Predicting the Output Power of a Photovoltaic Module Using an Optimized Offline Cascade-Forward Neural Network-Based on Genetic Algorithm Model

This paper aims to enhance the performance of a cascade-forward neural network (CFNN) model predict output power photovoltaic (PV) module. improvement is conducted by optimizing number hidden neurons using genetic algorithm (GA). The optimization carried out minimize value root mean square error (RMSE) between actual and predicted PV power. CFNN-based GA evaluated five statistical term terms; namely, RMSE, normalized RMSE (nRMSE), bias (MBE), MBE (nMBE), absolute percentage (MAPE). values for these terms are 0.0025 W, 0.0131%, 0.0011 0.0067% 0.8121%, respectively. proposed model’s compared with that obtained classical CFNN, feed-forward (FFNN), FFNN-based on in accuracy training testing time. results show performs better than models mentioned above based terms. MAPE which lower 98.63%, 98.77% 63.93% FFNF, FFNN-GA, Besides, it faster In which, FFNN-GA 89.68%, 59.50% 53.99% time, Moreover, also FFNN, 94.92%, 75.57% 78.95%, Accordingly, can be recommended modules.