此页面上的内容需要较新版本的 Adobe Flash Player。

获取 Adobe Flash Player

Short-term traffic flow online forecasting based on kernel adaptive filter


LI Jun, WANG Qiu-li


School of Automation & Electrical Engineering, Lanzhou Jiaotong University, Lanzhou 730070, China)


Abstract: Considering that the prediction accuracy of the traditional traffic flow forecasting model is low, based on kernel adaptive filter (KAF) algorithm, kernel least mean square (KLMS) algorithm and fixed-budget kernel recursive least-square (FB-KRLS) algorithm are presented for online adaptive prediction. The computational complexity of the KLMS algorithm is low and does not require additional solution paradigm constraints, but its regularization process can solve the problem of regularization performance degradation in high-dimensional data processing. To reduce the computational complexity, the sparse criterion is introduced into the KLMS algorithm. To further improve forecasting accuracy, FB-KRLS algorithm is proposed. It is an online learning method with fixed memory budget, and it is capable of recursively learning a nonlinear mapping and changing over time. In contrast to a previous approximate linear dependence (ALD) based technique, the purpose of the presented algorithm is not to prune the oldest data point in every time instant but it aims to prune the least significant data point, thus suppressing the growth of kernel matrix. In order to verify the validity of the proposed methods, they are applied to one-step and multi-step predictions of traffic flow in Beijing. Under the same conditions, they are compared with online adaptive ALD-KRLS method and other kernel learning methods. Experimental results show that the proposed KAF algorithms can improve the prediction accuracy, and its online learning ability meets the actual requirements of traffic flow and contributes to real-time online forecasting of traffic flow.


Key words: traffic flow forecasting; kernel adaptive filtering (KAF); kernel least mean square (KLMS); kernel recursive least square (KRLS); online forecasting


 

CLD number: TP273+.2              Document code: A


Article ID: 1674-8042(2018)04-0326-09             doi: 10.3969/j.issn.1674-8042.2018.04.004


 

References


1] Vlahogianni E I, Golias J C, Karlaftis M G. Short-term traffic forecasting: overview of objectives and methods. Transport Reviews, 2004, 24(5): 533-557.

2] Lippi M, Bertini M, Frasconi P. Short-term traffic flow forecasting: an experimental comparison of time series analysis and supervised learning. IEEE Transactions on Intelligent Transportation Systems, 2013, 14(2): 871-882.

3] Xu C, Li Z, Wang W. Short-term traffic flow prediction using a methodology based on autoregressive integrated moving average and genetic programming. Transport, 2016, 31(3): 343-358.

4] Xu R, Zhou D, Jiang S Z, et al. Adaptive traffic flow forecasting model based on particle swarm neural network. Journal of Xi’an Jiaotong University, 2015, 49 (10): 103-108.

5] Messer C, Urbanik T. Short-term freeway traffic volume forecasting using radial basis function neural network. Transportation Research Record Journal of the Transportation Research Board, 1998, 1651(1): 39-47.

6] Zhang Y L, Xie Y C. Forecasting of short term freeway volume with support vector machines. Transportation Research Record, 2007, 12(2): 92-99.

7] Li Q R, Zhao R, Chen L. A short-term traffic flow forecasting model based on SVM and adaptive spatio-temporal data fusion. Beijing University of Technology, 2015 (4): 597-602.

8] Chen X B, Liu X. Short-term traffic flow forecasting of road network based on GA-LSSVR model. Journal of Transportation Systems Engineering and Information Technology, 2017, 17(1): 60-73.

9] Shang Q, Yang Z S, Zhang W, et al. Short term traffic flow forecasting based on singular spectrum analysis and CKF-LSSVM. Journal of Jilin University (Engineering), 2016, 46(6): 1792-1798.

10] Huang G B, Zhu Q Y, Siew C K. Extreme learning machine: a new learning scheme of feedforward neural networks. In: Proceedings of 2004 IEEE International Joint Conference on Neural Networks, Budapest, Hungary, 2004, 2: 985-990.

11] Shang Q, Yang Z S, Li Z, et al. Prediction of  short term traffic flow based on phase space reconstruction and RELM. Journal of South China University of Technology, 2016, 44(4): 109-114.

12] Li J, Li D C. Prediction of wind power time series based on optimized kernel learning machine. Chinese Journal of physics, 2016, 65(13): 33-42.

13] Rosipal R, Trejo L. Kernel partial least square regression in reproducing kernel Hilbert space. Journal of Machine Learning Research, 2001, 2(2): 97-123.

14] Sun Z, Fox G. Traffic flow forecasting based on combination of multidimensional scaling and SVM. International Journal of Intelligent Transportation Systems Research, 2014, 12(1): 20-25.

15] Chen Q S, Chen X W, Wu Y. Optimization algorithm with kernel PCA to support vector machines for time series prediction. Journal of Computers, 2010, 5(3): 380-387.

16] Ma W T, Duan J D, Man W S, et al. Robust kernel adaptive filters based on mean p-power error for noisy chaotic time series prediction. Engineering Applications of Artificial Intelligence, 2017, 58: 101-110.

17] Engel Y, Mannor S, Meir R. The kernel recursive least-square algorithm. IEEE Transactions on Signal Processing, 2004, 52(8): 2275-2285.

18] Liu W, Pokharel P P, Principe J C. The kernel least-mean-square algorithm. IEEE Transactions on Signal Processing, 2008, 56(2): 543- 554.

19] Tobar F A, Kung S Y, Mandic D P. Multikernel least mean square algorithm. IEEE Transactions on Neural Networks & Learning Systems, 2014, 25(2): 265.

20] Vaerenbergh S V, Santamaría I, Liu W, et al. Fixed-budget kernel recursive least-squares. In: Proceedings of IEEE International Conference on Aacoustics Speech and Signal Processing. Dallas, USA, 2010: 1882-1885.

21] Sun S, Xu X. Variational inference for infinite mixtures of Gaussian processes with applications to traffic flow prediction. IEEE Transactions on Intelligent Transportation Systems, 2011, 12(2): 466-475.

22] Hou Y. Traffic flow prediction based on improved T-S fuzzy neural network. Computer Science and Exploration, 2014, 8 (1): 121-126.


 

基于核自适应滤波的短时交通流量在线预测


  军, 王秋莉


(兰州交通大学 自动化与电气工程学院,  甘肃 730070)


  要:  针对传统交通流量预测模型准确率低的问题, 提出一类基于核自适应滤波算法的在线自适应预测模型, 包括核最小均方(kernel least mean square, KLMS)算法和固定预算核递推最小二乘(fixed-dudget kernel recursive least-square, FB-KRLS)算法。 KLMS算法的计算复杂度较低且具有鲁棒性, 能通过正则化过程, 克服高维数据处理时正则化性能降低的问题。 为进一步降低计算复杂度, 引入稀疏技术到KLMS算法。 为提高预测精度, 提出FB-KRLS算法。 FB-KRLS算法是一种固定内存预算的在线学习方法, 与以往的“近似线性依赖技术(approximate linear dependence, ALD)”技术不同, 每次时间更新时并不“修剪”最旧的数据, 而是“修剪”最无用的数据, 从而抑制核矩阵的不断增长。 为验证所提方法的有效性, 将其应用于北京市实测交通流数据的单步及多步预测中, 并在同等条件下, 与在线自适应ALD-KRLS方法以及其他核学习方法进行了比较。 实验结果表明, 本文所提的滤波算法方法可以提高预测精度, 其在线学习能力使得所提出的两种预测方法能自适应、 实时地进行在线交通流量预测, 从而满足了交通流量的实际特性,  具有很好的应用潜力。


关键词:  交通流量预测; 核自适应滤波; 核最小均方; 核递推最小二乘; 在线预测


 

引用格式:  LI Jun, WANG Qiu-li. Short-term traffic flow online forecasting based on kernel adaptive filter. Journal of Measurement Science and Instrumentation, 2018, 9(4): 326-334. [doi: 10.3969/j.issn.1674-8042.2018.04.004]


[full text view]