高能物理实验径迹系统流式数据获取框架的研究

Research on streaming data acquisition framework for high energy physics tracking system

高能物理实验研究基本粒子及其相互作用,需要获取和分析大量的实验数据以发现新粒子或测量已知粒子的特性。随着高能物理实验规模的不断扩大以及加速器能量和亮度的不断提升,海量数据的获取、处理及分析将更具挑战性。在高能物理实验中,径迹系统探测器的通道数和数据量尤为巨大,为了读出和在线处理径迹系统探测器产生的海量数据,本文结合Hadoop大数据框架,引入主流开源的大数据处理组件,研究实现了一种流式数据获取框架——BigDataDAQ,并应用于时间投影室探测器模型实验中。实验室性能测试结果表明:这一数据获取框架具有良好的数据吞吐和数据处理能力,且易于部署和管理,为未来高能物理径迹系统结合大数据框架研制处理海量数据的流式数据获取系统进行了有益的尝试,并提供了一种可行的解决方案。

[Background]High energy physics(HEP)experiments aimed at studying elementary particles and their interactions need to acquire and analyze large amount of experimental data to discover new particles or measure the properties of known particles.With the development of HEP experiments,the energy and luminosity of accelerators are increasing,the scale of experiments is expanding,and consequently the acquisition,processing and analysis of large amount of data will be more challenging.[Purpose]This study aims to design and implement a more advanced or effective distributed data acquisition framework to acquire and process the huge amount of data generated by tracking detectors for future HEP experiments where the number of channels and data volume of tracking detectors is extremely large.[Methods]The distributed data acquisition framework in HEP experiment divided into data flow software and online software was redesigned by using Hadoop big data framework,mainstream open source big data processing components was adopted to develope a new data acquisition framework—BigDataDAQ.Finally,this framework was applied to the prototype of the time projection chamber for verification.[Results&Conclusions]Results of performance test show that the framework has high performance in data throughput and data processing,and can be easily deployed and managed.It provides a feasible solution for the future data acquisition system of high energy physics experiment.