BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Genetic Algorithm for determination of event collision time and pa
 rticle identification by Time-Of-Flight at SPD NICA
DTSTART;VALUE=DATE-TIME:20221201T161500Z
DTEND;VALUE=DATE-TIME:20221201T163000Z
DTSTAMP;VALUE=DATE-TIME:20260719T191454Z
UID:indico-contribution-3137@cern.ch
DESCRIPTION:Speakers: Semyon Yurchenko (Saint Petersburg State University\
 , Laboratory of ulta-high energy physics)\nStudying spin-related phenomena
  will help uncover information about internal structure of nucleon. It can
  be done using polarized beams of protons and deuterons. To study those ef
 fects on NICA collider it is proposed to install the Spin Physics Detector
 (SPD) in one of two interaction points.\n\nIn this work two problems were 
 studied. First is developing a fast and accurate algorithm for determinati
 on of $pp$-collision time($t_{0}$) for the SPD experiment [1]. Second is d
 eveloping particle a identification(PID) procedure based on $TOF$ signals.
 \n\nUsage of $TOF$ detectors to identify particles is one of the most reli
 able technique. To reconstruct the particle mass only three values are nee
 ded: $\\vec{p}$ - momentum of the particle\, $L$ - arc length of the parti
 cle trajectory and $TOF$ - Time-Of-Flight. To measure $TOF$ one needs star
 t($t_{0}$) and stop signals. While stop signals are obtained with $TOF$ de
 tector\, $t_{0}$ cannot be measured directly. Main idea for $t_{0}$ determ
 ination: for tracks in the event find a set of masses that minimises sum o
 f the squares of the residuals ($\\chi^{2}$). \n\nTo solve $\\chi^{2}$ min
 imisation problem Differential Evolution-inspired [2] Genetic Algorithm(GA
 ) was developed and it's results were compared to Brute Force Algorithm(BF
 A). BFA have very slow run time\, but it provides the exact solution to th
 e $\\chi^{2}$ minimisation problem. So main requirements to GA was to decr
 ease run time while keeping high accuracy. \n\nWith a reliable method for 
 $t_{0}$ determination\, particle identification can be performed. There ar
 e several approaches that can be used for PID procedure [3]. In this work 
 Bayesian approach\, "n-sigma" criteria and solution for $\\chi^{2}$ minimi
 sation problem were compared.\n\n\n  1. Conceptual design of the Spin Phys
 ics Detector. arXiv:2102.00442\n  2. E. Zhabitskaya and M. Zhabitsky\,\n  
 Proceedings of the 15th annual conference on Genetic and evolutionary comp
 utation\, GECCO'13\, ACM\, 455--462 (2013)\n  3. The ALICE Collaboration.\
 , Adam\, J.\, Adamová\, D. et al. Particle identification in ALICE: a Bay
 esian approach. Eur. Phys. J. Plus 131\, 168 (2016). https://doi.org/10.11
 40/epjp/i2016-16168-5\n\nhttps://indico.particle.mephi.ru/event/275/contri
 butions/3137/
LOCATION:Hotel Intourist Kolomenskoye 4* Moskvorechye 1
URL:https://indico.particle.mephi.ru/event/275/contributions/3137/
END:VEVENT
END:VCALENDAR