[1]李根.基于梯度提升决策树的高速公路交织区汇入模型[J].东南大学学报(自然科学版),2018,48(3):563-567.[doi:10.3969/j.issn.1001-0505.2018.03.027]
 Li Gen.Merging model in freeway weaving section based on gradient boosting decision tree[J].Journal of Southeast University (Natural Science Edition),2018,48(3):563-567.[doi:10.3969/j.issn.1001-0505.2018.03.027]
点击复制

基于梯度提升决策树的高速公路交织区汇入模型()
分享到:

《东南大学学报(自然科学版)》[ISSN:1001-0505/CN:32-1178/N]

卷:
48
期数:
2018年第3期
页码:
563-567
栏目:
交通运输工程
出版日期:
2018-05-20

文章信息/Info

Title:
Merging model in freeway weaving section based on gradient boosting decision tree
作者:
李根
东南大学交通学院, 南京 210096
Author(s):
Li Gen
School of Transportation, Southeast University, Nanjing 210096, China
关键词:
公路运输 交织区 汇入行为 梯度提升决策树 超车时间
Keywords:
highway transportation weaving section merging behavior gradient boosting decision tree(GBDT) time-to-overtake
分类号:
U491.2
DOI:
10.3969/j.issn.1001-0505.2018.03.027
摘要:
为研究高速公路匝道车辆在交织区的汇入行为, 基于梯度提升决策树(GBDT)建立了车辆汇入模型, 引入超车时间T、拒绝间隙数N以及最大拒绝间隙GLR来分析匝道车辆拒绝相邻间隙并超越主线前车的行为, 并利用美国NGSIM项目中的车辆轨迹数据对模型进行训练和测试. 结果表明: GBDT的预测精度较分类回归树和二元Logit模型分别提高5.3%和13.3%; 引入变量T,N,GLR使GBDT、分类回归树和二元Logit模型的预测精度分别提高6.0%,6.7%和5.3%; GBDT模型中超车时间T在所有变量中重要性值最高. GBDT模型能够准确地预测汇入行为, 获得变量与汇入行为间隐藏的非线性关系;引入变量T,N,GLR能够有效提高汇入模型的预测精度.
Abstract:
To investigate the merging behaviors in weaving sections of vehicles in freeway ramp, a merging model based on the gradient boosting decision tree(GBDT)was proposed. The time-to-Overtake T, the number of rejected gaps N and the largest rejected gap GLR were used to analyze the behaviors of rejecting adjacent gaps and overtaking mainline vehicles. Data extracted from NGSIM(next generation simulation)dataset of America were used to train and test the model. The results show that the prediction accuracies of the proposed GBDT are 5.3% and 13.3% higher than those of the classification and regression tree(CART)model and the binary logit model, respectively. Considering T, N and GLR in the GBDT, CART and binary logit models can improve the prediction accuracy by 6.0%, 6.7% and 5.3%, respectively. T is the most important variable among all variables in the GBDT model. The GBDT model can accurately predict merging behaviors and obtain the hidden nonlinear relationships between the variables and the merging behaviors. The introduction of T, N and GLR can effectively improve the prediction accuracy.

参考文献/References:

[1] 张元良. 基于Stackelberg博弈理论的自主性车道变换模型[J]. 交通运输系统工程与信息, 2014, 14(5):67-73. DOI: 10.3969/j.issn.1009-6744.2014.05.010.
Zhang Yuanliang. Discretionary lane changing modeling based on Stackelberg game theory[J]. Journal of Transportation Systems Engineering and Information Technology, 2014, 14(5):67-73. DOI:10.3969/j.issn.1009-6744.2014.05.010. (in Chinese)
[2] Sarvi M, Kuwahara M. Microsimulation of freeway ramp merging processes under congested traffic conditions[J]. IEEE Transactions on Intelligent Transportation Systems, 2007, 8(3): 470-479. DOI:10.1109/tits.2007.895305.
[3] Toledo T,Koutsopoulos H N, Ben-Akiva M. Estimation of an integrated driving behavior model[J]. Transportation Research Part C: Emerging Technologies, 2009, 17(4): 365-380. DOI:10.1016/j.trc.2009.01.005.
[4] Marczak F, Daamen W, Buisson C. Merging behaviour: Empirical comparison between two sites and new theory development[J]. Transportation Research Part C: Emerging Technologies, 2013, 36: 530-546. DOI:10.1016/j.trc.2013.07.007.
[5] Wan X, Jin P J, Yang F, et al. Merging preparation behavior of drivers: How they choose and approach their merge positions at a congested weaving area[J]. Journal of Transportation Engineering, 2016, 142(9): 05016005. DOI:10.1061/(asce)te.1943-5436.0000864.
[6] Wan X, Jin P J,Gu H, et al. Modeling freeway merging in a weaving section as a sequential decision-making process[J]. Journal of Transportation Engineering Part A: Systems, 2017, 143(5): 05017002. DOI:10.1061/jtepbs.0000048.
[7] Meng Q, Weng J. Classification and regression tree approach for predicting drivers’ merging behavior in short-term work zone merging areas[J]. Journal of Transportation Engineering, 2012, 138(8): 1062-1070. DOI:10.1061/(asce)te.1943-5436.0000412.
[8] 邱小平, 刘亚龙, 马丽娜, 等. 基于贝叶斯网络的车辆换道模型[J]. 交通运输系统工程与信息, 2015, 15(5): 67-73,95. DOI: 10.3969/j.issn.1009-6744.2015.05.010.
Qiu Xiaoping, Liu Yalong, Ma Lina, et al. A lane change model based on Bayesian networks[J]. Journal of Transportation Systems Engineering and Information Technology, 2015, 15(5):67-73,95. DOI:10.3969/j.issn.1009-6744.2015.05.010. (in Chinese)
[9] Wang E G, Sun J, Jiang S, et al. Modeling the various merging behaviors at expressway on-ramp bottlenecks using support vector machine models[J]. Transportation Research Procedia, 2017, 25: 1327-1341. DOI:10.1016/j.trpro.2017.05.157.
[10] Ahmed M, Abdel-Aty M. Application of stochastic gradient boosting technique to enhance reliability of real-time risk assessment: Use of automatic vehicle identification and remote traffic microwave sensor data[J]. Transportation Research Record: Journal of the Transportation Research Board, 2013,2386: 26-34. DOI:10.3141/2386-04.
[11] Friedman J H. Stochastic gradient boosting[J]. Computational Statistics & Data Analysis, 2002, 38(4): 367-378. DOI:10.1016/s0167-9473(01)00065-2.
[12] Friedman J H, Meulman J J. Multiple additive regression trees with application in epidemiology[J]. Stat Med, 2003, 22(9): 1365-1381. DOI:10.1002/sim.1501.
[13] Alexiadis V, Colyar J, Halkias J, et al. The next generation simulation program[J]. Institute of Transportation Engineers, 2004, 74(8): 22-26.

备注/Memo

备注/Memo:
收稿日期: 2017-11-28.
作者简介: 李根(1989—),男,博士生, gilg4226307@aliyun.com.
引用本文: 李根.基于梯度提升决策树的高速公路交织区汇入模型[J].东南大学学报(自然科学版),2018,48(3):563-567. DOI:10.3969/j.issn.1001-0505.2018.03.027.
更新日期/Last Update: 2018-05-20