Photovoltaic (PV) Cell Parameter Objective Optimization Using Novel Convolutional Neural Network (CNN) Technique

This paper presents a novel technique from the family of artificial intelligence to optimize the essential modeling parameters of Photovoltaic (PV) cell. Though many parameter extraction techniques are available in the literature, still there is a scope to improve the modeling efficiency due to the incorporation of advanced materials in PV cell, low response time, and mainly the wide difference between the extracted values and actual values. Therefore, the proposed Convolutional Neural Network (CNN) technique addresses and overcomes the above issues by accepting the nonlinear synthetic data. The proposed CNN technique architecture has six layers, technically these layers are classified as Convolution layer, Rectified Linear Unit (ReLU) layer, Padding layer, Pooling layer, Flattening layer, fully connected layer, and Output layer. The effectiveness of the proposed CNN technique in terms of response time, and closeness to the actual value is compared with Lightning Search Algorithm (LSA), Simulated Annealing (SA), Differential Evolution (DE), Artificial Bee Colony (ABC), and Advanced Particle Swarm Optimization (APSO) using 85W, 125W and 200W solar module.