ANN-Based CTR Modelling and Validation Results

Artificial neural network(ANN) is an efficient computing algorithm in MATLAB that emulates the biological neurons performance for the basic functions such as the human brain. Compared to other traditional methods, the ANN soft computing technique provides wide information in multi-dimensional information domains, accurate to solve complex and nonlinear problems, and less time consumed. Presently, ANN technique has been widely used in several applications of renewable energy technologies, particularly solar energy systems. It is used to effectively model, simulate, control, optimize and analyze solar energy systems. Consequently, this chapter presents the most important models of ANN technique that used in the solar energy fields and the criteria for selecting the optimal model. Also, it offers a new, simple, and accurate method for modeling CTRPP. This technique can control the flow rate of HTF from CST to the tower receiver. Thus, the receiver outlet temperature can be controlled at the required value regardless of the change in solar radiation or the receiver inlet temperature. Additionally, it contains a detailed explanation of the creation steps of the neural network model. Additionally, it presents the results of the proposed model described in Chaps. 3 and 4. Comparisons between ANN models to select the optimal model are discussed in this chapter. It was found that the MLP model the optimal model for controlling the receiver outlet temperature by adjusting the flow rate of the HTF. The proposed model results were compared with the results of SAM simulation program. The results showed full compatibility between them.