A robust predictive energy management for plug-in hybrid vehicles based on hybrid optimal control theory

This paper presents a predictive energy management for plug-in hybrid vehicles, feasible for the implementation on an electronic control unit. An optimal reference trajectory for the battery's state of charge is calculated by solving a hybrid optimal control problem that considers continuous controls as well as discrete decisions, such as the switching between drive modes. The problem is solved via an indirect variation of extremals approach, expanded for hybrid systems. To decrease calculation time, information necessary for solving the problem is calculated offline and stored in lookup tables. The information necessary for the formulation of the hybrid optimal control problem are obtained using information from modern navigation systems for each segment of the route. A driver model transforms this set of information into a continuous driving profile that contains all necessary information for solving the hybrid optimal control problem. A PI-controller is implemented to follow the reference trajectory of the battery's state of charge. The controller output is an offset to the costate obtained from the solution of the hybrid optimal control problem. The energy management was implemented in a pilot-production plug-in-hybrid vehicle. Results obtained from real world measurements are presented as well as simulation results to investigate the energy management's robustness.