過去號誌化道路旅行時間之推估方法係將路段旅行時間與路口延滯時間予以加總而行,此一做法並無立論依據與文獻證明其合理性,因兩者獨立之模式相加可能會產生無法處理之介面。 透過定義並估計期望路口影響長度,本研究將號誌化道路區分為「路段部分」與「路口部份」兩種車流型態,並以流量為決策變數進行旅行時間模式之參數校估。 實證分析結果顯示,本研究模式在下游路口號誌之G/C大於0.7時,其MAPE績效評比優於Oh + Webster模式,且MAPE皆低於百分之二十;而G/C小於 0.7之績效則以Oh + Webster較佳。在實際應用上僅需偵測器之流量與速率做為輸入項目,資料取得與計算過程較簡單,亦可降低時間與成本。
In the past, some studies found that the travel time of signalized road could be solved by summing up link travel times and delay times of related intersections. The rationality of this concept has not been proved. In this paper, we defined the Expectation of Intersection Influenced Length for demarcated two different states of traffic flow on signalized road. The travel time estimation model comprised “link part” and “intersection part”. The flow rate has been treated as an independent variable in regression analysis. The empirical study is shown that the value of MAPE of the established model is lower than 20% when the downstream signal G/C is greater than 0.7. But the value of MAPE is somewhat higher than Oh + Webster model when the downstream signal G/C is lower than 0.7. The research only applies flow rate and speed data from vehicle detectors. Compare with Oh + Webster model, the merits of this model are simple computation and less data collection.