TSIS模擬軟體參數校估方法之研究-以高快速道路為例

Abstract

交通管理策略實際應用於現況中風險過大，故通常以模擬的方式先行進行績效評估，然而於模擬分析前，須先針對車流模擬軟體進行參數校估以反映現況，本研究欲以TSIS為例建立一套車流模擬軟體參數校估的方法與流程，提供未來相關研究人員參考及使用。本研究以國道五號車系統為例，其原因為國道五號車輛偵測器布設甚為密集，車輛偵測器資料妥善率高，故歷史資料庫建置完整。由於國道五號車輛偵測器布設甚為密集，本研究首先以簡單的檢定方式，篩選出較穩定的車輛偵測器，並利用Line Search Method、粒子群演算法等方法結合模擬軟體進行參數校估之作業。

Traffic simulator has been widely used in the fields of transportation planning and traffic management. However, it must calibrate parameters of traffic model to reflect real traffic situation before using traffic simulator. The purpose of this study is to establish a calibration framework for traffic managers and researchers using in the feature. This study takes Freeway No.5 in Taiwan as an example. The reason is that the historical traffic database is well constructed. This study proposes a simple method to filter worse vehicle detectors, and then use Line Search Method and Particle Swarm Optimization method to calibrate parameters of traffic mode in TSIS.