面向行人空间认知的路径描述自动生成算法与模拟实验

An Automatic Generation Algorithm of Route Description for Pedestrian Spatial Cognition and Simulation Experiment

路径描述是人类心理空间表征外化的方式之一,与人们的日常生活息息相关。针对目前行人导航无法满足行人空间认知偏好多样性的问题,首先通过认知实验验证了参与者在确定自身位置时选取地标所考虑的主要因素为可视性和距离;然后结合行人的步行方向发展了一种基于地标可视性、地标与行人之间的距离以及步行方向的地标选取模型,并以该模型为基础提出一种以行人为中心的定性路径描述自动生成算法;最后,采用C++编程语言实现了基于路径描述自动生成算法的模拟程序。模拟实验表明,该算法能根据行人步行方向和位置的变化,实时调整相应的路径指示信息。因此,该文提出的算法可提供以行人为中心的路径描述动态生成服务,以满足行人空间认知偏好多样性的需求。

Route descriptions are the externalization of human mental spatial representation,and are so common in our daily life.However,the existing route descriptions are insufficient to satisfy the diversity of pedestrians′cognitive preferences,such as generate route descriptions based on properties of landmark but not on pedestrians′cognition.Several studies show that landmarks play an important role in route descriptions,hence,landmark selection is a key element for automatic generation algorithm of route description.In this paper,the main factors（i.e.,the visibility of landmarks and distances between landmarks and test positions）that effecting landmark selection are firstly verified through an in-situ cognitive experiment,then the landmark selection model,considering these two factors and the heading directions of pedestrians,is proposed.Based on the model,an automatically generated algorithm of pedestrian-centered qualitative route descriptions is devised.For the purpose of implementing the automatic generation algorithm for route descriptions,a simulation program is developed by C language.The simulation results indicate that the proposed algorithm can adjust route instructions in real time,as the heading directions and the locations of pedestrians change.Traditional navigation systems provide route descriptions at pre-set locations.In contrast,the algorithm proposed in this paper could provide a dynamic generating service of pedestrian-centered route descriptions,which compliances the diversity of spatial cognition preferences of pedestrians.