Determination of the parameters of two-beam laser cleaning of quartz raw materials using artificial neural networks and the finite element method

With the help of artificial neural networks, the process of two-beam laser cleaning of quartz raw materials has been modeled. For the formation of training data sets and data sets for testing neural networks, the ANSYS finite element analysis program was used. The calculations were performed for 500 variants of input parameters, 40 of which were used to test neural networks. The influence of the parameters of neural network models on the accuracy of determining the maximum temperatures in quartz particles formed as a result of two-beam exposure were studied. The parameters of neural networks were determined that provided acceptable results when predicting temperatures in the laser treatment zone. The results obtained can be used in determining the technological parameters of the processes of two-beam laser cleaning of quartz raw materials.