量子視覺流跟車行為模擬模式建構

Abstract

駕駛在進行跟車行為時，往往會受到周遭環境的影響，而過去的文獻大多只針對和前方車輛的相互關係來改變跟車行為。而本文嘗試引用以駕駛者的視覺資訊帶給內心刺激來改變駕駛行為的觀點之文獻，並定義解釋其中心理質量的意義，並以實際資料進行模式參數的校估和驗證。而驗證結果也在一穩定的誤差範圍內。 本模擬模式主要是建立一直線三車道道路，有大型車和小型車兩種車種，再以建構出之模擬模式設立幾種情境進行模擬，模擬的結果可發現，模式可以解釋當駕駛行駛時，鄰近車道車輛的有無造成駕駛的分心，的確會對於整體車流的速度和間距有顯著的影響，又模式可解釋當車輛的車種大小不同時，會對於整體車流的速度造成明顯的影響。

Drivers are often affected by the surrounding environment when driving. But most existing car-following models are assume that the following vehilce is affected by the front vehicle. This study use the car following model based quantum optical flow. Define and explain the psychophysical mass for the drivers. And this study construct the car following simulator. The simulation model are calibrated and test with the video-based car following data. The simulate results are as follows：The models can explain the drivers’ distraction which would affect the speed and spacing of the flowing traffic. And can explain when the car size which would affect the speed of the flowing traffic.