Modeling and Prediction of Laser Generation in UV Copper Bromide Laser via MARS

摘要

A case study of the multidimensional dependence of average output laser power on basic input laser parameters in UV Cu+ Ne-CuBr laser is presented. A nonparametric statistical analysis of large amount of experimental data is carry out by multuvariate adaptive regression splines (MARS) technique. The obtained results are in good agreement with practical issues. It is shown that the constructed best MARS models can be applied in estimation and prediction of current and future experiments in order to improve the laser generation.