【作者】 徐敏；

【导师】 张子平； 王贻芳； 何康林；

【作者基本信息】 中国科学技术大学 ， 粒子物理与原子核物理， 2009， 博士

【摘要】 北京正负电子对撞机(BEPC)及其大型通用探测器——北京谱仪(BES)改造工程BEPCⅡ/BESⅢ已经成功联调运行,并于2008年获取首批对撞数据。事例顶点重建是BESⅢ数据分析系统的重要组成部分,是物理分析的前提条件。BESⅢ的整体设计中未包含顶点探测器,粒子在探测器上的最内层击中信息位于主漂移室内,共振态粒子的衰变顶点大部分不能被直接探测。事例顶点重建过程中需要解决的主要问题是,如何利用共振态粒子衰变顶点的条件优化末态粒子径迹参数,如何高精度地重建共振态粒子的质量和寿命,如何有效去除前期径迹重建过程中产生的伪径迹,获得高精度的事例初级顶点。高效和精确的顶点重建能够提高径迹的质量分辨,有效地压低本底事例,以便观察信号事例特征。高精度的事例初级顶点重建能够准确地定位对撞束团的中心位置和分辨。本论文针对上述问题,围绕BESⅢ上的顶点和共振态粒子重建展开以下研究工作:(1)相对长寿命共振态粒子K_s~0,∧/(?)和Ξ重建研究共振态粒子K_s~0和∧/(?)是BESⅢ上大量产生的次级衰变粒子,它们的重建关键是衰变顶点的重建和粒子在衰变之前飞行过程中的运动学拟合。本论文详细讨论了全局顶点拟合方法的原理,研究了顶点限定性条件、粒子在飞行过程中满足的运动学限定性条件,以及共振态粒子参数的估算方程;描述了共振态粒子重建软件的设计和实现。利用蒙特卡罗数据进行重建,给出共振态粒子重建的质量,并通过对衰变长度的研究给出重建K_s~0,人的本底压制条件。详细分析了共振态粒子K_s~0寿命测量过程中探测器分辨、探测效率函数,设计开发了不借助模拟信息而直接运用数据估算探测效率的算法。蒙特卡罗数据研究表明拟合的寿命结果在误差范围内是合理的,从而为真实数据的共振态粒子寿命测量构建了基础框架。还讨论了级联衰变粒子Ξ的重建,通过对多次衰变粒子的重建研究,进一步验证了算法设计的合理性与可行性。(2)基于Kalman滤波方法的顶点重建算法及顶点寻找算法研究基于Kalman滤波方法,在BESⅢ离线软件框架BOSS下开发了顶点拟合算法。事例初级顶点重建的精确度直接影响对撞束团的离线重建位置和分辨,并影响产生自初级顶点的次级粒子重建。提出基于Kalman滤波方法的BESⅢ顶点拟合算法,通过对径迹的过滤、预测和平滑三个步骤,有效去除主漂移室径迹重建过程中的伪径迹信息,直接改善事例初级顶点拟合结果。详细讨论了Kalman顶点拟合软件设计与实现,描述了算法在BESⅢ上的具体运用,并结合蒙特卡罗样本优化了算法中相关的参数。提出了BESⅢ上的几何顶点寻找算法,其目的是将事例的末态粒子径迹进行归类。详细阐述了该算法的原理并描述了相应的软件设计和实现,蒙特卡罗数据研究表明经过顶点寻找之后的顶点拟合,其顶点分辨有明显改善。(3)BESⅢ实验数据的事例初级顶点和次级顶点重建及数据分析BESⅢ实验数据分析是对本论文中顶点重建算法,包括顶点寻找、全局顶点拟合以及基于Kalman滤波的顶点拟合算法的可靠检验。详细描述了针对BESⅢ取数初期,顶点重建的具体方案,通过实验数据中事例初级顶点重建分析了束团对撞中心位置的偏移结果,并根据实际情况改进了物理分析中径迹选择的条件。同时给出了实验数据中共振态粒子K_s~0,人和Ξ的初步重建结果。此外,为了满足BESⅢ数据质量监测系统(DQM)对事例顶点重建运行时间性能的要求,讨论了快速顶点拟合算法,设计并实现了相应的软件包。蒙特卡罗样本测试的结果表明该算法完全可以满足DQM对顶点重建的要求。更多还原

【Abstract】 The Beijing Electron Positron Collider(BEPC) and Beijing Spectrometer (BES) have been upgraded to BEPCⅡ/BESⅢ.The raw collision data have been collected at a high luminosity since July,2008.Event reconstruction,and especially vertex reconstruction,is an indispensable and important part of BESⅢdata analysis system,and also a precondition for subsequent physics analysis.BESⅢdoes not include a vertex detector in its overall design.The hit information of tracks that closest to the origin locates at the Main Drift Chamber(MDC) of BESⅢ,thus the decay vertices of resonance particles cannot be detected directly. The key problems of vertex reconstruction are:how to optimize track parameters of daughter particles based on the fact that they derive from a common decay point;how to precisely reconstruct the mass and lifetime of relatively long lifetime resonance particles;how to determine strict primary vertex positions and effectively remove the ghost tracks generated from track reconstruction,to suppress backgrounds so as to analysis signal events.High-precision primary vertex reconstruction will be able to determine the average of beam profile.In order to solve the above problems on vertex and resonance particle reconstruction at BESⅢ,The research works of this dissertation are:(1) Reconstruction of relatively long lifetime resonance particlesThe relatively long lifetime resonance particles,such as K_s~0 and∧/(?),etc,are abundantly produced atτ-charm energy region at BESⅢ.The reconstruction of resonance particles is actually reconstruction of decay vertices and kinematic fitting between the production and decay vertices.In this thesis,a global vertex fitting algorithm based on the Least Squares Method is presented,in which we also study the effects of vertex constraints,kinematic equations and the calculation of track parameters of resonance particles.A global vertex reconstruction program is designed and implemented using C++ language under the framework of BESⅢ offline Software System,BOSS.The results of resonance particle reconstruction based on Monte Carlo simulation are given.The cut on decay length is studied in order to suppress the backgrounds in physics analysis,and the reconstruction efficiency for their lifetime measurements is discussed.In addition,the reconstruction of cascading decay particleΞ,is also studied.It further verifies the applicability and feasibility of the algorithm.(2) Vertex reconstruction based on Kalman filter technique and vertex finding algorithmThe purpose of primary vertex reconstruction is to determine the interaction points of events,and then further determine the average of beam profile in high energy physics experiments.The information of a general interaction point is a critical parameter for reconstruction of secondary particles decayed from that point.In the thesis,a BESⅢvertex fitting algorithm based on Kalman filter technique is developed for handling the problems of ghost tracks and badly fitted tracks.The corresponding program is discussed,and its algorithm parameters are optimized using simulation data.A vertex finding algorithm which is to sort a set of tracks into subsets that share a common point is also proposed.(3) Analysis of vertex reconstruction using raw collision data at BE-SⅢFor checking the performance of vertex reconstruction algorithms,the beam bunch positions reconstructed from raw collision data collected at BESⅢare discussed.The procedure of raw collision data processing is described in detail, and for physics analysis,an optimized condition of selecting fitted tracks based on the results of primary vertex reconstruction is proposed.The preliminary results of resonance particle reconstruction from raw data is also given and analyzed.In order to achieve the time requirements in BESⅢData Quality Monitoring(DQM) system,a fast vertex fitting algorithm is developed and its results are compared with those of the global vertex fitting using simulation data.更多还原