SMuK 2023 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

T: Fachverband Teilchenphysik

T 127: Exp. Methods III

T 127.2: Vortrag

Donnerstag, 23. März 2023, 16:05–16:20, WIL/C129

Matrix inversion in the context of a novel track reconstruction algorithm for the ATLAS Event Filter — •Antara Paul¹ and Andre Schöning² for the ATLAS collaboration — ¹Physikalisches Institut, Universität Heidelberg, Germany — ²Physikalisches Institut, Universität Heidelberg, Germany

The High Luminosity LHC project is expected to provide a tenfold increase of the integrated luminosity compared to the LHC. To cope with the resulting high pile-up from proton-proton collisions, the ATLAS detector, and its trigger and DAQ systems are undergoing major upgrades. As a part of the upgrade of the online event filter, a fast triplet track reconstruction algorithm is being developed based on a broken line fit. The track parameters are estimated by minimizing a χ² function, which includes the multiple scattering and spatial hit uncertainties at each layer. This minimization involves the inversion of a matrix.
In this context, the talk focuses on different algorithms of matrix inversion, including but not limited to LDL^T decomposition and the partition method. The speed and accuracy of each method will be presented and compared, in view of their implementation in the track reconstruction algorithm.