基于浮动车的城市道路交通异常事件检测的研究
|
摘要
随着ITS(Intelligence Transportation System,智能交通系统)的进步和发展,人们逐渐认识到交通对于日常生活以及城市建设和发展的重要性,越来越多研究者将先进的技术和手段融合到交通管理和控制领域中,希望可以借此优化和改善城市道路交通状况,交通异常事件检测就是其中的重要内容。目前,国内外存在的交通异常事件检测的模型和算法很多,体系庞大,各种算法针对不同的检测器和道路环境具有其不同的优缺点,适用范围也不同。另一方面,随着浮动车技术的发展,国外越来越多的研究者看到了它在交通信息采集方面的潜力,开始开发和研究其应用价值,并将它应用到交通异常事件检测中。这使得城市道路交通异常事件检测有了新的方向。但在国内对于浮动车技术的开发和应用却并不多,主要集中在北京、杭州等大城市中,有待进一步的发展。因此,将浮动车技术融合到交通异常事件检测中,将使城市道路的交通异常事件检测有更长足的发展。
基于上述内容,本论文的研究目的是利用浮动车采集的交通信息数据,通过有效的交通异常事件检测算法,及时发现道路上的异常事件,使有关部门获得有效信息,并做出快速处理,令交通流得到及时疏导,尽快恢复道路交通的畅通。
本论文完成的主要工作包括以下五个部分:
(1)本论文在大量的文献阅读和算法研究的基础上,提出了一种交通异常事件检测的综合模型,其主要由车速预测模型和预测偏差分析模型组成。
(2)阐述车速预测模型——改进型ARIMA模型的基本框架,并推导了模型中相关的计算公式。
(3)介绍预测偏差分析模型的基本原理,并分析了模型关键参数的确定方法。
(4)对车速预测模型和预测偏差分析模型的关键计算步骤进行了程序编写,并利用程序对数据进行计算和分析。
(5)利用浮动车采集的实际数据,结合Flowsim交通仿真的数据,对综合模型进行了分析和验证,并对模型有效性做出了评价。
With the progress and development of ITS (Intelligence Transportation System), there is a growing awareness of the importance of traffic, as well as the daily lives of urban construction and development. Advanced technology and tools are integrated into the management and control of traffic by more and more researchers. It is wished that we can take this path to optimize and improve the urban traffic conditions. Traffic abnormal incident detection is one of the important elements in the field. At present, there are many of models and algorithms and is a huge system on traffic incident detection at home and abroad. Each algorithm has its advantages, disadvantages and different scope of application for different detector and the environment. On the other hand, with the development of floating car, more and more foreign researchers apply floating car to detect traffic abnormal incident while they discover its potential in traffic information collection and work on researching the development and application of floating car. However, in the domestic, the development and application for floating car mainly concentrate in Beijing, Hangzhou, and other big cities, and further development is pending. Therefore, there will be a more substantial development on the urban road traffic abnormal incident detection after floating car technology is integrated into traffic abnormal incident detection.
The purpose of this paper is to discover unusual events on the road using data collected by floating cars and effective incident detection algorithm in order to take an effective and rapid treatment to receive ease traffic flow timely and restore free trafficflow.
The completed major work in this paper includes the following five sections:
First of all, on the basis of a great deal of reading and researching literature and algorithm, a abnormal traffic incident detection integrated model that is mainly composed by short-term traffic flow forecasting model and prediction error analysis model is presented.
Second, the basic framework of short-term traffic flow forecasting model, which is the ARIMA model to improve, is elaborated, and calculation formula related to the model is derived.
Third, the basic principle of prediction error analysis model is introduced, and the key parameters of the model are analyzed.
Fourth, the procedures for the key steps of model calculation are prepared and used to calculate and analyze data.
Finally, with the actual data collected by floating vehicles and data obtained through Flowsim simulation, analysis and verification of the integrated model are made, and model validity is evaluated.
基于上述内容,本论文的研究目的是利用浮动车采集的交通信息数据,通过有效的交通异常事件检测算法,及时发现道路上的异常事件,使有关部门获得有效信息,并做出快速处理,令交通流得到及时疏导,尽快恢复道路交通的畅通。
本论文完成的主要工作包括以下五个部分:
(1)本论文在大量的文献阅读和算法研究的基础上,提出了一种交通异常事件检测的综合模型,其主要由车速预测模型和预测偏差分析模型组成。
(2)阐述车速预测模型——改进型ARIMA模型的基本框架,并推导了模型中相关的计算公式。
(3)介绍预测偏差分析模型的基本原理,并分析了模型关键参数的确定方法。
(4)对车速预测模型和预测偏差分析模型的关键计算步骤进行了程序编写,并利用程序对数据进行计算和分析。
(5)利用浮动车采集的实际数据,结合Flowsim交通仿真的数据,对综合模型进行了分析和验证,并对模型有效性做出了评价。
With the progress and development of ITS (Intelligence Transportation System), there is a growing awareness of the importance of traffic, as well as the daily lives of urban construction and development. Advanced technology and tools are integrated into the management and control of traffic by more and more researchers. It is wished that we can take this path to optimize and improve the urban traffic conditions. Traffic abnormal incident detection is one of the important elements in the field. At present, there are many of models and algorithms and is a huge system on traffic incident detection at home and abroad. Each algorithm has its advantages, disadvantages and different scope of application for different detector and the environment. On the other hand, with the development of floating car, more and more foreign researchers apply floating car to detect traffic abnormal incident while they discover its potential in traffic information collection and work on researching the development and application of floating car. However, in the domestic, the development and application for floating car mainly concentrate in Beijing, Hangzhou, and other big cities, and further development is pending. Therefore, there will be a more substantial development on the urban road traffic abnormal incident detection after floating car technology is integrated into traffic abnormal incident detection.
The purpose of this paper is to discover unusual events on the road using data collected by floating cars and effective incident detection algorithm in order to take an effective and rapid treatment to receive ease traffic flow timely and restore free trafficflow.
The completed major work in this paper includes the following five sections:
First of all, on the basis of a great deal of reading and researching literature and algorithm, a abnormal traffic incident detection integrated model that is mainly composed by short-term traffic flow forecasting model and prediction error analysis model is presented.
Second, the basic framework of short-term traffic flow forecasting model, which is the ARIMA model to improve, is elaborated, and calculation formula related to the model is derived.
Third, the basic principle of prediction error analysis model is introduced, and the key parameters of the model are analyzed.
Fourth, the procedures for the key steps of model calculation are prepared and used to calculate and analyze data.
Finally, with the actual data collected by floating vehicles and data obtained through Flowsim simulation, analysis and verification of the integrated model are made, and model validity is evaluated.
引文
[1]邓世雄,赵新才.中国道路交通事故流行趋势与特征.重庆医学.2004.7.第33卷第7期.起止页:1074-1076.
[2]孙棣华,董均宇,廖孝勇.基于GPS探测车的道路交通状态估计技术.计算机应用研究.2007.第二期.起止页:243-245.
[3]薛明,吕卫锋,诸彤宇.浮动车信息处理系统关键技术的研究.微计算机信息(测控自动化).2006.第22卷第11-1期.起止页:244-246.
[4]韩直.交通异常检测系统冲国交通工程.1999.第一期.起止页:7-9.
[5]Karl Frazier Petty.Incidents on the Freeway:Detection and Management.The University of California at Berkeley.[Dissertation]1997.Page(s):9-70.
[6]陈斌,金炜东.四类交通事故检测算法的分析.交通科技与经济,2005.第三期.起止页:50-52.
[7]史新宏,蔡伯根.高速公路自动事件检测算法.交通运输系统工程与信息.2001.11.第1卷第4期.起止页:306-310.
[8]杨耀华,李昕,江芳泽.高速公路事件检测系统及算法.公路交通科技.2003.6.第20卷第3期.起止页:133-136.
[9]谭光莉,姜紫峰.高速公路自动检测算法的探究.西安公路交通大学学报.1999.7.第19卷第3期起止页:55-57.
[10]张敬磊,王晓原.交通事件检测算法研究进展.武汉理工大学学报.2005.第29卷第2期.起止页:215-218.
[11]Shunsuke K,Liu Xiaolu.Semantic Hierarchy Based Reasoning Chain Algorithm for Event Detection on an Intersection.Intelligent Transportation Systems.2006.Page(s):343-348
[12]Zhou,Dingshan Sam.An integrated traffic incident detection model[Dissertation].Dissertation Abstracts International.Vol61-11,SecB.2000.8.Page(s):6025.
[13]王晓原,隽志才等.交通流突变分析的变点统计方法研究.中国公路学报.2002.10.第15卷第4期.起止页:69-72.
[14]覃频频,牙韩高,黄大明.基于Logit模型的城市道路交通事件检测仿真.公路交通科技.2006.12.第23卷第12期.起止页:137-141.
[15]张存保,杨晓光,严新平.基于浮动车的高速公路交通事件自动判别方法的研究.武汉理工大学学报.2006.12.第25卷第2期.起止页:1-5.
[16]向红艳,刘伟,朱顺应.基于偏差分析的交通事件自动检测算法.中南公路工程.2007.4.第32卷第2期起止页:142-144.
[17]Seung-Heon Lee,Jin-Woo Choi,Nam-Kwan Hong,Viswanathan M,Young-Kyu Yang.Development of incident detection model using neuro-fuzzy algorithm.Computer and Information Science.2005.Fourth Annual ACIS International Conference on 2005.Page(s):364-368.
[18]Balraj,Navaneethakrishnan.Automated accident detection in intersections via digital audio signal processing.Masters Abstracts International.Vol42-04.2003.Page(s):1328
[19]宫晓燕,汤淑明.基于非参数回归的短时交通流量预测与事件检测综合算法.中国公路学报.2003.第16卷第1期.起止页:82-86.
[20]Palmer J.P.,Bowers D.J.,Wall G.T..Automatic incident detection and improved traffic control in urban areas.Urban Congestion Management.IEE Colloquium on 16 Nov 1995.Page(s):411- 415.
[21]李彬,刘安.交通异常贝叶斯动态检测方法.中国公路学报.1999.7.第12卷第3期.起止页:68-72.
[22]王小华.基于模糊算法的高速公路事件检测及主线控制算法的研究与实现[学位论文].河南:河南大学.2003.4.起止页:8-21.
[23]翁剑成,荣建,任福田,魏中华.基于非参数回归的快速路形成速度短期预测算法.公路交通科学.2007.3.第24卷第3期.起止页:93-97.
[24]黄洪琼,杨天浩.基于最优梯度自适应优化算法的交通流预测.计算机工程.2007.4.第33卷第7期.起止页:23-25.
[25]杨胜,李莉,胡福乔,施鹏飞.基于决策树的城市短时交通流预测.计算机工程.2005.4.第31卷第8期.起止页:35-36.
[26]况爱武,黄中祥.基于RBF神经网络的短时交通流预测.系统工程.2004.2.第22卷第2期.起止页:63-65.
[27]王进,史其信.短时交通流预测模型综述.中国公共安全学术卷.2005.6.第一卷.起止页:92-98.
[28]周小鹏,冯奇,孙立军.基于最近邻法的短时交通流预测.同济大学学报(自然科学版).2006.11第34卷第11期.起止页:1494-1498.
[29]Brain L.Smith,Billy M.Williams,R.Keith Oswald.Comparison of parametric and nonparametric models for traffic flow forecasting.Transportation Research Part C 10.2002.Page(s):303-321.
[30]韩超,宋苏,王成红.基于ARIMA模型的短时交通流实时自适应预测.系统仿真学报.2004.7.第16卷第7期.起止页:1530-1535.
[31]韩超.时间序列分析的短时交通流量实时自适应预测[学位论文].北京:北京工业大学.2004.起止页:21-34.
[32]Astrom K.J.,Wittenmark B..On Self-turning Regulators.Automatica.Vol.9,No.2,1973.Page(s):185-199.
[33]韩曾晋.自适应控制[M].北京:清华大学出版社.1995.6.起止页:101-133.
[34]向红艳.短期交通流实时预测及交通事件检测算法研究[学位论文].重庆:重庆交通学院。2006.起止页:70-80.
[35]安明颖.基于车流波状态的微观尾气排放问题研究[学位论文].北京:北京交通大学.2007.起止页:12-13.
[36]Wu J,Brackstone M,McDonald M.Fuzzy Sets and Systems for a Motorway Microscopic Simulation Model.Fuzzy sets and systems,n(116).2000.Page(s):65-76.
[37]Jianping Wu,Mark Brackstone,Mike Mcdonald.The Validation of a Microscopic Simulation Model:A Methodological Case Study.Transportation Research Part C:Emerging Technologies.v(11),n(6).2003.Page(s):463-479.
[38]张敬磊.基于数据挖掘的交通流状态是别的算法研究[学位论文].山东:山东理工大学.2006.5.起止页:18-21.
[39]Justin Black,Indu Sreedevi.Automatic Incident Detection Algorithms[M].2001
[2]孙棣华,董均宇,廖孝勇.基于GPS探测车的道路交通状态估计技术.计算机应用研究.2007.第二期.起止页:243-245.
[3]薛明,吕卫锋,诸彤宇.浮动车信息处理系统关键技术的研究.微计算机信息(测控自动化).2006.第22卷第11-1期.起止页:244-246.
[4]韩直.交通异常检测系统冲国交通工程.1999.第一期.起止页:7-9.
[5]Karl Frazier Petty.Incidents on the Freeway:Detection and Management.The University of California at Berkeley.[Dissertation]1997.Page(s):9-70.
[6]陈斌,金炜东.四类交通事故检测算法的分析.交通科技与经济,2005.第三期.起止页:50-52.
[7]史新宏,蔡伯根.高速公路自动事件检测算法.交通运输系统工程与信息.2001.11.第1卷第4期.起止页:306-310.
[8]杨耀华,李昕,江芳泽.高速公路事件检测系统及算法.公路交通科技.2003.6.第20卷第3期.起止页:133-136.
[9]谭光莉,姜紫峰.高速公路自动检测算法的探究.西安公路交通大学学报.1999.7.第19卷第3期起止页:55-57.
[10]张敬磊,王晓原.交通事件检测算法研究进展.武汉理工大学学报.2005.第29卷第2期.起止页:215-218.
[11]Shunsuke K,Liu Xiaolu.Semantic Hierarchy Based Reasoning Chain Algorithm for Event Detection on an Intersection.Intelligent Transportation Systems.2006.Page(s):343-348
[12]Zhou,Dingshan Sam.An integrated traffic incident detection model[Dissertation].Dissertation Abstracts International.Vol61-11,SecB.2000.8.Page(s):6025.
[13]王晓原,隽志才等.交通流突变分析的变点统计方法研究.中国公路学报.2002.10.第15卷第4期.起止页:69-72.
[14]覃频频,牙韩高,黄大明.基于Logit模型的城市道路交通事件检测仿真.公路交通科技.2006.12.第23卷第12期.起止页:137-141.
[15]张存保,杨晓光,严新平.基于浮动车的高速公路交通事件自动判别方法的研究.武汉理工大学学报.2006.12.第25卷第2期.起止页:1-5.
[16]向红艳,刘伟,朱顺应.基于偏差分析的交通事件自动检测算法.中南公路工程.2007.4.第32卷第2期起止页:142-144.
[17]Seung-Heon Lee,Jin-Woo Choi,Nam-Kwan Hong,Viswanathan M,Young-Kyu Yang.Development of incident detection model using neuro-fuzzy algorithm.Computer and Information Science.2005.Fourth Annual ACIS International Conference on 2005.Page(s):364-368.
[18]Balraj,Navaneethakrishnan.Automated accident detection in intersections via digital audio signal processing.Masters Abstracts International.Vol42-04.2003.Page(s):1328
[19]宫晓燕,汤淑明.基于非参数回归的短时交通流量预测与事件检测综合算法.中国公路学报.2003.第16卷第1期.起止页:82-86.
[20]Palmer J.P.,Bowers D.J.,Wall G.T..Automatic incident detection and improved traffic control in urban areas.Urban Congestion Management.IEE Colloquium on 16 Nov 1995.Page(s):411- 415.
[21]李彬,刘安.交通异常贝叶斯动态检测方法.中国公路学报.1999.7.第12卷第3期.起止页:68-72.
[22]王小华.基于模糊算法的高速公路事件检测及主线控制算法的研究与实现[学位论文].河南:河南大学.2003.4.起止页:8-21.
[23]翁剑成,荣建,任福田,魏中华.基于非参数回归的快速路形成速度短期预测算法.公路交通科学.2007.3.第24卷第3期.起止页:93-97.
[24]黄洪琼,杨天浩.基于最优梯度自适应优化算法的交通流预测.计算机工程.2007.4.第33卷第7期.起止页:23-25.
[25]杨胜,李莉,胡福乔,施鹏飞.基于决策树的城市短时交通流预测.计算机工程.2005.4.第31卷第8期.起止页:35-36.
[26]况爱武,黄中祥.基于RBF神经网络的短时交通流预测.系统工程.2004.2.第22卷第2期.起止页:63-65.
[27]王进,史其信.短时交通流预测模型综述.中国公共安全学术卷.2005.6.第一卷.起止页:92-98.
[28]周小鹏,冯奇,孙立军.基于最近邻法的短时交通流预测.同济大学学报(自然科学版).2006.11第34卷第11期.起止页:1494-1498.
[29]Brain L.Smith,Billy M.Williams,R.Keith Oswald.Comparison of parametric and nonparametric models for traffic flow forecasting.Transportation Research Part C 10.2002.Page(s):303-321.
[30]韩超,宋苏,王成红.基于ARIMA模型的短时交通流实时自适应预测.系统仿真学报.2004.7.第16卷第7期.起止页:1530-1535.
[31]韩超.时间序列分析的短时交通流量实时自适应预测[学位论文].北京:北京工业大学.2004.起止页:21-34.
[32]Astrom K.J.,Wittenmark B..On Self-turning Regulators.Automatica.Vol.9,No.2,1973.Page(s):185-199.
[33]韩曾晋.自适应控制[M].北京:清华大学出版社.1995.6.起止页:101-133.
[34]向红艳.短期交通流实时预测及交通事件检测算法研究[学位论文].重庆:重庆交通学院。2006.起止页:70-80.
[35]安明颖.基于车流波状态的微观尾气排放问题研究[学位论文].北京:北京交通大学.2007.起止页:12-13.
[36]Wu J,Brackstone M,McDonald M.Fuzzy Sets and Systems for a Motorway Microscopic Simulation Model.Fuzzy sets and systems,n(116).2000.Page(s):65-76.
[37]Jianping Wu,Mark Brackstone,Mike Mcdonald.The Validation of a Microscopic Simulation Model:A Methodological Case Study.Transportation Research Part C:Emerging Technologies.v(11),n(6).2003.Page(s):463-479.
[38]张敬磊.基于数据挖掘的交通流状态是别的算法研究[学位论文].山东:山东理工大学.2006.5.起止页:18-21.
[39]Justin Black,Indu Sreedevi.Automatic Incident Detection Algorithms[M].2001