Nature-inspired MPPT algorithms for solar PV and fault classification using deep learning techniques

In recent years, renewable energy attracts the researchers interest due to its environment free nature and abundant availability. Solar photovoltaic (PV) is widely used to generation power from the sun light. Major issue in solar PV power generation is tracking of the peak power from the available multiple power peaks in the operating points. A proper MPPT algorithm is required to capture the maximum power point (MPP) from the characteristic curves of a solar PV under partial shaded conditions (PSC). An optimized maximum power point tracking (MPPT) and fault classification in solar PV systems are presented in this research work. To select the best optimization model for MPPT under PSC, the nature-inspired dragonfly algorithm (DA), moth flame optimization algorithm (MFOA), grasshopper optimization algorithm (GOA), and salp swarm optimization algorithm (SSOA) are used in this work to evaluate the tracking efficiency (TE) of the solar PV systems. From the simulation results, SSOA exhibits a supreme TE of 98.38%, which is better than the other algorithms like DA, GOA, and MFOA. To further classify the faults in solar PV systems, random forest (RF), artificial neural network (ANN), support vector machine (SVM), and convolutional neural network (CNN) models are employed. Among all, CNN provides a maximum accuracy of 94.11% in fault classification. Simulation analysis demonstrates the proof-of-concept for maximum TE and classification accuracy for all the methods. Thus, the optimized MPPT and fault classification models can be combined to enhance the overall performance of solar PV systems.