Computer methods for ITER-like materials LIBS diagnostics

摘要

Recent development of Laser-Induced Breakdown Spectroscopy (LIBS) caused that this method is considered as the most promising for future diagnostic applications for characterization of the deposited materials in the International Thermonuclear Experimental Reactor (ITER), which is currently under construction. In this article the basics of LIBS are shortly discussed and the software for spectra analyzing is presented. The main software function is to analyze measured spectra with respect to the certain element lines presence. Some program operation results are presented. Correct results for graphite and aluminum are obtained although identification of tungsten lines is a problem. The reason for this is low tungsten lines intensity, and thus low signal to noise ratio of the measured signal. In the second part artificial neural networks (ANNs) as the next step for LIBS spectra analyzing are proposed. The idea is focused on multilayer perceptron network (MLP) with backpropagation learning method. The potential of ANNs for data processing was proved through application in several LIBS-related domains, e.g. differentiating ancient Greek ceramics (discussed). The idea is to apply an ANN for determination of W, A1, C presence on ITER-like plasma-facing materials.