Search for chargino and neutralino pair production in final states with one lepton, two b-jets consistent with a Higgs boson and missing transverse momentum with the ATLAS detector at the LHC Run2

The R-parity conserving SUSY is one of the most motivated theories beyond the Standard Model. During the LHC 8 TeV Run1, many searches were dedicated to this model. Among those, the search for the lightest chargino and next-to-lightest neutralino (\chinoonepm, \ninotwo) pair production in final states with one lepton, two b-jets consistent with a Higgs boson and missing transverse energy. This search had negative outcome resulting in excluding \chinoonepm, \ninotwo masses up to 250 GeV for a massless lightest neutralino (\ninoone) at 95\% C.L.. The Run1 search used 20.3 fb$^{-1}$ of 8 TeV ATLAS data collected in 2012. In this thesis, this search with ATLAS at LHC 13 TeV Run2 is presented, it uses 36.1 fb$^{-1}$ ATLAS data collected in 2015 and 2016. This Run 2 search takes advantage of the sensitivity enhancement as the gain in the signal cross section is higher than the gain in the cross section of the main background, ttbar. Therefore, it provides an unprecedented sensitivity to high mass \chinoonepm, \ninotwo production. However, no evidence of new physics is observed and tighter new limits are placed on the \chinoonepm, \ninotwo production, with significant improvements over previous searches: \chinoonepm, \ninotwo masses up to 680 GeV for a massless neutralino \ninoone are excluded at 95\% C.L. The identification of jets containing B-hadrons, called b-tagging, plays a significant role in this and many other analyses at the LHC, in particular in the Higgs measurements, as h$\rightarrow$bb dominates the total width (58\%). The ATLAS b-tagging algorithms rely on B-hadron properties (long lifetime, high mass) and they are based on jets-associated tracks properties. In this thesis, a study on the optimisation of b-jet identification algorithms in ATLAS is presented as well as studies of the robustness of b-tagging performance with various conditions of the pixel detector, being the most important sub-detector for b-tagging. As most SUSY searches at the LHC, this analysis is carried out using Simplified Model of SUSY, i.e. the mass spectrum is chosen by hand and the considered SUSY decays are forced (branching fractions of 100\%). This way, the experimental sensitivity to the signal can be factorized out of the actual interpretation of the search result within realistic SUSY models. Suspect3 is one of the theory tools utilized to map the search result into realistic SUSY models, it is a SUSY spectrum calculator which calculates SUSY mass spectra and couplings. This thesis presents the post Higgs discovery SuSpect3 version we have been working on. This version takes advantage of the very important new Higgs boson mass input to constrain other basic model parameter typically most important one of the soft SUSY-breaking parameters, the top trilinear coupling, currently given as an input. The same idea also applies to the SUSY particles mass limits resulting from the SUSY searches at the LHC. This new Suspect3 version will facilitate generating SUSY spectra compatible with the experimental constraints, and is also more natural given the new fundamental input now available with the Higgs boson mass, which is moreover determined with a good accuracy.