Battery SOC constraint comparison for predictive energy management of plug-in hybrid electric bus

In this paper, model predictive control (MPC) is employed to resolve the energy management problem of a plug-in hybrid electric bus (PHEB). Dynamic programming (DP), as a global optimization method, is inserted at each time step of the MPC, to solve the optimization problem regarding the prediction horizon. A multi-step Markov prediction model is constructed to forecast the near future driving velocities for the MPC. The battery SOC is restrained to fluctuate near a reference trajectory to ensure the global performance of MPC. Three novel restraining methods are proposed and compared in this paper. The resultant fuel economy performance with different SOC constraint methods are evaluated. Simulation results indicate that by restraining the battery SOC adaptively to the control variables gains the best fuel economy performance, and the fuel consumption of MPC is 8.7% less than a ruled based strategy.