1 / 28

Track reconstruction in high density environment

Track reconstruction in high density environment. M.Ivanov 3-7 Oct 2005 TIME’05. Outlook. The ALICE detector description Combined reconstruction in high flux environment Combined tracking algorithm Reconstruction algorithm examples TPC ITS V0 finder Results. Solenoid magnet B<0.5 T.

cardea
Download Presentation

Track reconstruction in high density environment

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Track reconstruction in high density environment M.Ivanov 3-7 Oct 2005 TIME’05

  2. Outlook • The ALICE detector description • Combined reconstruction in high flux environment • Combined tracking algorithm • Reconstruction algorithm examples • TPC • ITS • V0 finder • Results M. Ivanov

  3. Solenoid magnet B<0.5 T TPC (the largest ever…): 88 m3 , 510 cm length, 250 cm radius Ne (90%) + CO2 (10%) 88 μs drift time 160 pad rows 570312 pads - channels main tracking device, dE/dx ALICE detectors • 6 Layers, 3 technologies • Material budget < 1% of X0 per layer! • Silicon Pixels vertices resolution in xy • (0.2 m2, 9.8 Mchannels) • Silicon Drift  resolution in z • (1.3 m2, 133 kchannels) • Double-sided Strip  connection w/TPC • (4.9 m2, 2.6 Mchannels) • Central tracking system: • Inner Tracking System • Time Projection Chamber M. Ivanov 2 * 1.8 units of pseudo-rapidity

  4. 2 * 1.8 units of pseudo-rapidity ALICE detectors • 6 layers for: • electron/pion separation at pt>1 GeV • factor 100 rejection at 2 GeV • tracking complement, mass resolution • 100 MeV/c2 @  • high pt trigger • Central tracking system: • Transition Radiation Detector • Time Of Flight Multigap Resistive Plate Chambers 5 years R&D, and  < 100 ps pions, kaons, protons separation electrons/pions at low pt M. Ivanov

  5. Reconstruction strategy • Main challenge - Reconstruction in the high flux environments (occupancy in the TPC detector up to 40 %) requires a new approach to tracking • Maximum Information Approach • Algorithms and data structures optimized for fast access and usage of all relevant information • Localize relevant information • Keep this information until it is needed ~dNch/dy=8000 ~15-40% occupancy ~50 million ADC amplitudes ~3 million clusters M. Ivanov

  6. Kalman filter (0) • Parallel Kalman Filter tracking approach chosen • To use optimal combination of local and global information about track's and clusters • Space points - clusters reconstructed before tracking • Advantages: • simultaneous track recognition and reconstruction • natural way to take into account multiple scattering, magnetic field unhomogenity • possibility to take into account mean energy losses • efficient way to match tracks between several detectors • Main assumptions - Space points used for Kalman filtering • Described by a Gaussian distribution with known width • Errors between layers are not correlated M. Ivanov

  7. Kalman filter (1) • Cluster overlaps problem: • Cluster unfolding based on the cluster shape for clusters with extended shape • Cluster shape parameterization depending on the track parameters • Multidimensional error parameterization in the space of observables • Space points characterized by the position, charge, shape characteristic and charge ratio for unfolded clusters • Space point errors assigned to the point only at the moment when the parameters of the tracks and tracklets are known • Implemented for all tracking detectors –TPC, TRD, ITS M. Ivanov

  8. Kalman filter (2) • Reduction of the influence of wrongly associated clusters on the track parameters • If the error of the prolongation comparable with the mean distance between of the clusters – building the tree of the hypothesis and choose the most probable - ITS • Fast, track segment finding algorithms (seeding algorithms) over the full detector system were implemented. For tracks with doubtful behavior (according to chi2 and cluster ratio criteria), multiple hypotheses are considered and tracked. (Used mainly in the TPC) • For Kink and V0 topologies refit the track towards the vertex • Reduced influence of the wrongly associated clusters • Optimal local parameters and covariance – not influenced by energy loss correction and multiple scattering M. Ivanov

  9. Tracking strategy – Primary tracks • Iterative process • Forward propagation towards to the vertex –TPC-ITS • Back propagation –ITS-TPC-TRD-TOF • Refit inward TOF-TRD-TPC-ITS • Continuous seeding –track segment finding in all detectors TOF TRD TPC ITS M. Ivanov

  10. Sources of information • spatial characteristic of a track and sets of tracks • px,py,pz,y,z parameters and covariance • chi2 • number of points on the track • number of shared clusters on the track • overlaps between tracks • DCA for V0s, Kinks and Cascades • … • dEdx • mean, sigma, number of points, number of shared points… • derived variables • Mass • Causality - Probability that particle “ really exists” in some space interval (used for causality cuts) • Based on clusters occurrence, and chi2 before – after vertex • Invariant mass • Pointing angle of neutral mother particle • … M. Ivanov

  11. TPC forward tracking • Cluster finding • Loop over seeding region • Seeding - non signed clusters • Tracking – all clusters • Space point error parameterization • Updating of the statistical information • Mean and RMS of the chi2, number of point, cluster ratio • Registering of the “gold tracks” with characteristic over to mean-N*RMS threshold • Signing the space points belonging to the gold tracks • Keeping all tracks • Kink Finding • Double track removal • dEdx calculation • Shared clusters not used • Correction of the amplitudes for clusters with extended shape (comparing with expected) M. Ivanov

  12. Tracking strategy – Kink topology • Iterative process for mother and daughter particles (Continuous seeding –track segment finding in all detectors) • Forward propagation towards to the vertex –TPC-ITS • Back propagation –ITS-TPC-TRD-TOF • Refit inward TOF-TRD-TPC-ITS • Kink topology finding, storing and updating (refitting) with the optimal local Kalman parameters TOF Mother particle TRD TPC ITS Kink vertex Daughter particle M. Ivanov

  13. TPC – dEdx resolution dN/dy~6000 dN/dy~10 dN/dy~10 dN/dy~6000 • Number of associated clusters and relative dEdx resolution M. Ivanov

  14. ITS - Cluster shape parameterization • Cluster shape parameterization • dependence on the local track angle in θ and φ (known only during tracking procedure) • dependence on the layer width and the granularity – different for each layer M. Ivanov

  15. ITS - Cluster sharing probability parameterization • cluster can belong to more than one track • probability as a function of distortion from the expected cluster shape parameterized non shared cluster shared clusters N digitsM-NdigitExp M. Ivanov

  16. ITS - Cluster error parameterization constant error constant error parameterized error parameterized error • If a constant effective errors are used - pulls have long tail – χ2 criteria for the selection of most probable track candidates can fail • Using the information about the distortion of the measured cluster shape from the expected one, and about the distortion of the measured cluster charge from the expected one • different types of the parameterization for each layer and cluster types Y pulls Z pulls M. Ivanov

  17. layer 5 – 2 hypothesis layer 4 – 6 hypothesis layer 3 – 18 hypothesis layer 2 –0 too many hypothesis Best track ITS – Parallel tracking (0) • Starting point track seeds from the TPC • Follow the tree of track hypotheses connecting reconstructed clusters: • After each layer track hypothesis sorted according to χ2 • Only gold track branches and a restricted amount of non gold tracks propagated further down M. Ivanov

  18. ITS -Parallel tracking (1) • For each seed, tree of the track hypotheses constructed • Most probable track candidates according to χ2 chosen • restricted amount of tracks kept for further parallel tracking procedure • for secondary tracks also short best tracks kept, for further V0 study • Overlap factor between selected track and all other tracks (from other TPC seed) is calculated • If the overlap factor is higher than a critical value, conflict resolving algorithm is applied M. Ivanov

  19. ITS - Parallel tracking (2) • Double loop over all possible pair of branches • Weighted χ2 of two tracks calculated • effective probability of cluster sharing and for secondary particles the probability not to cross given layer taken into account Conflict ! Best track 2 Best track 1 M. Ivanov

  20. ITS V0 finding strategy • V0 finding - Before deletion of the tree of hypotheses • DCA calculation in the ITS for most probable tracks defined by parallel tracking • Application of rough cuts • DCA, pointing angle, causality (defined by chi2 of the clusters found before the vertex) • Take the best pair of track candidates at the point after V0 vertex obtained in the first approximation • Application of stronger cuts • V0 object registration • V0 refitting in the 3rd reconstruction iteration M. Ivanov

  21. V0- Cut on Causality • Assumption • For different V0 topologies there can exist forbidden detection layer • Example : tracks from V0 decay at radius 10 should not have associated space point in ITS layer 0 and 1 • But, in high flux environment there is significant probability to catch space points produced by another tracks • The likelihood that tracks “exist” in the forbidden layers is function of the chi2 of the tracks in forbidden layers M. Ivanov

  22. Causality likelihood • Left side – background (dN/dy ~ 6000) • Right side – signal (dN/dy ~6000) M. Ivanov

  23. V0 - Efficiency as function of the likelihood cut • Cumulative function of the log-likelihood (normalization to all pre-selected candidates) • Left side – background • Right side – signal • Using new V0 combined with cut on the likelihood • Improvement of the V0 finding efficiency by factor of 2 in comparison with previous one (keeping the same signal/background) M. Ivanov

  24. Combined tracking efficiency – (TPC and ITS) • Efficiency and the fake ratio as the function of the particle momentum M. Ivanov

  25. Transverse momentum resolution Low flux Central Pb-Pb collision • Transverse momentum resolution as a function of momenta with and without TRD detector M. Ivanov

  26. Conclusion • Alice reconstruction based on MIA implemented • Performance of the reconstruction greatly improved • Acceptable also in the highest expected multiplicities (dN/dy~8000) • TPC tracking efficiency from 89% to 99% • TPC dEdx resolution improved from 9.5% to 6.9% • ITS tracking efficiency improved • Kink finder efficiency improvement by factor 2.5 • V0 finding efficiency improved • ALICE experience from combined tracking in the high flux environment indicates that the MIA approach leads to considerable improvements compared with tracking based on zero-level approximation M. Ivanov

  27. Simulation • Transport • Virtual Monte Carlo- Geant3(default), Gean4, Fluka • Gas detectors • The hitsare produced at every ionizing collision. The tracking step is calculated for every collision from the exponential distribution. • Microscopic response simulation • Secondary ionization • For each electron: • Drift: diffusion, electron attachment, ExB effect • Response: Gas gain fluctuation, Time response function, 2D pad response function • Detector response: electronic noise and conversion gain, digitization and zero suppression M. Ivanov

  28. Combined likelihood • Combined likelihood • Ldca*Lpa*Lcausality • Left side – background • Right sight - signal M. Ivanov

More Related