Stochastic self-optimizing power management for fuel cell hybrid scooters of different sized components

Abstract

Electric-powered scooters supplied by hybridizing a fuel-cell system (FCS) with lithium-ion batteries (LIB) need a good power management strategy (PMS) for attaining long driving distance and preventing from damage to the FCS or LIB while supplying adequate power to the vehicle. This paper proposes a stochastic self-optimizing power management strategy (SSOPMS) for the scooters to achieve this aim in various driving cycles. The power train of the scooter contains a lithium-ion battery and a boost converter interfacing an FCS to the DC-bus. The boost converter is under control of a power controller executing the trajectory of the SSOPMS. Various driving cycles are modeled as the discrete Markov process which extracts the statistical features from standard driving cycles. The SSOPMS determines the fuel-cell power trajectory depending on minimum fuel consumption and the constraints of FCS power, FCS power slope, LIB power, and LIB state of charge (SoC). To demonstrate the performance of the SSOPMS, two fuel-cell hybrid scooters of distinct FCS and LIB sizes are designed. Simulation results show the effectiveness of the SSOPMS under different components sizing and driving conditions.