本研究藉由新竹市2011年至2016年道路交通事故資料,研究交通事故與道路環境因子之關係,透過類神經網路建立事故危險程度模式,結果顯示分向設施、號誌動作及號誌種類為影響事故危險程度重要性較高之因素:分向設施中,最危險之分向設施型式為單向禁止超車線,光復路二段為最易發生之路段;號誌動作中,最危險之號誌動作狀態為不正常,民富街與林森路為最易發生之路段;號誌種類中,最危險之號誌種類為行車管制號誌(附行人專用),光復路二段為最易發生之路段。 本研究結果希望提供政府機關作為工程道路環境設計參考,更希望降低本市各路段因道路環境因子所致之肇事案件發生,以維用路人之生命財產安全。
This study uses road traffic accident data from 2011 to 2016 in Hsinchu City to study the relationship between traffic accidents and road environmental factors. Establishing accident risk patterns through artificial neural network. The results showed that the road separation facilities, signing movements, and signage types are important factors that affect the level of accident risk. In the road separation facilities, the most dangerous type of separation facility is One-way prohibited overtaking line type, and the section most likely to happen is Section 2 of Guangfu Road. In the signing movements, the most dangerous state of singing movements is abnormal, and the section most likely to happen is Minfu Street and Linsen Road. In the signage types, the most dangerous type of sign is traffic control signs (with pedestrians only), and the section most likely to happen is Section 2 of Guangfu Road. The results of this study hope to provide government agencies with reference to the design of the road environment of the project, and hope to reduce the occurrence of accidents caused by road environment factors in each section of Hsinchu City in order to maintain the safety of life and property.