Maximum Power Tracking Using Machine Learning Technique in Hybrid Solar-Wind Renewable Energy Power Generation

Hybrid renewable energy sources (HRES), particularly solar and wind, have gained interest and grown rapidly to address the energy crisis. At the same time, the problem arose due to the nonlinearity of the sources over time. Regardless of the environmental conditions, HRES can operate at the maximum power point using Maximum Power Point Tracking (MPPT) technology. Due to the economic advantages, several controllers are being developed to maximize the performance of the HRES. The traditional controllers had limitations: slower response speed and different step sizes. To ensure the maximum availability of power from the HRES under varying system conditions, the hybrid Gradient Boosting Decision Tree (GBDT) controller with Walrus Optimization Algorithm (WOA) is proposed in this work. The proposed WOA hybrid GBDT (WOHGD) studies the optimal duty cycle of the boost converter in a solar wind connected power system. The proposed work is implemented on the MATLAB/SIMULINK tool, and the results are verified with existing approaches.