VR技术对地下步行商业街心理环境辅助设计初探

文章力图以VR（虚拟现实）技术手段应用于城市地下步行商业街规划设计,来解决当前传统规划手段地下步行商业街规划中普遍存在的对其心理环境构筑的忽视和考虑不完善的诟病,以期最大限度消除规划设计中消极空间的产生,并对其关键技术进行了研究和探讨     

Personality subtypes of young moped drivers, their relationship to risk-taking behavior and involvement in road crashes in an Austrian sample

Objective

This study aimed to identify subtypes of young moped drivers and analyze how these subtypes are involved in risk preferences and road crashes.

Design

A group of Austrian teenage moped drivers (213, 28% girls) completed an online questionnaire about moped usage, injuries, driving style, inattention, impulsivity, and personality according to the Five-Factor Model of Personality and Cloninger's model of personality.

Results

A cluster analysis yielded four types of moped drivers. One type was characterized by a high level of neuroticism, the second type was characterized by a risky personality and a risky driving style, the third type had a more cautious driving style, and the fourth type was characterized by a risky personality and high levels on inattention and impulsivity.

Conclusions

Our conclusions suggest that young moped drivers should not be perceived as a homogenous group, according to measures of injury prevention and intervention.     

基于众包的嵌套流形匹配室内定位方法

随着普适应用的兴起,室内定位变得越来越重要。传统的基于指纹的定位方法需要现场勘测,所需时间及工作量巨大,且需实时更新,以适应室内变化,这大大限制了其应用范围。采用众包形式进行室内信息采集,并记录其在室内的大量路径信息,利用嵌套在路径中的低维流形一致性进行地理位置匹配,以建立位置指纹库。通过高斯粒子滤波器对传感器数据进行去噪,进而解决步长差异问题。定位时,根据用户位置的连续性和路径信息筛选出合理的近邻点,继而实现精确定位。在84m²的会议室进行实验,在不需要现场勘测的情况下,所提方法可达到与传统方法可比的定位精度。该方法可以实时适应环境变化,在2周甚至1个月之后,其定位准确性优于传统定位方法。     

基于Revit+Dynamo模式的异形人行斜拉桥BIM参数化建模技术

随着经济快速发展,我国对桥梁景观需求日益迫切,因此空间结构复杂、曲线优美的桥梁应运而生。对于造型独特的桥梁,传统的BIM建模手段冗杂且建模效率低下,因此造型独特的景观桥梁设计建模过程一直是亟待解决的问题。通过将Revit软件与可视化编程插件Dynamo结合,针对某异形人行斜拉桥各构件的特点,创建了异形人行斜拉桥参数化族库,开发了异形人行斜拉桥相关的快速建模技术。该方式通过获取桥梁中心线的三维坐标点,编制桥梁构件参数化Excel表格,编制Dynamo程序导入各构件的参数和位置信息,实现空间异形桥梁各构件的快速布置及参数赋值,为复杂异形人行斜拉桥的建模方式提供了新思路。     

结合语义和多层特征融合的行人检测

遮挡及背景中相似物干扰是行人检测准确率较低的主要原因. 针对该问题, 提出一种结合语义和多层特征融合(Combining semantics with multi-level feature fusion, CSMFF)的行人检测算法. 首先, 融合多个卷积层特征, 并在融合层上添加语义分割, 得到的语义特征与相应的卷积层连接作为行人位置的先验信息, 增强行人和背景的辨别性. 然后, 在初步回归的基础上构建行人二次检测模块(Pedestrian secondary detection module, PSDM), 进一步排除误检物体. 实验结果表明, 所提算法在数据集Caltech和CityPersons上漏检率(Miss rate, MR)为7.06 %和11.2 %. 该算法对被遮挡的行人具有强鲁棒性, 同时可方便地嵌入到其他检测框架.     

村落空间形态与步行运动——以婺源汪口村为例

该文应用空间句法理论和方法,通过实地观测汪口街道的步行人流密度,统计分析了村落空间形态的诸多变量与步行密度之间的系统关联、研究的目的是探讨村落的形态结构本身会多大程度地影响步行人流。初步的分析结果表明居民的步行运动模式主要受空间系统的整体组织关系特别是最短路径的选择因素的影响,而商业对人流的吸引作用对这种关系的影响有限。     

Development and validation of a pedestrian deformable finite element model

Pedestrian protection has become an increasingly important consideration in vehicle crash safety. Pedestrian-vehicle crashes cause a significant number of pedestrian fatalities and injuries globally. Computer models are powerful tools for understanding how to reduce the severity of injuries in such crashes. Real-world studies of pedestrians provide an important source of information for evaluating pedestrian model dynamic performance and ability to reconstruct injury-causing events. This study describes the validation process of deformable pedestrian model using published postmortem human subject (PMHS) trajectory and head resultant velocity corridors, and demonstrates its applicability to pedestrian — vehicle impact research. We implemented the deformable pedestrian model using LS-DYNA finite element code. Based on PMHS data, the pedestrian model is used to validate the displacement trajectories of the head, pelvis, knee and foot. The finite element pedestrian model thus obtained can help assess the friendliness of vehicles with pedestrians in traffic crashes and assist in the future development of pedestrian safety technologies. This paper was recommended for publication in revised form by Associate Editor Young Eun Kim Tso-Liang Teng is a Professo r in the Department of mechanical and automation engi-neer ing and the Dean of engineeri ng college at the Da-Yeh Uni versity, Taiwan. He received a BS (1981), MS (1986) and PhD (1994) from the Chung Cheng Institute of Technology. His research intere sts include design of passive safety systems in ve hicles, crash tests simulation, passenger and pedest rian injuries analysis, design of pedestrian protecti on systems.     

Modeling crowd evacuation of a building based on seven methodological approaches

Crowd evacuation of a building has been studied over the last decades. In this paper, seven methodological approaches for crowd evacuation have been identified. These approaches include cellular automata models, lattice gas models, social force models, fluid-dynamic models, agent-based models, game theoretic models, and approaches based on experiments with animals. According to available literatures, we discuss the advantages and disadvantages of these approaches, and conclude that a variety of different kinds of approaches should be combined to study crowd evacuation. Psychological and physiological elements affecting individual and collective behaviors should be also incorporated into the evacuation models.