本研究以國內某客運公司於台北市及新北市聯營公車共計20條路線所屬之125輛使用Jasslin JAS301型數位式行車紀錄器低地板公車以及151位駕駛員為研究對象，依據存活分析法建構Cox等比例危險模式並經校估後發現，「平均站距」、「公里營收」、「連續駕駛時間」及「平均時速」4項變數對Cox模式具有顯著性的，表示以上變數可以有效解釋發生異常之存活時間。 ;The cause of the traffic accidents are drivers, vehicles, roads and environment. The related researches show that main factor for accidents is the driver and mainly caused by improper driving operation and violations.
This study uses data of domestic bus company’s 125 buses and 151 drivers which collected by digital tachograph for the analysis. After calibrating the Cox proportional hazard model originated from the survival analysis, we found that a total of 4 factors: i.e. "average station distance", "revenue per kilometer", "Continuous driving time" and "average speed", which possess prominence demonstration to Cox model and also could be explain effectively to the interval period of abnormal incident.
