KMS of Xinjiang Astronomical Observatory, CAS
| FRB搜寻数据处理平台搭建与测试 | |
| Alternative Title | Establishment and Test of FRB SearchPlatform |
王军
| |
| Subtype | 硕士 |
| Thesis Advisor | 陈卯蒸 |
| 2017 | |
| Degree Grantor | 中国科学院大学 |
| Place of Conferral | 北京 |
| Degree Discipline | 天文技术与方法 |
| Keyword | Frb搜寻 Gpu 望远镜 |
| Abstract | 快速射电暴（Fast Radio Burs， FRB)是一种瞬时射电爆发，具有持续时间短、色散值大、基本不重复等特性。自2007年发现首个FRB至今，天文学家仅发现21个FRB现象，而据天文学家早期估测，FRB每天至少会爆发2500次。FRB的起源及其应用领域等都是目前天文学家的研究热点，这些问题的进一步研究都依赖于发现更多的FRB。 搭建FRB搜寻系统是进行FRB搜寻研究的基础和前提，在早期的FRB搜寻过程中，天文学家一般采用脉冲星搜寻软件SIGPROC进行FRB的线下搜寻。但是SIGPROC的搜寻速度非常低，难以满足天文学家对FRB的搜寻要求。目前天文学家常用的FRB搜寻系统主要分为两种，一种是基于CPU的FRB搜寻系统，它仅依赖CPU的处理能力进行FRB的搜寻；另一种是基于CPU和GPU混合架构的FRB搜寻系统，这种系统充分发挥CPU在串行控制领域以及GPU在并行处理方面的潜力，可以极大地提高FRB的搜寻速度。 为配合新疆天文台的FRB及脉冲搜寻终端XPB （Xinjiang astronomical ob-servatory Pulsar Backend）的研制，本文在总结国内外FRB 搜寻算法的基础上， 分别搭建了基于CPU和GPU的FRB搜寻系统，并利用现己公开的部分FRB数据对两套系统进行了对比测试。 对比结果表明：相比基于CPU的FRB搜寻系统，基于GPU的FRB搜寻系统在同时保证搜寻精度的条件下，可以极大地提高了FRB搜寻的速度，GPU系统 计算速度可提高35 - 45倍左右。 拟建于新疆奇台的110 m 口径全可动射电望远镜（QTT）在建成之后将是全世界最大的全向可动射电望远镜，相比目前的南山25 m望远镜，QTT将具有 更大的反射面口径和接收机带宽，对FRB的持续搜寻和观测分析极其需要具有超强跟踪能力、超高灵敏度的可动式射电望远镜，QTT的高灵敏度结合FRB数字搜寻终端将可以开展高精度的FRB搜寻和观测研究。 |
| Other Abstract | Fast Radio Burst (FRB) is a kind of instantaneous radio burst, with short duration, large dispersion value, and non-repetitive characteristics. Since 2007, the first FRB has been discovered, astronomers only found 21 FRB phenomenon. However, according to astronomers’ early estimates, FRB will burst at least 2,500 times per day. The origin and application of FRB are the hotspots of astronomers, and further research on these issues relies on the discovery of more FRBs. The establishment of an FRB search system is the basis and prerequisite for FRB research. In the early process of FRB search, astronomers generally use the pulsar search software SIGPROC to search FRB offline. But the search speed of SIGPROC is very slow, which is difficult to meet the requirements of astronomers. At present, the FRB search system commonly used by astronomers is divided into two kinds, one is CPU-based FRB search system, which only depends on the processing capacity of CPU to carry out FRB search; the other FRB search system is based on CPU and GPU hybrid architecture, This system can give full play to the CPU in the field of serial control and GPU in parallel processing potential, and thus can greatly improve the FRB search speed. In this paper, based on the FRB search algorithm at home and abroad, this paper builds a FRB search system based on CPU and GPU, and uses the FRB search system based on the FRB and the pulse computing terminal XPB (Xinjiang astronomical ob-servatory Pulsar Backend) of Xinjiang Observatory. Part of the FRB data on the two systems was compared test. The comparison results show that the GPU-based FRB search system can im¬prove the speed of FRB search and improve the speed of GPU system by about 35-45 times compared with CPU-based FRB search system. The QTT, which is going to be built in QiTai, Xinjiang, will be the world largest fully steerable telescope after its completion. Compared to the current Nanshan 25m telescope, QTT will have a greater reflector diameter and receiver bandwidth. Con¬tinuous search and observation analysis of FRB especially need steerable radio tele-scopes with super tracking capability and ultra-high sensitivity, The high sensitivity of QTT combined with FRB digital search terminal will be contributable to carry out high-precision FRB search and observation studies. |
| Language | 中文 |
| Document Type | 学位论文 |
| Identifier | http://ir.xao.ac.cn/handle/45760611-7/1708 |
| Collection | 研究生学位论文 射电天文研究室 |
| Affiliation | 中国科学院新疆天文台 |
| Recommended Citation GB/T 7714 | 王军. FRB搜寻数据处理平台搭建与测试[D]. 北京. 中国科学院大学,2017. |
| Files in This Item: | ||||||
| File Name/Size | DocType | Version | Access | License | ||
| 2017届硕士学位论文-王军.pdf（3621KB） | 暂不开放 | CC BY-NC-SA | Application Full Text | |||
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.
Edit Comment