基于变量选择-神经网络模型的复杂路网短时交通流预测

摘要/Abstract

摘要：

摘要：针对传统交通流预测模型正在由单断面历史数据处理向多断面、多时刻历史数据处理转变,但在考虑各断面间的影响时,多变的交通状况往往会使预测模型复杂化的问题,引入一种多元线性回归最小绝对收缩和选择算子方法(Lasso),并利用其优秀的变量选择能力,在复杂路网多断面中选出相关性较高的断面；结合神经网络(NN)的非线性特性,提出了LassoNN组合模型.结果表明：LassoNN模型在路网交叉口对未来15 min交通流数据预测的误差率低于9.2%；在非交叉口的误差率低于6.7%,总体优于各自单独使用得出的结果.

关键词: 短时交通流预测, 最小绝对收缩和选择算子, 变量选择, 神经网络

Abstract:

Abstract： Traditional traffic forecasting models are transforming from single section historical data processing to multi-section and multi-timing historical data processing. However, when considering the influence between each section, the fickle traffic condition tends to complicate the forecasting model. Therefore, this paper introduced the Lasso method used in multivariable linear regression and utilized its excellent ability of variable selection. It selected partial high correction sections from a complex multi-section road network. Combined with the non-linear neural network, a new LassoNN model was proposed. The result shows that the LassoNN model has an overall lower error rate. In the intersection region of the road network, the error rate is less than 9.2% and in the nonintersection region, it is less than 6.7%.

Key words: short-term traffic forecasting, least absolute shrinkage and selection operator (Lasso), variable selection, neural network

中图分类号:

U 491

蒋士正，许榕，陈启美. 基于变量选择-神经网络模型的复杂路网短时交通流预测[J]. 上海交通大学学报（自然版）, 2015, 49(02): 281-286.

JIANG Shizheng,XU Rong,CHEN Qimei. Complex Network Short-Term Traffic Forecasting
Based on Lasso-NN Model[J]. Journal of Shanghai Jiaotong University, 2015, 49(02): 281-286.

参考文献

［1］王进, 史其信. 短时交通流预测模型综述［J］. ITS通讯, 2005, 7(1): 1013.

WANG Jin, SHI Qixin. Survey on shortterm traffic flow forecasting modelling［J］. ITS Communication, 2005, 7(1): 1013.

［2］Thomas T, Weijermars W, Berkum E V. Predictions of urban volumes in single time series［J］. IEEE Transactions on Intelligent Transportation Systems, 2010, 11(1): 7180.

［3］Castillo E, Nogal M, Menéndez J M, et al. Stochastic demand dynamic traffic models using generalized betaGaussian Bayesian networks［J］. IEEE Transactions on Intelligent Transportation Systems, 2012, 13(2): 565581.

［4］Adeli H. Neural networks in civil engineering: 19892000［J］. ComputerAided Civil and Infrastructure Engineering, 2001, 16(2): 126142.

［5］Vlahogianni E I, Karlaftis M G, Golias J C. Shortterm traffic forecasting: Where we are and where we’re going［J］. Transportation Research Part C: Emerging Technologies, 2014, 43(1): 319.

［6］Tibshirani R. Regression shrinkage and selection via the lasso［J］. Journal of the Royal Statistical Society. Series B(Methodological), 1996, 58(1): 267288.

［7］Efron B, Hastie T, Johnstone I, et al. Least angle regression［J］. The Annals of statistics, 2004, 32(2): 407499.

［8］Zou H, Hastie T, Tibshirani R. Sparse principal component analysis［J］. Journal of Computational and Graphical Statistics, 2006, 15(2): 265286.

［9］Trendafilov N T, Jolliffe I T. DALASS: Variable selection in discriminant analysis via the LASSO［J］. Computational statistics & Data Analysis, 2007, 51(8): 37183736.

［10］Kamarianakis Y, Shen W, Wynter L. Realtime road traffic forecasting using regimeswitching spacetime models and adaptive LASSO［J］. Applied Stochastic Models in Business and Industry, 2012, 28(4): 297315.

［11］Sun S, Huang R, Gao Y. Networkscale traffic modeling and forecasting with graphical lasso and neural networks［J］. Journal of Transportation Engineering, 2012, 138(11): 13581367.

［12］张文龙, 于洪浩. 一种星形神经网络的混沌同步［J］. 上海交通大学学报, 2013, 47(2)： 220225.

ZHANG Wenlong, YU Honghao. Chaotic synchronization of a type of starshaped neural network［J］. Journal of Shanghai Jiaotong University, 2013, 47(2)： 220225.

［13］Zou H, Hastie T. Regularization and variable selection via the elastic net［J］. Journal of the Royal Statistical Society: Series B(Statistical Methodology), 2005, 67(2): 301320.

［14］California department of transportation. The freeway performance measurement system(PeMS) ［DB/OL］.(20130606)［20140410］. http://pems. dot.ca.gov/

[1]	潘美琪, 贺兴. 基于零样本学习的风力机故障诊断方法[J]. 上海交通大学学报, 2025, 59(5): 561-568.
[2]	黄逸翔, 窦迅, 李林溪, 杨函煜, 于建成, 霍现旭. 基于全局灵敏度分析的综合能源设备响应价值量化方法[J]. 上海交通大学学报, 2025, 59(5): 569-579.
[3]	. 基于表面肌电信号的BP-LSTM混合模型肘部运动实时预测[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(3): 455-462.
[4]	. 基于基尼不纯度结构优化物理引导神经网络的薄膜型声学超材料传声损失预测[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(3): 613-624.
[5]	. 基于毫米波雷达的智能心率提取方法[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(3): 493-498.
[6]	覃浩, 苏立伟, 伍广斌, 蒋崇颖, 徐智鹏, 康峰, 谭火超, 张勇军. 基于集成学习和卷积神经网络的电网客服短期话务量预测[J]. 上海交通大学学报, 2025, 59(2): 266-273.
[7]	. 基于RGB-D图像的机器人抓取检测高效全卷积网络和优化方法[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(2): 399-416.
[8]	. 重复零和博弈中对无遗憾对手进行盘剥的研究[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(2): 385-398.
[9]	Sahaya Anselin Nisha1, NARMADHA R.1, AMIRTHALAKSHMI T. M.2, BALAMURUGAN V.1, VEDANARAYANAN V.1. LOBO优化的深度卷积神经网络用于脑肿瘤分类[J]. J Shanghai Jiaotong Univ Sci, 2025, 30(1): 107-114.
[10]	孙栋一, 蒲宇亭, 章建榜. 基于GA-BP神经网络的防空导弹实时目标分配方法[J]. 空天防御, 2025, 8(1): 62-70.
[11]	刘俊城, 谭勇, 张生杰. 地铁车站深基坑开挖变形智能多步预测方法[J]. 上海交通大学学报, 2024, 58(7): 1108-1117.
[12]	李利娟, 刘海, 刘红良, 张青松, 陈永东. 融合外部注意力机制的序列到点非侵入式负荷分解[J]. 上海交通大学学报, 2024, 58(6): 846-854.
[13]	李明爱1, 2, 魏丽娜1. 基于朴素卷积神经网络和线性插值的运动想像分类[J]. J Shanghai Jiaotong Univ Sci, 2024, 29(6): 958-966.
[14]	蒋伊琳1, 2, 李向1, 2, 张昊平3. 基于卷积神经网络和接收信号强度的波束状态感知方法[J]. J Shanghai Jiaotong Univ Sci, 2024, 29(6): 1017-1022.
[15]	张国栋, 刘凯, 蒲海涛, 姚福强, 张帅帅. 基于长短时记忆神经网络的励磁涌流与故障电流识别方法[J]. 上海交通大学学报, 2024, 58(5): 730-738.