Enhancing MPPT Efficiency in Standalone Photovoltaic Systems: Integration of Finite Control Set Model Predictive Control with Optimized Artificial Neural Network

This paper presents an advanced Finite Control Set Model Predictive Control (FSC-MPC) strategy tailored for optimizing Maximum Power Point Tracking (MPPT) in standalone photovoltaic systems. Leveraging a dynamic model of a boost converter, the FSC-MPC strategy integrates a unique cost function, which incorporates the absolute difference between the reference and predicted current to effectively regulate the boost switch, thereby enhancing reference tracking and minimizing losses. An advancement in this strategy involves the inclusion of an additional term in the cost function aimed at optimizing the switching frequency, thus reducing losses and attenuating undesirable oscillations. The FSC-MPC's current reference generation relies on an optimized Artificial Neural Network (OP-ANN) architecture, wherein parameters such as the number of hidden layers, neurons per layer, and activation function are dynamically adjusted using a pattern-search algorithm (PSA). By minimizing the loss function through adjusting these parameters, the optimization algorithm ensures more effective learning and enhances the ANN's adaptability to changing environmental conditions. The proposed model is implemented, modeled, and simulated within the MATLAB/Simulink environment, with extensive discussion of simulation results providing a comprehensive evaluation of the performance of the proposed algorithms.