| Titre : | Intelligent predictive energy management for smart homes and buildings : a reinforcement learning and KNX-Based approach |
| Auteurs : | Imene Sedkaoui, Auteur ; Omar Stihi, Directeur de thèse |
| Type de document : | document électronique |
| Editeur : | [S.l.] : [s.n.], 2026 |
| Format : | 1 fichier PDF (8.5 Mo) / ill. |
| Note générale : |
Mode d'accès : accès au texte intégral par intranet.
Mémoire de Projet de Fin d'Études : Automatique : Alger, École Nationale Polytechnique : 2026 Appendis p. 99-107 .- Bibliogr. p. 108-109 |
| Langues : | Anglais |
| Index. décimale : | PA01026 |
| Tags : | Energy Reinforcement Learning Prediction Optimization Peak Shaving LSTM Smart Grid Load Shifting Artificial Intelligence |
| Résumé : |
This end-of-study project proposes a predictive energy management system aimed at optimizing electricity consumption and reducing costs in residential and industrial sectors.
A hybrid artificial intelligence approach is developed, combining LSTM forecasting with reinforcement learning decision-making. The LSTM model predicts electrical load and photovoltaic production, enabling anticipation of future grid behavior. Based on these forecasts, reinforcement learning agents optimize energy usage strategies including battery scheduling and flexible load management. Different RL algorithms are applied depending on the context: SAC for industrial microgrids, PPO for residential optimization, and DQN for discrete load shifting. This integrated approach enables peak shaving and demand smoothing. Simulation results show up to 30% cost reduction, reduced peak demand, and improved grid stability. The system is validated using a KNX-based hardware setup with Python control. |
Exemplaires (1)
| Code-barres | Cote | Support | Localisation | Section | Disponibilité | Spécialité | Etat_Exemplaire | ||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| EP01059 | PA01026 | Ressources électroniques | Bibliothèque centrale | Projet Fin d'Etudes | Disponible | Automatique | Téléchargeable |
