A Systematic Study on Reinforcement Learning Based Applications

Energies, Journal Year: 2023, Volume and Issue: 16(3), P. 1512 - 1512

Published: Feb. 3, 2023

We have analyzed 127 publications for this review paper, which discuss applications of Reinforcement Learning (RL) in marketing, robotics, gaming, automated cars, natural language processing (NLP), internet things security, recommendation systems, finance, and energy management. The optimization use is critical today’s environment. mainly focus on the RL application Traditional rule-based systems a set predefined rules. As result, they may become rigid unable to adjust changing situations or unforeseen events. can overcome these drawbacks. learns by exploring environment randomly based experience, it continues expand its knowledge. Many researchers are working RL-based management (EMS). utilized such as optimizing smart buildings, hybrid automobiles, grids, managing renewable resources. contributes achieving net zero carbon emissions sustainable In context technology, be optimize regulation building heating, ventilation, air conditioning (HVAC) reduce consumption while maintaining comfortable atmosphere. EMS accomplished teaching an agent make judgments sensor data, temperature occupancy, modify HVAC system settings. has proven beneficial lowering usage buildings active research area buildings. used electric vehicles (HEVs) learning optimal control policy maximize battery life fuel efficiency. acquired remarkable position gaming applications. majority security-related operate simulated recommender provide good suggestions accuracy diversity. This article assists novice comprehending foundations reinforcement

Language: Английский

---

Improving building occupant comfort through a digital twin approach: A Bayesian network model and predictive maintenance method

Energy and Buildings, Journal Year: 2023, Volume and Issue: 288, P. 112992 - 112992

Published: March 23, 2023

This study introduces a Bayesian network model to evaluate the comfort levels of occupants two non-residential Norwegian buildings based on data collected from satisfaction surveys and building performance parameters. A Digital Twin approach is proposed integrate information modeling (BIM) with real-time sensor data, occupant feedback, probabilistic detect predict HVAC issues that may impact comfort. The also uses 200000 points as historical various sensors understand previous systems' behavior. presents new methods for using BIM visualization platform predictive maintenance identify address problems in system. For maintenance, nine machine learning algorithms were evaluated metrics such ROC, accuracy, F1-score, precision, recall, where Extreme Gradient Boosting (XGB) was best algorithm prediction. XGB average 2.5% more accurate than Multi-Layer Perceptron (MLP), up 5% other models. Random Forest around 96% faster XGBoost while being relatively easier implement. paper novel method utilizes several standards determine remaining useful life HVAC, leading potential increase its lifetime by at least 10% resulting significant cost savings. result shows most important factors affect are poor air quality, lack natural light, uncomfortable temperature. To challenge applying these wide range buildings, proposes framework ontology graphs different systems, including FM, CMMS, BMS, BIM. study's results provide insight into influence comfort, help expedite identifying equipment malfunctions point towards solutions, sustainable energy-efficient buildings.

Language: Английский

---

Scalable energy management approach of residential hybrid energy system using multi-agent deep reinforcement learning

Applied Energy, Journal Year: 2024, Volume and Issue: 367, P. 123414 - 123414

Published: May 14, 2024

Language: Английский

---

Digital Twin framework for automated fault source detection and prediction for comfort performance evaluation of existing non-residential Norwegian buildings

Energy and Buildings, Journal Year: 2022, Volume and Issue: 281, P. 112732 - 112732

Published: Dec. 28, 2022

Numerous buildings fall short of expectations regarding occupant satisfaction, sustainability, or energy efficiency. In this paper, the performance in terms comfort is evaluated using a probabilistic model based on Bayesian networks (BNs). The BN founded an in-depth analysis satisfaction survey responses and thorough study building parameters. This also presents user-friendly visualization compatible with BIM to simplify data collecting two case studies from Norway 2019 2022. paper proposes novel Digital Twin approach for incorporating information modeling (BIM) real-time sensor data, occupants' feedback, comfort, HVAC faults detection prediction that may affect comfort. New methods as platform, well predictive maintenance method detect anticipate problems system, are presented. These will help decision-makers improve conditions buildings. However, due intricate interaction between numerous equipment absence integration among FM systems, CMMS, BMS, integrated into framework utilizing ontology graphs generalize so it can be applied many results aid facility management sector by offering insight aspects influence speeding up process identifying malfunctions, pointing toward possible solutions.

Language: Английский

---

Comparative study of model-based and model-free reinforcement learning control performance in HVAC systems

Journal of Building Engineering, Journal Year: 2023, Volume and Issue: 74, P. 106852 - 106852

Published: May 19, 2023

Language: Английский

---

Dynamic energy management with thermal comfort forecasting

Building and Environment, Journal Year: 2023, Volume and Issue: 237, P. 110341 - 110341

Published: April 24, 2023

Reducing greenhouse gas emissions and energy cost in the building sector largely relies on effective management. Yet, when it comes to heating or cooling, savings may translate uncomfortable conditions for users. To ensure thermal comfort with a minimal consumption, this paper we propose modular methodology that dynamically schedules operating hours of heating/cooling system using forecasts. decide time mode operation, our approach utilizes indoor air temperature relative humidity forecasting models, as well data-driven algorithm predicts level based said forecasts employs pre-heating/cooling necessary. Our empirical evaluation, conducted Casal Mira-sol culture center Sant Cugat, Spain, suggests can improve satisfaction users, while achieving significant savings.

Language: Английский

---

Predictive control optimization of chiller plants based on deep reinforcement learning

Journal of Building Engineering, Journal Year: 2023, Volume and Issue: 76, P. 107158 - 107158

Published: June 22, 2023

Language: Английский

---

AI-Driven Innovations in Building Energy Management Systems: A Review of Potential Applications and Energy Savings

Energies, Journal Year: 2024, Volume and Issue: 17(17), P. 4277 - 4277

Published: Aug. 27, 2024

Despite the tightening of energy performance standards for buildings in various countries and increased use efficient renewable technologies, it is clear that sector needs to change more rapidly meet Net Zero Emissions (NZE) scenario by 2050. One problems have been analyzed intensively recent years operation much than they were designed to. This problem, known as gap, found many often attributed poor management building systems. The application Artificial Intelligence (AI) Building Energy Management Systems (BEMS) has untapped potential address this problem lead sustainable buildings. paper reviews different AI-based models proposed applications with intention reduce consumption. It compares evaluated reviewed papers presenting accuracy error rates model identifies where greatest savings could be achieved, what extent. review showed offices (up 37%) when employ AI HVAC control optimization. In residential educational buildings, lower intelligence existing BEMS results smaller 23% 21%, respectively).

Language: Английский

---

Reinforcement Learning-Based Tracking Control for Networked Control Systems With DoS Attacks

IEEE Transactions on Information Forensics and Security, Journal Year: 2024, Volume and Issue: 19, P. 4188 - 4197

Published: Jan. 1, 2024

This paper is concerned with the reinforcement learning-based tracking control problem for a class of networked systems subject to denial-of-service (DoS) attacks. Taking effects DoS attacks into consideration, novel value function proposed, which considers cost input, external disturbance and error. Then, using structure function, Bellman equation Hamilton are defined. By employing optimality theory, optimal strategy game algebraic Riccati (GARE) solved function. Next, desired performance guaranteed as solution GARE found. Furthermore, an attacks-based Q-learning algorithm projected find without system dynamics convergence given. Finally, F-404 aircraft engine given verify effectiveness proposed strategy.

Language: Английский

---

Towards various occupants with different thermal comfort requirements: A deep reinforcement learning approach combined with a dynamic PMV model for HVAC control in buildings

Energy Conversion and Management, Journal Year: 2024, Volume and Issue: 320, P. 118995 - 118995

Published: Sept. 4, 2024

Language: Английский

---