Optimization of the nonlinear model of neural network training in predicting thermal efficiency of solar concentrator with simulated annealing algorithm

Nowadays, artificial neural networks are widely used to solve large-scale and complex problems. The purpose of this study is to use artificial intelligence techniques such as artificial neural networks and simulated annealing algorithm, to optimize the solar energy system in order to maximize its economic benefits. Here, a new nonlinear computational model has been presented to predict the thermal performance of compound parabolic concentrator (CPC). In this regard, three models of artificial neural network (ANN) including radial-basis function (RBFANN), multi-layer perception (MLPANN) as well as adaptive neuro fuzzy inference system (ANFIS) are used to identify the nonlinear relationship between input and output parameters of the system. The optimal nonlinear structure of the model is determined through the simulated annealing (SA) method. Validation of the proposed models is performed on a CPC. The results show that all the three models are efficient. In particular, statistical analyses show that the ANFIS model is more accurate in the process of predicting thermal efficiency. So, the given models can be effectively applied in practical fields.