AI-Based MPC Controller for Energy-Efficient HVAC Systems

Model Predictive Control (MPC) is gaining popularity for real-time applications in different fields, particularly in embedded technologies. This work investigates the use of predictive algorithms to regulate a thermal system in real-time. Specifically, we propose employing the MPC strategy to control a light bulb system with a Raspberry Pi while minimizing power consumption. This study details the procedural framework and thoroughly compares MPC with conventional methods such as the Proportional-Integral-Derivative (PID) controller. This comparison includes energy efficiency, cost estimations, and hardware performance metrics, highlighting the advantages and disadvantages of each approach.