The most relevant result of this thesis work is the development of a firstlevel trigger to fast filtering events for the Cherenkov Telescope Array (CTA) project. After a depth study of the discrimination events problem, which allowed to develop a filter working with two original algorithms called FCTAR and FCTAN. The first one, FCTAR, consists essentially in a criteria to ``cut'' the absolute timeline, in intervals of time, each of them containing at least one ore more events but with the strong condition that no event is divided into different timebox. The second one, FCTAN, is a original Network Filter Algorithm, which trasforms a compatibility matrix, well defined between the delays matrices in a network and than evolves the systems obtaining a clustering effect. The cluster corresponds at the simulated physical events. All this procedure is innovative and probably never used first in CTA data analysis. Finally we tested our filter onto Official ASTRIMC data. Obtaining a very good performance with a purity of about 99.8% and the highest efficiency obtainable with the same purity. The deployment of the algorithm on a GPU system, made possible by the FCTAR condition and it is probably the our next step.
Fast filtering events for the Cherenkov Telescope Array
Urbani, Michele
2016/2017
Abstract
The most relevant result of this thesis work is the development of a firstlevel trigger to fast filtering events for the Cherenkov Telescope Array (CTA) project. After a depth study of the discrimination events problem, which allowed to develop a filter working with two original algorithms called FCTAR and FCTAN. The first one, FCTAR, consists essentially in a criteria to ``cut'' the absolute timeline, in intervals of time, each of them containing at least one ore more events but with the strong condition that no event is divided into different timebox. The second one, FCTAN, is a original Network Filter Algorithm, which trasforms a compatibility matrix, well defined between the delays matrices in a network and than evolves the systems obtaining a clustering effect. The cluster corresponds at the simulated physical events. All this procedure is innovative and probably never used first in CTA data analysis. Finally we tested our filter onto Official ASTRIMC data. Obtaining a very good performance with a purity of about 99.8% and the highest efficiency obtainable with the same purity. The deployment of the algorithm on a GPU system, made possible by the FCTAR condition and it is probably the our next step.File  Dimensione  Formato  

Tesi_LM_Urbani2.pdf
accesso aperto
Dimensione
4.15 MB
Formato
Adobe PDF

4.15 MB  Adobe PDF  Visualizza/Apri 
The text of this website © Università degli studi di Padova. Full Text are published under a nonexclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/24663