A genetic algorithm approach for multiplet deconvolution in γ-ray spectra

The assay of radioactive material by γ-ray spectrometry often involves processing overlapping photopeaks. In this paper, we propose a new method for deconvolution of γ- and X-peak complex multiplets consisting of a genetic algorithm followed by a grid search. Although this approach is more costly in computer time than traditional techniques, it is more robust and minimises user intervention. We use artificially generated four-peak multiplets with varying peak height ratios to test the performance of the method. Furthermore, to illustrate its capabilities on real spectra, we apply it to the deconvolution of the K–X spectral region from natural sediment samples.