Predictive Energy Management for Electric Variable Transmission HEV

This paper introduces predictive energy management for a hybrid electric powertrain with an electric variable transmission. The optimization of the energy management in hybrid mode is formulated as a bi-level program, in which one static level optimizes the engine speed of a compound unit consists of the engine and electric variable transmission (EVT). The other layer optimizes a dynamic program by splitting power between the compound unit and the battery using equivalent consumption minimization strategy (ECMS). By combining ECMS with dynamic programming (ECMS-DP), the proposed strategy allows the engine on/off control to be also taken into account by iteratively updating the costate. Finally, the efficient ECMS-DP optimization is incorporated into the framework of model predictive control (MPC) and solved in a moving horizon fashion. Computation efficiency and optimization results are presented by simulation.