Combination Forecast Model of Traffic Flow Probability Based on Similarity Clustering
-
摘要: 针对交通流呈现出周期性动态变化特征，充分挖掘并利用交通流数据潜在的时段相似性特征，提出了一种基于相似性聚类的交通流概率组合模型。首先采用自适应k-means++聚类方法对历史交通流数据进行聚类，对具有周期相似性的交通流数据进行分类，然后针对同类特征序列数据集构建子组合模型。针对新输入的交通流状态数据，分析其与已分类数据的相似度计算组合模型的概率权重，然后通过概率加权融合组合模型预测输出。仿真实验验证了本文模型的有效性与准确性。Abstract: Aiming at the periodic dynamic characteristics of traffic flow, a probabilistic combination model of traffic flow based on similarity clustering is proposed, which fully excavates the similarity characteristics of traffic flow in different periods. Firstly, the adaptive k-means + + clustering method is used to cluster the historical traffic flow data and classify the traffic flow data with time similarity. Then, the combination model is constructed for different sequence feature data sets. Furthermore, according to the new traffic flow state data, the similarity between the new traffic flow state data and the classified data is analyzed, and the probability weight of the combined model is calculated. Then, the prediction output is obtained by fusing the probability weight of the combined model results. Finally, the validity and accuracy of the proposed prediction model are verified by simulation experiments.
-
图 1 基于相似性聚类的交通流概率组合预测模型
Figure 1. Combination forecasting model of traffic flow probability based on similarity clustering
图 3 基于在线误差的自适应权重调节流程图
Figure 3. Flow chart of adaptive weight adjustment based on online error
表 1 预测指标对比
Table 1. Evalution of prediction models
Model RMSE MAE MAPE/% PLS-LSTM 30.90 21.99 8.98 PLS-LSTM(without clustering) 32.89 23.17 9.35 ARIMA-BP 32.97 23.67 9.51 LSTM 33.50 23.92 9.62 SVR 36.24 29.12 15.67 
下载: 导出CSV
表 2 计算时间与精度的关系
Table 2. Relationship between calculation time and accuracy
Epoch t/s RMSE 20 30.56 31.96 40 42.29 31.01 60 50.75 30.90 80 61.07 30.80
下载: 导出CSV
-
[1] HUANG W, JIA W, GUO J, et al. Real-time prediction of seasonal heteroscedasticity in vehicular traffic flow series[J]. IEEE Transactions on Intelligent Transportation Systems, 2017, 19(10): 3170-3180. [2] MOREIRA-MATIAS L, GAMA J, FERREIRA M, et al. Predicting taxi–passenger demand using streaming data[J]. IEEE Transactions on Intelligent Transportation Systems, 2013, 14(3): 1393-1402. doi: 10.1109/TITS.2013.2262376 [3] LEE S, LEE Y I, CHO B. Short-term travel speed prediction models in car navigation systems[J]. Journal of Advanced Transportation, 2010, 40(2): 122-139. [4] SAFIKHANI A, KAMGA C, MUDIGONDA S, et al. Spatio-temporal modeling of Yellow Taxi Demands in New York City using generalized STAR models[J]. International Journal of Forecasting, 2020, 36(3): 1138-1148. doi: 10.1016/j.ijforecast.2018.10.001 [5] WANG JIN, SHI QIXIN. Short-term traffic speed forecasting hybrid model based on Chaos-wavelet analysis-support vector Machine theory[J]. Transportation Research: Part C. Emerging Technologies, 2013, 27: 219-232. doi: 10.1016/j.trc.2012.08.004 [6] WU CHUN-HISIN, HO JAN-MING, LEE D T. Travel-time prediction with support vector regression[J]. IEEE Transactions on Intelligent Transportation Systems, 2005, 5(4): 276-281. [7] YANG Z, MEI D, YANG Q, et al. Traffic flow prediction model for large-scale road network based on cloud computing[J]. Mathematical Problems in Engineering, 2014(3): 1-8. [8] 许岩岩, 翟希, 孔庆杰, 等. 基于分类回归树的交通流短时预测[C]//第七届中国智能交通年会优秀论文集——智能交通技术. 北京: 中国智能交通协会, 2012: 35-36. [9] SONG C, LEE H, KANG C, et al. Traffic speed prediction under weekday using convolutional neural networks concepts[C]//Proceedings of the 2017 IEEE Intelligent Vehicles Symposium (IV). USA: IEEE, 2017: 1293-1298, . [10] ZHAO Z, CHEN W H, WU X M, et al. LSTM network: A deep learning approach for short-term traffic forecast[J]. IET Intelligent Transport Systems, 2017, 11(2): 68-75. doi: 10.1049/iet-its.2016.0208 [11] LIU B Y, CHENG J R, CAI K Q, et al. Singular point probability improve LSTM network performance for long-term traffic flow prediction[C]//Proceedings of the National Conference of Theoretical Computer Science. Singapore: Springer, 2017: 328-340. [12] 沈国江, 朱芸, 钱晓杰, 等. 短时交通流组合模型预测[J]. 南京理工大学学报(自然科学版), 2014, 38(2): 246-251. [13] 李建森, 沈齐, 范馨月. 城市道路短时交通流量预测[J]. 数学的实践与认识, 2019, 49(5): 192-197. [14] 钱伟, 车凯, 李冰锋. 基于组合模型的短时交通流量预测[J]. 控制工程, 2019, 26(1): 125-130. [15] GUO Y, YANG H, CHEN M, et al. Ensemble prediction-based dynamic robust multi-objective optimization methods[J]. Swarm and Evolutionary Computation, 2019, 48: 156-171. doi: 10.1016/j.swevo.2019.03.015 [16] YANG L, GUO Y, CHENG J. Manifold distance-based over-sampling technique for class imbalance learning[C]//Proceedings of the AAAI Conference on Artificial Intelligence. Honolulu, Hawaii, USA: AAAI, 2019, 10071-10072. [17] ARTHUR D, VASSILVITSKII S. k-Means++: The advantages of careful seeding[C]// Proceedings of the Eighteenth Annual ACM-SIAM Symposium on Discrete Algorithms. New Orleans, Louisiana, USA: ACM, 2007: 1027-1035. [18] DAI G W, MA C X, XU X C. Short-term traffic flow prediction method for urban road sections based on space-time analysis and GRU[J]. IEEE Access, 2019, 7: 143025-143035. doi: 10.1109/ACCESS.2019.2941280 -
点击查看大图
图(7) / 表(2)
计量
- 文章访问数: 700
- HTML全文浏览量: 515
- PDF下载量: 29
- 被引次数: 0
