摘 要:为了实时判断城市交通及快速路交通状况,并充分考虑到快速路及交通系统的模糊性、差异性、波动性,文章首先给出了一套基于模糊 C 均值聚类法(Fuzzy C-Means)的快速路交通状况判断算法,将道路的交通状况区分为通畅、基本顺畅、缓行、堵塞 4 类。其次,利用了长短期记忆网络(LSTM)的长短时预测技术,给出了一个基于深度学习的模式,对交通运行三参数进行预测,并利用路侧检测设备不断获取新的真实数据,对预测模型进行迭代训练。然后利用已形成的模糊聚类模型,得到快速路交通运行状态的预测结果。最后以天津市黑牛城道的真实交通数据进行实验验证,结果表明,文章提出的交通流状态预测模型准确性达到 93.22%,能为快速路交通控制提供有效的状态预测。
关键词:城市快速路;模糊 C 均值聚类;深度学习;LSTM 短时预测方法;交通运行三参数
DOI:10.19850/j.cnki.2096-4706.2023.01.001
基金项目:天津市科技计划项目(22YDTPJC00120,XC202028)
中图分类号:TP391.9 文献标识码:A 文章编号:2096-4706(2023)01-0001-08
Traffic Operation State Discrimination and Prediction of Urban Expressway Based on Deep Learning
ZHAO Biao, LIU Xiaofeng, FU Tian
(Tianjin University of Technology and Education, Tianjin 300222, China)
Abstract: In order to judge the urban traffic and traffic condition of expressway in real time, and to fully consider the fuzziness, variability, and fluctuation of expressway and traffic system, a set of judgment algorithms of expressway traffic condition based on Fuzzy C-Means is given in this paper firstly, which distinguishes the traffic conditions of roads into four categories: smooth, basically smooth, slow and blocked. Secondly, a model based on deep learning is given to predict the three parameters of traffic operation by using the Long Short-Term Memory (LSTM) long short-term prediction technique, and the prediction model is trained iteratively by continuously acquiring new real data using roadside detection equipments. Then using the developed fuzzy clustering model, the prediction results of the traffic operation state of the expressway are obtained. Finally, the real traffic data of Heiniu Road in Tianjin is used for experimental validation, and the results show that the accuracy of the traffic flow state prediction model proposed in this paper reach 93.22%, which can provide effective state prediction for expressway traffic control.
Keywords: urban expressway; Fuzzy C-Means; deep learning; LSTM short-time prediction method; three parameters of traffic operation
参考文献:
[1] GUO J H,HUANG W,WILLIAMS B M.Adaptive Kalman Filter Approach for Stochastic Short-Term Traffic Flow Rate Prediction and Uncertainty Quantification [J].Transportation Research Part C, 2014,43:50-64.
[2] WANG Z C,LU M,YUAN X R,et al.Visual Traffic Jam Analysis Based on Trajectory Data [J].IEEE Transactions on Visualization & Computer Graphics,2013,19(12):2159-2168.
[3] 焦朋朋,安玉,白紫秀,等 . 基于 XGBoost 的短时交通流预测研究 [J]. 重庆交通大学学报:自然科学版,2022,41(8):17-23+66.
[4] 褚瑞娟 . 高速公路交通运行状态判别与预测方法研究 [D].长春:吉林大学,2021.
[5] 张亮亮 . 城市快速路交通运行状态评价及预测方法研究[D]. 北京:北京交通大学,2016.
[6] LIN X F,HUANG Y Z. Short-Term High-Speed Traffic Flow Prediction Based on ARIMA-GARCH-M Model [J].Wireless Personal Communications,2021,117(4):3421-3430.
[7] SHAHRIARI S,GHASRI M,SISSON S A,et al.Ensemble of ARIMA:Combining Parametric and Bootstrapping Technique for Traffic Flow Prediction [J].Transportmetrica a-Transport Science, 2020,16(3):1552-1573.
[8] 黄艳国,许伦辉,邝先验 . 基于模糊 C 均值聚类的城市道路交通状态判别 [J].重庆交通大学学报:自然科学版,2015,34(2):102-107.
[9] 李晓璐 . 基于模糊 C 均值聚类的城市道路交通状态判别[J]. 交通科技与经济,2016,18(4):32-36+42.
[10] 陈治亚,王小军 . 基于多维度 LSTM 模型的短时交通流预测 [J]. 铁道科学与工程学报,2020,17(11):2946-2952.
[11] 王斌 . 基于 LSTM 循环神经网络的交通流量预测 [D]. 昆明:云南大学,2017.
[12] 刘易诗,关雪峰,吴华意,等 . 基于时空关联度加权的LSTM 短时交通速度预测 [J]. 地理信息世界,2020,27(1):41-47.
[13] 符锌砂,梁中岚,郑伟,等 . 二分 K-FCM 结合算法在交通运行状态判别中的应用 [J]. 公路工程,2018,43(2):118-123.
[14] 刘川 . 基于 LSTM 及其变体的短时交通流预测 [J]. 交通节能与环保,2022,18(4):99-105.
作者简介:赵彪(1994.08—),男,汉族,甘肃静宁人,硕士研究生在读,研究方向:城市交通控制与仿真。