Adaptive Time-Dependent Traffic Signal Control Scheme with Variable Cycle Length Based on Signaling data

NCUIR > Engineering College > Graduate Institute of Civil Engineering > Electronic Thesis & Dissertation > Item 987654321/89044

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/89044

Title:	Adaptive Time-Dependent Traffic Signal Control Scheme with Variable Cycle Length Based on Signaling data
Authors:	顏傑;Yen, Chieh
Contributors:	土木工程學系
Keywords:	依時性交通量指派;延伸性交通量指派;雙層規劃模型;號誌最佳化;手機信令;time-dependent traffic assignment;extended traffic assignment;bi-level programming model;signal optimization;mobile phone signal data
Date:	2022-09-07
Issue Date:	2022-10-04 10:49:13 (UTC+8)
Publisher:	國立中央大學
Abstract:	在傳統的交通量指派當中往往缺乏考慮依時性以及號誌的影響，且在進行交通量指派前所需進行的交通量調查過程中需要花費相當多時間進行調查。為了更加符合實際的交通環境並提升交通量調查之效率，在本研究中不只將導入雙層規劃模型並納入依時性交通量指派以及號誌對道路的影響，更會將手機信令資料進行導入做為一個更加便捷的資料蒐集方法。在旅運需求預測模型中，不同的模組間彼此將會存在有介面，為了消除介面所造成的誤差在本研究中將會採用超級路網的概念將旅次分布以及交通量指派進行整合並視為延伸性交通量指派之問題。本研究率先採用手機信令資料進行交通量指派之研究。並進行比較其中包括傳統交通分區以及信令資料所採用之網格資料之差異，由於本研究並未執行長時間之預測，且設立傳統交通分區需要進行許多的調查作業，故以網格作為交通分區即足以滿足本研究之需求且能夠提升建構之效益並藉此使交通量指派導入大數據之領域。此外為檢視測試之合理性，本研究將以台北內湖之路網作為研究之場域，並透過指派後之成果與實際道路感測器之資料進行比對。在測試的結果中可以發現透過號誌控制的調整將可以使總旅行時間下降，並能符合交通量指派之最佳化條件。本研究亦透過Webster formula 與Hooke and Jeeves method去進行號誌週期之調整。測試之結果顯示使用Hooke and Jeeves method 將可以獲得較Webster formula 更佳之結果，但須耗費大量的時間進行計算。最後在道路感測器流量與指派結果之比較分析中，本研究將其分成三種類型，並可分別利用依時性旅次分布與指派問題之特性進行解釋。而其中有部分路段在道路感測器流量與指派結果有極高的相似程度，顯示本研究之模型具有一定之參考價值。 ;In the traditional traffic assignment process, the time dependency and influences of signals are neglected; thus, it is necessary to conduct a time-consuming traffic survey to obtain accurate traffic data. To represent real-world traffic situations and improve the survey efficiency, we construct a bi-level programming model (in which time-dependent traffic assignment can be implemented under optimal signal conditions) and introduce mobile phone signal data to enhance traffic data collection. In addition to constructing a time-dependent traffic assignment model, we combine trip distribution and traffic assignment and use a supernetwork representation to view this problem as an extended traffic assignment toward the reduction of the interface between travel demand forecasting models. As a pioneer study in introducing mobile phone signal data as input data in traffic assignment, this study also distinguishes and clarifies the concept of traffic analysis zones, which is used in traditional traffic assignment with grid which is constituted the storage format of mobile phone signal data. Our model does not consider the influence of long-term demand prediction; thus, it is sufficient for us to use grids to conduct traffic assignment. To demonstrate that our idea can be implemented in real-world scenarios, we select Neihu, Taipei, Taiwan, as our testing field and use actual demands obtained from signal data to facilitate comparison with the existing vehicle detector data. The test results show that by adjusting the signal, the total travel time can be reduced, and the optimal condition can be matched. For improved realism, we also use Webster’s formula and the Hooke and Jeeves method to adjust each cycle length. The results show that the use of the Hooke and Jeeves method improves the result more than Webster’s formula, but renders the calculation process time consuming. By comparing the link flows between the vehicle detector data and our assignment results, we find that it is able to sort these flows into three patterns, owing to the principles of the TD-TDTA model. The results indicate that our assignment resembles that obtained from the vehicle detector data.
Appears in Collections:	[Graduate Institute of Civil Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	82	View/Open

社群 sharing

Loading...