Online Energy Management Strategy of FC/UC Hybrid Electric Vehicle Based on Neural Network

Fuel cell (FC) vehicles seem to be the solution for clean and efficient transportation. However, the coupling of the fuel cell with another power source remains essential to overcome its low power density and preserve its lifetime as long as possible. Thanks to its fast response and high power density, the ultra-capacitor (UC) can meet a fast and short-duration engine power demand. This paper presents a power management strategy between the FC and the UC, based on a neural network (NN). The proposed NN model is trained with wavelet decomposition outputs. To ensure that both sources operate optimally, the fuel cell power limits as well as the state of charge (SOC) of the UC are considered.