Intelligent MPPT control for high-gain converter design to enhance PV energy system power extraction efficiency

The energy extraction efficiency of photovoltaic (PV) systems operating under dynamic environmental conditions is achieved via an effective converter design and Maximum Power Point Tracking (MPPT) algorithms driving it. The work has proposed a Diode-Capacitor Multiplier-based DC-DC converter for a solar energy system, which is tested under variable irradiation levels between 1000 W/m 2 and 400 W/m 2 and partial shading scenarios. The MPPT algorithm proposed is Global–Local Optimization Reinforcement Learning, which is a hybrid intelligent control framework that integrates Grey Wolf Optimization (GWO) with Proximal Policy Optimization (PPO) as a reinforcement learning agent. GWO executes global optimization to find the best PPO hyperparameters and reward-function coefficients, which enhance the system's training starting point and its ability to reach solutions. The PPO agent learns an optimal control policy by interacting with the PV and new converter design using MATLAB, where partial shading conditions with variable irradiation inputs are modeled. This enables it to create adaptive duty cycle commands for converter control. The results found that the tracking time t GLORL − PPT of the proposed converter was a minimum of approximately 0.255 s, outperforming t GLORL − PPT = 0.34 sec of the PPO algorithm and t GWO − PPT = 0.42 sec of GWO method, which resulted in the reduction of transients in the power outputs significantly.