Techniques for Estimating Intrinsic Resolution Cern估计本征分解欧洲核子的研究中心的技术的研究.pptVIP

August 26, 2003 P. Nilsson, SPD Group Meeting Contents: Intrinsic Resolution The 2003 test beam Preprocessing and tracking Assessment of intrinsic resolution Calculation of the “tracking error” Measurement error and Multiple Scattering (2 different methods) Residuals ? intrinsic resolution (3 different methods) Iterative Method The 2003 Test Beam Proton/pion beam at 120 GeV/c Heavy ion beam (In) at 158 GeV/c The study presented here is based on proton beam data Setup: Two minibuses, each with two single chip assemblies, constituting four reference planes (0, 1, 3 and 4). Plane 2 is the plane under study Preprocessing and Reconstruction Raw data Converter Data Cleaning Used dataset: c0r0_tilt0_thXXX Several noisy pixels were removed “Typical” track quality ?2 distributions Assessment of Intrinsic Resolution General strategy: Start from measured residual distribution Convolution of intrinsic detector resolution (flat pdf) and “tracking error” (Gauss pdf) Calculate “tracking error” Knowing tracking error ? get intrinsic resolution “Tracking Error” Tracking Error : Contribution of the measurement error to the difference (xproj – xhit) In our case: x = az +b, z = 0 in test plane, b = xproj Tracking Error (2) Tracking error will depend on σi σ2i = σ2intrinsic + σ2multiple scattering Multiple Scattering (1) Multiple Scattering (2) Multiple Scattering (3) The χ2 Method (Tracking error cont.) Analytic Calculation (Tracking error cont.) Intrinsic Resolution Traditional approximation: σ2intrinsic = σ2 residual – σ2tracking “Hypothesis test” - method “Hypothesis Test” Method Toy MC to convolute: Gauss (estimated tracking error) Flat (intrinsic resolution = ) Loop over parameter w and test hypothesis: “Data compatible with simulated distribution” ? Reject hypothesis if test statistics critical value (Kolmogorov-Smirnov / χ2 test) Result: sample output of routine Results: Intrinsic resolut