Robust DC–DC and Inverter Control for PEM Fuel-Cell/Battery/Supercapacitor Hybrid Electric Vehicles Using Fuzzy Logic and MOPSO under Load Transients and Parameter Uncertainty
Adel Elgammal
*
Utilities and Sustainable Engineering, The University of Trinidad & Tobago UTT, Trinidad and Tobago.
*Author to whom correspondence should be addressed.
Abstract
This paper presents a reliable power-electronics control and energy-management system for PEM fuel-cell electric vehicles (FCEVs) equipped with hybrid energy storage (battery–supercapacitor) to guarantee stability of the system in the presence of sudden load variations and parameter intervals. A unidirectional DC–DC interface for the PEMFC, bidirectional DC–DC stages for the battery and supercapacitor, and a traction inverter feeding the electric drive are included in this system. Real-time power split of the fuel cell and supercapacitor is achieved by a fuzzy-logic supervisory controller, which employs rule-based decision making with respect to NBC traction demand, DC-link voltage error, PEMFC ramp constraints, battery/supercapacitor SOC/SOE limits. A multi-objective particle swarm optimization (MOPSO) layer optimizes fuzzy membership parameters and weighting factors with multiple objectives that enable the minimization of hydrogen consumption, DC-link ripple, peak converter currents, and inverter stress as well as the imposition of fuel-cell current-slew rate limits and storage health constraints. Stability is tested with dual uncertainties in fuel- cell polaristic/ohmic parameters and converter/inverter elements (e.g., ±20% fluctuations in the main electrical parameters), and under aggressive step-load perturbations.
Simulations with representative drive cycles along with step-load disturbances (e.g., 0.3–0.5 pu torque/power steps) reveal that the fuzzy–MOPSO approach reduces DC-link voltage overshoot by 40–60% and settling time by 30–45% compared to cascaded PI control and non-optimized fuzzy supervision were observed. The DC-link ripple amplitude was reduced by 35–55%, and the peak currents in the supercapacitor and battery utilization were decreased by 20–35% demonstrating less thermal stress and increased reliability. The supercapacitor provided the high-frequency transients, while the battery the midfrequency dynamics, and the PEMFC supplied nominal power with safe ramp limits, achieving a reduction in hydrogen consumption of 5-12% across tested cycles. In summary, the results demonstrate that fuzzy–MOPSO co-design offers a practical method to robust, efficient and durable FCEV powertrain control under real driving fluctuations and uncertainties.
Keywords: PEM fuel-cell hybrid electric vehicle (FCEV), battery–supercapacitor hybrid energy storage, fuzzy-logic energy management, multi-objective particle swarm optimization (MOPSO), robust DC–DC and inverter control, load transients, parameter uncertainty