Optimization of a photovoltaic-battery system using deep reinforcement learning and load forecasting
Home Energy Management Systems (HEMS) are increasingly relevant for demand-side management at the residential level by collecting data (energy, weather, electricity prices) and controlling home appliances or storage systems. This control can be performed with classical models that find optimal solut...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2024-05-01
|
| Series: | Energy and AI |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2666546824000132 |
| _version_ | 1827256568882659328 |
|---|---|
| author | António Corte Real G. Pontes Luz J.M.C. Sousa M.C. Brito S.M. Vieira |
| author_facet | António Corte Real G. Pontes Luz J.M.C. Sousa M.C. Brito S.M. Vieira |
| author_sort | António Corte Real |
| collection | DOAJ |
| description | Home Energy Management Systems (HEMS) are increasingly relevant for demand-side management at the residential level by collecting data (energy, weather, electricity prices) and controlling home appliances or storage systems. This control can be performed with classical models that find optimal solutions, with high real-time computational cost, or data-driven approaches, like Reinforcement Learning, that find good and flexible solutions, but depend on the availability of load and generation data and demand high computational resources for training. In this work, a novel HEMS is proposed for the optimization of an electric battery operation in a real, online and data-driven environment that integrates state-of-the-art load forecasting combining CNN and LSTM neural networks to increase the robustness of decisions. Several Reinforcement Learning agents are trained with different algorithms (Double DQN, Dueling DQN, Rainbow and Proximal Policy Optimization) in order to minimize the cost of electricity purchase and to maximize photovoltaic self-consumption for a PV-Battery residential system. Results show that the best Reinforcement Learning agent achieves a 35% reduction in total cost when compared with an optimization-based agent. |
| first_indexed | 2024-03-08T04:06:47Z |
| format | Article |
| id | doaj.art-a939dbd56fa244458b70baec9fc9bf4d |
| institution | Directory Open Access Journal |
| issn | 2666-5468 |
| language | English |
| last_indexed | 2025-03-22T01:40:59Z |
| publishDate | 2024-05-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Energy and AI |
| spelling | doaj.art-a939dbd56fa244458b70baec9fc9bf4d2024-05-09T04:37:03ZengElsevierEnergy and AI2666-54682024-05-0116100347Optimization of a photovoltaic-battery system using deep reinforcement learning and load forecastingAntónio Corte Real0G. Pontes Luz1J.M.C. Sousa2M.C. Brito3S.M. Vieira4IDMEC, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, 1049-001, PortugalInstituto Dom Luiz, Faculdade de Ciências, Universidade de Lisboa, Campo Grande, 1749-016, Lisboa, Portugal; Corresponding author.IDMEC, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, 1049-001, PortugalInstituto Dom Luiz, Faculdade de Ciências, Universidade de Lisboa, Campo Grande, 1749-016, Lisboa, PortugalIDMEC, Instituto Superior Técnico, Universidade de Lisboa, Lisboa, 1049-001, PortugalHome Energy Management Systems (HEMS) are increasingly relevant for demand-side management at the residential level by collecting data (energy, weather, electricity prices) and controlling home appliances or storage systems. This control can be performed with classical models that find optimal solutions, with high real-time computational cost, or data-driven approaches, like Reinforcement Learning, that find good and flexible solutions, but depend on the availability of load and generation data and demand high computational resources for training. In this work, a novel HEMS is proposed for the optimization of an electric battery operation in a real, online and data-driven environment that integrates state-of-the-art load forecasting combining CNN and LSTM neural networks to increase the robustness of decisions. Several Reinforcement Learning agents are trained with different algorithms (Double DQN, Dueling DQN, Rainbow and Proximal Policy Optimization) in order to minimize the cost of electricity purchase and to maximize photovoltaic self-consumption for a PV-Battery residential system. Results show that the best Reinforcement Learning agent achieves a 35% reduction in total cost when compared with an optimization-based agent.http://www.sciencedirect.com/science/article/pii/S2666546824000132Energy management systemDeep reinforcement learningDemand-side managementLoad forecastingDeep learningEnergy storage systems |
| spellingShingle | António Corte Real G. Pontes Luz J.M.C. Sousa M.C. Brito S.M. Vieira Optimization of a photovoltaic-battery system using deep reinforcement learning and load forecasting Energy and AI Energy management system Deep reinforcement learning Demand-side management Load forecasting Deep learning Energy storage systems |
| title | Optimization of a photovoltaic-battery system using deep reinforcement learning and load forecasting |
| title_full | Optimization of a photovoltaic-battery system using deep reinforcement learning and load forecasting |
| title_fullStr | Optimization of a photovoltaic-battery system using deep reinforcement learning and load forecasting |
| title_full_unstemmed | Optimization of a photovoltaic-battery system using deep reinforcement learning and load forecasting |
| title_short | Optimization of a photovoltaic-battery system using deep reinforcement learning and load forecasting |
| title_sort | optimization of a photovoltaic battery system using deep reinforcement learning and load forecasting |
| topic | Energy management system Deep reinforcement learning Demand-side management Load forecasting Deep learning Energy storage systems |
| url | http://www.sciencedirect.com/science/article/pii/S2666546824000132 |
| work_keys_str_mv | AT antoniocortereal optimizationofaphotovoltaicbatterysystemusingdeepreinforcementlearningandloadforecasting AT gpontesluz optimizationofaphotovoltaicbatterysystemusingdeepreinforcementlearningandloadforecasting AT jmcsousa optimizationofaphotovoltaicbatterysystemusingdeepreinforcementlearningandloadforecasting AT mcbrito optimizationofaphotovoltaicbatterysystemusingdeepreinforcementlearningandloadforecasting AT smvieira optimizationofaphotovoltaicbatterysystemusingdeepreinforcementlearningandloadforecasting |