【作者】 谢征宇；

【导师】 董宝田；

【作者基本信息】 北京交通大学 ， 交通运输规划与管理， 2012， 博士

【摘要】 随着中国高速铁路步入迅速发展阶段,与高铁相配套的现代化综合客运枢纽的新建和改扩建工程项目也大量开展,高铁综合客运枢纽逐步发展成为高速铁路、航空、公路、水运与城市轨道交通、公共交通、小汽车、出租车等多种交通方式立体交叉衔接为主要特征的综合客运枢纽。作为综合客运网络的重要节点,高铁综合客运枢纽是多种客运交通方式和大规模客流的集散场所,枢纽的旅客集散量将会随着高速铁路运营里程的不断增加持续急剧增长,这对高铁综合客运枢纽客流安全预警提出了更高的要求。目前铁路客运枢纽的视频监控系统功能比较单一,对视频图像的监视和分析主要依靠工作人员持续地监视屏幕,人为解读视频信息,做出相应的决策。当监控点增多、监控强度加大时,人工操作容易产生失误,不能及时发现客流安全隐患,导致铁路客运枢纽客流安全应急响应处置只能被动回应,并未充分发挥枢纽视频监控系统的主动性。鉴于高铁综合客运枢纽安全预警的需求和当前铁路客运枢纽视频监控系统的不足,本文依托国家高技术研究发展计划“高速铁路综合客运枢纽集散服务网络行人安全状态识别技术”项目,开展对高速铁路综合客运枢纽客流安全预警关键技术的研究。通过合理布设视频监控采集点,快速、准确提取客流参数实时信息,精确预测瓶颈点客流参数信息,有效把握客流状态变化趋势,达到及早发现客流安全隐患,及时发布预警信息的目的。本文首先从信息采集、信息处理和信息融合三个角度对高铁综合客运枢纽客流安全预警的需求进行了分析,根据需求设计了高铁综合客运枢纽客流安全预警系统结构和预警流程,指出了高铁综合客运枢纽客流安全预警的关键技术。根据高铁综合客运枢纽安全预警对信息采集的需求,提出高铁综合客运枢纽视频监控采集点分级布设策略,分析了采集点分级布设流程,针对无障碍和有障碍功能区域分别提出基于正六边形区域剖分的视频监控采集点布设模型和基于集合覆盖的视频监控采集点布设模型。并运用该策略对北京南站地下一层进行了实例布设。根据高铁综合客运枢纽安全预警对信息处理的需求,结合高铁综合客运枢纽的客流图像特征,提出基于证据推理的均值背景模型。在对经过背景处理后的客流前景图像的识别研究中,提出基于标记自动获取的改进分水岭算法和基于客流图像连通域特征的识别算法,并利用基于相关模型和算法建立的客流人数识别系统对北京南站某检票口的客流图像进行了识别。根据高铁综合客运枢纽安全预警对信息融合的需求,结合高铁综合客运枢纽视频监控采集点的关联结构,通过对枢纽视频监控采集点间的关联度和客流参数预测方法的研究,提出基于空间的客流参数预测算法和基于时空的客流参数预测模型。利用提供的预测信息,获取瓶颈点的客流状态安全等级,发布有效的预警处理指示。并利用北京南站某检票点及其空间相连的视频监控采集点的客流密度及客流速度数据对算法和模型进行了验证。通过本文对高铁综合客运枢纽客流安全预警关键技术的研究,能够为预警处理提供正确的判断信息,发布及时有效的预警指示,提高枢纽客流安全预警的速度、准度和精度,确保高速铁路综合客运枢纽的客流高效安全集散,降低事故的发生率和严重性,对提升高速铁路客运安全保障能力具有指导意义和参考价值。更多还原

【Abstract】 With the rapid development of high-speed railway in China, the rebuilding and extension of modern comprehensive passenger transport hubs which are built for matching with high-speed railway have been implemented largely. The high-speed railway comprehensive passenger transport hub has become the crossing and interface of multi-transportation which includes high-speed railway, civil aviation, highway, waterway, urban rail transit, public transport, motor vehicle and taxi. As the vital node of passenger transport net, the high-speed railway comprehensive passenger transport hub is the important distribution place of various passenger transportations and massive passenger flow. With the increase of high-speed railway operation mileage, the distribution quantity of passengers will be sustained to increase sharply. So there is a higher demand for security forewarning of passenger flow in high-speed railway comprehensive passenger transport hub.As the video surveillance system functions of railway passenger hubs are relatively simple, the surveillance and analysis mainly depend on monitoring the screen continuously, interpreting the video information and making relevant decisions artificially. When the quantity of monitor points and the strength of monitor increase, manual operations make mistakes easily and can not find the security risks of passenger flow immediately. It leads to the passive response of railway passenger hub emergency handling and underplaying initiative of video surveillance systems. For the demand of security forewarning of passenger flow in high-speed railway comprehensive passenger transport hub, and the disadvantages of video surveillance system in railway passenger hubs, this paper focuses on the key technologies study of passenger flow safety forewarning in high-speed railway comprehensive passenger transport hub based on Hi-Tech Researcher and Development Program of China "Passenger Security Status Recognition of High-speed Railway Comprehensive Passenger Transport Hub Distribution Network." By the rational layout of video monitor point, rapidly and precisely abstracting information of passenger flow, exactly forecasting parameters of passenger flow parameters and effectively grasping the variation tendency of passenger flow status, achieves to detect the security risks of passenger flow and releases the information of security forewarning promptly.Firstly, we analyze the demand of passenger flow security forewarning in high-speed railway comprehensive passenger transport hub from information acquisition, information processing and information fusion. According to the demand, we design the architecture of passenger flow security forewarning, describe the procedure of security forewarning and present key technologies of passenger flow safety forewarning in high-speed railway comprehensive passenger transport hub.Secondly, according to the information acquisition demand of safety forewarning in high-speed railway comprehensive passenger transport hub, we present the hierarchical layout strategy of video monitor points, analyze the hierarchical layout procedure of monitor points. For the function areas with or without obstacle, we respectively present different layout model. By using the layout strategy, we take Beijing south railway station for example to implement layout.Thirdly, according to the information processing demand of safety forewarning and the characteristics of passenger flow image in high-speed railway comprehensive passenger transport hub, we present the average background model bases on the Dempster-Shafer theory for building background image. After the subtraction of background image, we present the improved watershed algorithm based on marker automatic acquisition and recognition algorithm based on the connected domain characteristic of passenger flow image. We take one ticket entrance of Beijing south railway station for example to recognize the quantity of passenger flow by the passenger quantity recognizing system based on the algorithms mentioned above.Finally, according to the information fusion demand of safety forewarning and association structure among video monitor points in high-speed railway comprehensive passenger transport hub, by the studies of association degree among video monitor points and the forecast mode of passenger flow parameters, we present the passenger flow parameter forecast algorithm based on space and passenger flow parameter forecast model base on space-time. By the forecast information, to get the passenger status security level of choke point and release the order of security forewarning. We test the algorithm and model by the passenger flow density and velocity from the ticket entrance and association points in Beijing south railway station.Through the key technologies study of passenger flow safety forewarning in high-speed railway comprehensive passenger transport hub, we can improve the speed and precision of safety forewarning, ensure the efficient and safe distribution of passenger flow and reduce the incidence rate and severity of accidents. The study in this paper has the guidance and reference for improving the security capabilities of railway passenger transport.更多还原