Neural network simulation for selective dissociation of SF6 molecules by single-mode and multi-mode carbon dioxide lasers

Abstract

Selective dissociation of sulfur hexafluoride molecules was simulated using single-mode and multi-mode carbon dioxide lasers using a two-layer feedforward neural network with 4 and 2nodes in the first and second layers, respectively. Existing experimental results (Sajad et al., 2004 [1]) were used to train and verify neural networks. Simulation was performed using MATLAB software neural network toolbox. Using the trained neural network, the results of non-existent physical conditions in the previous experimental results were predicted and presented.