Forecasting of Photovoltaic Power by Means of Non-Linear Auto-Regressive Exogenous Artificial Neural Network and Time Series Analysis

In this research paper, a nonlinear autoregressive with exogenous input (NARX) model of the system based on neural network and time series analysis is proposed to deal one-month forecast produced power from photovoltaic modules (PVM). The PVM monocrystalline cell rated production 175 watts that placed at Heliopolis University, Bilbéis city, Egypt. NARX considered powerful enough emulate dynamic state-space model. It extensively performed resolve variety problems mainly important in complex process control. Moreover, method selected because its quick learning completion times, as well high appropriateness, distinguished by advantageous dynamics interference resistance. (NN) trained optimized three algorithms, Levenberg–Marquardt Algorithm (NARX-LMA), Bayesian Regularization (NARX-BRA) Scaled Conjugate Gradient (NARX-SCGA), attain best performance. forecasted results using algorithms are compared experimentally measured data. NARX-LMA, NARX-BRA NARX-SCGA models validated statistical criteria. general, weather conditions have significant impact execution quality results.