Real-Time State-Constrained Power Management Based on Pontryagin’s Minimum Principle: Application to Fuel Cell Vehicles

The contribution of this article is to propose a power management system (PMS) for a hybrid vehicle based on an adaptive Pontryagin’s minimum principle (A-PMP) approach that optimally handles state constraints in an online setting. We show that the resulting algorithm is implementable and can run on a real-time target with performance very close to offline computation. Optimally enforcing state constraints [in this case, the battery state of charge (SoC)] allows the battery to be used more efficiently, enabling a reduction in its capacity without increasing energy consumption. The proposed algorithm, therefore, provides greater flexibility in the design and sizing of the vehicle’s energy architecture.