Development of RBFNN Machine Learning MPPT Algorithm for Improved KSK Converter

Abstract

Solar energy systems are less stable and less efficient because of the variation in irradiation and temperature. Therefore, an advanced control and power conversion method is required to ensure efficient energy extraction. This paper describes a solar-powered DC–DC conversion system that uses an Improved KSK Converter and a smart Maximum Power Point Tracking (MPPT) strategy. The solar panel is the main source of power, it makes different amounts of voltage and current depending on the amount of sunlight and heat it gets. A Machine Learning (ML) MPPT controller based on Radial Basis Function Neural Network (RBFNN) is used to generate the optimal control signals for maximum power extraction. The Improved KSK converter offers high voltage gain, low ripple and stable output performance. A PWM generator processes these signals to boost and regulate the voltage effectively. The proposed method is better than traditional methods because it has a conversion efficiency of 95% and an MPPT tracking efficiency of 95%.