An artificial intelligence approach for prediction of hole taper in Nd: YAG laser drilling on thermoplastic material reinforced by glass fiber

Laser drilling is an efficient technique for producing intrinsic holes with a diameter of less than 1mm. The critical parameters affecting laser power in a pulse mode of Nd: YAG laser drilling are pulse current, pulse duration, assist gas pressure and workpiece thickness. This study explores an artificial intelligence (AI) technique to forecast the hole taper of Nd: YAG laser drilling. The data acquired from Taguchi's L27 orthogonal array experiment were employed to construct an artificial neural network (ANN) model. The hole taper was represented mathematically as an explicit nonlinear function of the chosen input parameters. Studies have shown that ANNs can be successfully used to analyze the influence of laser drilling parameters on the hole taper produced by Nd: YAG laser drilling on GFRP material. Compared to the experimental outcome data set, the predictive ANN model based on 4-6-1 architecture had the lowest error (MSE=0.001) and an absolute average percentage error of less than 8.94%.