Adaptive neuro fuzzy control of a high gain bidirectional power converter for photovoltaic-hydrogen renewable electric vehicles with enhanced lifespan and reliability

Abstract

The demand for green energy and application of hydrogen or photovoltaic (PV) for electrical vehicles (EVs) are enhancing steadily each day. DC-DC converters are critical power conversion systems that regulate voltage and current levels for battery packs in electric vehicles (EVs) powered by fuel cells (FCs) or PV panels and set the voltage for electric motor through an inverter circuit. The longevity of renewable energy sources (RESs) such as the FCs and PV arrays is heavily influenced by the current drawn by the DC-DC converter. Additionally, the converter topology must be cost-effective, minimize voltage and current stresses on semiconductor devices, offer ease of control, and provide flexible voltage outputs to meet the dynamic demands of the battery pack. This study introduces a switching DC-DC power converter designed specifically for FC-based electric vehicles (FCEVs), controlled by an innovative adaptive neuro fuzzy controller (ANFC). The high gain of the proposed converter enables the energy obtained from FCs and PV cells to be stored in a high-voltage battery pack and subsequently used to drive the electric motor and other electric vehicle components (such as lighting, heating, or cooling). This implies that, in an electric vehicle, it is sufficient to use only the proposed power converter instead of employing separate DC-DC converters for different energy sources such as PV or FCs. Subsequently, the stored energy can be used to operate the motor by providing the input voltage to the inverter. This approach makes the overall system more efficient and cost-effective. At the end of the simulation studies, it was observed that the proposed controller successfully ensures the control of the DC-DC converter, that no overshoot or oscillation occurs at the converter output, that an extremely short settling time of 0.016 s is achieved, and that a very low steady-state error of 0.7 is obtained. Experimental results for the proposed power converter are presented, thereby validating the theoretical findings.