自動駕駛車輛可透過電腦系統自動安全地控制車輛並行駛於道路上， 未來我們將面臨如何適當地使該車輛通過事件路段中，可供自動駕駛車輛 與手動駕駛車輛混合行駛之車道的挑戰。因此本研究假設在單一車道自動 公路系統發生事件下，設計一套跟車邏輯，判斷其鄰近之混合車道上自動 駕駛車輛針對前方車流之適當反應。該跟車邏輯以避免加減速影響後方車 流的不穩定為目標，其中設計兩個指標，動態距離設定最大值及最小值， 並考量搭乘者的舒適性，利用程式語言，模擬各種跟車情境，再與車流模 擬軟體中之通用汽車跟車指標相比較，以判定自動駕駛車輛跟車穩定性。 本研究歸納出4 種跟車控制邏輯，可依前方兩輛車的相對距離以及速度，判斷各種加速或是減速的可能。此研究成果可作為非正常車流自動駕駛車輛跟車控制邏輯之依據。

While automatic driving vehicles can be controlled automatically by a safety vehicle computer system and driven on a road, we will face the challenge of how to properly make the automatic driving vehicle passing through the lane where automatic and manual driven vehicles mix near the event area and adjacent to automated highway system. The main subject of research topic is to design an automatic car-following logic by the judgment of the front vehicle behavior under the assumption of an incidence occurring on in single-automated-lane highway system. The purpose of automatic car-following logic is to avoid instability in the rear traffic. There are two indicators, maximum dynamic setting distance and minimum one designed in the logic. We also considered the rider comfort and used the programming languages to simulate various scenarios of car-following. Then we compared with the traffic simulation software in the General Motors car-index to determine the stability of automatic driven vehicles under car-following situations. The study concludes four kinds of automatic car-following logic, to judge the acceleration or deceleration of automatic driven vehicle possible by the relative distance and speed of the front two vehicles. The results of this research can support automatic driven vehicle control logic under abnormal traffic.