{"id":{"repo_id":"cambridge","oai_identifier":"oai:www.repository.cam.ac.uk:1810/331414"},"canonical_url":"https://search.dev.ndltd.org/etd/cambridge/oai:www.repository.cam.ac.uk:1810/331414","repository":{"repo_id":"cambridge","name":"Cambridge University","base_url":"https://api.repository.cam.ac.uk/server/oai/request"},"display":{"title":"Convolutional neural networks and photonic crystals for particle identification at high energy collider experiments","abstract":"The LHCb RICH detection system at CERN has been established as an effective way to identify particles in high energy physics experiments. Future increases in collision rate and luminosity make limitations of both the particle identification (PID) algorithm and the Cherenkov radiators apparent. These two challenges are addressed in this thesis. In the first part, a convolutional neural network (CNN) approach to particle identification in LHCb RICH is introduced. Changes in the LHCb computing infrastructure towards multi-core architectures provide the opportunity to benefit from the quickly evolving field of machine learning. The calculations of the presented neural network are well suited for quasi- real-time reconstruction through parallel processing. It takes binary input images from the two RICH detectors to classify particle species. The model is hyperparameter-optimised and trained on classification accuracy with simulated collision data for the upcoming LHC opera- tion starting in 2022, reaching a PID performance comparable to the conventional algorithm. Under conditions of reduced combinatorial background, as expected from the introduction of timing resolution to the RICH detectors in future upgrades, the network achieves a particle identification performance close to 100 %, with simultaneous misclassification of the most prevalent particle species approaching 0 %. The second part of the thesis addresses the limitations of conventional Cherenkov detectors, which require large gas chambers to distinguish between particles at high energies and to reach sufficient photon yield. Photonic crystals have been shown to emit photons under specific angles similar to the conventional Cherenkov effect. The properties of this novel Cherenkov radiator offer the possibility to distinguish particles in a much wider range of energies while requiring significantly less space as a Cherenkov radiator. The work presented in this chapter introduces a framework to manufacture, simulate, and study photonic crystals under high-energy beam conditions at CERN. In particular, the manufacturing process of a one-dimensional photonic crystal, suitable for high-energy applications, is described, as well as its tests in a 200 MeV electron beam at the CERN CLEAR facility.","abstract_html":"The LHCb RICH detection system at CERN has been established as an effective way to identify particles in high energy physics experiments. Future increases in collision rate and luminosity make limitations of both the particle identification (PID) algorithm and the Cherenkov radiators apparent. These two challenges are addressed in this thesis. In the first part, a convolutional neural network (CNN) approach to particle identification in LHCb RICH is introduced. Changes in the LHCb computing infrastructure towards multi-core architectures provide the opportunity to benefit from the quickly evolving field of machine learning. The calculations of the presented neural network are well suited for quasi- real-time reconstruction through parallel processing. It takes binary input images from the two RICH detectors to classify particle species. The model is hyperparameter-optimised and trained on classification accuracy with simulated collision data for the upcoming LHC opera- tion starting in 2022, reaching a PID performance comparable to the conventional algorithm. Under conditions of reduced combinatorial background, as expected from the introduction of timing resolution to the RICH detectors in future upgrades, the network achieves a particle identification performance close to 100 %, with simultaneous misclassification of the most prevalent particle species approaching 0 %. The second part of the thesis addresses the limitations of conventional Cherenkov detectors, which require large gas chambers to distinguish between particles at high energies and to reach sufficient photon yield. Photonic crystals have been shown to emit photons under specific angles similar to the conventional Cherenkov effect. The properties of this novel Cherenkov radiator offer the possibility to distinguish particles in a much wider range of energies while requiring significantly less space as a Cherenkov radiator. The work presented in this chapter introduces a framework to manufacture, simulate, and study photonic crystals under high-energy beam conditions at CERN. In particular, the manufacturing process of a one-dimensional photonic crystal, suitable for high-energy applications, is described, as well as its tests in a 200 MeV electron beam at the CERN CLEAR facility.","abstract_has_math":false,"creators":["Blago, Michele Piero"],"institution":"University of Cambridge","degree_name":"Doctor of Philosophy (PhD)","degree_level":"Doctoral","degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":["Wotton, Stephen","D'Ambrosio, Carmelo"],"committee_chairs":[],"committee_members":[],"year":2021,"date_issued":"2021-09-30","date_published":"2021-09-30","updated_at":"2026-07-22T22:24:23Z","subjects":["CERN","LHCb","Ring-imaging Cherenkov detectors","Convolutional Neural Network","Deep Learning","Particle Identification","Photonic Crystal"],"languages":["eng"],"rights":[],"rights_urls":["https://www.rioxx.net/licenses/all-rights-reserved/"],"identifier_entries":[]},"links":{"outbound_url":"https://doi.org/10.17863/CAM.78867","outbound_label":"DOI","outbound_source":"dc:identifier.doi"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Wotton, Stephen","D'Ambrosio, Carmelo"]},{"key":"dc:contributor.sponsor","label":"Sponsor","values":["CERN Doctoral Student Program"]},{"key":"dc:creator","label":"Author","values":["Blago, Michele Piero"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.issued","label":"Date","values":["2021-09-30"]},{"key":"dc:publisher.institution","label":"Dc Publisher Institution","values":["University of Cambridge"]},{"key":"dc:relation.isreferencedby.uri","label":"Dc Relation Isreferencedby URI","values":["https://www.repository.cam.ac.uk/handle/1810/331414"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]},{"key":"dc:type.qualificationlevel","label":"Dc Type Qualificationlevel","values":["Doctoral"]},{"key":"dc:type.qualificationname","label":"Dc Type Qualificationname","values":["Doctor of Philosophy (PhD)"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["CERN","LHCb","Ring-imaging Cherenkov detectors","Convolutional Neural Network","Deep Learning","Particle Identification","Photonic Crystal"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["https://www.rioxx.net/licenses/all-rights-reserved/"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.doi","label":"DOI","values":["10.17863/CAM.78867"]},{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://apollo8-f-pro.lib.cam.ac.uk/bitstreams/61133054-d090-48a8-9a4d-439ab8560641/download"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["The LHCb RICH detection system at CERN has been established as an effective way to identify particles in high energy physics experiments. Future increases in collision rate and luminosity make limitations of both the particle identification (PID) algorithm and the Cherenkov radiators apparent. These two challenges are addressed in this thesis. In the first part, a convolutional neural network (CNN) approach to particle identification in LHCb RICH is introduced. Changes in the LHCb computing infrastructure towards multi-core architectures provide the opportunity to benefit from the quickly evolving field of machine learning. The calculations of the presented neural network are well suited for quasi- real-time reconstruction through parallel processing. It takes binary input images from the two RICH detectors to classify particle species. The model is hyperparameter-optimised and trained on classification accuracy with simulated collision data for the upcoming LHC opera- tion starting in 2022, reaching a PID performance comparable to the conventional algorithm. Under conditions of reduced combinatorial background, as expected from the introduction of timing resolution to the RICH detectors in future upgrades, the network achieves a particle identification performance close to 100 %, with simultaneous misclassification of the most prevalent particle species approaching 0 %. The second part of the thesis addresses the limitations of conventional Cherenkov detectors, which require large gas chambers to distinguish between particles at high energies and to reach sufficient photon yield. Photonic crystals have been shown to emit photons under specific angles similar to the conventional Cherenkov effect. The properties of this novel Cherenkov radiator offer the possibility to distinguish particles in a much wider range of energies while requiring significantly less space as a Cherenkov radiator. The work presented in this chapter introduces a framework to manufacture, simulate, and study photonic crystals under high-energy beam conditions at CERN. In particular, the manufacturing process of a one-dimensional photonic crystal, suitable for high-energy applications, is described, as well as its tests in a 200 MeV electron beam at the CERN CLEAR facility."]},{"key":"dc:format.checksum.md5","label":"Dc Format Checksum Md5","values":["83b309931c5eb2f20e681f2a7b6d4d4a"]},{"key":"dc:title","label":"Title","values":["Convolutional neural networks and photonic crystals for particle identification at high energy collider experiments"]}]}],"canonical_facts":{"dc:contributor.advisor":["Wotton, Stephen","D'Ambrosio, Carmelo"],"dc:contributor.sponsor":["CERN Doctoral Student Program"],"dc:creator":["Blago, Michele Piero"],"dc:date.issued":["2021-09-30"],"dc:description.abstract":["The LHCb RICH detection system at CERN has been established as an effective way to identify particles in high energy physics experiments. Future increases in collision rate and luminosity make limitations of both the particle identification (PID) algorithm and the Cherenkov radiators apparent. These two challenges are addressed in this thesis. In the first part, a convolutional neural network (CNN) approach to particle identification in LHCb RICH is introduced. Changes in the LHCb computing infrastructure towards multi-core architectures provide the opportunity to benefit from the quickly evolving field of machine learning. The calculations of the presented neural network are well suited for quasi- real-time reconstruction through parallel processing. It takes binary input images from the two RICH detectors to classify particle species. The model is hyperparameter-optimised and trained on classification accuracy with simulated collision data for the upcoming LHC opera- tion starting in 2022, reaching a PID performance comparable to the conventional algorithm. Under conditions of reduced combinatorial background, as expected from the introduction of timing resolution to the RICH detectors in future upgrades, the network achieves a particle identification performance close to 100 %, with simultaneous misclassification of the most prevalent particle species approaching 0 %. The second part of the thesis addresses the limitations of conventional Cherenkov detectors, which require large gas chambers to distinguish between particles at high energies and to reach sufficient photon yield. Photonic crystals have been shown to emit photons under specific angles similar to the conventional Cherenkov effect. The properties of this novel Cherenkov radiator offer the possibility to distinguish particles in a much wider range of energies while requiring significantly less space as a Cherenkov radiator. The work presented in this chapter introduces a framework to manufacture, simulate, and study photonic crystals under high-energy beam conditions at CERN. In particular, the manufacturing process of a one-dimensional photonic crystal, suitable for high-energy applications, is described, as well as its tests in a 200 MeV electron beam at the CERN CLEAR facility."],"dc:format.checksum.md5":["83b309931c5eb2f20e681f2a7b6d4d4a"],"dc:identifier.doi":["10.17863/CAM.78867"],"dc:identifier.uri":["https://apollo8-f-pro.lib.cam.ac.uk/bitstreams/61133054-d090-48a8-9a4d-439ab8560641/download"],"dc:language":["eng"],"dc:publisher.institution":["University of Cambridge"],"dc:relation.isreferencedby.uri":["https://www.repository.cam.ac.uk/handle/1810/331414"],"dc:rights":["https://www.rioxx.net/licenses/all-rights-reserved/"],"dc:subject":["CERN","LHCb","Ring-imaging Cherenkov detectors","Convolutional Neural Network","Deep Learning","Particle Identification","Photonic Crystal"],"dc:title":["Convolutional neural networks and photonic crystals for particle identification at high energy collider experiments"],"dc:type":["Thesis"],"dc:type.qualificationlevel":["Doctoral"],"dc:type.qualificationname":["Doctor of Philosophy (PhD)"]},"updated_at":"2026-07-22T22:24:23Z"}