Advanced AI-driven MPPT and control algorithms for performance improvisation in dual-stage SEPIC-Buck converters for solar-powered DC drives

• Theoretical Design and simulation of a solar-powered dual-stage SEPIC-Buck converters for DC drive applications using advanced AI-based MPPT and control algorithms. • Hardware implementation of AI algorithms optimised dual-stage SEPIC-Buck converters using VHDL-code in FPGA platform for real-time control and monitoring. • Analysis of transient and steady state performance metrics of the proposed SEPIC-Buck converters with real world challenges. • Load regulation analysis of the proposed SEPIC-Buck converters for 12 V and 8.33 A output from a 100 W solar panel. Single-Ended Primary-Inductor Converter (SEPIC) topology, combined with optimization algorithms, plays a crucial role in renewable energy systems due to its ability to perform both step-up and step-down DC voltage conversion with smooth input current ideal for solar energy maximum power tracking. This study introduces a dual-stage SEPIC-Buck converter to optimize solar power extraction and regulate output voltage for a DC motor load. Traditional microcontroller-based control methods offer voltage regulation but struggle with transient response. To overcome this, the proposed system employs Field-Programmable Gate Arrays (FPGAs) for faster feedback and enhanced dynamic performance. The SEPIC stage integrates Perturb and Observe (P&O), Particle Swarm Optimization (PSO), and Artificial Neural Network Gauss-Seidel (ANN-GS) algorithms for Maximum Power Point Tracking (MPPT). The buck converter stage uses Dragonfly Algorithm (DF), Fuzzy Logic (FL), and Proportional-Integral-Derivative (PID) control to maintain a stable 12 V output for the DC Drive. Control logic was developed and simulated prior to hardware implementation. Experimental results validated the system’s efficiency, optimization algorithms extracted up to 94.32 W from the 100 Watts solar panel, while control strategies ensured consistent 12 V output. The converter system, using ANN-FLC algorithms, achieved a DC link voltage of 21.5 V and delivered 94.32 W output power (≈ 94% efficiency) to the DC drive. This research demonstrates the effectiveness of intelligent control techniques in improving solar-powered converter performance, offering promising applications in industrial and automotive sectors.