Modelos lineares para estimação da energia em tempo real do calorímetro de telhas do ATLAS

Abstract

The LHC (Large Hadron Collider) is the largest particle accelerator in the world, providing an environment full of challenges for various fields of knowledge. Given the LHC’s ongoing upgrade to high luminosity (HL-LHC), expected to be completed by 2029, a new reading electronic architecture will be implemented to enable operation in freerunning mode for digital signal processing. In this HL-LHC operating context, a considerable increase in the number of interactions per collision is expected, which brings significant challenges to processing algorithms, especially in the calorimeters, where signal stacking can compromise energy resolution. The tile calorimeter of the ATLAS experiment at the LHC, currently used to identify and measure particles generated in high-energy proton collision events, employs the Optimal Filter 2 method, which aims to minimize the variance of electronic noise. However, in high luminosity scenarios, where calorimeter response signals overlap due to high occupancy, this method may introduce systematic errors, making it difficult to accurately estimate the signal amplitude. In this context, this study aims to explore and compare the performance of different linear signal estimation methods: Optimal Filter 2, Optimal Filter 1, and least squares with and without constraints, considering both fixed pedestal and variable pedestal conditions. The results show that the least squares method with constraints is the best alternative under pedestal variation conditions.