標題: 應用情境分析方法研究台灣地區單一小客車交通事故
The Study of Single Automobile Accidents in Taiwan with Scenario Analysis
作者: 蘇宥宜
Yu-Yi Su
汪進財
Dr. Tinn-Tsai Wong
運輸與物流管理學系
關鍵字: 單一小客車交通事故;事故情境分析;順序迴歸分析;single automobile accident;scenario analysis;ordinal regression analysis
公開日期: 2004
摘要: 歷年道路交通事故分析易受資料限制,以致無法深入了解道路交通事故的影響因素,故本研究選以情境方法分析道路交通事故,擴展影響因素的解釋範圍。情境分析方法可以描述道路交通事故發生過程,有助於檢討影響駕駛者發生事故的原因,而得以有效減少道路交通事故。本研究從各種角度如:駕駛、道路與環境等因素來考量行車情境,以期透過描述道路交通事故發生的情境,來發掘道路交通事故發生的原因。為了簡化事故情境的複雜程度,故選擇研究單一小客車的道路交通事故。經由整理2003年警政署道路交通事故資料,獲得一連串事故特性,並配合文獻歸納的肇因鏈,來描述單一小客車的駕駛過程。 研究方法是利用順序迴歸分析與事故情境分析二者,探討單一小客車的駕駛過程。藉由觀察特定事故類型與影響因素間的關係,找尋關鍵影響原因。由順序迴歸分析結果可知影響事故嚴重度的關鍵因素。在衝出路外的單一小客車事故中,關鍵因素為行動電話使用情形、安全帶使用情形、飲酒情形、違規行為、旅次目的、時段、道路能見度和鋪面品質;在撞路中物體的單一小客車事故中,關鍵因素為駕駛執照、駕駛資格、行動電話使用情形、安全帶使用情形、旅次目的、道路視距、道路能見度、鋪面品質和速限;在撞路旁物體的單一小客車事故中,關鍵因素為行動電話使用情形、安全帶使用情形、駕駛執照、駕駛資格、飲酒情形、違規行為、年齡、時段、旅次目的和地區。 由事故情境分析結果可知,影響駕駛者表現的行為與決策和事故嚴重度的風險因子在事故發生前為駕駛本身生理、觀念、認知和情緒;在事故發生當時則為道路的可辨視距離、反應時間與空間、道路安全設計和防護設施。在衝出路外的單一小客車事故中,主要的風險因子為道路安全設計;在撞路中物體的單一小客車事故中,主要的風險因子為駕駛認知;在撞路旁物體的單一小客車事故主要的風險因子為道路能見路。
Most previous studies on traffic accidents were so limited by information relative to traffic accidents that those studies could not help people understand how the factors influence traffic accidents deeply. In order to overcome the difficulty, this study chooses scenario analysis to examine a single automobile accident. Since scenario analysis is useful to describe the matter of traffic accidents, it could also help researchers to evaluate the possible causes leading to the happenings of traffic accidents and figure out strategies to reduce the frequency of accidents. This study takes factors such as the physical status and mentality of the driver, the condition of the road, and the climatic environment into consideration to analyze the situation how drivers drive. To simplify the complexity of the situation of traffic accidents, this study focuses on a single automobile accident. Combining the data collected from Police Policy Administration’s information about traffic accidents in 2003 with the conclusion acquired from the scenario analysis of traffic accidents organized in the literature, this study illustrates the driving process of a single automobile driver. Moreover, this study applies ordinal regression analysis and scenario analysis to systematize a single automobile driver’s driving process. We can discover the key factors by observing the influential degree of every factor. By means of the results of ordinal regression analysis, this study discovers phenomena as follows: in a single automobile which runs off the road, the main factors are such as the use of cell phone and seat belt, drinking alcohol, violative behavior, trip purpose, timing, visibility, and road quality; in a single automobile which hits islands, the key factors are consist of driver’s license and qualification, the use of cell phone and seat belt, trip purpose, visible distance, road visibility, road quality, and limited speed; in a single automobile which hits the roadside objects, the primary factors include the use of cell phone and seat belt, driver’s license and qualification, drinking alcohol, violative behavior, age, timing, trip purpose, and region. The results from scenario analysis show phenomena as follows: before accidents, the latent risk of a driver’s behavior and a serious accident comes from a driver’s physical condition, conception, sensation, and emotion. When accidents happen, the factors are such as visible distance, reactive space and time, road safety design, and protective facilities. In a single automobile which runs off the road, the latent risk lies in road safety design. In a single automobile which hits safety islands, the main risk lies in a driver’s sensation. In a single automobile which hits the roadside objects, the primary risk lies in road visibility.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT009236514
http://hdl.handle.net/11536/77242
Appears in Collections:Thesis


Files in This Item:

  1. 651401.pdf

If it is a zip file, please download the file and unzip it, then open index.html in a browser to view the full text content.