Please use this identifier to cite or link to this item:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/74886
Title: | 透過自行車與行人之衝突分析探索號誌路口之自行車車流型態 Exploring Cyclist Flow Patterns at Signalized Crossing: Perspective of Cyclist-pedestrian Conflict Analysis |
Authors: | Ming-Yi Tseng 曾明儀 |
Advisor: | 許聿廷(Yu-Ting Hsu) |
Keyword: | 自行車與行人之衝突,自行車專用道,行人行為模式,自行車行為模式,號誌路口,泊松回歸模型, Cyclist-pedestrian Conflict,Dedicated Bike Lane,Pedestrian Behavior,Cyclist Behavior,Signalized Crossing,Poisson Regression Model, |
Publication Year : | 2019 |
Degree: | 碩士 |
Abstract: | 近年來,自行車的日益普及引發針對自行車騎士道路使用權的爭論。考量永續運輸發展及安全因素,許多城市發展出自行車專用道,將自行車騎士與其他交通運具分離,但自行車專用道的效率與使用率常受到質疑。本研究旨在探索有關自行車專用道的存在對於自行車與行人間之衝突與自行車之行為模式,為當前的設計手冊提供經驗和定量的分析。考慮到數據收集便利性,本研究選擇以號誌路口發生的自行車與行人間之衝突為調查主軸,因號誌路口具高密度人流與自行車流,可預期發生衝突的頻率更高。影像攝於國立臺灣大學校總區周圍之號誌路口,本研究提取影像中所選週期之每個自行車騎士與行人的移動路徑,用以分析穿越行為模式。本研究將衝突分為八類,並觀察到283起衝突事件。在所有衝突中,以對向自行車與行人之對撞衝突和對向自行車與行人之側撞衝突最多。研究使用泊松回歸模型,結果顯示衝突傾向於發生在綠燈周期的前20秒,且衝突數量會隨著行人流量的增加而增加。本研究認為拓寬穿越道、號誌分離、劃設行人等待區與自行車等待區、重新劃設機車待轉區位置有助於降低自行車與行人間之衝突,增進穿越效率。 The growing popularity of cycling induces the argument over the right of way for cyclists. Many cities have deployed dedicated bike lanes to separate cyclists from other traffic flows in light of sustainable transportation development and safety concerns. However, the effectiveness of dedicated bike lanes is questioned in terms of their utilization rates. This research seeks to explore the behavioral patterns related to cyclist-pedestrian conflicts against the presence of dedicated bike lanes, so as to provide both empirical and quantitative insights for the current design guidelines. Upon the consideration of data collection, this research chooses to focus on the cyclist-pedestrian conflicts occurring at signalized crossings, where higher frequencies of conflicts can be expected. Video streams were filmed over the signalized crossings around a university campus in Taipei, Taiwan. Movements of individual cyclist and pedestrian were abstracted to identify the aggregate behavioral pattern of crossing streets with the presence of dedicated bike lanes. Eight conflict types have been classified, and 283 conflict events were observed. Among all the conflicts, cyclist-pedestrian head-on conflict in opposite directions and cyclist-pedestrian sideswipe conflict in opposite directions are the most. The Poisson regression model was employed, it showed that conflicts tend to appear in the first 20s of the green interval, and they increase as the pedestrian flow increase. This research suggest that widening the crossings, using pedestrian scramble, marking separated waiting areas, and rearranging the position of the two-stage left-turn area will help reduce the conflict between cyclists and pedestrians and improve the efficiency of crossing. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/74886 |
DOI: | 10.6342/NTU201904206 |
Fulltext Rights: | 有償授權 |
Appears in Collections: | 土木工程學系 |
Files in This Item:
File | Size | Format | |
---|---|---|---|
ntu-108-1.pdf Restricted Access | 23.03 MB | Adobe PDF |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.