Hybrid model predictive control tracking of a sawtooth driving profile for an HEV

Based on a two-mode operation model of a parallel hybrid electric vehicle (PHEV) developed in previous work, this paper presents a hybrid optimal control solution for the power management problem of a PHEV. The optimal power flows between the vehicle's main subsystems are computed as solutions of a switching-system optimization (nonlinear programming problem) formulated at the supervisory level assuming a hierarchical vehicle control structure. The resulting equations are numerically solved using collocation methods. The approach is illustrated using the optimal and MPC tracking of a sawtooth velocity driving profile. Unlike other work, the assumption that there is always sufficient power to achieve perfect tracking is not made in this work.