Le Modèle Standard de la physique des particules a jusqu’alors extrêmement bien réussi à décrire les particules élémentaires et leurs interactions. Malgré cela, il demeure toujours des questions ouvertes. La possibilité de répondre à ces questions grâce la Supersymétrie est actuellement à l’étude dans les collisions proton-proton à 13 TeV dans le cadre de l’expérience ATLAS au LHC. Cette thèse présente la recherche de la production de paires de particules colorées dans ATLAS, ces dernières se désintégrant en paires de jets. Pour ce faire, les données de 2016, 2017 et 2018 ont été utilisées. De telles particules échappent aux recherches standards de la Supersymétrie du fait de l’absence d’énergie transverse manquante dans l’état final. Deux signatures furent considérées, la désintégration de stops via des couplages violant la R-parité et la production de sgluon, le partenaire scalaire du gluino. En l’absence de signal, une amélioration de 200 GeV sur la masse maximum exclue est attendue. Le HL-LHC augmentera la luminosité intégrée délivrée afin de nous permettre de rechercher des particules plus massives et d'améliorer les mesures de précision du Modèle Standard. La luminosité instantanée augmentera d’un facteur 5 et une luminosité intégrée de 4000 fb⁻¹ devrait pouvoir être atteinte à la fin du LHC en 2037.Cette thèse présente également une étude des perspectives de mesure des couplages du Higgs au HL-LHC effectuée à l’aide de SFitter. Il est démontré que dans le cadre des Delta et d’une EFT, l’augmentation de la luminosité génère une amélioration de la précision de la mesure des couplages. Finalement, le Détecteur de temps fortement segmenté, qui sera installé dans ATLAS au HL-LHC, est présenté. La simulation de ce détecteur a été développée pour prendre en compte la résolution temporelle du détecteur et fut utilisée pour optimiser sa géométrie. Les performances de ce détecteur ont été étudiées, plus de 80 % des traces ont leurs temps correctement associés avec une résolution de 20 ps avant irradiation de 50 ps après. En utilisant les informations temporelles, l’isolation des électrons peut être amélioré de 10 %. / The Standard Model of particle physics has been extremely successful in describing the elementary particles and their interactions. Nevertheless, there are open questions that are left unanswered. Whether supersymmetry can provide answers to some of these is being studied in 13 TeV proton-proton collisions in the ATLAS experiment at the LHC. In this thesis a search for pair produced colored particles in ATLAS decaying into pairs of jets using data from 2016, 2017 and 2018 is presented. Such particles would escape standard Supersymmetry searches due to the absence of missing transverse energy in the final state. Stops decaying via a R-parity violating coupling and sgluon, scalar partners of the gluino, were considered. In the absence of a signal, an improvement of 200 GeV on the limit on the stop mass is expected. The HL-LHC will increase the integrated luminosity delivered to probe even higher mass ranges as well as improving the precision of Standard model measurements. The instantaneous luminosity will be increased by a factor 5 and an integrated luminosity of 4000 fb⁻¹ should be reached by the end of the LHC in 2037.A study of the Higgs coupling measurement prospects at the HL-LHC using SFitter is performed. Using the Delta and EFT framework shows that the increase in luminosity will result in a significant improvement of the precision of the measurement of the couplings. The High granularity timing detector detector will be installed in ATLAS for the HL-LHC. A simulation of the detector that takes into account the timing resolution was developed and used to optimize its layout. The detector performance was studied. More than 80 % of the tracks have their time correctly reconstructed with a resolution of 20 ps before irradiation and 50 ps after. Using the timing information, the electron isolation efficiency is improved by 10 %.

Les trois premières années de fonctionnement de l'expérience ATLAS au LHC, aux résultats fructueux, ont ouvert l'accès à la physique à l'échelle du TeV. Les travaux effectués dans ce contexte ont porté à la fois sur des aspects de performance (reconstruction des électrons) et d'analyse (recherche de nouvelle physique). La première partie est consacrée aux mesures in-situ des efficacités de reconstruction et d'identification des électrons. La méthodologie employée pour réaliser ces mesures dans les données est présentée en détails ainsi que les méthodes développées pour l'estimation du bruit de fond résiduel. La deuxième partie traite de la recherche de Supersymétrie dans les événements avec deux leptons (électrons ou muons) de charges identiques, accompagnés de jets et éventuellement d'énergie transverse manquante, basée sur l'ensemble des données collectées par ATLAS en 2012 (soit 21 fb-1 à sqrt{s}= 8 TeV). Ces états finaux sont particulièrement appropriés à la recherche de gluinos ou de squarks de troisième génération. Les principaux aspects de l'analyse sont décrits, notamment les méthodes d'estimation des différents bruits de fond. L'absence d'excès dans les régions de signal par rapport aux prédictions du Modèle Standard est interprétée en termes de limites sur les masses des superpartenaires, dans le cadre de modèles aux spectres de masse simplifiés. Ainsi, dans les modes de désintégrations étudiés, l'existence de gluinos de masse inférieure à 1 TeV, ou de sbottoms de masse inférieure à 500 GeV, est exclue. / The first three years of operation of the ATLAS experiment at LHC, with fruitful results, have opened the way to physics at the TeV scale. The work realized in this context included some performance aspects (electron reconstruction) as well as analysis (search for new physics). The first part is devoted to in-situ measurements of electron reconstruction and identification efficiencies. The methodology employed to perform these measurements in data is presented in details, as well as the methods developped for estimating the residual background. The second part deals with a search for Supersymmetry in events with two leptons (electrons or muons) of identical charges, together with jets and in some cases missing transverse energy, based on the whole dataset colledcted by ATLAS in 2012 (namely 21 fb-1 at 8 TeV). These final states are particularly adapted to searches for gluinos or third generation squarks. The main aspects of the analysis are described, in particular the methods used to estimate the various backgrounds. The absence of excess in the signal regions with respect to the Standard Model predictions is interpreted in terms of limits on the masses of superpartners, in the frame of models with simplified mass spectra. Thus, in the studied decay modes, the existence of gluinos with masses below 1 TeV or sbottoms with masses below 500 GeV is excluded.

Diese Dissertation beschäftigt sich mit der Suche nach dem skalaren Top Quark (stop) und dem skalaren Bottom Quark (sbottom) innerhalb des Minimal Supersymmetric Standard Model (MSSM) unter der Annahme der R-Paritätserhaltung. Suchen nach den folgenden Zerfallsmoden des Stop-Quark wurden durchgeführt: stop -> c neutralino_1, stop -> b l sneutrino (wobei l mit gleichen Wahrscheinlichkeiten entweder electron, muon oder tau-lepton ist) und stop -> b tau sneutrino (nur das Tau-Lepton wird berücksichtigt). Zusätzlich wurde der Dreikörperzerfall stop -> b W neutralino_1 im erlaubten Massenbereich M_stop > M_b + M_W + M_neutralino1 >= 86 GeV gesucht. Für das Sbottom-Quark wurde der Zerfall sbottom -> b neutralino_1 studiert. Jede dieser Zerfallsmoden wurde voneinander unabhängig unter der Annahme eines 100 %-igen Verzweigungsverhältnisses untersucht. Für diese Suche wurden Daten aus electron-positron-Kollisionen bei Schwerpunktsenergien im Bereich von 202-208 GeV benutzt. Die Daten wurden im Jahr 2000 von dem L3 Detektor am Large Electron Positron Collider (LEP) am CERN aufgenommen. Ferner wurden die Resultate der Datenanalyse aus dem Jahr 2000 mit Resultaten der Squark-Suche kombiniert, die die L3 Kollaboration in vorhergehenden Jahren bei Schwerpunktsenergien von 161 bis 202 GeV durchgeführt hat. Die untersuchten Squark Zerfallskanäle bestimmen die Topologie der für uns interessanten Ereignisse: 2 Jets (oder b-Jets) + fehlende Energie (+ 2 Leptonen für die Stop-Dreikörperzerfälle). Die stop -> b W neutralino_1 Zerfallstopologie hängt signifikant von den weiteren Zerfällen des W-Bosons ab und kann bis zu 6 Jets im Endzustand haben. Die Annahme der R-Paritätserhaltung impliziert die Stabilität des leichtesten supersymmetrischen Teilchens (des LSP), das das leichteste Neutralino ist. Das LSP wechselwirkt nur schwach und entweicht deswegen unentdeckt. Ein besonderes Merkmal der Signal-Ereignisse ist somit eine erhebliche Menge fehlender Energie. Die sichtbare Energie ist in etwa proportional zu der Massendifferenz zwischen dem Squark und dem LSP. Weil die Standardmodell-Untergrundzusammensetzung vom Anteil der sichtbaren Energie abhängt, hängt die Analyse auch vom Wert von der Massendifferenz ab. Abhängig von der Menge fehlender Energie kann der Standardmodell-Untergrund in drei Kategorien eingeteilt werden: - die zwei-Fermion-Prozesse sind e e -> e e, e e -> mu mu, e e -> tau tau und e e -> e e q q; - die vier-Fermion-Kategorie besteht aus e e -> W W, e e -> W e nu, e e -> Z Z und e e -> Z e e Prozessen; - die zwei-Photon-Untergrundprozesse sind e e -> e e e e, e e -> e e mu mu, e e -> e e tau tau und e e -> e e q q. Der letzte Prozess, e e -> e e q q, trägt den grössten Anteil zu den SM-Untergrundprozessen bei (wegen sehr hohem und stark schwankendem E_miss und dem grössten Wirkungsquerschnitt). Im ersten Schritt der Analyse wurden Events mit der gewünschten Topologie (2 Jets und hohes E_miss) vorselektiert. Die Selektion von Stop- und Sbottom-Ereignissen wurde durch die Minimierung der mit 95 % Confidence Level (C.L.) erwarteten oberen Grenze des Squark-Wirkungsquerschnitts - berechnet aus MC-Vorhersagen - optimiert, wobei der kleine theoretisch vorhergesagte Produktionswirkungsquerschnitt des Squarks berücksichtigt wurde. In allen für den jeweiligen Squark Zerfallskanal optimierten Selektionen stimmt die Anzahl von Daten Events mit der erwarteten Anzahl von Standardmodellprozessen überein: - für den stop -> c neutralino_1 Zerfall wurden 29 Daten-Events beobachtet, wobei 26.5 +- 2.7 Events von den SM-Prozessen erwartet wurden; - für den Dreikörperzerfall stop -> b l sneutrino, wurden 4 Daten-Events selektiert bei einer Standardmodell-Erwartung von 4.0 +- 1.0 Events; - für den Zerfall stop -> b tau sneutrino sind die Daten- und SM-Eventzahlen 5 bzw. 3.9 +- 1.0; - in der Selektion für stop -> b W neutralino_1, wurden 184 Daten Events beobachtet und 181.6 +- 3.0 Events wurden vom Standardmodell vorhergesagt; - für den Bottom Squark Zerfall sbottom -> b neutralino_1 entsprachen die beobachteten 6 Events der SM-Erwartung von 7.7 +- 1.3 Events. Es wurden keine MSSM-Skalar-Quarks in den Daten des Experiments beobachtet und das Resultat der Suche ist negativ. Die modellunabhängige 95 % C.L. obere Grenze für den Squark-Produktionswirkungsquerschnitt wurde aus der gemessenen Anzahl von Daten-Events und der aus dem Standardmodell erwarteten Eventanzahl berechnet. Für die Berechnung der oberen Grenzen der Produktionswirkungsquerschnitte wurden die Resultate der Squark-Suchen aus den L3-Daten bei Schwerpunktsenergien von c.m.s. Energie 202 - 208 GeV mit den Resultaten aus vorherigen Suchen der L3-Kollaboration bei 161 GeV - 202 GeV kombiniert. Eine neue Methode wurde entwickelt, um die kombinierten Grenzen zu berechnen. Die Methode berücksichtigt die statistische Unabhängigkeit jeder Messung und die Abhängigkeit des Squark-Produktionswirkungsquerschnittes von der Schwerpunktsenergie. In der Berechnung wurde den systematischen Unsicherheiten in der Standardmodell-Untergrundabschätzung und der Signal-Selektionseffizienz Rechnung getragen. Für die hier betrachteten Squark-Zerfälle werden typisch folgende oberen Grenzen mit 95 % C.L. für den Squark Produktionswirkungsquerschnitt erhalten: ~ 0.05-0.2 pb (für stop) und ~ 0.05-0.1 pb (für sbottom). Bei den Suchen nach dem Stop-Dreikörperzerfall stop -> b W neutralino_1 wurden die Produktionswirkungsquerschnitte über 0.7-1.0 pb mit 95 % C.L. ausgeschlossen. Innerhalb des Minimal Supersymmetrischen Standard Modells mit R-Paritätserhaltung wurden die unabhängigen Wirkungsquerschnittsgrenzen für den Ausschluss von MSSM Parametern benutzt, insbesondere für die Stop- und Sbottom-Massen. Die Squark-Massen wurden für jeden betrachteten Zerfallskanal in zwei möglichen Szenarien ausgeschlossen: für den maximalen und den (näherungsweise) minimalen theoretischen Wirkungsquerschnitt. Der erste Fall korrespondiert zur maximalen Mischung zwischen den links- und rechtshändigen Squark-Eigenzuständen, $\cos\theta_{LR}$ = 1; der zweite Fall ist definiert durch den Wert von $\cos\theta_{LR}$, bei dem die Squarks vom $Z^0$ Boson entkoppeln. Abhängig vom Wert $\Delta M$ wurden die Squark Massen mit 95 % C.L. bis zu den folgenden Werten ausgeschlossen: - für stop -> c neutralino_1: M_stop < 90-93 GeV (min. Wirkungsquerschnitt), M_stop < 95-96 GeV (max. Wirkungsquerschnitt), - für stop -> b l sneutrino: M_stop < 87-89 GeV (min. Wirkungsquerschnitt), M_stop < 90-91 GeV (max. Wirkungsquerschnitt), - für stop -> b tau sneutrino: M_stop < 83-88 GeV (min. Wirkungsquerschnitt), M_stop < 88-91 GeV (max. Wirkungsquerschnitt), - für sbottom -> b neutralino_1: M_stop < 76-83 GeV (min. Wirkungsquerschnitt), M_stop < 94-97 GeV (max. Wirkungsquerschnitt), In beiden Fällen werden die experimentell beobachteten 95 % C.L. Massen Ausschlussgrenzen mit den aus Monte Carlo Simulationen ohne SUSY Teilchen erwarteten verglichen. Die experimentallen Ausschlussgrenzen Sind verträglich mit den erwarteten. Die mit 95 % C.L. erhaltene obere Grenze für den Stop-Produktionquerschnitt ist im Zerfall stop -> b W neutralino_1 grösser als die zugehörige theoretische Vorhersage. Der Ausschluss mit 95 % C.L. auf Massen war mit dem zur Verfügung stehenden Datensatz aus diesen Grund nicht möglich. Unter der Annahme, dass die Zerfallstopologie der skalaren Quarks der ersten zwei Generationen ähnlich dem Zweikörperzerfall des Stop ist, wurden die Resultate der Suche nach dem Zerfall stop -> c neutralino_1 auch für die Berechnung der Massenausschlussgrenzen für die Squarks der ersten beiden Familien benutzt. Zwei Möglichkeiten wurden hier in Erwägung gezogen: die Massenentartung zwischen vier (scalar u, d, c, s) und fünf (sbottom zusätzlich) Squarks. Die Ausschlussgrenzen mit 95 % C.L. auf die massenentarteten skalaren Quarks in den Fällen der "nur-rechts" oder "links-und-rechts" Eigenzustände sind die folgenden: - für die Massenentartung zwischen vier Squarks: M_squark < 95-96 GeV ("nur-rechts"), M_squark < 99-100 GeV ("links-und-rechts"); - für die Massenentartung zwischen fünf Squarks: M_squark < 96-97 GeV ("nur-rechts"), M_squark < 99-101 GeV ("links-und-rechts"); Mit der Annahme der Gaugino-Vereinigung an der GUT-Skala im MSSM wurden die Grenzen für die vierfach massenentarteten Squarks erneut in der Squark-Gluino Ebene interpretiert. Ferner wurde das absolute Limit auf den MSSM-Parameter M_2, der für tan(beta) = 4 aus anderen L3-SUSY-Suchen (für Chargino, Neutralino und skalare Leptonen) ermittelt worden ist, in ein Gluino-Massenlimit übersetzt. Die mit 95 % C.L. erhaltenen Ausschlussgrenzen in der Squark-Gluino Massenebene sind - M_gluino > 267-314 GeV, - M_squark > 99-100 GeV. / This thesis is devoted to searches for the scalar top and the scalar bottom quarks within the framework of the Minimal Supersymmetric Standard Model (MSSM) with the assumption of R-parity conservation. Searches for the following decay modes of the stop quark have been performed: stop -> c neutralino_1, stop -> b l sneutrino, (where l is either electron, muon or tau-lepton with equal probabilities) and stop -> b tau sneutrino (where only the tau-lepton is considered). In addition, a three body decay stop -> b W neutralino_1 has been searched for in the allowed mass region of M_stop > M_b + M_W + M_neutralino1 >= 86 GeV. For the sbottom quark the decay sbottom -> b neutralino_1 was considered. Each of these decay modes was considered independently assuming a branching ratio of 100 %. For this search, the experimental data of electron-positron collisions at center-of-mass energies (c.m.s.) in the range of 202-208 GeV have been used. These data were collected in the year 2000 by the L3 detector at the Large Electron Positron Collider (LEP) at CERN. The results of the year 2000 data analysis were also combined with results of the squark searches performed by the L3 Collaboration in previous years at center-of-mass energies from 161 up to 202 GeV. The analyzed squark decay channels determine the topology of the events of our interest: 2 jets (or b-jets) + missing energy (+ 2 leptons for stop three body decays). The stop -> b W neutralino_1 decay topology depends significantly on the further decay of the W boson and can have up to 6 jets in the final state. The assumed conservation of R-parity implies stability of the lightest supersymmetric particle (the LSP), which is the lightest neutralino. The LSP interacts only weakly and thus escapes undetected. This leads to a large missing energy as a feature of the signal events. The visible energy is roughly proportional to the difference between the masses of the squark and the LSP, and since the Standard Model background composition depends on the visible energy fraction, the whole analysis depends also on the value of this mass difference. Depending on the magnitude of visible energy, the Standard Model background can be grouped into three categories: - the two-fermion processes are e e -> e e, e e -> mu mu, e e -> tau tau and e e -> q q; - the four-fermion category is composed of e e -> W W, e e -> W e nu, e e -> Z Z and e e -> Z e e processes; - the two-photon background processes are e e -> e e e e, e e -> e e mu mu, e e -> e e tau tau and e e -> e e q q. The last process, e e -> e e q q, constitutes the largest fraction of all SM background processes (due to very high and highly fluctuating missing energy and the highest cross section). At the very first step of the analysis, only the events of interesting topology (with 2 jets and high missing energy) were preselected. Then, taking into account the small value of the theoretically predicted production cross section of the scalar quarks, the selection of stop and sbottom events was optimized by minimization of the 95 % confidence level expected upper limit on the squark cross section using calculated Monte Carlo events. In all selections optimized for each particular squark decay channel, the number of selected data events statistically agrees with the number of events expected from the Standard Model processes: - for stop -> c neutralino_1 decay, 29 data evens were observed, while 26.5 +- 2.7 were expected from the SM processes; - for the three body decay stop -> b l sneutrino, 4 data events were selected and the expectation from the Standard Model is 4.0 +- 1.0 events; - for the decay stop -> b tau sneutrino, the data and SM event numbers are 5 and 3.9 +- 1.0, respectively; - in the selection for stop -> b W neutralino_1, 184 data events were observed and 181.6 +- 3.0 were expected from the Standard Model; - for the bottom squark decay sbottom -> b neutralino_1 the observed 6 events correspond to the SM expectation of 7.7 +- 1.3. Thus, the MSSM scalar quarks were not observed in the experimental data and the search results are negative. The model independent 95 % C.L. upper limits on the squark production cross section have been derived from the numbers of the observed data events and numbers of events expected from the Standard Model. For calculation of the upper cross section limits, the results of the squark searches performed in the L3 data of c.m.s. energy 202 - 208 GeV were combined with results of searches performed by the L3 Collaboration previously in the data of c.m.s. energy from 161 up to 202 GeV. A new method has been developed for calculating such combined limits. This method takes into account the statistical independence of each measurement and the dependency of the squark production cross section on the center-of-mass energy. In this calculation, the systematic uncertainties in the Standard Model background estimation and in the signal selection efficiency have been also accounted for. For the considered squark decays, the typical obtained 95 % C.L. upper limits on the squark production cross section are: ~ 0.05-0.2 pb (for stop) and ~ 0.05-0.1 pb (for sbottom). In the searches for the stop three body decay stop -> b W neutralino_1, the cross sections above 0.7-1.0 pb have been excluded at 95 % C.L. Within the framework of MSSM with conserved R-parity, the experimental model independent cross section limits have been used for exclusion of the MSSM model parameters, in particular, exclusion of the stop and the sbottom masses. For each considered decay channel, the squark masses have been excluded in two possible scenarios: for the maximal and for the (approximately) minimal theoretical cross section. The first case corresponds to the maximal mixing between the left and right squark eigenstates, cos(theta) = 1; the second case is defined by the cos(theta) value, where squarks decouple from the Z boson. Depending on the mass difference between squark and the LSP, the squark masses have been excluded at 95 % C.L. up to the following values: - for stop -> c neutralino_1: M_stop < 90-93 GeV for minimal cross section, M_stop < 95-96 GeV for maximal cross section; - for stop -> b l sneutrino: M_stop < 87-89 GeV for minimal cross section, M_stop < 90-91 GeV for maximal cross section; - for stop -> b tau sneutrino: M_stop < 83-88 GeV for minimal cross section, M_stop < 88-91 GeV for maximal cross section; - for sbottom -> b neutralino_1: M_stop < 76-83 GeV for minimal cross section, M_stop < 94-97 GeV for maximal cross section. For both cases, the experimentally observed 95 \% C.L. mass exclusions are compared to the expected ones, which have been obtained from the Monte-Carlo assuming no SUSY particles. The observed exclusions of the squark masses are at the same level as the expected ones. The obtained 95 % C.L. upper limits on the stop production cross section in the decay stop -> b W neutralino_1 are bigger than the corresponding theoretical predictions, so, the exclusion of masses at 95 % C.L. was not possible with the available data sample. Assuming the topology of decays of the scalar quarks of the first two generations to be similar to the two body decay of the stop, the results of the searches for the decay stop -> c neutralino_1 have been also used for calculation of the mass exclusion limits for the squarks of the first two families. Two possibilities were considered here: the mass degeneracy between four (scalar u, d, c, s) and five (scalar b in addition) squarks. The 95 % C.L. exclusion limits on the mass degenerate scalar quarks for the cases of the "right-only" or "left-and-right" eigenstates are the following: - for the mass degeneration between 4 squarks: M_squark < 95-96 GeV ("right-only"), M_squark < 99-100 GeV ("left-and-right"); - for the mass degeneration between 5 squarks: M_squark < 96-97 GeV ("right-only"), M_squark < 99-101 GeV ("left-and-right"). Using the MSSM assumption about gaugino unification at the GUT scale, the limits on the four mass degenerate squarks have been reinterpreted on the squark-gluino mass plane. Moreover, the absolute limit on the MSSM parameter M_2, obtained for tan(beta) = 4 from other L3 SUSY searches (for chargino, neutralino and scalar leptons), has been translated into a gluino mass limit. The obtained 95 % C.L. exclusions in the squark-gluino mass plane are - M_gluino > 267-314 GeV, - M_squark > 99-100 GeV.

Ce mémoire de thèse présente une étude de l’identification des candidats tau par différents algorithmes dans les collisions pp au LHC et impliquant différents processus physiques : électrofaibles, production de quarks top, SUSY et QCD. Les leptons tau sont identifiés géométriquement dans la voie de désintégration hadronique car les jets issus de ces désintégrations sont généralement plus étroits et mieux isolés que les jets QCD. Une discrimination supplémentaire est appliquée en exigeant la présence d’un des modes de désintégration hadronique dominants du tau contenant un ou trois pions chargés et jusqu’à deux pions neutres. Les jets QCD produits en très grande quantité dans les collisions proton-proton du LHC constituent une source de bruit de fond importante pour les analyses impliquant l’identification de leptons tau. L’analyse des jets QCD produits dans le LHC pour les études de mauvaise identification porte sur 36 pb−1 de données enregistrées par le détecteur CMS pendant l’année 2010. Les efficacités d’identification et les taux de mauvaise identification nous révèlent une forte dépendance en les caractéristiques de l’évènement étudié : les efficacités sont globalement plus basses dans les évènements présentant une activité hadronique importante comme les évènements SUSY et tt que dans les évènements QCD mous et électrofaibles. Les effets de l'environnement sur l’identification de tau sont paramétrés à l’aide de deux variables globales : somme(ET) et la multiplicité de hadrons chargés. Les efficacités chutent typiquement de 80% entre les multiplicités de hadrons les plus faibles et le plus fortes. / The work presented in this thesis deals with tau identified by different algorithms in pp collision at LHC and in various kind of physics processes : Electroweak, top quark production, SUSY and QCD. Tau leptons are geometrically identified in their hadronic decay mode because the jets coming from those decays are generally more collimated and isolated than QCD jets. An additional discrimination is applied in asking the presence of one of the dominant # decay modes containing one or three charged pions and up to two neutral pions. The QCD jets produced in large quantities in LHC proton-proton collisions constitute a large background source for analyses involving tau-ID. The analysis of QCD jets produced in LHC for fake-rate studies processed 36pb−1 of data registered during year 2010. Identification efficiencies and fake-rate distributions reveals a strong dependency on event type : distributions are globally lower in events having a large hadronic activity like SUSY and tt events than in electroweak and QCD events. The environment effects of the collision on the identification of tau leptons can be parameterized by two global variables : somme(ET) and charged hadron multiplicity. Efficiencies drop down by 80% between the lowest and the highest charged hadron multiplicities.

This thesis dives into the three main aspects of today's experimental high energy physics: detector operation and data preparation, reconstruction and identification of physics objects, and physics analysis.The symbiosis of these is the key to reach a better understanding of the underlying principles of nature. Data from proton-proton collisions at a centre-of-mass energy of 13 TeV collected by the ATLAS detector during 2015-2018 are used. In the context of detector operation and data preparation, the data quality assessment for the Liquid Argon calorimeter of the ATLAS experiment is improved by an adaptive monitoring of noisy channels and mini noise bursts allowing an assessment of their impact on the measured data at an early stage. Besides data integrity, a precise energy calibration of electrons, positrons and photons is essential for many physics analyses and requires an excellent understanding of the detector. Corrections of detector non-uniformities originating from gaps in between the Liquid Argon calorimeter modules and non-nominal high voltage settings are derived and successfully recover the homogeneity in the energy measurement. A further enhancement is reached by introducing the azimuthal position of the electromagnetic cluster in the calibration algorithm. Additionally, a novel approach to exploit tracking information in the historically purely calorimeter-based energy calibration for electrons and positrons is presented. Considering the track momentum results in an about 30% better energy resolution for low-pT electrons and positrons. The described optimisation of the energy calibration is especially beneficial for precision measurements which are one way to test and challenge our current knowledge of the Standard Model of particle physics. Another path is the hunt for new particles, here presented by a search for stopped long-lived particles suggested by many theoretical models. This analysis targets gluinos which are sufficiently long-lived to form quasi-stable states and come to rest in the detector. Their eventual decay results in large energy deposits in the calorimeters. The special nature of the expected signature requires the exploration of non-standard datasets and reconstruction methods. Further, non-collision backgrounds are dominant for this search which are investigated in detail. In the context of simplified supersymmetric models an expected signal sensitivity of more than 3σ for gluinos with a mass up to 1.2 TeV and a lifetime of 100 μs is achieved.

The Standard Model (SM) of particle physics is widely taken as an elegant effective theory of nature at the electroweak scale, with new physics expected at higher energy. Collider searches and other experimental inputs play a vital role in our hunt for the unknown physics, offering great insights along the way and eventually establishing the extension to the SM. Here we present our studies on prospects of direct and indirect searches for three types of models beyond the SM. The Inert Doublet Model (IDM) extends the SM electroweak sector by an extra Higgs doublet with a Z₂-symmetry. We first examine the IDM dilepton signal at the LHC with a center-of-mass energy of 14 TeV and find it exceeding SM backgrounds at 3σ–12σ significance level, with 100 fb⁻¹ integrated luminosity. We further show that it is possible to obtain the IDM trilepton signal at the 5σ significance level, with an integrated luminosity of 300 fb⁻¹. The Left-Right Twin Higgs (LRTH) model solves the little Hierarchy problem by taking the SM Higgs as a pseudo-Goldstone boson from the spontaneous breaking of a global symmetry. We focus on the discovery potential of the heavy top quark partner in the LRTH model at the LHC. With a luminosity of 30 fb⁻¹ at the early stage of the LHC operation, we conclude that the heavy top partner could be observed at a significance level above 5σ. Supersymmetric extensions of the SM enable cancellations among loop corrections to the Higgs mass from bosonic and fermionic degrees of freedom, leading to a solution to the well-known Hierarchy problem. However, the supersymmetry has to be broken by certain mechanism. We present an exploration of the B-physics observables and electroweak precision data in three distinct soft supersymmetry-breaking scenarios. Projection for future sensitivities of the precision data is also explored.

Searches for the supersymmetric partner of the top quark (stop) are motivated by natural supersymmetry, where the stop has to be light to cancel the large radiative corrections to the Higgs boson mass. This thesis presents three different searches for the stop at √s = 8 TeV and √s = 13 TeV using data from the ATLAS experiment at CERN’s Large Hadron Collider. The thesis also includes a study of the primary vertex reconstruction performance in data and simulation at √s = 7 TeV using tt and Z events. All stop searches presented are carried out in final states with a single lepton, four or more jets and large missing transverse energy. A search for direct stop pair production is conducted with 20.3 fb−1 of data at a center-of-mass energy of √s = 8 TeV. Several stop decay scenarios are considered, including those to a top quark and the lightest neutralino and to a bottom quark and the lightest chargino. The sensitivity of the analysis is also studied in the context of various phenomenological MSSM models in which more complex decay scenarios can be present. Two different analyses are carried out at √s = 13 TeV. The first one is a search for both gluino-mediated and direct stop pair production with 3.2 fb−1 of data while the second one is a search for direct stop pair production with 13.2 fb−1 of data in the decay scenario to a bottom quark and the lightest chargino. The results of the analyses show no significant excess over the Standard Model predictions in the observed data. Consequently, exclusion limits are set at 95% CL on the masses of the stop and the lightest neutralino.

In this thesis, the general framework of supersymmetric quantum mechanics and the path integral approach will be presented (as well as the worked out example of the supersymmetric harmonic oscillator). Then the theory will be specialized to the case of supersymmetric quantum mechanics on Riemannian manifolds, which will start from a supersymmetric Lagrangian for the general case and the special case for S2. Afterwards, there will be a discussion on the superfield formalism. Concluding this thesis will be the Hamiltonian formalism followed by the inclusion of deforma- tions by potentials.

Ce travail de thèse s'attache à simuler le potentiel de découverte de la supersymétrie (SUSY) au collisionneur LHC du CERN avec le détecteur ATLAS dans le canal pp->w1+z2->3l+X. Les résultats sont fournis pour les phases de basse et de haute luminosité du LHC. Dans un premier temps, nous montrons que la SUSY induit un excès à 5 écart-standards par rapport aux pédictions du Modèle Standard (MS) dans la topologie inclusive trilepton dans un grand domaine de l'espace des paramètres SUSY. Nous indiquons que les contributions des processus SUSY autres que pp->w1+z2 (fond supersymétrique) sont non négligeables. Ensuite, nous affinons l'analyse pour augmenter la pureté du signal pp->w1+z2->3l+X. Ceci nous permet d'étudier les propriétés spécifiques de ce processus. En particulier nous reconstruisons la différence de masse m(z2)-m(z1). Nous proposons une méthode originale pour reconstruire tout le secteur des gauginos à partir de cette observable expérimentale (cette méthode suppose l'unification de la masse des gauginos). Un chapitre est également consacré à l'étude de l'uniformité spatial e de la réponse des photomultiplcateurs Hamamatsu R5900 utilisés pour la lecture du calorimètre hadronique (Tilecal) d'ATLAS.

I present the results of a search for pair production of scalar top quarks (~t1) in an R-parity violating supersymmetric scenario using 322 pb_1 of pp collisions at ps = 1.96 TeV collected by the upgraded Collider Detector at Fermilab. I assume each ~t1 decays into a tau lepton and a b quark, with branching ratio B, and search for final states containing either an electron or a muon from a leptonic tau decay, a hadronically decaying tau lepton, and two or more jets. Two candidate events pass my final selection criteria, consistent with the expectation from standard model processes. I present upper limits on the cross section times branching ratio squared (~t1~t1)B2 as a function of the stop mass m(~t1). Assuming B = 1, I set a 95% confidence level limit m(~t1) > 153 GeV=c2. These limits are also fully applicable to the case of a pair produced third generation scalar leptoquark that decays into a tau lepton and a b quark.