| 247 | 5 | 1105 |
| 下载次数 | 被引频次 | 阅读次数 |
城市交叉口的道路通行能力目前已成为影响居民出行效率的重要因素之一。通过正向雷达道路车辆检测器采集过车数据，利用深度学习技术Tensorflow框架构建平稳化长短期记忆网络（long short-term memory,LSTM）交通量预测模型。该模型根据预测精度自适应机制更新层数和隐藏单元数，同时为了减小交通流序列的随机性，通过季节性时间差分方法对原始交通时间序列的输出特征序列进行了平稳化处理。在使用相同数据集条件下，将所提算法的预测结果与传统算法的结果进行对比，结果表明：平稳化的LSTM网络算法能有效提高交通流量预测的准确性，实际值与预测值拟合度超过95.5%，且在均方根误差和平均百分比误差上有较大的提升，该模型可为治理交通拥堵提供依据。
Abstract：The road capacity of urban intersections has become one of the important factors affecting residents′ travel efficiency. Through the forward radar road vehicle detector to collect passing data, the data mining and analysis can gather the road condition information, so as to provide a basis for the management of traffic congestion. This paper uses the deep learning technology to build a smooth long short-term memory(LSTM) traffic volume prediction model to solve the problem of time series traffic volume prediction at intersections. The model uses the Tensorflow framework of in-depth learning to construct the hierarchical structure of LSTM network, and updates the number of layers and hidden units according to the prediction accuracy adaptive mechanism. At the same time, in order to reduce the randomness of traffic flow sequence, the output characteristic sequence of the original traffic time series is smoothed by seasonal time difference method. Under the same data set, the comparison with the prediction results of GRU model and support vector regression model shows that the smoothed LSTM network algorithm can effectively improve the accuracy of traffic flow prediction. The fitting degree between the actual value and the predicted value is more than95.5%. Moreover, the root mean square error and the average percentage error are greatly improved. This model can provide reference for easing the traffic congestion.
[1]沈夏炯，张俊涛，韩道军.基于梯度提升回归树的短时交通流预测模型[J].计算机科学，2018, 45(6):222-227.SHEN X J, ZHANG J T, HAN D J. Short-term traffic flow prediction model based on gradient boosting regression tree[J]. Computer Science, 2018, 45(6):222-227.(in Chinese)
[2]罗向龙，焦琴琴，牛力瑶，等.基于深度学习的短时交通流预测[J].计算机应用研究，2017, 34(1):91-93.LUO X L, JIAO Q Q, NIU L Y, et al. Short-term traffic flow prediction based on deep learning[J]. Application Research of Computers, 2017, 34(1):91-93.(in Chinese)
[3]刘明宇，吴建平，王钰博，等.基于深度学习的交通流量预测[J].系统仿真学报，2018, 30(11):4100-4105.LIU M Y, WU J P, WANG Y B, et al. Traffic flow prediction based on deep learning[J]. Journal of System Simulation, 2018, 30(11):4100-4105.(in Chinese)
[4] LIPPI M, BERTINI M, FRASCONI P. Short-term traffic flow forecasting:an experimental comparison of time-series analysis and supervised learning[J]. IEEE Transactions on Intelligent Transportation Systems, 2013, 14(2):871-882.
[5] ZHANG H, WANG X M, CAO J, et al. A multivariate short-term traffic flow forecasting method based on wavelet analysis and seasonal time series[J]. Applied Intelligence,2018, 48(10):3827-3838.
[6] ZHOU T, HAN G Q, XU X M, et al. A learning-based multimodel integrated framework for dynamic traffic flow forecasting[J]. Neural Processing Letters, 2019, 49(1):407-430.
[7]康军，段宗涛，唐蕾，等.一种平稳化短时交通流预测方法[J].测控技术，2018, 37(2):33-37.KANG J, DUAN Z T, TANG L, et al. Stationary shortterm traffic flow prediction method[J]. Measurement and Control Technology, 2018, 37(2):33-37.(in Chinese)
[8]谈苗苗，成孝刚，周凯，等.基于ARIMA和灰色模型加权组合的短期交通流预测[J].计算机技术与发展，2016,26(11):77-81.TAN M M, CHENG X G, ZHOU K, et al. Short-term traffic flow forecasting based on combination of ARIMA and gray model[J]. Computer Technology and Development,2016, 26(11):77-81.(in Chinese)
[9]黄慧琼.基于模糊理论的城市道路短时交通流量预测研究[J].公路工程，2016, 41(1):172-175.HUANG H Q. Research on the urban road short-term traffic flow forecast based on fuzzy theory[J]. Highway Engineering, 2016, 41(1):172-175.(in Chinese)
[10]韩志聪，樊彦国，吴会胜，等.基于GA-SVR模型的短期交通流量预测方法研究[J].公路交通科技，2017, 34(1):130-136.HAN Z C, FAN Y G, WU H S, et al. Study on methods of short-time traffic volume forecasting based on GA-SVR model[J]. Journal of Highway and Transportation Research and Development, 2017, 34(1):130-136.(in Chinese)
[11]韦凌翔，陈红，王永岗，等.基于RVM和ARIMA的短时交通流量预测方法研究[J].武汉理工大学学报（交通科学与工程版），2017, 41(2):349-354.WEI L X, CHEN H, WANG Y G, et al. Study on the methods of short-term traffic flow forecasting based on RVM and ARIMA[J]. Journal of Wuhan University of Technology(Transportation Science and Engineering), 2017, 41(2):349-354.(in Chinese)
[12]高波.基于PLS的交通流量预测模型的研究[J].交通科学与工程，2017, 33(2):97-101.GAO B. Study on the prediction model of traffic flow based on partial least squares method[J]. Journal of Transport Science and Engineering, 2017, 33(2):97-101.(in Chinese)
[13]王体迎，时鹏超，刘蒋琼，等.基于门限递归单元循环神经网络的交通流预测方法研究[J].重庆交通大学学报（自然科学版），2018, 37(11):76-82.WANG T Y, SHI P C, LIU J Q, et al. Research on traffic flow prediction method based on gated recurrent unit recurrent neural network[J]. Journal of Chongqing Jiaotong University(Natural Science), 2018, 37(11):76-82.(in Chinese)
[14]张红.基于大数据的城市公路交通流短时预测研究[D].兰州：兰州理工大学，2018.ZHANG H. Research on short-term traffic flow forecasting of urban highway based on big data[D]. Lanzhou:Lanzhou University of Technology, 2018.(in Chinese)
[15]刘剑，刘丽华，赵悦.基于KPCA与SVM的混合核交通流数据检测[J].沈阳建筑大学学报（自然科学版），2018,34(5):921-928.LIU J, LIU L H, ZHAO Y. A mixed kernel traffic flow data detection based on KPCA and SVM[J]. Journal of Shenyang Jianzhu University(Natural Science), 2018, 34(5):921-928.(in Chinese)
[16]陆文琦，谷远利，陈伦.基于时空融合的城市快速路短时交通流预测[J].计算机仿真，2018, 35(9):136-140.LU W Q, GU Y L, CHEN L. Short-term traffic flow forecasting of urban expressway based on spatial and temporal traffic characteristics[J]. Computer Simulation, 2018, 35(9):136-140.(in Chinese)
[17] HABTEMICHAEL F G, CETIN M. Short-term traffic flow rate forecasting based on identifying similar traffic patterns[J]. Transportation Research Part C:Emerging Technologies, 2016, 66:61-78.
[18] POLSON N G, SOKOLOV V O. Deep learning for shortterm traffic flow prediction[J]. Transportation Research Part C:Emerging Technologies, 2017, 79:1-17.
[19] ABOUA?SSA H, FLIESS M, JOIN C. On short-term traffic flow forecasting and its reliability[J]. IFAC-PapersOnLine, 2016, 49(12):111-116.
[20] HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation, 1997, 9(8):1735-1780.
基本信息:
DOI：10.12194/j.ntu.20200327002
中图分类号:U491.1
引用信息:
[1]吕心钰,施佺.基于平稳化长短期记忆网络模型的交叉口交通流量预测[J],2022,21(01):37-43.DOI:10.12194/j.ntu.20200327002.
基金信息:
国家自然科学基金面上项目（61771265）;; 江苏省“333工程”项目(BRA2017475);; 江苏省“青蓝工程”项目;; 南通市科技计划项目（CP12017001）