Energy Management in Fuel cell, Battery and Super Capacitor driven Hybrid Electric Vehicle using Reinforcement learning Algorithm

Hybrid Electric Vehicles (HEVs) integrating fuel cells, batteries, and super capacitors require an efficient Energy Management System (EMS) to optimize power distribution for improved performance, longevity, and efficiency. This paper proposes a Reinforcement Learning (RL)-based EMS that dynamically adjusts power allocation among energy sources based on real-time driving conditions. Unlike conventional rule-based methods, RL continuously learns and adapts to varying load demands, ensuring optimal energy utilization while minimizing fuel consumption and component degradation. The proposed EMS operates based on the vehicle’s driving cycle, responding to power fluctuations efficiently. Simulation results demonstrate that the fuel cell supplies peak power of 8.95 kW between 90s to 130s, while the battery transitions from charging to discharging based on energy demand. The super capacitor responds instantly to sudden power fluctuations, charging during excess power availability and discharging at critical moments (e.g., 136s, 186s, and 270s) when the fuel cell and battery alone cannot meet demand. The RL-based approach enhances overall system efficiency by improving power balancing and reducing energy losses compared to traditional methods. The proposed EMS leads to a reduction in fuel consumption and an extension of battery and super capacitor lifespan, making it a viable solution for sustainable and intelligent HEV operation.