Related references
Note: Only part of the references are listed.
Article
Green & Sustainable Science & Technology
Fan He et al.
Summary: Adding metal fins can effectively improve the thermal conductivity of Phase-Change Material (PCM). The use of spiral fins is proposed to overcome the poor natural convection caused by annular fins. Experimental results show that employing spiral fins can enhance heat transfer performance, especially compared to annular fins.
Article
Construction & Building Technology
Yuan Gao et al.
Summary: This study proposes a hybrid prediction model and MPC framework to optimize the combination of solar PV and batteries in building energy systems. Validated with real building data, the results show significant improvements in battery safety and combined heat and power operation using the proposed framework.
SUSTAINABLE CITIES AND SOCIETY
(2023)
Article
Energy & Fuels
Hongwei Chen et al.
Summary: This paper proposes an emergency load shedding method based on data-driven strategies and deep reinforcement learning. The traditional event-driven load shedding strategy faces the risk of mismatching in high renewable energy penetrated power systems. In order to address this challenge, the proposed method transforms the emergency load shedding strategy into a Markov Decision Process and optimizes it using deep reinforcement learning.
Article
Energy & Fuels
Jiawei Feng et al.
Summary: This paper proposes an optimal dispatch strategy for an industrial park based on a deep reinforcement learning approach. The strategy models renewable generations and flexible loads using random fuzzy theory and introduces dynamic line rating to improve power flow capacity. The dispatch process is modeled as a Markov decision process to avoid linearization processes. The paper also proposes a variant of the distributed proximal policy optimization algorithm, where distributed agents are trained to explore optimal strategies in different environments.
SUSTAINABLE ENERGY GRIDS & NETWORKS
(2023)
Article
Construction & Building Technology
Jide Niu et al.
Summary: Numerous studies have focused on energy efficiency in chiller plants, primarily through equipment retrofits and optimal operation strategies. However, certain measures may disrupt building operations, such as conflicting with production schedules. This study developed a comprehensive simulation model for a chiller plant and implemented an optimization framework to investigate the energy-saving potential. By optimizing setpoints in the control strategy, the chiller plant achieved an average global energy-saving rate of 8.5%. Water pumps were found to contribute the most to energy savings, despite a decrease in chiller efficiency. This work aids engineers in selecting easily deployable optimization variables for significant energy savings.
JOURNAL OF BUILDING ENGINEERING
(2023)
Article
Construction & Building Technology
Hanbei Zhang et al.
Summary: This study develops a coordinated optimal load scheduling strategy for building cluster load management that considers dynamic electricity price and marginal emission factor simultaneously. The strategy effectively solves the conflicts of minimizing electricity cost, carbon emissions, and peak load while ensuring user satisfaction. The results show that the strategy can achieve a compromise between conflicting objectives in different correlation scenarios.
ENERGY AND BUILDINGS
(2023)
Article
Thermodynamics
Yanxue Li et al.
Summary: An efficient and flexible energy management strategy is crucial for energy conservation in the building sector. This study proposes a hybrid model-based reinforcement learning framework that uses short-term monitored data to optimize indoor thermal comfort and energy cost-saving performance. Simulation results demonstrate the efficiency and superiority of the proposed framework, with the D3QN agent achieving over 30% cost savings compared to measurement results.
Article
Zhe Chen et al.
Advances in Applied Energy
(2023)
Article
Construction & Building Technology
Shanshan Li et al.
Summary: Ensuring occupants' work or cognitive performance and maintaining thermal comfort are important targets of indoor thermal environment management. Advanced sensing technologies based on physiological measurements and machine learning (ML) approaches have provided a more precise and efficient way to assess the link between the indoor thermal environment and the performances of occupants. This review work integrates the results of cognitive tests related to the thermal environment and performance, summarizes the application of existing physiological indicators, and the practice of using sensing technologies and ML technology to assess occupant performance and predict indoor thermal comfort.
JOURNAL OF BUILDING ENGINEERING
(2023)
Article
Engineering, Electrical & Electronic
Ying Wu et al.
Summary: This paper discusses the positive roles of buildings and electric vehicles in establishing sustainable transactive energy communities, focusing on their potential for optimization and balance of supply and demand within a full-scale flexibility framework.
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS
(2022)
Article
Engineering, Electrical & Electronic
Sara Abedi et al.
Summary: This study focuses on developing an intelligent and real-time battery energy storage control system based on reinforcement learning to efficiently manage residential solar energy systems. By using a uniquely designed cyclic time-dependent Markov Process in the Q-learning algorithm, the system effectively saves electricity costs while adapting to changing demand and pricing conditions. Additionally, the performance of the Q-learning algorithm is shown to be superior to the One-step Roll-out algorithm, with variations in electricity prices having the most significant impact on performance.
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS
(2022)
Article
Energy & Fuels
Christopher J. Bay et al.
Summary: This study modifies the Limited-Communication (LC) DMPC algorithm to make it applicable to systems with multiple buildings and enables grid-interactive buildings by adding a grid-aggregator subsystem. The research demonstrates that adequate power tracking can be achieved for different simulation scenarios involving heterogeneous buildings.
Article
Energy & Fuels
Giuseppe Pinto et al.
Summary: The increasing use of renewable energy has the potential to reduce carbon emissions in the building energy sector. However, integrating these energy sources brings challenges, such as grid instability due to their variable nature. Demand side management, which exploits demand flexibility, is seen as a promising solution to improve grid stability. Multi-agent systems offer a tool for managing energy in buildings at district and grid scales. In this study, the energy management of four buildings with thermal energy storage and PV systems is formulated as a multi-agent problem. Two multi-agent reinforcement learning methods outperformed a rule-based controller in terms of cost and peak demand.
Article
Energy & Fuels
Heeyun Lee et al.
Summary: This study proposes a reinforcement learning-based energy efficient speed planning strategy for autonomous electric vehicles to minimize energy consumption in diverse driving situations.
Article
Energy & Fuels
Rendong Shen et al.
Summary: Under the backdrop of high global building energy consumption, utilizing renewable energy to meet the increasing demand of building energy systems can promote clean energy transformation and carbon neutrality. However, the complexity of BES control is increased with the introduction of renewable energy, and addressing the mismatch between supply and demand sides is challenging. The proposed multi-agent deep reinforcement learning framework optimized energy management in buildings, improving device control efficiency and renewable energy utilization in BES.
Article
Thermodynamics
Alexander Dreher et al.
Summary: This study presents a Reinforcement Learning-based energy management approach for CO2-neutral hydrogen production and storage in the industrial combined heat and power sector. Simulation results show the potential of this approach in hydrogen production planning and improvement through integrating energy and price forecasts. However, balancing multiple conflicting objectives remains a challenge.
ENERGY CONVERSION AND MANAGEMENT
(2022)
Article
Thermodynamics
Rui Tang et al.
Summary: This study developed a data-driven model predictive control using support vector regression for fast demand response events in commercial buildings. By optimizing SVR hyperparameters and shortening the genetic algorithm search range, the proposed SVR-based MPC successfully achieved simultaneous control of power demand and indoor temperature. Compared with RC-based MPC, the SVR-based MPC reduced time/labor costs without sacrificing control performance in fast DR events.
BUILDING SIMULATION
(2022)
Article
Energy & Fuels
Chia-Shing Tai et al.
Summary: The demand-side management (DSM) system has become an important research topic due to its ability to reduce total electricity cost and peak-to-average ratio. By utilizing sensor data and machine learning skills, a multi-agent DSM system based on HAN was proposed to reduce energy costs in smart homes. The system successfully reduced peak value, PAR value, and electricity cost, and achieved significant cost reduction when applied to the REDD dataset.
SUSTAINABLE ENERGY GRIDS & NETWORKS
(2022)
Review
Construction & Building Technology
Wenjie Zhang et al.
Summary: This paper provides a comprehensive overview of the key issues and methods for enhancing the interaction performance between zero-energy buildings and power grids. It first introduces the types and applications of grid-interaction (GI) indicators for performance evaluation, and then reviews the optimization measures for enhancing the interaction performance. Finally, it summarizes the limitations of current research and the direction of future work on the interaction between zero-energy buildings and power grids.
SUSTAINABLE CITIES AND SOCIETY
(2022)
Article
Energy & Fuels
Yuan Gao et al.
Summary: This study focuses on the off-grid operation and battery safety optimization of a renewable building energy system using reinforcement learning algorithms. Through training and validation with real measured data, the proposed reinforcement learning design achieves the optimization goals of off-grid operation and battery safety.
Article
Construction & Building Technology
Max Bird et al.
Summary: This paper presents a case study on improving the energy and carbon usage of a food-retail building through the use of a generalisable model predictive control (MPC) framework. The study shows that the MPC approach can achieve a lower overall cost for the HVAC system while ensuring thermal comfort of occupants.
ENERGY AND BUILDINGS
(2022)
Article
Green & Sustainable Science & Technology
Yaling Wu et al.
Summary: In this study, a new operation strategy and algorithm for optimizing building energy systems with PV-battery integration were proposed. Compared with traditional optimization methods, this approach reduces calculation time and decreases total cost.
Article
Engineering, Multidisciplinary
Mohammed H. Alabdullah et al.
Summary: This paper proposes a deep reinforcement learning-based approach to optimally manage different energy resources within a microgrid. The proposed methodology considers the stochastic behavior of the main elements and formulates the energy management problem as a finite horizon Markov Decision Process (MDP) without prior knowledge of transition probabilities. An efficient reinforcement learning algorithm based on deep Q-networks is implemented to solve the problem. The results of a case study based on a real microgrid demonstrate the effectiveness of the proposed methodology in achieving optimal cost-effective scheduling of energy resources under stochastic conditions.
ALEXANDRIA ENGINEERING JOURNAL
(2022)
Article
Energy & Fuels
Xiaoyu Jin et al.
Summary: The spatial feature of building energy consumption is crucial for urban energy planning and policymaking, and machine learning has shown its potential in predicting urban building energy consumption. However, current building energy benchmarking datasets only cover large buildings, which do not represent the entire city. This study proposes a novel framework based on semi-supervised learning to utilize unlabeled datasets and develop more generic urban-level data-driven building energy prediction models with higher spatial resolution.
Article
Energy & Fuels
Lissy Langer et al.
Summary: This study analyzes the potential of using deep reinforcement learning to control multiple energy devices in a smart home and transforms a mixed-integer linear program into a DRL implementation. The experimental results show that this DRL approach outperforms rule-based methods and achieves a high level of self-sufficiency with limited comfort violations.
Article
Energy & Fuels
Kun Zhang et al.
Summary: This paper investigates the application of MPC in small commercial buildings and evaluates its performance using field tests. The results show that the MPC controller can achieve cost savings and peak demand reduction while maintaining comfortable indoor temperature and food safety.
ADVANCES IN APPLIED ENERGY
(2022)
Article
Energy & Fuels
Jingjing Liu et al.
Summary: Building demand flexibility (DF) research is gaining attention, and a quantitative understanding of DF based on large empirical datasets is needed for it to become a predictable grid resource. This study presents a novel DF benchmarking framework focused on load shedding and shifting, using simple single-event metrics to analyze and visualize the performance and trends of DF. The feasibility and scalability of the framework are tested through case studies of office buildings and big-box retail buildings with demand response participation data, providing insights for building stakeholders.
ADVANCES IN APPLIED ENERGY
(2022)
Article
Energy & Fuels
Samir Touzani et al.
Summary: The study highlights the effectiveness of distributed energy resources in supporting carbon reduction goals and increasing grid reliability. Traditional control algorithms are inadequate for optimizing across systems in commercial buildings, while deep reinforcement learning offers a potentially effective solution. The research applied a deep reinforcement learning-based control method in practice, resulting in significant cost savings.
Article
Thermodynamics
Simone Totaro et al.
Summary: This paper addresses the lifelong control problem of isolated microgrids, proposing a novel model-based reinforcement learning algorithm that can handle system changes and outperforms traditional algorithms.
Review
Thermodynamics
Hong Tang et al.
Summary: This paper discusses the importance of building energy flexibility in smart grids, proposes a method to categorize building energy flexibility, and analyzes the potential and limitations of buildings in providing energy flexibility.
Article
Energy & Fuels
M. Eslami et al.
Summary: This study investigates the effects of different policies on residential photovoltaic and wind turbine investment in various cities of Iran, finding that increasing feed-in tariffs in 2019 is effective in some cities and proposing other investment incentives and optimization methods. Granting loans to domestic investors can enhance the net present value for photovoltaic and wind systems, and wind turbine installations in Bushehr can function as off-grid systems.
Article
Energy & Fuels
Sepehr Sanaye et al.
Summary: A novel energy management method based on Deep Q Network algorithm for an integrated hybrid system is proposed, eliminating the need for historical data or online trial and error. Compared to two rule-based methods, the RL-DQN method significantly reduces operating costs and CO2 emissions, while increasing the proportion of electricity supplied by renewable sources.
Article
Energy & Fuels
Prateek Munankarmi et al.
Summary: Demand-side management (DSM) strategies, such as energy efficiency (EE) and demand flexibility (DF), contribute to cost-effective operation of the electricity grid by reducing costs, enhancing reliability, and helping customers reduce utility bills. Utilizing EE upgrades and enabling DF through a home energy management system (HEMS) in homes can reduce HVAC energy use, utility bills, and peak demand while increasing load flexibility in the community.
Article
Energy & Fuels
Pietari Puranen et al.
Summary: Solar photovoltaic systems are popular in prosumer households for increasing renewable energy share. By analyzing data from two real-life cases in Finland, it was found that switching to hourly net metering can increase self-sufficiency, while network energy storage is more economically feasible. However, physical battery storage may not be economically competitive in northern climates without significant cost reductions.
Review
Energy & Fuels
Han Li et al.
Summary: This paper reviews recent studies on residential building demand side management, focusing on the characterization and quantification of energy flexibility in various types of flexible loads. The studies show that shifting loads is the dominant flexibility type, with control strategies and mathematical formulations playing a key role in achieving flexibility. The review reveals research opportunities to address significant gaps in the existing literature and emphasizes the importance of standardizing definitions, representation of flexibility resources, integrating occupant impacts, and developing requirements and credits of energy flexibility in building energy codes and standards.
ADVANCES IN APPLIED ENERGY
(2021)
Article
Green & Sustainable Science & Technology
Mohammad Jafari et al.
Article
Engineering, Electrical & Electronic
Marcelo M. Morato et al.
INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS
(2020)
Review
Automation & Control Systems
Jan Drgona et al.
ANNUAL REVIEWS IN CONTROL
(2020)
Article
Thermodynamics
Elena Bee et al.
BUILDING SIMULATION
(2019)
Article
Energy & Fuels
Yu Wang et al.
Review
Energy & Fuels
Gianluca Serale et al.
Article
Green & Sustainable Science & Technology
Romeo Pacudan
Article
Green & Sustainable Science & Technology
Yanxue Li et al.
Article
Multidisciplinary Sciences
Volodymyr Mnih et al.