A data-driven predictive energy management strategy for plug-in hybrid vehicles

Plug-In Hybrid Electric Vehicles show great potential for decreasing the fuel consumption on specified routes. However, in many cases the trip destination or the distance until the next charge is unknown for the vehicle. This paper presents a data-driven, online energy management strategy that is based on a trip and speed profile prediction for a receding horizon, which takes personal points of interest or upcoming charging stations into consideration. Pontryagin's Minimum Principle including a reduced shooting algorithm is applied to optimize the vehicle state. We evaluated the method for multiple trips of varying length and expect an estimated fuel saving of 8.0% compared to a non-predictive approach.     «

Plug-In Hybrid Electric Vehicles show great potential for decreasing the fuel consumption on specified routes. However, in many cases the trip destination or the distance until the next charge is unknown for the vehicle. This paper presents a data-driven, online energy management strategy that is based on a trip and speed profile prediction for a receding horizon, which takes personal points of interest or upcoming charging stations into consideration. Pontryagin's Minimum Principle including a...     »