基于复杂网络的地铁多班组人因失误的研究

发布时间：2018-03-11 03:38

本文选题：多班组　切入点：人因失误　出处：《首都经济贸易大学》2017年硕士论文　论文类型：学位论文

【摘要】：地铁,因其便利、快捷、经济、高效、安全等优势,成为人们的出行首选,在缓解城市道路交通压力和减少道路交通安全事故方面发挥着极为重要的作用。毫无疑问,地铁运营是一个庞大的系统工程,要依靠包括调度指挥、设备设施、运营管理等子系统来维持地铁的正常运营,依靠调度班组、车站班组等多班组的各司其职及其相互间的密切协调配合。特别是当地铁在正常运营过程中发生突发事件导致部分区间或者全线无法正常运营,这种多班组之间的协调配合更是至关重要。如何减少多班组在突发事件应急响应流程中的人因失误来提高应对突发事件的有效性,是地铁运营管理中亟待解决的安全问题之一,也是本论文的核心研究内容。人因失误主要分为操作性失误、认知性失误和沟通失误等。以往对于人因失误的研究大都着重于人的认知失误和行为失误,忽略沟通失误。然而事实证明,在对突发事件紧急应对失败的绝大多数原因恰恰是紧急状态下多班组协同作业过程中出现的沟通不畅或沟通错误导致的。因此,本文在认知失误和行为失误的基础之上,还要加入沟通失误进行研究。主要涉及的人员有行车调度员、司机、车站值班站长、行车值班员等。可以将其在各种突发事件应急处理流程中的相互间协调配合问题视为一个复杂网络的问题进行研究。为了构建该复杂网络,首先从任务分析和人因特征分析入手,通过描述地铁系统运行多班组的角色及各自职责,得到地铁各岗位工作人员人误行为分类及人误影响因素,根据VACCP分析地铁人因失误的等级。之后以屏蔽门故障为例,分析确定针对这种突发事件的应急处理流程中各班组的相互协作模式,并进而对十三种发生可能性较大的地铁运行突发事件都绘制应急响应流程图。将流程图中的结构分为“节点”和“边”,节点代表应急处理流程中的事件或设备的状态,节点存在失误表示的是存在人误影响因素;边代表应急处理流程中人的行为动作,边存在失误表示有人因失误发生。借鉴VACP模型对信息处理资源的要求等级来决定边权的大小,即应急处理流程中人的失误概率,并对此边权进行赋值。针对节点、边以及边权,选择NetworkX软件运用Python语句编程完成复杂网络的构建。再采用NetworkX自带的节点重要度分析方法,对所构建的复杂网络中的节点重要度(即人误影响因素的影响程度)进行排序及分析;选择边的重要度分析方法得出网络中相对重要的边(即容易发生人因失误的地方)。根据节点和边的重要度排序,提出地铁多班组协同作业应对突发事件过程中减少人因失误的对策建议。
[Abstract]:Subway, because of its advantages of convenience, speed, economy, efficiency, safety and so on, has become the first choice for people to travel. It plays an extremely important role in relieving urban road traffic pressure and reducing road traffic safety accidents. There is no doubt that, Subway operation is a huge system engineering. It depends on subsystems such as dispatching command, equipment and facilities, operation management and so on to maintain the normal operation of the subway, and depends on the dispatch teams. Station teams and other multi-shift teams perform their respective duties and closely coordinate with each other. Especially when a sudden event occurs during the normal operation of the subway, part of the section or the entire line is unable to operate normally. How to reduce the number of human errors in the emergency response process to improve the effectiveness of emergency response is one of the urgent safety problems in subway operation and management. It is also the core research content of this thesis. Human error is mainly divided into operational error, cognitive error and communication error. In the past, the study of human error mostly focused on human cognitive error and behavior error. Ignoring communication errors. However, it turns out that most of the reasons for the failure of emergency response to emergencies are precisely due to poor communication or communication errors in the course of cooperative work of multi-shift teams in a state of emergency. On the basis of cognitive and behavioral errors, this paper also includes the study of communication errors. The main personnel involved are traffic dispatchers, drivers, station watchmen, In order to construct the complex network, the problem of coordination and coordination between them in the process of dealing with emergencies can be considered as a problem of a complex network. First of all, by describing the roles and responsibilities of multi-shift teams in subway system, the classification of human error behavior and the influencing factors of human error are obtained from task analysis and human factor analysis. According to the VACCP analysis of the subway human error grade. Then taking the shield door fault as an example, this paper analyzes and determines the mode of cooperation among the teams in the emergency handling process of this kind of emergency. Furthermore, the flow chart of emergency response is drawn for 13 kinds of subway emergency that are more likely to occur. The structure of the flow chart is divided into "node" and "edge", and the node represents the state of the event or equipment in the emergency treatment process. The presence of errors in the node indicates that there are human error factors, while the side represents the human behavior in the emergency processing process, while the error indicates that someone is wrong. The level of information processing resources required by VACP model is used for reference to determine the size of the edge weight. For node, edge and edge weight, NetworkX software is selected to complete the construction of complex network with Python statement. Then, the analysis method of node importance is adopted in NetworkX. The degree of node importance (that is, the influence of human error factors) in the complex network is sorted and analyzed. The analysis method of importance of selected edges can get the relatively important edges in the network (that is, the places where human errors are prone to occur.) according to the importance degree of nodes and edges, The countermeasures and suggestions on how to reduce human error in the course of dealing with unexpected events are put forward.
【学位授予单位】：首都经济贸易大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U298

【相似文献】