基于投影寻踪的快速路交织区交通状态判别方法
|
摘要
为了准确判别城市快速路交织区的交通状态,实现交通控制策略的优化决策,基于投影寻踪模型与k-means聚类算法,研究了一种新的交通状态判别方法。以交通状态的量化分析为目标,考虑投影寻踪模型的特性,定义了交通状态系数;根据类内聚集度与异类间散度的分析,建立了聚类效果评价系数表达式;应用推导的改进式遗传算法,结合k-means聚类算法,计算获得最优投影方向与聚类中心;应用最优投影方向将新观测的交通流数据转化为交通状态系数,判定欧式距离最小的聚类中心,获得相对应的交通流状态。新方法克服了传统方法对专家经验的依赖性,解决了熵权法对小概率事件信息熵的过量估计问题,并改进了投影寻踪模型的聚类效果评价系数。仿真实验结果表明,新方法状态判别准确率为96.63%,较神经元网络和决策树算法分别提高了5.58%和7.01%,能够准确判别交织区交通流状态。
In order to accurately identify traffic states of urban expressways for optimizing traffic control decision, a new identification method for traffic states is proposed based on projection pursuit model and k-means algorithm. The new method regards quantitative analysis of traffic states as the goal. Considering characteristics of projection pursuit model, coefficients of traffic states are defined. An equation for clustering effect evaluation coefficient is developed based on an analysis of concentration degree and dispersion degree. An improved genetic algorithm and k-means clustering algorithm are applied to calculate the optimal projection direction and cluster center. Traffic data is converted into traffic state coefficient by optimal projection direction, to identify the minimal Euclidean distance, and obtain the corresponding traffic flow states. The new method less relies on expert experiences than traditional methods; and can solve the problem of overestimation on small probability events when using entropy method; and improves clustering effect evaluation coefficient from projection pursuit. Simulation results show that accuracy of identification of the new methodology is 96.63%, which is 5.58% and 7.01% higher than Neural Network and Decision Tree, respectively. The results verify that the new method can accurately identify traffic states of weaving areas.
In order to accurately identify traffic states of urban expressways for optimizing traffic control decision, a new identification method for traffic states is proposed based on projection pursuit model and k-means algorithm. The new method regards quantitative analysis of traffic states as the goal. Considering characteristics of projection pursuit model, coefficients of traffic states are defined. An equation for clustering effect evaluation coefficient is developed based on an analysis of concentration degree and dispersion degree. An improved genetic algorithm and k-means clustering algorithm are applied to calculate the optimal projection direction and cluster center. Traffic data is converted into traffic state coefficient by optimal projection direction, to identify the minimal Euclidean distance, and obtain the corresponding traffic flow states. The new method less relies on expert experiences than traditional methods; and can solve the problem of overestimation on small probability events when using entropy method; and improves clustering effect evaluation coefficient from projection pursuit. Simulation results show that accuracy of identification of the new methodology is 96.63%, which is 5.58% and 7.01% higher than Neural Network and Decision Tree, respectively. The results verify that the new method can accurately identify traffic states of weaving areas.
引文
[1] 李国豪.中国土木建筑百科辞典[M].北京:中国建筑工业出版社,2006.LI Guohao.An encyclopedia of architecture & civil engineering of China[M].Beijing:China Architecture & Building Press,2006.(in Chinese)
[2] 尹胜超,许润民,张毅,等.城市快速路瓶颈区域控制策略研究[J].交通运输系统工程与信息,2012,12(02):27-33.YIN Shengchao,XU Runmin,ZHANG Yi,et al.Signal control strategies for bottleneck area on urban expressway[J].Journal of Transportation Systems Engineering and Information Technology,2012,12(2):27-33.(in Chinese)
[3] 郭义荣,董宝田,吴蕾.基于速度的交通状态识别及动态评价研究[J].公路交通科技,2012,29(S1):26-31.GUO Yirong,DONG Baotian,WU Lei.Traffic state recognition and dynamic evaluation based on velocity Study[J].Journal of Highway and Transportation Research and Development,2012,29(S1):26-31.(in Chinese)
[4] TRB.Highway capacity manual[R].Washington,D.C.:TRB Business Office,2000.
[5] KRAUSE B,von AITROCK C,POZYBILL M.Intelligent highway by fuzzy logic:Congestion detection and traffic control on multi-lane roads with variable road signs[C].New Orleans:Fuzzy Systems,1996(3):1832-1837.
[6] LU J,CAO L.Congestion evaluation from traffic flow information based on fuzzy logic[C].IEEE International Conference on Systems,Shanghai:IEEE,2003.
[7] MONTAZERI-GH M,FOTOUHI A.Traffic condition recognition using the k-means clustering method[J].Scientia Iranica,2011,18(4):930-937.
[8] 陈征,刘钊,王静,等.基于RFID和视频监测的城市交通运行状态模糊判别[J].交通信息与安全,2018,36(2):68-75.CHEN Zheng,LIU Zhao,WANG Jing,et al.A fuzzy method for identifying urban traffic state using RFID and video monitoring system[J].Journal of Transport Information and Safety,2018,36(2):68-75.
[9] 李晓璐,于昕明,杜崇,等.基于权值优化的FCM-MSVM算法及其在高速公路状态判别中的应用[J].北京交通大学学报,2018,42(4):72-78,84.LI Xiaolu,YU Xinming,DU Chong,et al.FCM-MSVM algorithm based on weight optimization and its application in state identification on freeway[J].Journal of Beijing Jiaotong University,2018,42(4):72-78,84.
[10] FRIEDMAN J H,STUETZLE W.Projection pursuit regression[J].Journal of the American statistical Association,1981,76(376):817-823.
[11] CROUX C,RUIZ-GAZEN A.High breakdown estimators for principal components:the projection-pursuit approach revisited[J].Journal of Multivariate Analysis,2005,95(1):206-226.
[12] 金菊良,魏一鸣,丁晶.水质综合评价的投影寻踪模型[J].环境科学学报,2001(4):431-434.JIN Juliang,WEI Yiming,DING Jing.Projection pursuit model for comprehensive evalustion of water quality[J].Acta Scientiae Circumstantiae,2001(4):431-434.(in Chinese)
[13] 邴其春,龚勃文,杨兆升,等.基于投影寻踪动态聚类的快速路交通状态判别[J].西南交通大学学报,2015,50(6):1164-1169.BING Qichun,GONG Bowen,YANG Zhaosheng,et al.Traffic state identification for urban expressway based on projection pursuit dynamic cluster model[J].Journal of Southwest Jiaotong University,2015,50(6):1164-1169.(in Chinese)
[14] DIACONIS P,SALZMAN J.Projection pursuit for discrete data[M].Bethesda:Institute of Mathematical Statistics,2008.
[15] BRADLEY P S,BENNETT K P,DEMIRIZ A.Constrained k-means clustering[J].Microsoft Research,Redmond,2000:1-8.
[16] 姜桂艳.道路交通状态判别技术与应用[M].北京:人民交通出版社,2004.JIANG Guiyan.Technologies and applications of the identification of road traffic conditions[M].Beijing:China Communication Press,2004.(in Chinese)
[17] 张琳琳,李雪玮,李振龙,等.基于两阶段K-means聚类的道路运行状况评价方法[J].交通信息与安全,2017,35(5):99-105.ZHANG Linlin,LI Xuewei,LI Zhenlong,et al.An evaluation method of road operation condition based on two-stage k-means clustering[J].Journal of Transport Information and Safety,2017,35(5):99-105.(in Chinese)
[18] ROESS R P,PRASSAS E S,MCSHANE W R.Traffic engineering[M].Upper Saddle River:Pearson,2010.
[19] MADGWICK S O H,HARRISON A J L,VAIDYANATHAN R.Estimation of IMU and MARG orientation using a gradient descent algorithm[C].Switzerland:Rehabilitation Robotics (ICORR),2011 IEEE International Conference on.IEEE,2011:1-7.
[20] PFROMMER B G,C?Té M,LOUIE S G,et al.Relaxation of crystals with the quasi-Newton method[J].Journal of Computational Physics,1997,131(1):233-240.
[21] 张文修,梁怡.遗传算法的数学基础[M].西安:西安交通大学出版社,2000.ZHANG Wenxiu,LIANG Yi.Mathematical foundation of genetic algorithms[M].Xi′an:Xi′an Jiaotong University Press,2000.(in Chinese)
[22] 段玉倩,贺家李.遗传算法及其改进[J].电力系统及其自动化学报,1998(1):43-56.DUAN Yuqian,HE Jiali.Genetic algorithm and improvement[J].Proceedings of the CSU-EPSA,1998(1):43-56.(in Chinese)
[23] 房丽侠,魏连雨,闫伟阳.基于决策树的交通拥挤状态预测[J].河北工业大学学报,2010,39(2):105-110.FANG Lixia,WEI Lianyu,YAN Weiyang.Prediction of traffic congestion based on decision tree[J].Journal of Hebei University of Technology,2011,29(4):71-74.(in Chinese)
[2] 尹胜超,许润民,张毅,等.城市快速路瓶颈区域控制策略研究[J].交通运输系统工程与信息,2012,12(02):27-33.YIN Shengchao,XU Runmin,ZHANG Yi,et al.Signal control strategies for bottleneck area on urban expressway[J].Journal of Transportation Systems Engineering and Information Technology,2012,12(2):27-33.(in Chinese)
[3] 郭义荣,董宝田,吴蕾.基于速度的交通状态识别及动态评价研究[J].公路交通科技,2012,29(S1):26-31.GUO Yirong,DONG Baotian,WU Lei.Traffic state recognition and dynamic evaluation based on velocity Study[J].Journal of Highway and Transportation Research and Development,2012,29(S1):26-31.(in Chinese)
[4] TRB.Highway capacity manual[R].Washington,D.C.:TRB Business Office,2000.
[5] KRAUSE B,von AITROCK C,POZYBILL M.Intelligent highway by fuzzy logic:Congestion detection and traffic control on multi-lane roads with variable road signs[C].New Orleans:Fuzzy Systems,1996(3):1832-1837.
[6] LU J,CAO L.Congestion evaluation from traffic flow information based on fuzzy logic[C].IEEE International Conference on Systems,Shanghai:IEEE,2003.
[7] MONTAZERI-GH M,FOTOUHI A.Traffic condition recognition using the k-means clustering method[J].Scientia Iranica,2011,18(4):930-937.
[8] 陈征,刘钊,王静,等.基于RFID和视频监测的城市交通运行状态模糊判别[J].交通信息与安全,2018,36(2):68-75.CHEN Zheng,LIU Zhao,WANG Jing,et al.A fuzzy method for identifying urban traffic state using RFID and video monitoring system[J].Journal of Transport Information and Safety,2018,36(2):68-75.
[9] 李晓璐,于昕明,杜崇,等.基于权值优化的FCM-MSVM算法及其在高速公路状态判别中的应用[J].北京交通大学学报,2018,42(4):72-78,84.LI Xiaolu,YU Xinming,DU Chong,et al.FCM-MSVM algorithm based on weight optimization and its application in state identification on freeway[J].Journal of Beijing Jiaotong University,2018,42(4):72-78,84.
[10] FRIEDMAN J H,STUETZLE W.Projection pursuit regression[J].Journal of the American statistical Association,1981,76(376):817-823.
[11] CROUX C,RUIZ-GAZEN A.High breakdown estimators for principal components:the projection-pursuit approach revisited[J].Journal of Multivariate Analysis,2005,95(1):206-226.
[12] 金菊良,魏一鸣,丁晶.水质综合评价的投影寻踪模型[J].环境科学学报,2001(4):431-434.JIN Juliang,WEI Yiming,DING Jing.Projection pursuit model for comprehensive evalustion of water quality[J].Acta Scientiae Circumstantiae,2001(4):431-434.(in Chinese)
[13] 邴其春,龚勃文,杨兆升,等.基于投影寻踪动态聚类的快速路交通状态判别[J].西南交通大学学报,2015,50(6):1164-1169.BING Qichun,GONG Bowen,YANG Zhaosheng,et al.Traffic state identification for urban expressway based on projection pursuit dynamic cluster model[J].Journal of Southwest Jiaotong University,2015,50(6):1164-1169.(in Chinese)
[14] DIACONIS P,SALZMAN J.Projection pursuit for discrete data[M].Bethesda:Institute of Mathematical Statistics,2008.
[15] BRADLEY P S,BENNETT K P,DEMIRIZ A.Constrained k-means clustering[J].Microsoft Research,Redmond,2000:1-8.
[16] 姜桂艳.道路交通状态判别技术与应用[M].北京:人民交通出版社,2004.JIANG Guiyan.Technologies and applications of the identification of road traffic conditions[M].Beijing:China Communication Press,2004.(in Chinese)
[17] 张琳琳,李雪玮,李振龙,等.基于两阶段K-means聚类的道路运行状况评价方法[J].交通信息与安全,2017,35(5):99-105.ZHANG Linlin,LI Xuewei,LI Zhenlong,et al.An evaluation method of road operation condition based on two-stage k-means clustering[J].Journal of Transport Information and Safety,2017,35(5):99-105.(in Chinese)
[18] ROESS R P,PRASSAS E S,MCSHANE W R.Traffic engineering[M].Upper Saddle River:Pearson,2010.
[19] MADGWICK S O H,HARRISON A J L,VAIDYANATHAN R.Estimation of IMU and MARG orientation using a gradient descent algorithm[C].Switzerland:Rehabilitation Robotics (ICORR),2011 IEEE International Conference on.IEEE,2011:1-7.
[20] PFROMMER B G,C?Té M,LOUIE S G,et al.Relaxation of crystals with the quasi-Newton method[J].Journal of Computational Physics,1997,131(1):233-240.
[21] 张文修,梁怡.遗传算法的数学基础[M].西安:西安交通大学出版社,2000.ZHANG Wenxiu,LIANG Yi.Mathematical foundation of genetic algorithms[M].Xi′an:Xi′an Jiaotong University Press,2000.(in Chinese)
[22] 段玉倩,贺家李.遗传算法及其改进[J].电力系统及其自动化学报,1998(1):43-56.DUAN Yuqian,HE Jiali.Genetic algorithm and improvement[J].Proceedings of the CSU-EPSA,1998(1):43-56.(in Chinese)
[23] 房丽侠,魏连雨,闫伟阳.基于决策树的交通拥挤状态预测[J].河北工业大学学报,2010,39(2):105-110.FANG Lixia,WEI Lianyu,YAN Weiyang.Prediction of traffic congestion based on decision tree[J].Journal of Hebei University of Technology,2011,29(4):71-74.(in Chinese)