【摘要】：旅行時間是智能交通系統(tǒng)的重要基礎性數據，交通流的動態(tài)、隨機性及偶發(fā)事件等影響了車輛在高速公路網絡中的旅行時間，，而旅行時間的可靠性會影響到出行者對出發(fā)時間、路徑和出行方式的選擇，因此旅行時間可靠性研究日益受到重視。旅行時間可靠性研究必須有長期的數據作為支撐，目前國外針對旅行時間的研究主要基于車輛檢測器數據，但是全面安裝檢測器需要增加大量的投資且數據穩(wěn)定性不高，很難收集到全面、完整的數據；我國的高速公路采用封閉式收費管理方式，收費管理系統(tǒng)記錄了豐富的信息，這為高速公路的旅行時間可靠性研究提供了可能。因此，基于收費數據的高速公路旅行時間可靠性研究具有現(xiàn)實意義。然而，目前基于數據的研究還存在若干關鍵問題有待解決，主要包括：收費數據存在噪聲干擾；收費數據只記錄出入口信息，無法全程跟蹤車輛的行駛過程。針對這些問題，本文完成了以下工作：（1）針對異常數據干擾問題，基于高速路限速和正態(tài)分布概率篩選性質，提出了一種兩步數據預處理方法。在數據預處理的基礎上，在一個較長時間跨度內，分別對一周之中每一天不同時間區(qū)間出發(fā)車輛的平均旅行時間進行分布擬合，證明其服從對數正態(tài)分布。（2）提出了一種依據軌跡法計算車輛時空坐標軌跡，以路段長度為權重進行路段速度修正的路段旅行時間估計算法。該算法較直接使用收費數據進行路段旅行時間估計的精度提高了30%以上，直接提高了路段斷面流量估算的準確性。（3）選定緩沖指數作為本文中旅行時間可靠性測度指標。利用回歸分析的方法，以緩沖指數為因變量，重車比、V/C和路段長度為自變量，建立不同車型的旅行時間可靠性經驗方程，并分析各個影響因素的影響程度。（4）基于本文研究成果，提出了基于收費數據的路段斷面流量估算方法的應用場景，并與其他幾種基于車輛檢測器的流量檢測方法進行了全面比較，論證其在成本效益上較其他方法的優(yōu)勢。此外，介紹了旅行時間可靠性在先進交通管理系統(tǒng)中的應用。 此外，本文中以豐富的算例分析，驗證了以上所提方法的可行性和準確性，算例分析中獲得的眾多有益結論不僅充實了研究成果，更為研究成果投入實際應用提供了思路和借鑒。研究成果的應用將為提高高速公路服務水平、降低運輸成本做出重要的貢獻。
[Abstract]:Travel time is an important basic data of intelligent transportation system. The dynamics of traffic flow, randomness and random events affect the travel time of vehicles in highway network, and the reliability of travel time will affect the departure time of travelers. Because of the choice of route and travel mode, more and more attention has been paid to the study of travel time reliability. The research of travel time reliability must be supported by long-term data. At present, the research of travel time abroad is mainly based on vehicle detector data, but the complete installation of detectors needs to increase a lot of investment and the data stability is not high. It is difficult to collect comprehensive and complete data; In our country, the closed toll management mode is adopted, and the toll management system records abundant information, which makes it possible to study the reliability of highway travel time. Therefore, the study of highway travel time reliability based on toll data has practical significance. However, at present, there are still some key problems to be solved in the data-based research, such as: there is noise interference in the charging data; the charging data only records the information of the entrance and exit, and can not track the driving process of the vehicle in the whole course. To solve these problems, this paper has completed the following work: (1) aiming at the abnormal data interference problem, a two-step data preprocessing method is proposed based on the property of high speed limit and normal distribution probability screening. On the basis of data pre-processing, the average travel time of vehicles starting from different time intervals of each day in a week is fitted in a long time span. It is proved that it obeys log-normal distribution. (2) A road travel time estimation algorithm based on trajectory method to calculate the vehicle space-time coordinate trajectory and to modify the road section speed based on the length of the road section is proposed. This algorithm improves the accuracy of road section travel time estimation by more than 30% compared with the direct use of toll data. (3) the buffer index is selected as the measure index of travel time reliability in this paper. Taking buffer index as dependent variable, heavy-to-vehicle ratio, V / C and length of road section as independent variables, the empirical equations of travel time reliability of different models are established by using regression analysis method. And analyze the influence degree of each influence factor. (4) based on the research results of this paper, put forward the application scenario of section flow estimation method based on toll data. Compared with other flow detection methods based on vehicle detector, the cost-effectiveness of this method is proved to be better than that of other methods. In addition, the application of travel time reliability in advanced traffic management system is introduced. In addition, the feasibility and accuracy of the above-mentioned method are verified by a wealth of examples in this paper, and many useful conclusions obtained in the case analysis not only enrich the research results, but also prove the feasibility and accuracy of the proposed method. More research results are put into practical application to provide ideas and reference. The application of the research results will make an important contribution to the improvement of highway service level and the reduction of transportation cost.
【學位授予單位】：華南理工大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：U495

【參考文獻】

相關期刊論文 前9條

1 高愛霞;陳艷艷;;菜單基于神經網絡的快速路交通流運行狀態(tài)可靠性研究[J];北京工業(yè)大學學報;2010年03期

2 楊曉光,張汝華,儲浩,伍速峰;基于高速公路收費系統(tǒng)的交通信息采集與處理基本問題研究[J];系統(tǒng)工程;2004年11期

3 薛文;郭艷梅;;天津市高速公路電子收費交通流量數據分析[J];公路交通科技(應用技術版);2011年12期

4 王海燕,項喬君,陸鍵,高存厚,梁辰光;高速公路出入口OD矩陣推算方法[J];公路交通科技;2005年02期

5 沈強;;基于高速公路收費數據的路網運行狀態(tài)評價[J];公路交通科技;2012年08期

6 于德新;高學英;楊兆升;;基于GPS數據及車輛運行特性分析的單車路段行程時間估計[J];吉林大學學報(工學版);2010年04期

7 蒲云;劉海旭;;基于部分隨機用戶平衡的可靠性網絡設計[J];西南交通大學學報;2011年02期

8 李先;溫慧敏;高永;陳琨;;北京市路網單位距離行程時間可靠性評價[J];交通運輸系統(tǒng)工程與信息;2007年02期

9 許良;高自友;;基于路段能力可靠性的城市交通網絡設計[J];中國公路學報;2006年02期

本文編號：2457630