Novel String Banana Template Method of Track Reconstruction for high Multiplicity Events with Significant Multiple Scattering

Novel String Banana Template Method (SBTM) for track reconstruction in high multiplicity events in non-uniform magnetic field spectrometer with emphasis on the lowest momenta tracks with significant Multiple Scattering (MS) is described. Two steps model of track with additional parameter/s which takes into account MS for this particular track is introduced. SBTM is time efficient and demonstrates better resolutions than another method equivalent to the Least Squares method (LSM). 1 SBTM description Detailed Monte Carlo (MC) simulations (using GEANT [1], for example) could provide database of track’s characteristics (templates) for later fast use. The MS cone (as on Fig. 1a)) for an ensemble of particles with the same vector momentum ~ P has a significant width (volume) at low momenta. At fitting stage in high multiplicity event it results in heavy computation with covariance matrices [2]. For pattern recognition in such difficult cases it’s crucial to have narrow Search Windows (SW) what requires an a priori knowledge of momentum. It was found [3] that for another ensembles SWs are more narrow. The main idea of SBTM is in use of ensembles with 3 fixed points. Where the first point is initial vertex V0. As two other points two strips (pixels) i, j in two reference planes (RP) close to the middle and end of the track are used. It results in 2-dimensional {i, j} templates (see Fig. 1b)). For such ensemble which geometrical image in magnetic field has a Banana-like shape all necessary characteristics (centre of Bananas, their widths, angles, lengths ... ) could be saved a priori to track reconstruction. Two steps model of track is used: the first one – averaged over ensemble, gives rough estimations for SW (Banana center and width) and momentum of candidate. The second – per event dependent (takes into account MS for this particular track), permits to localize track in more narrow String SW and provides track parameters corrections depending on relative deflection of a String inside a Banana window. So SBTM uses more parameters in its model space than usually and it provides additional corrections to track’ parameters. At first track recognition stage one check different combinations of (i, j) signals in RPs and select such which has proper number of signals inside of Banana road. At the second stage relative positions of signals inside Banana are checked and if they are inside of a more narrow window String – track candidate is recognized. 1