Adaptive Variable Step Size Photovoltaic MPPT Algorithm Based on Light Intensity Prediction

To enhance the efficiency and stability of maximum power point tracking (MPPT) in photovoltaic (PV) power generation systems, and to address the issue of misjudgment in the traditional perturbation and observation method (P&O), an adaptive MPPT control approach based on light intensity prediction is proposed. Firstly, the correlation between the input and output of the system is analyzed to determine the necessary input terms. Secondly, the multi-time series LSTM algorithm is utilized to select pertinent historical PV sequence information according to the correlation analysis results, enabling short-term prediction of PV output. Finally, the Boost circuit is employed as the control object, with the light intensity prediction result serving as the input to adaptively adjust the step size of the P&O. Through the MATLAB/Simulink simulation platform, the results demonstrate that the adaptive variable step size photovoltaic MPPT algorithm based on light intensity prediction can effectively mitigate power oscillation in steady state while exhibiting good tracking speed and accuracy.