Résumé:
The increasing demand for energy efficiency and sustainability in smart
home environments has spurred the development of advanced energy
management systems (EMS). This thesis proposes a novel approach
integrating multi-agent reinforcement learning (MARL) with fuzzy logic
for efficient energy management in smart homes. The system employs a
distributed architecture where autonomous agents interact with various
smart devices to optimize energy consumption while considering user
preferences and comfort levels.
The use of MARL enables the system to adapt and learn from dynamic
environments, allowing for real-time decision-making and optimization of
energy usage. Each agent operates independently, yet collaboratively, to
achieve the overarching goal of minimizing energy consumption and costs
while maintaining user comfort. Fuzzy logic is incorporated to handle
uncertainties and imprecise data inherent in smart home environments,
providing robustness and flexibility to the decision-making process.
The proposed system demonstrates significant improvements in energy
efficiency compared to traditional approaches. Furthermore, the
integration of fuzzy logic enhances the system's ability to handle complex
and uncertain environments, resulting in more reliable and adaptive
energy management solutions for smart homes. This research contributes
to advancing the field of smart home automation by offering a scalable and
intelligent energy management system capable of optimizing energy usage
while ensuring user satisfaction and comfort.
