UWB室内定位多标签防碰撞算法研究

发布时间：2018-09-05 15:44

【摘要】：随着人工智能、无人工业等新兴产业的出现,室内定位技术也日益受到人们的关注,室内定位技术指在室内环境中对目标物体的位置完成定位或目标物体移动路径跟踪,该技术主要应用基站、无线通讯技术等多种技术的结合,由此形成一个适用于高精度定位需求的室内定位系统。本文的研究内容是基于超宽带多标签防碰撞的室内实时定位技术。首先阐述目前海内外对超宽带室内定位技术的研究近况,在认识到超宽带室内定位技术对多定位标签防碰撞的需求的情况。明确了本文以基于超宽带室内定位系统为平台的多标签防碰撞算法的研究。详细介绍了超宽带定位系统的组成、定位算法以及基站间的时钟同步模型。随后,分析了二进制树算法和ALOHA算法两类标签防碰撞算法,详细阐述两类算法的数学模型,对纯ALOHA、时隙ALOHA、帧时隙ALOHA以及动态帧时隙ALOHA做了系统的分析,并通过算法比较,提出将动态帧时隙ALOHA算法应用于超宽带室内定位系统的想法。基于动态帧时隙ALOHA算法,文章详细分析了实现动态帧所涉及到的标签估计算法,介绍并通过仿真比较了 Low bound、Schoute、Vogt和贝叶斯估计算法。详述贝叶斯标签估计算法的实现概率分布函数。研究基于超宽带的动态帧时隙ALOHA算法的贝叶斯标签估计算法。针对超宽带室内定位系统的系统组成以及定位实现的原理,研究分析了一种基于超宽带室内多标签定位系统的标签防碰撞算法改进思路。在多标签定位系统中,由于定位标签数量多,为便于对多定位标签的性能做测试。在文章末尾使用成功识别率测试对多标签定位系统做测试,并提出了基于超宽带室内定位系统的圆概率误差算法R95用以同时对多标签做定位精度分析。全文通过基于MATLAB的仿真以及对定位系统的成功识别率和精度测试,论证了超宽带室内定位系统的高精度性能以及基于贝叶斯标签估计算法的动态帧时隙ALOHA的优势。本文的研究旨在为超宽带室内多标签定位系统的标签防碰撞提出一种可行的参考,用以实现定位系统的多标签防碰撞能力。
[Abstract]:With the emergence of artificial intelligence, unmanned industry and other emerging industries, indoor positioning technology has been paid more and more attention. Indoor positioning technology refers to the location of the target object or the tracking of the moving path of the target object in the indoor environment. This technology is mainly used in the combination of base station, wireless communication technology and other technologies, thus forming an indoor positioning system suitable for high precision positioning requirements. The research content of this paper is the indoor real-time positioning technology based on UWB multi-label anti-collision. This paper first describes the research status of UWB indoor positioning technology at home and abroad, and recognizes the need of UWB indoor positioning technology for anti-collision of multi-positioning tags. The research of multi-label anti-collision algorithm based on UWB indoor positioning system is made clear in this paper. The composition, location algorithm and clock synchronization model of UWB positioning system are introduced in detail. Then, two kinds of label anti-collision algorithms, binary tree algorithm and ALOHA algorithm, are analyzed in detail. The mathematical models of the two kinds of algorithms are described in detail. The pure ALOHA, slot ALOHA, frame slot ALOHA and the dynamic frame slot ALOHA are systematically analyzed, and the algorithms are compared. The idea of applying dynamic frame slot ALOHA algorithm to UWB indoor positioning system is presented. Based on the dynamic frame slot ALOHA algorithm, this paper analyzes the tag estimation algorithm involved in the implementation of dynamic frame in detail, and introduces and compares the Low bound,Schoute,Vogt and Bayesian estimation algorithms through simulation. The realization probability distribution function of Bayesian tag estimation algorithm is described in detail. Bayesian tag estimation algorithm based on UWB dynamic frame slot ALOHA algorithm is studied. According to the system composition of UWB indoor positioning system and the principle of location realization, an improved anti-collision algorithm based on UWB indoor multi-label positioning system is studied and analyzed. In the multi-label positioning system, because of the large number of localization tags, it is convenient to test the performance of multi-label. At the end of the paper, the multi-label positioning system is tested by using the successful recognition rate test, and the circle probability error algorithm R95 based on UWB indoor positioning system is proposed to analyze the positioning accuracy of multi-label at the same time. Through the simulation based on MATLAB and the successful recognition rate and precision test of the positioning system, the high precision performance of UWB indoor positioning system and the advantage of dynamic frame slot ALOHA based on Bayesian tag estimation algorithm are demonstrated. The purpose of this paper is to provide a feasible reference for the anti-collision of UWB indoor multi-label positioning system, so as to realize the multi-label anti-collision capability of the positioning system.
【学位授予单位】：海南大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN925

【相似文献】