者、消费者和分布式能源者（der）能够以协调和经济高效的方式共同管理能源。在这项工作中，我们为ec提出了一种能源管理系统（EMS），该系统将需求响应（DR）计划与基于密封投标拍卖和连续Stackelberg动态的点对点（P2P）市场集成在一起。买家根据他们的能源需求和风险厌恶来决定价格，而发电商则根据获得的回报和相关成本来决定出售的能源数量。在方法上，我们开发了三种算法来最大化社区的福利。第一种算法结合了DR程序和发电限制，以保持EC随着时间的推移与电网价格具有竞争力。第二和第三种算法使用复制因子动力学（RD）来找到优化系统福利的平衡点，使用拉格朗日松弛（LR）来处理模型约束。我们通过一个模拟Stackelberg博弈的微分方程系统来整合卖方和买方的模型。此外，采用滤波机制提高收敛性，减少计算时间。我们在一个案例研究中验证了EMS，表明与没有需求响应的系统相比，所提出的方法实现了更大的自给自足，并且与非分层和解耦模型相比，可以实现更好的资源管理，增强公平性和更公平的利益分配。

{"title":"An optimal peer-to-peer market in energy communities: A game-theoretic approach with replicator dynamics","authors":"Sofía Chacón , Katerine Guerrero , Germán Obando , Andrés Pantoja","doi":"10.1016/j.segan.2025.102114","DOIUrl":"10.1016/j.segan.2025.102114","url":null,"abstract":"<div><div>Energy communities (ECs) enable prosumers, consumers, and distributed energy resources (DERs) to jointly manage energy in a coordinated and economically efficient manner. In this work, we propose an energy management system (EMS) for ECs that integrates a demand response (DR) program with a peer-to-peer (P2P) market based on sealed-bid auctions and continuous Stackelberg dynamics. The buyers determine prices according to their energy demand and risk aversion, and generators decide on the amount of energy to sell based on the rewards received and their associated costs. Methodologically, we develop three algorithms to maximize the welfare of the community. The first algorithm incorporates a DR program and generation constraints to keep the EC competitive with grid prices over time. The second and third algorithms use replicator dynamics (RD) to find equilibria that optimize the system’s welfare, using Lagrangian relaxation (LR) to handle the model constraints. We integrate the models for sellers and buyers via a system of differential equations that simulate a Stackelberg game. Additionally, a filtering mechanism is employed to improve convergence and reduce computation time. We validate the EMS in a case study, showing that the proposed approach achieves greater self-sufficiency compared to a system without demand response and enables better resource management, enhanced fairness, and a more equitable distribution of benefits compared to a non-hierarchical and decoupled model.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102114"},"PeriodicalIF":5.6,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145926050","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Risk-based industrial load management with integrated distributed energy resources to enhance grid flexibility and market participation 基于风险的工业负荷管理与集成分布式能源资源，以提高电网灵活性和市场参与

IF 5.6 2区工程技术 Q2 ENERGY & FUELS

Sustainable Energy Grids & Networks

Pub Date : 2025-12-30 DOI: 10.1016/j.segan.2025.102115

Nasim Eslaminia , Sina Ghaemi , Amjad Anvari-Moghaddam

Industrial loads, with their inherent flexibility and growing integration of energy storage systems (ESSs) and photovoltaic (PV) generation, present a significant opportunity to enhance grid stability through ancillary services (ASs) and unlock financial benefits. However, uncertainties in renewable generation and market prices can complicate decision-making, particularly for risk-averse industrial entities. To address these challenges, this study develops a risk-based, mixed-integer linear programming (MILP) optimization model for day-ahead scheduling. This model leverages flexible assets and incorporates the Conditional Value-at-Risk (CVaR) technique to evaluate how risk preferences impact participation in energy and AS markets. Moreover, the proposed model integrates a detailed characterization of industrial sub-loads, assesses load flexibility, and accounts for the combined effects of renewable energy sources (RESs) and ESSs. Four case studies are used to analyze the participation of industrial loads, PV, and ESS in AS markets, investigating the influence of risk preferences and AS participation strategies. The case studies demonstrate that engaging in AS markets yields financial gains regardless of risk preference. The results emphasize the critical role of flexible assets in enhancing system flexibility, promoting greater involvement in energy and AS markets, and improving grid support capabilities.

工业负荷具有固有的灵活性，并且越来越多地集成了储能系统（ess）和光伏发电（PV），这为通过辅助服务（ASs）提高电网稳定性和释放经济效益提供了一个重要的机会。然而，可再生能源发电和市场价格的不确定性可能使决策复杂化，特别是对于厌恶风险的工业实体。为了解决这些挑战，本研究开发了一种基于风险的混合整数线性规划（MILP）优化模型，用于日前调度。该模型利用灵活资产并结合条件风险价值（CVaR）技术来评估风险偏好如何影响能源和AS市场的参与。此外，所提出的模型集成了工业子负荷的详细特征，评估了负荷灵活性，并考虑了可再生能源（RESs）和ess的综合影响。本文采用四个案例研究分析了工业负荷、光伏和ESS在AS市场中的参与情况，调查了风险偏好和AS参与策略的影响。案例研究表明，无论风险偏好如何，参与AS市场都能产生财务收益。研究结果强调了灵活资产在增强系统灵活性、促进更多地参与能源和AS市场以及提高电网支持能力方面的关键作用。

{"title":"Risk-based industrial load management with integrated distributed energy resources to enhance grid flexibility and market participation","authors":"Nasim Eslaminia , Sina Ghaemi , Amjad Anvari-Moghaddam","doi":"10.1016/j.segan.2025.102115","DOIUrl":"10.1016/j.segan.2025.102115","url":null,"abstract":"<div><div>Industrial loads, with their inherent flexibility and growing integration of energy storage systems (ESSs) and photovoltaic (PV) generation, present a significant opportunity to enhance grid stability through ancillary services (ASs) and unlock financial benefits. However, uncertainties in renewable generation and market prices can complicate decision-making, particularly for risk-averse industrial entities. To address these challenges, this study develops a risk-based, mixed-integer linear programming (MILP) optimization model for day-ahead scheduling. This model leverages flexible assets and incorporates the Conditional Value-at-Risk (CVaR) technique to evaluate how risk preferences impact participation in energy and AS markets. Moreover, the proposed model integrates a detailed characterization of industrial sub-loads, assesses load flexibility, and accounts for the combined effects of renewable energy sources (RESs) and ESSs. Four case studies are used to analyze the participation of industrial loads, PV, and ESS in AS markets, investigating the influence of risk preferences and AS participation strategies. The case studies demonstrate that engaging in AS markets yields financial gains regardless of risk preference. The results emphasize the critical role of flexible assets in enhancing system flexibility, promoting greater involvement in energy and AS markets, and improving grid support capabilities.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102115"},"PeriodicalIF":5.6,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884020","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Prosumers' peer-to-peer multi-energy transactions considering distributed energy resources and demand response 考虑分布式能源和需求响应的产消点多能源交易

IF 5.6 2区工程技术 Q2 ENERGY & FUELS

Sustainable Energy Grids & Networks

Pub Date : 2025-12-30 DOI: 10.1016/j.segan.2025.102113

Mohammad Bagher Moradi , Mohammad Hassan Nazari , Hamed Nafisi , Hossein Askarian Abyaneh , Seyed Hossein Hosseinian , Marco Merlo

Smart grids support the integration of renewable resources and enable demand response, transforming consumers into prosumers who both generate and use energy. Prosumers can improve their economic outcomes by selling surplus energy through peer‑to‑peer (P2P) transactions instead of relying solely on the upstream grid. In residential microgrids, this can reduce energy costs while increasing revenues from surplus energy sales. This study investigates P2P energy sharing as a mechanism for energy exchange among prosumers and examines how different optimization objectives affect individual benefits in smart grid environments. Two primary objectives are considered: maximizing revenues from energy sales and minimizing energy procurement costs. The approach determines transactive energy through prosumer self‑scheduling and formulates a P2P model that maximizes social welfare. The objective functions are assessed under three scenarios: P2P energy sharing within a resource‑constrained smart grid, an expanded‑resource setting that evaluates the influence of additional prosumer capacity, and a multi‑energy hub in which participants can trade both electrical and thermal energy. Mixed‑integer linear programming simulations are carried out under two pricing schemes, with and without differentiation between renewable and conventional energy prices, and are complemented by a demand sensitivity analysis. The results indicate that prosumers prioritizing cost minimization achieve substantially lower energy expenses, with reductions between 2.2 % and 67.8 % compared with prosumers focused on revenue maximization. Furthermore, increased prosumer resources and energy price variations significantly affect profitability under each objective. Appropriate adjustments to resources and prices can enhance profits when energy sales are prioritized over cost reduction, as confirmed by sensitivity analysis and comparison with prior work.

智能电网支持可再生资源的整合，实现需求响应，将消费者转变为生产和使用能源的生产消费者。生产消费者可以通过点对点（P2P）交易出售剩余能源，而不是仅仅依赖上游电网，从而提高经济效益。在住宅微电网中，这可以降低能源成本，同时增加剩余能源销售的收入。本研究探讨了P2P能源共享作为产消者之间能源交换的机制，并考察了智能电网环境中不同的优化目标如何影响个人利益。考虑两个主要目标：能源销售收入最大化和能源采购成本最小化。该方法通过产消自我调度来确定交易能量，并构建了社会福利最大化的P2P模型。目标函数在三种情况下进行评估：资源受限的智能电网中的P2P能源共享，评估额外产消能力影响的扩展资源设置，以及参与者可以交易电能和热能的多能中心。混合整数线性规划模拟是在两种定价方案下进行的，可再生能源和传统能源价格有和没有差别，并辅以需求敏感性分析。结果表明，优先考虑成本最小化的生产消费者实现了显著降低的能源支出，与专注于收入最大化的生产消费者相比，减少了2.2% %至67.8% %。此外，生产消费者资源的增加和能源价格的变化显著影响了每个目标下的盈利能力。敏感度分析和与之前工作的比较证实，当能源销售优先于降低成本时，适当调整资源和价格可以提高利润。

{"title":"Prosumers' peer-to-peer multi-energy transactions considering distributed energy resources and demand response","authors":"Mohammad Bagher Moradi , Mohammad Hassan Nazari , Hamed Nafisi , Hossein Askarian Abyaneh , Seyed Hossein Hosseinian , Marco Merlo","doi":"10.1016/j.segan.2025.102113","DOIUrl":"10.1016/j.segan.2025.102113","url":null,"abstract":"<div><div>Smart grids support the integration of renewable resources and enable demand response, transforming consumers into prosumers who both generate and use energy. Prosumers can improve their economic outcomes by selling surplus energy through peer‑to‑peer (P2P) transactions instead of relying solely on the upstream grid. In residential microgrids, this can reduce energy costs while increasing revenues from surplus energy sales. This study investigates P2P energy sharing as a mechanism for energy exchange among prosumers and examines how different optimization objectives affect individual benefits in smart grid environments. Two primary objectives are considered: maximizing revenues from energy sales and minimizing energy procurement costs. The approach determines transactive energy through prosumer self‑scheduling and formulates a P2P model that maximizes social welfare. The objective functions are assessed under three scenarios: P2P energy sharing within a resource‑constrained smart grid, an expanded‑resource setting that evaluates the influence of additional prosumer capacity, and a multi‑energy hub in which participants can trade both electrical and thermal energy. Mixed‑integer linear programming simulations are carried out under two pricing schemes, with and without differentiation between renewable and conventional energy prices, and are complemented by a demand sensitivity analysis. The results indicate that prosumers prioritizing cost minimization achieve substantially lower energy expenses, with reductions between 2.2 % and 67.8 % compared with prosumers focused on revenue maximization. Furthermore, increased prosumer resources and energy price variations significantly affect profitability under each objective. Appropriate adjustments to resources and prices can enhance profits when energy sales are prioritized over cost reduction, as confirmed by sensitivity analysis and comparison with prior work.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102113"},"PeriodicalIF":5.6,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925938","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Addressing system strength and reliability concerns in renewable energy-based weak grids using synchronous condensers determined by hybrid GRU-classical optimization method 采用gru -经典混合优化方法求解基于可再生能源的弱电网中同步冷凝器的系统强度和可靠性问题

IF 5.6 2区工程技术 Q2 ENERGY & FUELS

Sustainable Energy Grids & Networks

Pub Date : 2025-12-29 DOI: 10.1016/j.segan.2025.102111

Md Ohirul Qays , Iftekhar Ahmad , Daryoush Habibi , Mohammad A.S. Masoum , Thair Mahmoud

Owing to the higher-integration of renewable energy generators (REGs), conventional coal-based synchronous generators are being decommissioned from generation fleets, resulting in system strength and reliability concerns. Along with the increasing load demand, deficiency of system strength can be a huge risk to system stability and can eventually lead to blackouts by disconnecting REGs from grid systems. In the literature, researchers and power engineers have proposed to deploy synchronous condensers (SynCons) as a mitigation strategy to address the system strength and reliability challenges. SynCons are, however, expensive and require investigation for higher reliability results before installation. To address the concerns, SynCons’ optimal sizes, placement and reliability assessment are investigated in this paper. The proposed solution is achieved by modeling an optimization problem and retaining SynCons-related costs low while maintaining short circuit ratio and minimizing loss of load probability, measurement indexes of system strength and reliability analysis of a grid above a satisfactory level respectively. A hybrid data-driven gated recurrent unit (GRU)-classical optimization framework is developed for data processing and achieving the optimization results. The implemented learning model is capable of achieving higher accuracy 99.691 % and lower computation time 0.023 sec when compared with the existing learning models. Additionally, the obtained results, such as transient stability and economic analysis of SynCons-conducted weak-grid present that the proposed solution can significantly perform 21.581 % cost minimization and 6.391 % reliability enhancement.

由于可再生能源发电机（REGs）的高度集成化，传统的煤基同步发电机正在从发电机组中退役，导致系统强度和可靠性问题。随着负荷需求的增加，系统强度不足会给系统稳定性带来巨大风险，并最终导致reg与电网系统断开连接而导致停电。在文献中，研究人员和电力工程师建议部署同步冷凝器（SynCons）作为解决系统强度和可靠性挑战的缓解策略。然而，SynCons价格昂贵，并且在安装之前需要进行调查以获得更高的可靠性结果。为了解决这些问题，本文研究了SynCons的最优尺寸、放置和可靠性评估。该方案通过对优化问题进行建模，在保持较低的syncon相关成本的同时，使短路率和负荷损失率、系统强度测量指标和电网可靠性分析指标分别保持在满意水平以上。提出了一种混合数据驱动门控循环单元(GRU)-经典优化框架，用于数据处理并实现优化结果。与现有学习模型相比，所实现的学习模型的准确率提高了99.691 %，计算时间降低了0.023 秒。此外，对syncon系统的暂态稳定性和经济分析结果表明，该方案可显著降低21.581 %的成本，提高6.391 %的可靠性。

{"title":"Addressing system strength and reliability concerns in renewable energy-based weak grids using synchronous condensers determined by hybrid GRU-classical optimization method","authors":"Md Ohirul Qays , Iftekhar Ahmad , Daryoush Habibi , Mohammad A.S. Masoum , Thair Mahmoud","doi":"10.1016/j.segan.2025.102111","DOIUrl":"10.1016/j.segan.2025.102111","url":null,"abstract":"<div><div>Owing to the higher-integration of renewable energy generators (REGs), conventional coal-based synchronous generators are being decommissioned from generation fleets, resulting in system strength and reliability concerns. Along with the increasing load demand, deficiency of system strength can be a huge risk to system stability and can eventually lead to blackouts by disconnecting REGs from grid systems. In the literature, researchers and power engineers have proposed to deploy synchronous condensers (SynCons) as a mitigation strategy to address the system strength and reliability challenges. SynCons are, however, expensive and require investigation for higher reliability results before installation. To address the concerns, SynCons’ optimal sizes, placement and reliability assessment are investigated in this paper. The proposed solution is achieved by modeling an optimization problem and retaining SynCons-related costs low while maintaining short circuit ratio and minimizing loss of load probability, measurement indexes of system strength and reliability analysis of a grid above a satisfactory level respectively. A hybrid data-driven gated recurrent unit (GRU)-classical optimization framework is developed for data processing and achieving the optimization results. The implemented learning model is capable of achieving higher accuracy 99.691 % and lower computation time 0.023 sec when compared with the existing learning models. Additionally, the obtained results, such as transient stability and economic analysis of SynCons-conducted weak-grid present that the proposed solution can significantly perform 21.581 % cost minimization and 6.391 % reliability enhancement.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102111"},"PeriodicalIF":5.6,"publicationDate":"2025-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884017","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Low-carbon and robust economic scheduling of virtual power plants considering multiple uncertainties 考虑多重不确定性的虚拟电厂低碳稳健经济调度

IF 5.6 2区工程技术 Q2 ENERGY & FUELS

Sustainable Energy Grids & Networks

Pub Date : 2025-12-26 DOI: 10.1016/j.segan.2025.102104

Yongcan Zhu, Naying Wei, Junjun Kang, Yi Tian

The uncertainties of distributed generation, power load, and electricity price forecasts pose significant challenges for optimal dispatching of load virtual power plants (VPPs). This study addresses these issues by introducing a two-stage robust economic optimization scheduling model based on information gap decision theory (IGDT). Initially, a deterministic VPP objective function is formulated to minimize operating and carbon trading costs while defining constraints for each participating element. Subsequently, a robust VPP scheduling model is developed using IGDT to quantify uncertainties in wind power, solar power, load, and electricity market price predictions. The Karush–Kuhn–Tucker conditions are applied to simplify the optimization model under both risk aversion and risk pursuit behaviors. The effectiveness of the proposed model is validated through a comparative analysis of VPP scheduling results and total costs across different scenarios using real VPP case studies. The results indicate that participation in the carbon trading market led to a reduction in carbon emissions by 17.34 %–23.57 %. The introduction of demand response averaged a 39.18 % reduction in system total costs. The risk pursuit model, considering multiple uncertainties, reduced the total costs by 25.46 %–30.00 %.

分布式发电、电力负荷和电价预测的不确定性对负荷虚拟电厂的优化调度提出了重大挑战。本文通过引入基于信息差距决策理论的两阶段鲁棒经济优化调度模型来解决这些问题。首先，制定确定性VPP目标函数，以最小化运营和碳交易成本，同时定义每个参与元素的约束。随后，利用IGDT建立了稳健的VPP调度模型，量化风电、太阳能、负荷和电力市场价格预测中的不确定性。应用Karush-Kuhn-Tucker条件对风险规避行为和风险追求行为下的优化模型进行简化。通过实际VPP案例，对比分析了不同场景下VPP调度结果和总成本，验证了该模型的有效性。结果表明，参与碳交易市场导致碳排放量减少17.34% ~ 23.57%。需求响应的引入平均降低了系统总成本39.18%。考虑多重不确定性的风险追求模型使总成本降低了25.46% ~ 30.00%。

{"title":"Low-carbon and robust economic scheduling of virtual power plants considering multiple uncertainties","authors":"Yongcan Zhu, Naying Wei, Junjun Kang, Yi Tian","doi":"10.1016/j.segan.2025.102104","DOIUrl":"10.1016/j.segan.2025.102104","url":null,"abstract":"<div><div>The uncertainties of distributed generation, power load, and electricity price forecasts pose significant challenges for optimal dispatching of load virtual power plants (VPPs). This study addresses these issues by introducing a two-stage robust economic optimization scheduling model based on information gap decision theory (IGDT). Initially, a deterministic VPP objective function is formulated to minimize operating and carbon trading costs while defining constraints for each participating element. Subsequently, a robust VPP scheduling model is developed using IGDT to quantify uncertainties in wind power, solar power, load, and electricity market price predictions. The Karush–Kuhn–Tucker conditions are applied to simplify the optimization model under both risk aversion and risk pursuit behaviors. The effectiveness of the proposed model is validated through a comparative analysis of VPP scheduling results and total costs across different scenarios using real VPP case studies. The results indicate that participation in the carbon trading market led to a reduction in carbon emissions by 17.34 %–23.57 %. The introduction of demand response averaged a 39.18 % reduction in system total costs. The risk pursuit model, considering multiple uncertainties, reduced the total costs by 25.46 %–30.00 %.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102104"},"PeriodicalIF":5.6,"publicationDate":"2025-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884016","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Consensus clustering-based electric vehicle charging considering inaccurate user preferences and efficient charging operation zone 考虑不准确用户偏好和高效充电操作区域的共识聚类电动车充电

IF 5.6 2区工程技术 Q2 ENERGY & FUELS

Sustainable Energy Grids & Networks

Pub Date : 2025-12-26 DOI: 10.1016/j.segan.2025.102112

Shicong Zhang , Klaas Thoelen , Mohamed Yasko , Geert Deconinck

The rapid adoption of electric vehicles (EVs) necessitates smart charging solutions to prevent distribution grid overload. However, existing optimization frameworks often overlook critical real-world factors: (1) behavioral uncertainties from inaccurate user preferences (e.g., energy requirements), and (2) non-negligible energy losses during charging operations. This paper addresses critical inefficiencies in EV charging optimization through data-driven behavioral analysis and operational innovation. Using two real-world datasets—from the EnergyVille smart charging platform and the public ACN dataset at the Caltech campus—we quantify critical estimation gaps in user-provided preferences, such as energy demand and departure time. These genuine behavioral inputs are typically missing from synthetic data. Building on these insights, we develop a consensus-clustering forecasting framework that enhances preference prediction accuracy by 18 % (EnergyVille) and 85 % (ACN) versus user inputs. Furthermore, we propose an efficient charging operation zone (ECOZ), a dynamic constraint model that adapts to nonlinear charging efficiency characteristics. Integrated within a mixed-integer linear programming (MILP) optimization formulation, ECOZ maintains 85 % energy conversion efficiency during power allocation. Through simulation, we demonstrate the effectiveness of the proposed method on real-world data and achieve a 5 % reduction in daily EV charging energy losses compared to unconstrained scheduling approaches.

电动汽车的快速普及需要智能充电解决方案来防止配电网过载。然而，现有的优化框架往往忽略了现实世界的关键因素：(1)不准确的用户偏好（例如，能量需求）带来的行为不确定性；(2)充电过程中不可忽略的能量损失。本文通过数据驱动的行为分析和操作创新解决了电动汽车充电优化中的关键低效问题。使用两个真实世界的数据集——来自EnergyVille智能充电平台和加州理工学院校园的公共ACN数据集——我们量化了用户提供的偏好中的关键估计差距，比如能源需求和出发时间。这些真实的行为输入通常在合成数据中缺失。基于这些见解，我们开发了一个共识聚类预测框架，与用户输入相比，该框架将偏好预测精度提高了18% （EnergyVille）和85% （ACN）。在此基础上，提出了一种适应非线性充电效率特征的动态约束模型——高效充电操作区（ECOZ）。ECOZ集成在混合整数线性规划（MILP）优化配方中，在功率分配期间保持85%的能量转换效率。通过仿真，我们证明了该方法在实际数据上的有效性，与无约束调度方法相比，每日电动汽车充电能量损失减少了5%。

{"title":"Consensus clustering-based electric vehicle charging considering inaccurate user preferences and efficient charging operation zone","authors":"Shicong Zhang , Klaas Thoelen , Mohamed Yasko , Geert Deconinck","doi":"10.1016/j.segan.2025.102112","DOIUrl":"10.1016/j.segan.2025.102112","url":null,"abstract":"<div><div>The rapid adoption of electric vehicles (EVs) necessitates smart charging solutions to prevent distribution grid overload. However, existing optimization frameworks often overlook critical real-world factors: (1) behavioral uncertainties from inaccurate user preferences (e.g., energy requirements), and (2) non-negligible energy losses during charging operations. This paper addresses critical inefficiencies in EV charging optimization through data-driven behavioral analysis and operational innovation. Using two real-world datasets—from the EnergyVille smart charging platform and the public ACN dataset at the Caltech campus—we quantify critical estimation gaps in user-provided preferences, such as energy demand and departure time. These genuine behavioral inputs are typically missing from synthetic data. Building on these insights, we develop a consensus-clustering forecasting framework that enhances preference prediction accuracy by 18 % (EnergyVille) and 85 % (ACN) versus user inputs. Furthermore, we propose an efficient charging operation zone (ECOZ), a dynamic constraint model that adapts to nonlinear charging efficiency characteristics. Integrated within a mixed-integer linear programming (MILP) optimization formulation, ECOZ maintains 85 % energy conversion efficiency during power allocation. Through simulation, we demonstrate the effectiveness of the proposed method on real-world data and achieve a 5 % reduction in daily EV charging energy losses compared to unconstrained scheduling approaches.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102112"},"PeriodicalIF":5.6,"publicationDate":"2025-12-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884021","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A novel industrial load scheduling model to balance scheduling with virtual power plant regulation requirements 一种新的工业负荷调度模型，以平衡调度与虚拟电厂的调节需求

IF 5.6 2区工程技术 Q2 ENERGY & FUELS

Sustainable Energy Grids & Networks

Pub Date : 2025-12-25 DOI: 10.1016/j.segan.2025.102109

Sheng Xian Cao, Lin Yue, Gong Wang, Gui Chao Duan, Kun Li, Jun Li, Ying Zhe Kang, Hui Jing Sun

Within virtual power plants (VPPs), large industrial loads function as key controllable loads (CL), and precise load quantification is pivotal to achieving efficient VPP operation. For industrial loads subject to heterogeneous scheduling problem (SP) constraints alongside uncertain renewable outputs, production scheduling and energy use decisions are strongly coupled, and demand response (DR) execution can conflict with pre-established production plans. There is an urgent need to establish a VPP-oriented unified constraint set and feasible region. However, the coordination challenges of production scheduling optimization, economic performance improvement, and flexible participation in VPP operation remain insufficiently addressed. Accordingly, this paper proposes an industrial load scheduling model for virtual power plants (ILS-VPP), an industrial load scheduling model that reconciles factory-level scheduling with VPP-level regulation requirements. First, to address the difficulty of VPP participation under flexible job shop scheduling problem (FJSSP) constraints, we embed FJSSP into the optimization framework and unify the feasible region with participation constraints. Second, to overcome the limited adaptability of DR responses, we design a co-optimization mechanism that integrates production scheduling with DR. Third, to balance economic benefits and completion deadlines under high uncertainty, we develop a three-stage robust optimization (RO) strategy grounded in multi-polyhedral uncertainty sets, chance-constrained programming, and the

ϵ

-constraint method. An improved football team training algorithm (FTTA) is employed to solve the model, enhancing convergence stability and solution-set quality. A case study over a 90-day operating horizon shows that the proposed model improves economic performance by 21.3 %, can supply 11,003 kWh of energy to the VPP, achieves a load flexibility index (LFI) of 79.8 %, and increases the load factor (LF) from 0.708 to 0.807.

在虚拟电厂（VPP）中，大型工业负荷是关键可控负荷（CL），而精确的负荷量化是实现VPP高效运行的关键。对于受异构调度问题（SP）约束以及不确定可再生输出约束的工业负荷，生产调度和能源使用决策是强耦合的，并且需求响应（DR）的执行可能与预先建立的生产计划相冲突。迫切需要建立面向vpp的统一约束集和可行域。然而，在VPP操作中，优化生产调度、提高经济效益和灵活参与的协调挑战仍然没有得到充分解决。在此基础上，提出了虚拟电厂工业负荷调度模型（ILS-VPP），这是一种协调工厂级调度和vpp级调节需求的工业负荷调度模型。首先，针对柔性作业车间调度问题（FJSSP）约束下VPP参与困难的问题，将FJSSP嵌入到优化框架中，统一了具有参与约束的可行域；其次，为了克服生产调度响应的有限适应性，设计了一种集成生产调度和生产调度的协同优化机制。第三，为了平衡高不确定性下的经济效益和完工期限，我们开发了基于多多面体不确定性集、机会约束规划和ϵ-constraint方法的三阶段鲁棒优化（RO）策略。采用改进的足球队训练算法（FTTA）对模型进行求解，提高了收敛稳定性和解集质量。90天运行周期的实例研究表明，该模型提高了21.3%的经济效益，可为VPP提供11,003千瓦时的能源，实现了79.8%的负荷灵活性指数（LFI），并将负荷系数（LF）从0.708提高到0.807。

{"title":"A novel industrial load scheduling model to balance scheduling with virtual power plant regulation requirements","authors":"Sheng Xian Cao, Lin Yue, Gong Wang, Gui Chao Duan, Kun Li, Jun Li, Ying Zhe Kang, Hui Jing Sun","doi":"10.1016/j.segan.2025.102109","DOIUrl":"10.1016/j.segan.2025.102109","url":null,"abstract":"<div><div>Within virtual power plants (VPPs), large industrial loads function as key controllable loads (CL), and precise load quantification is pivotal to achieving efficient VPP operation. For industrial loads subject to heterogeneous scheduling problem (SP) constraints alongside uncertain renewable outputs, production scheduling and energy use decisions are strongly coupled, and demand response (DR) execution can conflict with pre-established production plans. There is an urgent need to establish a VPP-oriented unified constraint set and feasible region. However, the coordination challenges of production scheduling optimization, economic performance improvement, and flexible participation in VPP operation remain insufficiently addressed. Accordingly, this paper proposes an industrial load scheduling model for virtual power plants (ILS-VPP), an industrial load scheduling model that reconciles factory-level scheduling with VPP-level regulation requirements. First, to address the difficulty of VPP participation under flexible job shop scheduling problem (FJSSP) constraints, we embed FJSSP into the optimization framework and unify the feasible region with participation constraints. Second, to overcome the limited adaptability of DR responses, we design a co-optimization mechanism that integrates production scheduling with DR. Third, to balance economic benefits and completion deadlines under high uncertainty, we develop a three-stage robust optimization (RO) strategy grounded in multi-polyhedral uncertainty sets, chance-constrained programming, and the <span><math><mi>ϵ</mi></math></span>-constraint method. An improved football team training algorithm (FTTA) is employed to solve the model, enhancing convergence stability and solution-set quality. A case study over a 90-day operating horizon shows that the proposed model improves economic performance by 21.3 %, can supply 11,003 kWh of energy to the VPP, achieves a load flexibility index (LFI) of 79.8 %, and increases the load factor (LF) from 0.708 to 0.807.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102109"},"PeriodicalIF":5.6,"publicationDate":"2025-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925959","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Capacity allocation of pumped hydro storage under marketization process: A transitional strategy 市场化进程下抽水蓄能容量配置：一种过渡性策略

IF 5.6 2区工程技术 Q2 ENERGY & FUELS

Sustainable Energy Grids & Networks

Pub Date : 2025-12-25 DOI: 10.1016/j.segan.2025.102107

Yizhou Feng , Zhi Wu , Chen Chen , Liang Ma , Wei Gu , Suyang Zhou

To address the challenges posed by renewable energy integration in power systems, China is advancing the development of Pumped Hydro Storage (PHS). However, the rapid growth of PHS installations, coupled with strict regulations and a high reliance on capacity compensation, has led to increasing financial burdens on other utilities. One solution is to reduce PHS’s capacity compensation through its marketization. To this end, a ‘partial-regulated dispatch’ mechanism is proposed as a transitional strategy for gradual marketization. Also, an operational policy analysis framework is proposed based on evaluating dispatch mechanisms and business models. The dispatch mechanism evaluates the capacity support PHS provides to the power system, while the business models focus on enhancing PHS profitability to reduce the dependency on capacity compensation while ensuring long-term economic sustainability. Furthermore, the flexibility of PHS is introduced into the capacity compensation to incentivize PHS to support the power system during transitional stages. This flexibility is mathematically defined using the discrete Minkowski sum, considering both the vibration characteristics of individual units and the unit-commitment of PHS as a whole. The case study shows that through partial-regulated dispatch, PHS can reduce its reliance on capacity compensation by nearly 50 % while ensuring its regulatory service via flexibility compensation. This policy effectively balances economic viability with system support capabilities. Moreover, flexibility compensation provides PHS operators with a risk mitigation strategy in the complex power market environment. Under an appropriate operational strategy and policy incentives, flexibility can be enhanced by nearly 30 % in a fully marketized scenario, thereby contributing to both system stability and operational efficiency.

为了应对可再生能源在电力系统中的整合所带来的挑战，中国正在推进抽水蓄能（PHS）的发展。然而，小灵通安装的快速增长，加上严格的法规和对容量补偿的高度依赖，导致其他公用事业的财务负担不断增加。解决方案之一是通过市场化降低小灵通的容量补偿。为此，建议建立“部分管制调度”机制，作为逐步市场化的过渡战略。同时，提出了基于调度机制和业务模型评估的操作策略分析框架。调度机制评估小灵通为电力系统提供的容量支持，商业模式侧重于提高小灵通的盈利能力，以减少对容量补偿的依赖，同时确保长期的经济可持续性。此外，在容量补偿中引入小灵通的灵活性，以激励小灵通在过渡阶段支持电力系统。这种灵活性在数学上是用离散闵可夫斯基和来定义的，同时考虑了单个单元的振动特性和小灵通作为一个整体的单元承诺。案例研究表明，通过部分调节调度，小灵通在保证灵活性补偿的同时，可将其对容量补偿的依赖程度降低近50%。这一政策有效地平衡了经济可行性和系统支持能力。此外，灵活性补偿为小灵通运营商在复杂的电力市场环境中提供了一种降低风险的策略。在适当的运营战略和政策激励下，在完全市场化的情况下，灵活性可以提高近30%，从而有助于系统稳定性和运营效率。

{"title":"Capacity allocation of pumped hydro storage under marketization process: A transitional strategy","authors":"Yizhou Feng , Zhi Wu , Chen Chen , Liang Ma , Wei Gu , Suyang Zhou","doi":"10.1016/j.segan.2025.102107","DOIUrl":"10.1016/j.segan.2025.102107","url":null,"abstract":"<div><div>To address the challenges posed by renewable energy integration in power systems, China is advancing the development of Pumped Hydro Storage (PHS). However, the rapid growth of PHS installations, coupled with strict regulations and a high reliance on capacity compensation, has led to increasing financial burdens on other utilities. One solution is to reduce PHS’s capacity compensation through its marketization. To this end, a ‘partial-regulated dispatch’ mechanism is proposed as a transitional strategy for gradual marketization. Also, an operational policy analysis framework is proposed based on evaluating dispatch mechanisms and business models. The dispatch mechanism evaluates the capacity support PHS provides to the power system, while the business models focus on enhancing PHS profitability to reduce the dependency on capacity compensation while ensuring long-term economic sustainability. Furthermore, the flexibility of PHS is introduced into the capacity compensation to incentivize PHS to support the power system during transitional stages. This flexibility is mathematically defined using the discrete Minkowski sum, considering both the vibration characteristics of individual units and the unit-commitment of PHS as a whole. The case study shows that through partial-regulated dispatch, PHS can reduce its reliance on capacity compensation by nearly 50 % while ensuring its regulatory service via flexibility compensation. This policy effectively balances economic viability with system support capabilities. Moreover, flexibility compensation provides PHS operators with a risk mitigation strategy in the complex power market environment. Under an appropriate operational strategy and policy incentives, flexibility can be enhanced by nearly 30 % in a fully marketized scenario, thereby contributing to both system stability and operational efficiency.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102107"},"PeriodicalIF":5.6,"publicationDate":"2025-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145926019","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Supervised learning-driven dead band control of occupant thermostats for energy-efficient residential HVAC 节能住宅暖通空调使用人员恒温器的监督学习驱动死区控制

IF 5.6 2区工程技术 Q2 ENERGY & FUELS

Sustainable Energy Grids & Networks

Pub Date : 2025-12-25 DOI: 10.1016/j.segan.2025.102110

Alper Savasci , Oguzhan Ceylan , Sumit Paudyal

Heating, ventilation, and air conditioning (HVAC) systems play a crucial role in demand-side management (DSM) by shaping residential electricity consumption and enabling flexible, grid-responsive operation. Thermostats in HVAC systems regulate indoor temperature as part of a closed-loop control framework, typically incorporating a fixed temperature dead band–a range around the setpoint where no action is taken–to reduce energy use and prevent frequent cycling of the HVAC system. Although essential for efficiency and equipment longevity, fixed dead bands limit adaptability, as dynamically adjusting them under varying environmental conditions remains challenging for occupants. To address this limitation, we propose a machine learning (ML)-based dead band tuning framework that optimally adjusts thermostat settings in real time. The method integrates conventional optimization with data-driven modeling: a mixed-integer linear programming (MILP) model is first used to generate optimal dead band values under measured outdoor temperature records (diverse seasonal weather scenarios) which are then employed to train the ML-based predictor to learn a real-time discrete dead band decision policy that approximates the MILP-optimal hysteresis-aware decisions. Among the evaluated models, Random Forest demonstrates superior predictive performance, achieving a mean squared error (MSE) of 0.0399 and a coefficient of determination (

R^{2}

) of 95.75 %.

供暖、通风和空调（HVAC）系统在需求侧管理（DSM）中发挥着至关重要的作用，它塑造了住宅用电量，实现了灵活的电网响应式运行。暖通空调系统中的恒温器作为闭环控制框架的一部分调节室内温度，通常包含一个固定的温度死区-在设定值周围不采取任何行动的范围-以减少能源使用并防止暖通空调系统的频繁循环。虽然对于效率和设备寿命至关重要，但固定死区限制了适应性，因为在不同的环境条件下动态调整它们对使用者来说仍然是一个挑战。为了解决这一限制，我们提出了一个基于机器学习（ML）的死区调优框架，该框架可以实时优化调整恒温器设置。该方法将传统优化与数据驱动建模相结合：首先使用混合整数线性规划（MILP）模型在室外测量温度记录（不同季节天气情景）下生成最优死区值，然后使用该模型训练基于ml的预测器，以学习接近MILP最优迟滞感知决策的实时离散死区决策策略。在评价的模型中，随机森林模型的预测性能较好，均方误差（MSE）为0.0399，决定系数（R2）为95.75%。

{"title":"Supervised learning-driven dead band control of occupant thermostats for energy-efficient residential HVAC","authors":"Alper Savasci , Oguzhan Ceylan , Sumit Paudyal","doi":"10.1016/j.segan.2025.102110","DOIUrl":"10.1016/j.segan.2025.102110","url":null,"abstract":"<div><div>Heating, ventilation, and air conditioning (HVAC) systems play a crucial role in demand-side management (DSM) by shaping residential electricity consumption and enabling flexible, grid-responsive operation. Thermostats in HVAC systems regulate indoor temperature as part of a closed-loop control framework, typically incorporating a fixed temperature dead band–a range around the setpoint where no action is taken–to reduce energy use and prevent frequent cycling of the HVAC system. Although essential for efficiency and equipment longevity, fixed dead bands limit adaptability, as dynamically adjusting them under varying environmental conditions remains challenging for occupants. To address this limitation, we propose a machine learning (ML)-based dead band tuning framework that optimally adjusts thermostat settings in real time. The method integrates conventional optimization with data-driven modeling: a mixed-integer linear programming (MILP) model is first used to generate optimal dead band values under measured outdoor temperature records (diverse seasonal weather scenarios) which are then employed to train the ML-based predictor to learn a real-time discrete dead band decision policy that approximates the MILP-optimal hysteresis-aware decisions. Among the evaluated models, Random Forest demonstrates superior predictive performance, achieving a mean squared error (MSE) of 0.0399 and a coefficient of determination (<span><math><msup><mi>R</mi><mn>2</mn></msup></math></span>) of 95.75 %.</div></div>","PeriodicalId":56142,"journal":{"name":"Sustainable Energy Grids & Networks","volume":"45 ","pages":"Article 102110"},"PeriodicalIF":5.6,"publicationDate":"2025-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145884022","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0