Artificial Intelligence Hybrid Control Technique of DC/DC Boost Converter for EV Battery and Motor Load Systems

Electric vehicles (EVs) require high-performance traction battery systems capable of delivering fast and efficient energy to electric motors for rapid acceleration. To meet the increasing demand for efficient power conversion, this paper proposes an artificial intelligence-based hybrid control technique for a boost converter supplying an EV battery load. The proposed approach combines conventional proportional–integral–derivative (PID) control with pulse-width modulation (PWM) to ensure the system’s stability, while a fuzzy NARMA-L2 controller is employed to enhance dynamic performance under varying operating conditions. The intelligent controller utilizes artificial neural networks (ANNs) to predict system behavior and adaptively regulate switching actions, reducing SSE and improving efficiency. A hybrid switching mechanism is implemented to enable smooth transition between conventional and intelligent control modes, thereby enhancing system reliability and preventing potential failure conditions. The performance of the proposed controllers is evaluated using multiple criteria, including cross-correlation analysis, which proves the stability and disturbance-rejection capabilities. Simulation and experimental validation using an Arduino Uno platform demonstrate the effectiveness of the proposed method. The results show a significant reduction in total harmonic distortion (THD), from 9.6% under PID control to 1.67% under AI-based operation load conditions; steady state error (SSE) is 0.02, and rise time (tr) is 0.3 s. The 0.02 SSE with 0.5% ripple were established in simulation under ideal parameter models, whereas prototype hardware tests demonstrated steady-state regulation within 1.67% voltage ripple and an overall system conversion efficiency of 78.0, running on dual 16 MHz microcontrollers. Additionally, the proposed strategy improves overall system performance by up to six times compared to conventional control, and up to seven times under motor-load conditions and variable inputs (5,9,12,16,24) V.