| 摘要: | 隨著臺灣經濟蓬勃發展，交通科技之發達，汽機車數量急速成長且道路交通狀況日益複雜，道路交通安全更彰顯其重要性。另外因公路隧道為一封閉空間，對於交通事故的偵知與救援，明顯較一般開放路段更為困難。一旦發生事故時，往往會造成隧道內外連絡困難、救援人車可及性不易及狀況難以掌握等問題。尤其我國高快速公路網有幾個長隧道完工通車，未來隧道內之事故救援勢必成為公路管理之一項重要的挑戰。如何提高高速公路長隧道內駕駛安全及防止意外事故發生，進而開發能預防事故發生之警示系統，是近年來的主要研究議題。 本研究彙整文獻回顧結果，律訂前方車輛防撞警示系統(FVCWS)對駕駛行為影響之相關駕駛作業項目，包含了警示型態(無、語音、影像、震動)、系統警示時間(1.5 秒、2 秒、2.5 秒)、道路環境(晴天、霧天、直線隧道、彎曲隧道)、分心任務(無分心、撥打手機)、事件型態(前車急煞、旁車強切入急煞)等五種因子。希冀能夠利用駕駛模擬儀，設計合適的高速公路和長隧道場景與駕駛任務劇情， 進行主觀評量問卷以及14 種駕駛績效之量測與分析，以找出最適合我國交通地理及駕駛特性的FVCWS，並提出建議及改善之道。 在NASA-TLX 主觀評量結果顯示，道路型態和分心任務皆對受測者的工作負荷皆有所影響，而在警示系統的喜好接受度則以語音及震動功能為主。同時，實驗結束後，有八成以上的受測者認為加裝FVCWS 有助於行車安全。在四種實驗因子對於14 種駕駛績效的影響探討上，可明確得知分心任務、系統警示時間、環境型態、事件型態有著不同程度的影響。其中，在碰撞時間平均值、兩車相對距離標準差以及車道偏移量標準差等三種均達到顯著差異。未來在發展防撞警示系統的效能評估，可考慮此三種駕駛績效進行評比。 另外，在環境型態的影響比較上，可明顯得知因霧天所導致的視線不良，駕駛人會加大與前車的安全距離。而在彎曲隧道及直線隧道下的車道偏移量標準差的最大，顯示駕駛人在隧道的車輛漂移現象較為嚴重。在系統警示時間的影響比較上，可發現以2.5 秒的設定時間最大。這是因為當系統警示時間的設定愈大，駕駛人對於前車急煞或旁車強切入急煞的突發事件的時間壓力愈小，有較長的時間來採取與前車碰撞的危險規避。在四種警示系統對於駕駛績效的影響探討上，可明確得知有警示系統時比無警示系統要來的好。同時，在各種警示系統的功能性比較，可得知以震動防撞系統對於駕駛績效的提升效果最好。 綜觀以上，本研究針對FVCWS 在警示型態、系統警示時間、道路環境、分心任務、事件型態等不同因子，探討其對於駕駛績效的影響，相信本研究結果可提供未來在最適合國人FVCWS 開發的設計應用參考。 Following the development of economy, the number of motor vehicles has increased rapidly and traffic conditions have become more complex than ever. Traffic accidents also greatly increased year by year. Particularly on the tunnel, the obstruction of external circumstance enhances the rescue of accidents will more difficult than opened spaces. These problems are becoming even more important and challenging. How to improve driving safety on highway and long tunnels has become an interesting issue recently. The objective of this study is attempt to identify the best Forward Vehicle Collision Warning Systems (FVCWS) that most suitable for the driving characteristics and geographical features of our country. This research utilized a fixed-base driving simulator (DS) to execute the simulated driving task. Fourteen driving performance and two subjective questionaries were adopted to determined the effect of different FVCWS, including three warning functions (voice, image, vibration), and three system alert time (1.5 s, 2 s, 2.5 s). Furthermore, three on-road conditions were considered to test the efficiency of the most suitable FVCWS, including four road types (sunshine, fog, rectilineal tunnel, curvaceous tunnel), two distraction task (no distraction, use phone), and accident types (emergency brake, sidecar plug-in). Analytic results of NASA-TLX subjective questionary indicated that road types and distraction tasks greatly increase the workload while driving. The subjects reported that the voice function and vibration function of FVCWS were the most acceptable systems. After the experiment, over 80% of subjects have the willing to set up the FVCWS on their cars. Moreover, the ANOVA results of fourteen driving performance clearly showed that the time to collision (TTC), standard deviation of headway, and standard deviation of lateral position (SDLP) all have a significant difference in the distraction task, the system alerts time, the road type, and accident type. Furthermore, analytic results of independent t-test indicated that the road type of fog has the highest headway due to the poor visibility. The curvaceous tunnel and rectilineal tunnel have the highest SDLP that seriously increase the probably of vehicle drift. The 2.5 seconds is the best system alert time of FVCWS due to it have the smallest time pressure to react the emergency accidents. Meanwhile, experimental results clearly demonstrated that the driving performance of assembled FVCWS are superior to the without FVCWS and the vibration function of FVCWS has the best performance on three kinds of warning functions. We believe that the findings of this research would useful to the optimal design of FVCWS in our country. |
