Título: Estimação de Medidas de Altas Energias em Calorimetria utilizando Múltiplas Redes Neuronais
Autores: Vassali, Marcos da Rocha; Ribeiro, Cássio Barboza; Seixas, José Manoel de
Resumo: In high energy physics, calorimeters play an important role, as they are highly-segmented detectors that are able to measure the energy of incoming particles by combining hundreds of readout cells. In practical applications, calorimeters typically suffer from some degree of non-linearity, which must be compensated for to improve the final accuracy of their measurements. In this work, a neural technique for compensating for such non-linearities in calorimetry is presented. Combining the energy estimation of three artificial neural networks, each one dedicated to a specific energy range of the full dynamic range to be covered in the measurement, the energy of particles can be better estimated, when a comparison with classical linear combination of cells is performed. Experimental results show that such combined neural estimate achieves a linearity better than 4.5%.
Código DOI: 10.21528/CBRN2001-045
Artigo em pdf: 5cbrn_045.pdf
Arquivo BibTex: 5cbrn_045.bib