Proactive Control of Hybrid Electric Vehicles for Maximum Fuel Efficiency

Abstract

This paper proposes a Markov decision process based model for proactive control of hybrid electric vehicle that results in optimal control policy which maximizes fuel efficiency. Novel aspect of the model is inclusion of supplied power in the state space and inclusion of difference between supplied and demanded power in the reward function of the model. This inclusion allows for supplied and demanded power of the vehicle to be different which is not catered for in previous work. Another advantage of the proposed model over existing approaches is proactive nature of control, in that, the control policy restricts the supplied power when the vehicle is running low on fuel and/or battery charge level. Such control can save precious fuel and avoid the situation of being stranded on a highway with empty fuel tank. Simulation based case study has been included for demonstration of the results from optimal policy.