基于移动物联网的展会现场多源数据分析方法研究

发布时间：2018-03-26 21:15

本文选题：移动物联网　切入点：数据包络分析　出处：《河北师范大学》2017年硕士论文

【摘要】：随着信息和通信技术的不断提高,移动终端的普及和层出不穷的新技术,移动物联网的应用面日益广泛,不断向人们生活的方方面面渗透,改善生活及体验服务。以“智慧化”概念引导着经济和科技的竞相发展,展会“智慧化”的进程加快和水平的提高,使得展会业在社会经济发展和社会生活方面贡献越来越大。展会“智慧化”促进我国会展行业的快速信息化,简化展会工作流程提升展会效率。展会业应用移动物联网,使得展会业具有了“大数据”的基础服务设施。针对移动物联网构建的复杂性,高耦合性,终端设备的多样性,使得获取数据的来源和数据类型多种多样化。为了在运营过程中去发现可能存在的规律现象或者需要解决的问题。通过采集展会现场数据,结合数据特性,根据所要研究的问题对数据集进行分类。对独立展位的展商和多个展位的展团采用数据包络分析,观察其投入产出效率;根据质心法求解出的客流位置,用DBSCAN方法对客流位置点在以展馆平面图上进行空间聚类,为了给出高密度客流集在地图上的区域范围,通过凸包生成算法构建人流高密集区域凸多边形。根据凸多边形区域,应用地理围栏技术,设定地理围栏,并以地图展示给公众。采用中心点最近距离法计算其周边疏散通道的优先权,确定人流疏散方案。按以上方法对基于移动物联网下的现场多源数据进行分析,得到以下结果(1)采用包络分析法对展位和展团进行分析,发现多数样本并未达到有效的DEA单元,主要是由规模效率和技术效率不足引起的,参展商需要根据自身的参展需求重新选择展位位置,展位数量和展位面积,以达到技术效率和规模效率上的资源合理配置,实现效率最大化。(2)经过质心算法求解出客流中每个人的位置集,采用客流密度4人/m2得到高密度客流集簇,并按凸包算法提取出边界,应用地理围栏对观众进行信息告知或在地图上显示出人流密集区域。通过对基于移动物联网下的现场多源数据进行分析,包络分析的结果可以对目前参展商的投入方案更改和完善,达到高效率。在应用公共空间内客流上限值和人际距离作为参数,用DBSCAN方法确定高密集点簇即为高密度客流集,是引发客流踩踏拥挤时间的标志集,为客流疏散提供可支持的方案。对凸多边形区域结合热度图为客流密集区域的识别和定位提供了定量和定性的方法,并且实现了其高密度客流区域量化,可以为展会的参与者管理者提供安全及风险信息。
[Abstract]:With the continuous improvement of information and communication technology, the popularity of mobile terminals and the emergence of new technologies, the mobile Internet of things is increasingly widely used, and continues to permeate all aspects of people's lives. To improve life and experience services. With the concept of "wisdom" to guide the development of economy and science and technology, the exhibition "smart" process to accelerate and improve the level, It makes the exhibition industry contribute more and more to the social and economic development and social life. The "wisdom" of the exhibition promotes the rapid informatization of the exhibition industry in China, simplifies the work process of the exhibition, and promotes the efficiency of the exhibition. The exhibition industry applies the mobile Internet of things. The exhibition industry has the basic service facilities of "big data". Aiming at the complexity, high coupling and diversity of terminal equipment of mobile Internet of things, To diversify the sources and types of data to be obtained. In order to discover possible regular phenomena or problems to be solved in the course of operation, by collecting the on-site data of the exhibition, combining the characteristics of the data, The data sets are classified according to the problems to be studied. The data envelopment analysis is used to observe the input-output efficiency for the exhibitors of independent stands and the groups of several booths, and the location of the passenger flow is calculated according to the centroid method. In order to give the area range of high-density passenger flow set on the map, the convex polygon of high-density passenger flow area is constructed by the convex hull generation algorithm, according to the convex polygon region, according to the convex polygon region, which is based on the convex polygon region. Use geo-fencing technology, set geo-fencing, and display it to the public on a map. Use the centre-point nearest distance method to calculate the priority of its surrounding evacuation passage. According to the above method, the field multi-source data based on the mobile Internet of things are analyzed, and the following results are obtained: (1) Envelope analysis is used to analyze the booth and the exhibition group, and it is found that most of the samples do not reach the effective DEA unit. It is mainly caused by the shortage of scale efficiency and technical efficiency. Exhibitors need to re-select the booth position, the number of booths and the space of the booth according to their exhibitors' demand, in order to achieve the rational allocation of resources on technical efficiency and scale efficiency. To realize the maximum efficiency, the center of mass algorithm is used to solve the location set of each person in the passenger flow. The high-density passenger flow cluster is obtained by using the passenger flow density of 4 person / m2, and the boundary is extracted according to the convex hull algorithm. The geographic fence is used to inform the audience or to display the crowded area on the map. By analyzing the field multi-source data based on the mobile Internet of things, The result of envelopment analysis can change and perfect the investment scheme of exhibitors and achieve high efficiency. Using the upper limit value of passenger flow and the interpersonal distance in public space as parameters, the DBSCAN method is used to determine the high density point cluster as the high density passenger flow set. It is the symbol set that causes the passenger flow to stampede the crowded time, and provides the supporting scheme for the passenger flow evacuation. It provides a quantitative and qualitative method for the identification and location of the passenger flow dense area by combining the heat intensity map of the convex polygon region, and provides a quantitative and qualitative method for the identification and location of the passenger flow dense area. It can provide the safety and risk information for the participants in the exhibition.
【学位授予单位】：河北师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.44;TN929.5;TP274

【相似文献】