Model Reduction Methodology for Energy Management Strategy of Hybrid Electric Vehicles

Energy management strategy (EMS) is the key for efficient hybrid vehicles. To design EMS, the two main approaches are the: optimal-based algorithms and rule-based algorithms. Optimal-based EMS allows reaching of the best behaviour but required backward simulation and it is an off-line method. Rule-based EMS which is based on heuristic rules can be used on-line and with a forward simulation but are not very efficient. Many research has proposed to deduce rule-based EMS from optimal-based EMS analysis. The paper objective is to propose a methodology to structure the deduction of this design to avoid errors during the switching between backward and forward simulation of the deduced rule-based EMS.