Estimação de energia e qualidade de dados em condições de fina segmentação e alto ruído de empilhamento

• Ano de defesa: 2019
• Autor(a) principal: Pereira, Rodrigo Araujo
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Rio de Janeiro; Brasil; Instituto Alberto Luiz Coimbra de Pós-Graduação e Pesquisa de Engenharia; Programa de Pós-Graduação em Engenharia Elétrica; UFRJ
• Palavras-chave em Português: Engenharia elétrica; Processamento de sinais; Física de partículas; Estimação de sinais; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA
• Link de acesso: http://hdl.handle.net/11422/12290

The hadronic calorimeter (TileCal) of ATLAS (A Toroidal LHC ApparatuS), one of the major LHC (Large Hadron Collider) particle accelerator experiments at CERN, consists of more than 10,000 read channels that work at a 40 MHz event rate. The quality of the results obtained in this experiment depends on the correct estimation of the energy of the particles that interact with its material. The energy estimation can be compromised by a number of factors, such as noisy channels, the method chosen for the online or offline estimation of energy and, mainly, by electronic and pile-up noise. The present work presents a method that uses a minimum variance estimator to mitigate noise in groupings of reading channels of a calorimeter built with read redundancy. It is also be shown that this method can be used to identify and mask noisy channels of a calorimeter. We will also present measures to evaluate energy estimation algorithms using real particle collision data. The results show that the proposed method achieves better precision of energy estimation by up to 41% without compromising and, in some cases, improving its accuracy approximating the estimation to real value. The method is also independent of the estimation algorithms used for the channel, as well as being effective in several stacking noise scenarios. The algorithm evaluation measures were effective in evaluating an online estimation algorithm in TileCal.