本研究主要係利用實際車流之觀測與分析,建構一動態防撞門檻,之後利用JAVA 2程式語言,以行為門檻模式建構一車流模擬系統,設計不同車流情境模式,對防撞門檻進行績效評估。 實驗探討之範圍為路段上之跟車行為,但不包括變換車道之變數,欲探討之車流情境有五種,分別為:車流量差異、駕駛反應時間差異、駕駛對警告信賴程度差異、駕駛期望速度差異及車輛位於車隊位置不同差異。並分別探討各情境中,有無防撞系統之影響及進行績效評估。系統績效評估主要可分為:行車安全、行車平穩度、行車效率及跟車穩定程度四個項目分析。 對每個車流環境狀況進行30次電腦模擬後,由其模擬結果可知,當加裝防撞系統後,車輛之行車安全及跟車平穩度績效均較無防撞系統之車輛有顯著之進步。而行車平穩程度及行車效率怎無明顯差異。 而當反應時間較長、越靠近緊急剎車之車輛和期望速度越高之情況,其有無防撞系統之績效差異越大。
Driving behavior in every country is different from the other countries. The car following with dangerous headways will be recognized as different dangerous levels by different drivers. The threshold for warning driver to take action in a Collision Avoidance System (CAS) developed in developed countries will not be suitable for the other developing countries. The microscopic survey of dangerous car following headway and basing on it to identify the threshold of safety headway is necessary. In general, due to the high dense traffic condition in Taiwan, the driving behavior is seen as more aggressive. Therefore, this study conducted a field survey microscopically to analyze the car following headway and using the braking action to identify the possible reaction to the dangerous headway. Through the statistical analysis, the scenario for safety headway are generated and input to a simulation system developed for simulating the dangerous traffic risk in this study, and then the effect of the threshold on CAS system is investigated. According the study result, the threshold will depend on the speed and has stochastic feature. The over reaction due to the inadequate warning threshold will cause new dangerous situation on the traffic surrounding the observed vehicle. This result indicates the need to revise the decision threshold value in CAS system when it will be adopted in different countries.