Dynamic programming improved online fuzzy power distribution in a demonstration fuel cell hybrid bus

The optimal energy management strategy of power sources plays a vital role in enhancing the efficiency and extending the driving range of fuel cell electric vehicles. However, current studies predominantly focus on optimizing energy management for standard driving cycles, which may differ significantly from actual driving cycles. This paper addresses this gap by developing a dynamic programming-based fuzzy logic strategy for optimizing energy management in a demonstration fuel cell hybrid bus. It begins by introducing the propulsion system of the fuel cell hybrid bus and the specific driving cycles used in the demonstration. Subsequently, a dynamic programming-based fuzzy logic strategy is derived by considering the relationship between fuel cell power, battery state of charge, and demand power from the optimal allocation strategy of dynamic program -ming, which is tailored to the unique characteristics of the demonstration driving cycles instead of a generic standard driving cycle. The effectiveness and real-time performance of the dynamic programming-based fuzzy logic strategy are then validated through offline simulation and hardware-in-the-loop simulation. The results highlight that the proposed method outperforms the original rule-based strategy regarding energy efficiency, particularly when the initial state of charge is low and the energy consumption per unit mileage of the driving cycle is high.

Automated summary

This paper develops a dynamic programming-based fuzzy logic strategy for optimizing energy management in a demonstration fuel cell hybrid bus. The strategy considers the relationship between fuel cell power, battery state of charge, and demand power. It is tailored to the unique characteristics of the demonstration driving cycles and shows improved energy efficiency compared to the original rule-based strategy, especially under certain conditions.