The Research On CMS Outer Track Trigger And Parameter Reconstruction Algorithm

With the discovery of Higgs particle,the two general-purpose detectors ATLAS and CMS on the Large Hadron Collider(LHC)have achieved their design goals.Precision Higgs studies and the extended research for new physics present strong demands for higher luminosity of the LHC,therefore physicists propose the High Luminosity LHC(HL-LHC)upgrade,leading to a peak luminosity of 5×1034 cm-2s-1,with an average of about 140 proton-proton collisions(pileup,PU)per 25 ns bunch crossing(BX).High interaction rate associated with the pileup effects will be a particular concern for the CMS detector.A few years ago CMS has proposed the Phase-2 upgrade plan to prepare during LS3(2023-2025)for the upcoming HL-LHCIn the CMS phase-2 upgrade project,one plans to add outer track detector information in L1 trigger system to compress the massive data generated at 40MHz collision frequency to 750khz physical event rate.The total delay of L1 trigger system is 12.5us.The delay budget of the trigger algorithm detected in the outer diameter trace detection area is 5us.In addition,the online trigger in outer track detector need to provide trajectory parameter reconstruction results with particle transverse momentum Pt relative resolution better than 5%.Thus,for the track trigger algorithm,it needs to complete the identification of the valid charged particle trajectory amount the event and reconstruct the corresponding trajectory parameters as same time.This requires the designed track trigger algorithm to have high real-time performance and good resolution of reconstruction parameters as well.As the pre research of CMS phase Ⅱ L1 track trigger,this paper proposed the combination of retina algorithm and Kalman filter to handle the L1 trigger for CMS outer track detector,and designs the corresponding firmware module based on FPGA to realize the on-chip track search and parameter reconstruction of charged particles in the magnetic field.We aims to explore the feasibility of applying highly parallel retinal algorithm and high-precision Kalman filter algorithm to the L1 track trigger of CMS experimental outer diameter track detector,and provide a specific firmware design solution in the end.The specific research contents and innovations of this paper are as follows:1.Based on the application of retina algorithm in trajectory recognition and reconstruction in other high-energy physics experiments,a retina algorithm for identifying the trajectory of curved charged particles in magnetic field is designed and proposed.The function of retina algorithm is verified by using various types of single track events(simulating self generated track events and Single-Muon events).The feasibility study of retina algorithm as CMS experiment for trajectory recognition algorithm of charged particles in magnetic field is completed.2.The design of trajectory recognition algorithm based on retina algorithm for CMS phase II outer tracker detector--iterative retina algorithm(IR algorithm iterative retina)is proposed.Based on the retina algorithm,an iterative retina algorithm working in the way of iterative calculation is proposed for the first time,which greatly reduces the resource consumption required by the algorithm at the expense of a certain algorithm delay.Compared with the original retina algorithm,according to the scanning granularity of retina algorithm,the total computing overhead is reduced by several times to dozens of times.By designing the IR algorithm,the defect that the original retina algorithm itself can not identify tracks concurrently is solved,so that the algorithm itself has the function of identifying multiple particle tracks from samples at the same time,and there is no need to add the design of judgment algorithm.In addition,in the design of IR algorithm,through the control of threshold and algorithm granularity,the trajectory recognition efficiency of the algorithm in identifying the trajectory in t-tbar samples with CMS experimental event rate of pu200 is about 95%.3.The algorithm design of the combination of IR algorithm and Kalman filter is proposed.The Kalman filter is added after the IR algorithm to meet the requirements of CMS experiment for the accuracy of parameter reconstruction in online trajectory algorithm,and solve the problems of low accuracy and incomplete reconstruction parameters when using IR algorithm alone.After adding the Kalman filter,on the one hand,the algorithm can completely reconstruct the four 3D space parameters of the particle trajectory in the outer diameter trace space.On the other hand,the accuracy of the reconstructed parameters compared with the IR algorithm is greatly improved,especially the relative resolution of the particle transverse momentum Pt,which is an important index for identifying particles,is from the original 5-10%(PT<10Gev/c)/>10%in the low energy section(high energy Pt>10Gev/c)increased to 1-3%(low energy Pt<10Gev/c)/≈5%(high energy Pt>10Gev/c).The firmware corresponding to IR algorithm and Kalman filter algorithm is designed to meet the requirements of on-line algorithm delay and real-time in CMS experiment.In the firmware design of IR algorithm,in order to balance the consumption of various resources of FPGA,different schemes are used to realize the same firmware structure of retinal computing cells.In the firmware design of Kalman filter,an accurate state machine control algorithm is designed,and the same computing array is repeatedly used to calculate the system state transition.While keeping the algorithm delay unchanged,the resource overhead of Kalman filter is optimized,so that the number of DSP48 units consumed by each Kalman filter computing unit is reduced from 200 to 80.Taking KC705 evaluation board as the hardware platform,the firmware trajectory identification and parameter reconstruction online algorithm system of IR algorithm+Kalman filter is integrated.The t-tbar sample event(top antitop quark pair)of event rate pu200 is used as the input event to test the reliability of the algorithm and verify the performance of the algorithm.Finally,the trajectory recognition efficiency of the online algorithm is as high as 94%,and the purity is as high as 95%.In the whole range of particle transverse momentum,the relative reconstruction resolution of the algorithm for its transverse momentum PT is better than 5%.At the same time,the calculation delay of the online algorithm system is 4.35us,in which the IR algorithm is 2.45us and the Kalman filter is 1.86us,which meets the requirements of CMS experiment for algorithm delay.