Cellular Automaton Tracking for VXD Cellular Automaton Tracking for - - PowerPoint PPT Presentation

▶
cellular automaton tracking for vxd cellular automaton
SMART_READER_LITE
LIVE PREVIEW

Cellular Automaton Tracking for VXD Cellular Automaton Tracking for - - PowerPoint PPT Presentation

Sep 10, 2023 158 likes •365 views

Cellular Automaton Tracking for VXD Cellular Automaton Tracking for VXD Cellular Automaton Tracking for VXD Cellular Automaton Tracking for VXD based on Mini Vectors based on Mini Vectors based on Mini Vectors based on Mini

slide-1
SLIDE 1

1

AWLC14, Fermilab, May 2014

Cellular Automaton Tracking for VXD Cellular Automaton Tracking for VXD based on Mini – Vectors based on Mini – Vectors Cellular Automaton Tracking for VXD Cellular Automaton Tracking for VXD based on Mini – Vectors based on Mini – Vectors

  • Y. Voutsinas, F. Gaede
slide-2
SLIDE 2

2

  • Overview of the ILD tracking scheme
  • Silicon tracking status
  • Cellular automaton based on mini – vectors

➢

Cellular automaton algorithm

➢

Mini - vectors

➢

Performance

➢

Robustness

  • Outlook

Outline Outline Outline Outline

AWLC14, Fermilab, May 2014

slide-3
SLIDE 3

3

  • Various algorithms for standalone pattern recognition in TPC, forward region, barrel Si detectors
  • Track fitting with KalTest (c++ Kalman filter)
  • IMarlinTrk: provides loose coupling between track finding & track fitting

ILD Tracking Overview ILD Tracking Overview ILD Tracking Overview ILD Tracking Overview

slide-4
SLIDE 4

4

Pattern Recognition in ILD Pattern Recognition in ILD Pattern Recognition in ILD Pattern Recognition in ILD

  • Clupatra processor
  • Form seeds using Nearest Neighbours hit clustering
  • Propagate seeds both inwards & outwards using Kalman fitter
  • Associate best matching hit
  • Update track state
  • So on...
  • SiliconTracking
  • Divide VXD – SIT into angular sectors
  • brute force triplet search in phi sectors based on a set of seed-layer-triplets
  • Fit a helix to the seed triplets
  • Follow the seed inwards – attach hits according to the distance from the helix
  • Refit with Kalman fitting
  • Forward Tracking
  • Standalone tracking algorithm at FTD
  • Pattern recognition: Cellular automaton
  • Fitting: Kalman filter
  • Ambiguities resolution: Hopfield NN
  • FullLDCTracking
  • Combines track from TPC – FTD – Silicon

tracking

  • Based on track parameter compatibility
  • Adding spurious leftover hits
  • Final track fit

AWLC14, Fermilab, May 2014

slide-5
SLIDE 5

5

ILD Tracking Overall Performance ILD Tracking Overall Performance ILD Tracking Overall Performance ILD Tracking Overall Performance

AWLC14, Fermilab, May 2014

Plots from DBD – ttbar sample, pair bkg included ~ 99.7% eff, P≥ 1 GeV, ≥ 99.8%, cos(θ) < 0.95 Achieve ILD goals in resolution

slide-6
SLIDE 6

6

  • Std algorithm (used for DBD studies)

➢

Doesn't appear to have optimal performance under realistic conditions

➢

Can't cope with combinatorics induced by pair bkg

  • FPCCD Tracking

➢

Big step ahead, shows promising performance in the presence of pair bkg

➢

See Jan's talk

Silicon Tracking Status in ILD Silicon Tracking Status in ILD Silicon Tracking Status in ILD Silicon Tracking Status in ILD

AWLC14, Fermilab, May 2014

From Tatsuya Mori

slide-7
SLIDE 7

7

  • Mainly the standalone VXD tracking

➢

Track finding in the low PT range (~ 100 MeV)

  • Cellular automaton core tools already included in ilcsoft - used for FTD tracking

➢

Can we use them for another subdetector?

  • Added values of mini – vectors

➢

Exploitation of the double sided structure of the VXD ladders

➢

Can they reduce combinatorial bkg?

  • Tracking performances w.r.t. VXD configuration – sensor specifications

➢

Speed, robustness ...

Motivations Motivations Motivations Motivations

AWLC14, Fermilab, May 2014

slide-8
SLIDE 8

8

  • Detector studied through these slides

➢

DBD VXD, equipped with fast CMOS sensors

➢ ➢ ➢ ➢ ➢

Overall number of VXD hits

➢

DBD VXD: 160k

➢

Fast CMOS VXD: 120k

Detector Configuration Detector Configuration Detector Configuration Detector Configuration

DBD VXD Fast CMOS VXD Fast CMOS VXD

layer σspatial (μm) σtime(μs) σspatial(μm) σtime(μs) L1 3 / 6 50 / 10 3 / 6 50 / 2 L2 4 100 4 / 10 100 / 7 L3 4 100 4 / 10 100 / 7 AWLC14, Fermilab, May 2014

slide-9
SLIDE 9

9

  • Core tools are already there for the FTD tracking
  • Very flexible

➢

Appealing to be used for pattern recognition in other detectors

➢

See R. Glattauer Diploma thesis

http://www.hephy.at/fileadmin/user_upload/Publikationen/DiplomaThesis.pdf

✔

VXD & mini – vectors related definitions of KiTrack abstract classes have been created in KiTrackMarlin

✔

Set of criteria for mini – vector connections have been defined in KiTrack

✗

Minor modifications in core tools

➢

Pattern recognition is quite detector - specific...

Cellular Automaton Tools Cellular Automaton Tools Cellular Automaton Tools Cellular Automaton Tools

KiTrack Basic algos Abstract classes (hits, tracks, ...) CA criteria KiTrackMarlin lcio / Marlin implementation MarlinTrkProc MiniVector CA AWLC14, Fermilab, May 2014

slide-10
SLIDE 10

10

Mini – Vector Tracking Mini – Vector Tracking Mini – Vector Tracking Mini – Vector Tracking

  • Mini – vector formation

1) Hits in adjacent layers (dist 2mm) with max distance 5mm 2) Or δθ between hits in adjacent layers (cut can go up to 0.10)

  • Divide VXD into θ, φ sectors

➢

Try to connect mini – vectors in neighbouring sectors

  • Cellular automaton criteria

➢

φ, θ pointing direction of the mini – vectors

➢

No zig-zag (2 MV segments)

  • ttbar sample, pair bkg included for √s = 500GeV
  • Fast CMOS vertex detector

Dist < 5mm δΘ <0.50 δΘ <0.30 δΘ <0.10 VXD hits 105 105 105 105 MiniVectors 3x105 105 6x104 2x104 Connections O(105) O(105) < 105 ~ 104 Raw tracks O(106) O(106) O(105) < 105 Time ~10min ~ 2min ~ 1min ~ 20 s

ttbar, δθ of hits belonging to a MV based on MC info

δθ (deg)

AWLC14, Fermilab, May 2014

slide-11
SLIDE 11

11

  • FPCCD tracking

➢

Most performant algorithm for standalone Silicon tracking in ILD

  • Examined track sample

➢

All charged tracks inside the geometrical acceptance of the VXD

  • Definition of found track

➢

75% purity, ≥ 4 hits

  • "Ghost" tracks

➢

all tracks which does not correspond to a found MC particle

➢

Could be pair bkg particles or combinatorics or misreconstructed tracks

Comparison with FPCCD Tracking* Comparison with FPCCD Tracking* Comparison with FPCCD Tracking* Comparison with FPCCD Tracking*

* as it was at beginning of March 2014

slide-12
SLIDE 12

12

Comparison with FPCCD Tracking II Comparison with FPCCD Tracking II Comparison with FPCCD Tracking II Comparison with FPCCD Tracking II

  • Ghost tracks / evt (PT > 1 GeV)

➢

FPCCD: ~ 10

➢

CA: ~ 11

  • Time / evt

➢

FPCCD: ~ 75 s

➢

CA: ~ 25 s AWLC14, Fermilab, May 2014

Sample: ttbar, Sample: ttbar, √ √s = 500 GeV, fast CMOS VXD, pair bkg overlayed, 120 events s = 500 GeV, fast CMOS VXD, pair bkg overlayed, 120 events

slide-13
SLIDE 13

13

  • Efficiency ~ 99% for PT > 1 GeV
  • Why we can't find this ~1% of tracks?
  • Typical case of lost track, MC particle PT = 21

GeV

  • Particle doesn't create hits to all layers, in L4

and L6 crosses the insensitive electronic band

➢

Can form mini – vectors only in inner layer

➢

Need > 1 mini vector to reconstruct a track...

  • Marginal effect in tracking but...
  • ... what about alignment?

Search for the lost tracks Search for the lost tracks Search for the lost tracks Search for the lost tracks

AWLC14, Fermilab, May 2014

slide-14
SLIDE 14

14

  • Investigating SUSY scenario with light Higgsinos
  • Very soft fermions in the final state

➢

Ideal sample to test the CA mini – vector algorithm

➢

Replace the std Silicon tracking with the new algorithm

➢

No pair bkg overlayed

➢

Comparing the overall tracking performance for each Si tracking algorithm

➢ ➢ ➢ ➢ ➢ ➢ ➢ ➢

  • Significant improvement to low P

Significant improvement to low PT

T region!

region!

Light Higgsinos Study (Hale Sert) Light Higgsinos Study (Hale Sert) Light Higgsinos Study (Hale Sert) Light Higgsinos Study (Hale Sert)

PT (GeV) PT distribution of stable and charged MC particles (cosθ < 0.9397)

slide-15
SLIDE 15

15

  • Mini – vector tracking can be sensitive to missing hits

➢

What will happen if we don't have 100 % sensor detection efficiency?

–

Track finding eff. as a function of hit detection eff.

–

Studied values for hit detection efficiencies for the sensors: 99.5%, 99%

  • Robustness vs combinatorics
  • Up to which hit density the C.A. Algorithm can cope with?
  • Is it performant for the DBD assumed sensors specifications (time resolution)
  • One should account for the uncertainties in pair bkg simulations
  • Also: changes in ILD configuration may have a significant impact on pair bkg hit

densities

  • Anti – DID field, beamcal design ...

Robustness Robustness Robustness Robustness

AWLC14, Fermilab, May 2014

slide-16
SLIDE 16

16

Robustness vs Missing Hits Robustness vs Missing Hits Robustness vs Missing Hits Robustness vs Missing Hits

AWLC14, Fermilab, May 2014

slide-17
SLIDE 17

17

Robustness Robustness Robustness Robustness

AWLC14, Fermilab, May 2014

  • Mini – vector tracking can be sensitive to missing hits

➢

What will happen if we don't have 100 % sensor detection efficiency?

–

Track finding eff. as a function of hit detection eff.

–

Studied values for hit detection efficiencies for the sensors: 99.5%, 99%

  • Robustness vs combinatorics
  • Up to which hit density the C.A. Algorithm can cope with?
  • Is it performant for the DBD assumed sensors specifications (time resolution)
  • One should account for the uncertainties in pair bkg simulations
  • One should account for hits due to electronic noise (but probably marginal effect...)
  • Also: changes in ILD configuration may have a significant impact on pair bkg hit

densities

➢

Anti – DID field, BeamCal design ...

slide-18
SLIDE 18

18

  • Severely compromised performance (CPU and efficiency) observed for DBD

VXD option

➢

FPCCD tracking performs better

  • Why CA mini – vector is suffering?

➢

For each MV, too many candidate MV to connect with in neighboring sectors

  • Approach

➢

Smarter selection of neighboring sectors

➢

MV are small tracks – can point to the candidate sector

➢

Fully exploit the MV concept

➢

Work on going...

Performance for Higher Hit Densities Performance for Higher Hit Densities Performance for Higher Hit Densities Performance for Higher Hit Densities

AWLC14, Fermilab, May 2014

slide-19
SLIDE 19

19

  • For a fast VXD

➢

CA MV tracking shows very good perf. In terms of efficiency and CPU time

➢

Ghost tracks

–

High PT: reject via matching with TPC (on going)

–

Low PT: more complicated...

  • For slower detectors / higher hit densities

➢

Smarter sector connection (on going)

  • Integration to overall tracking sw

➢

CA MV shows promising performance as a part of the overall tracking

➢

Improves significantly the efficiency on low PT tracks

➢

Few technical issues need to be resolved

  • Do some physics studies!

Summary & Outlook Summary & Outlook Summary & Outlook Summary & Outlook

AWLC14, Fermilab, May 2014

georgios.voutsinas@desy.de georgios.voutsinas@desy.de