Performance qualification of the DIRICH readout system and development of a novel electron reconstruction scheme for the CBM experiment / vorgelegt von Pavish Subramani. Wuppertal, 2025
Content
- Abstract
- kurzfassung
- Contents
- Introduction
- Electromagnetic probes
- Standard model
- Quantum Chromo Dynamics (QCD)
- “Recipes for the fireball”
- Signatures of QGP
- Di-lepton production in a thermal medium
- Theoretical estimation of di-lepton production in medium and vacuum
- Di-electron productions from heavy-ion collisions
- Challenges in di-electron reconstruction
- CBM experiment
- FAIR
- CBM detector
- Beam Monitor (BMON)
- CBM Magnet
- Micro-Vertex Detector (MVD)
- Silicon Tracking System (STS)
- Muon Chamber (MUCH)
- Transition Radiation Detector (TRD)
- Time of Flight detector (TOF)
- Forward Spectator Detector (FSD)
- Data acquisition (DAQ) infrastructure
- Compute Cluster
- Ring Imaging CHerenkov detector (RICH)
- CBM simulation and reconstruction framework
- CBM RICH reconstruction
- Performance and quality testing of CBM RICH front-end electronics
- Ring Imaging Cherenkov detector front-end electronics
- Motivation
- Laboratory setup
- Charge sharing crosstalk measurement
- High occupancy analysis
- Leading edge- and Time-over-Threshold timing characteristics of the DIRICH FEB
- Test of DIRICH maximum hit rate capability
- Data flow in the TRB-based lab setup compared to later (m)CBM operation
- Analysis method
- DIRICH performance at high rate
- Data quality test
- Results and discussions
- Investigation of noise induced by DC/DC converters in a new DIRICH Power module
- Electron identification in CBM
- Simulation setup
- Electron Identification methods
- Figure of merit
- Optimization of track extrapolation
- Systematic correction for track projection
- Enhanced RICH ML model
- Fundamentals of the XGBoost model
- Supervised learning
- Decision trees
- Gradient boosting
- XGBoost
- Output of XGBoost model
- Importance of input features
- Training strategy
- Understanding input variables to the ML model
- Electron identification using the XGBoost model
- Inclusion of differential ring-track distance variable
- Backtracking TRD fitted tracks to RICH
- Ring-backtrack fit performance
- Inclusion of TRD backtrack-ring distance parameter into forward ML model
- Backtracking conversion electrons
- Understanding the origin of Cherenkov rings in the RICH detector
- An approach to account for the conversion electrons
- Characteristics of the conversion electrons
- Proof of concept
- General application of the method
- Background contribution
- Additional variables for reducing background
- RICH as secondary timer
- Machine Learning model for tagging the RICH rings
- Including the XGBoost response into the RICH electron identifier models
- Reconstruction of omega mesons
- Implementing XGBoost models into CBMROOT
- Electron candidate selection for omega meson reconstruction
- Topological cuts on candidate tracks
- Omega meson signal reconstruction
- Signal to background ratio estimation
- S/B ratio at the maximum significance
- Summary and Outlook
- Additional applications
- Particle Identification via time of flight using the RICH ring time
- Removing track fragments in Di-electron analysis
- Reconstructing low-momentum primary electrons
- Additional figures
- Additional figures for Chapter 4
- Calibration of filter with ToT cut
- Simulation of the double photons
- Scaler rate vs. rate at DAQ for different threshold values
- Median noise bandwidth
- Direct comparison of noise bandwidth due to two power modules at mRICH
- Effect of change in power module to time over threshold measurement
- Additional figures for Chapter 5
- Y component of the ring-track distance for different extrapolation layers
- Ring-track absolute distance for positrons
- Ring-track distance in X after applying corrections
- Ring-track distance in Y after applying corrections
- Additional figures for Chapter 6
- Response from XGBoost model for the training sample with inclusion of differential distance measure
- Transformation of input features
- Ring-backtrack fit performance for positrons
- Response from XGBoost model for the training sample with inclusion of TRD backtracked variables
- Addressing missing values in the data
- Additional figures for Chapter 7
- Angular acceptance as a function of conversion vertex for conversion electrons
- Response from XGBoost model trained for deriving conversion probability for the ring
- Response from the XGBoost model for the training sample with inclusion of differential distance measure and conversion probability
- Response from the upgraded XGBoost model for the training sample with inclusion of TRD backtracked ring-track distances and conversion probability
- Additional figures for Chapter 8
- Omega meson signal estimation using ANN in CBMROOT as electron identifier
- Omega meson signal estimation using XGBoost model variant 1 as electron identifier
- Omega meson signal estimation using XGBoost model variant 2 as electron identifier
- Omega meson signal estimation using XGBoost model variant 3 as electron identifier
- Omega meson signal estimation using XGBoost model variant 4 as electron identifier
- Omega meson signal estimation using XGBoost model variant 5 as electron identifier
- Non-electron background estimation using ANN in CBMROOT as electron identifier
- Non-electron background estimation using XGBoost model variant 1 as electron identifier
- Non-electron background estimation using XGBoost model variant 2 as electron identifier
- Non-electron background estimation using XGBoost model variant 3 as electron identifier
- Non-electron background estimation using XGBoost model variant 4 as electron identifier
- Non-electron background estimation using XGBoost model variant 5 as electron identifier
- Omega meson signal estimation using conventional ANN and XGBoost variants as electron identifiers at the point of maximum significance
- Non-electron background using conventional ANN and XGBoost variants as electron identifiers at the point of maximum significance
- References
- List of Figures
- List of Tables
- Acknowledgements
- Declaration