Adaptive Hybrid ANN and Incremental Conductance Approach for Maximum Power Point Tracking in Solar PV Systems

Abstract

Current research in solar photovoltaic (PV) energy is primarily directed toward improving the efficiency and output performance of PV arrays. The main objective is to reduce energy losses caused by changing weather patterns and irregular load variations. Traditional single Maximum Power Point Tracking (MPPT) methods frequently fail to strike an effective compromise between fast tracking response and high precision. To overcome this challenge, a novel hybrid approach is proposed that combines an Artificial Neural Network (ANN) with the Incremental Conductance (INC) method for effective MPPT. This paper introduces and demonstrates the implementation of the ANN-INC hybrid technique within photovoltaic systems for enhancing MPPT performance. The proposed hybrid approach utilizes the accuracy of the INC algorithm along with the adaptive learning abilities of the ANN to enhance tracking efficiency under diverse environmental conditions. This method is thoroughly assessed and compared with commonly used techniques such as Perturb and Observe (P&O), standalone INC, ANN, and Fuzzy Logic controllers. Simulation results demonstrate the effectiveness of the hybrid method, showing quicker convergence to the maximum power point, fewer oscillations, and enhanced performance under fluctuating irradiance levels. The ANN-INC technique achieves a high efficiency of approximately 99% and outperforms other approaches by reaching the maximum power point more rapidly with minimal energy losses and reduced fluctuations. Moreover, the hybrid method enhances power extraction efficiency, ensuring reliable and consistent operation of solar PV systems. This comparative analysis emphasizes the practical advantages of combining ANN and INC in addressing the limitations associated with traditional MPPT techniques. Finally, validation through real-time simulation using the  OPAL-RT platform confirms the applicability and effectiveness of the proposed approach in real-world conditions.