Abstract
Because vehicular traffic is affected by weather conditions, knowledge of the relationship between weather and traffic enables attempts to improve social services through applications such as situation-aware anomaly vehicle detection and snow-removal planning in snowy countries. We propose a weather-aware traffic state model for vehicular traffic analysis in consideration of weather conditions. The model is a probabilistic latent variable model that integrates weather and traffic data, whereby the characteristics of the traffic according to location, time, and weather condition are obtained automatically. After we observe both weather and travel times along road segments, we derive the expectation–maximization algorithm for model parameter estimation and the predictive distribution of travel time given the weather observation values. We evaluated the model qualitatively and quantitatively using winter traffic and weather data for the city of Sapporo, Japan, which is a large city that suffers heavy snowfalls. The empirical analysis with model visualization outcomes demonstrated the relationship between the expected vehicular speed and weather conditions, and showed the potential bottleneck segments for given weather conditions. The quantitative evaluation showed that our model fits the data better than a linear regression model, which suggests the potential for anomaly detection from vehicular observation data.
This is a preview of subscription content, log in via an institution.
References
Bishop, C.M.: Mixture Models and EM. In: Pattern Recognition and Machine Learning, pp. 423–460. Springer, New York (2006)
Cao, P., Miwa, T., Morikawa, T.: Modeling distribution of travel time in signalized road section using truncated distribution. In: Proceedings of 9th International Conference on Traffic and Transportation Studies (ICTTS 2014), vol. 138, pp. 137–147. Elsevier (2014). doi:10.1016/j.sbspro.2014.07.189
City of Sapporo: Actual budget for the counterplan against snow. http://www.city.sapporo.jp/kensetsu/yuki/jigyou/budget.html. Accessed 30 Nov 2016
Guessous, Y., Aron, M., Bhouri, N., Cohen, S.: Estimating travel time distribution under different traffic conditions. Transp. Res. Procedia 3(July), 339–348 (2014). doi:10.1016/j.trpro.2014.10.014
Hall, F.L., Barrow, D.: Effect of weather on the relationship between flow and occupancy on freeways. Transp. Res. Record 1194, 55–63 (1988)
Japan Meteorological Agency: Observations. http://www.jma.go.jp/jma/en/Activities/observations.html. Accessed 23 Nov 2016
Keay, K., Simmonds, I.: The association of rainfall and other weather variables with road traffic volume in Melbourne, Australia. Accid. Anal. Prev. 37(1), 109–124 (2005). doi:10.1016/j.aap.2004.07.005
Kinoshita, A., Takasu, A., Adachi, J.: Real-time traffic incident detection using a probabilistic topic model. Inf. Syst. 54, 169–188 (2015). doi:10.1016/j.is.2015.07.002
Lee, J., Hong, B., Lee, K., Jang, Y.J.: A prediction model of traffic congestion using weather data. In: 2015 IEEE International Conference on Data Science Data Intensive System, pp. 81–88. IEEE (2015). doi:10.1109/DSDIS.2015.96
Li, W., McCallum, A.: Pachinko allocation: DAG-structured mixture models of topic correlations. In: Proceedings of the 23rd International Conference on Machine Learning (ICML 2006), pp. 577–584. ACM (2006). doi:10.1145/1143844.1143917
Tanimura, R., Hiromori, A., Yamaguchi, H., Higashino, T., Umedu, T.: Prediction of deceleration amount of vehicle speed in snowy urban roads using weather information and traffic data. In: 2015 IEEE 18th International Conference on Intelligent Transportation Systems, pp. 2268–2273. IEEE (2015). doi:10.1109/ITSC.2015.366
Turksma, S.: The various uses of floating car data. In: Tenth International Conference on Road Transportation Information and Control, pp. 51–55. IET, London (2000). doi:10.1049/cp:20000103
Xu, X., Su, B., Zhao, X., Xu, Z., Sheng, Q.Z.: Effective traffic flow forecasting using taxi and weather data. In: Li, J., Li, X., Wang, S., Li, J., Sheng, Q.Z. (eds.) ADMA 2016. LNCS, vol. 10086, pp. 507–519. Springer, Cham (2016). doi:10.1007/978-3-319-49586-6_35
Yang, Q., Wu, G., Boriboonsomsin, K., Barth, M.: Arterial roadway travel time distribution estimation and vehicle movement classification using a modified Gaussian mixture model. In: 16th International IEEE Conference on Intelligent Transportation Systems (ITSC 2013), pp. 681–685. IEEE, The Hague (2013). doi:10.1109/ITSC.2013.6728310
Acknowledgment
This work was supported by the CPS-IIP Project in the research promotion program for national-level challenges “Research and development for the realization of next-generation IT platforms” of the Ministry of Education, Culture, Sports, Science and Technology, Japan. Weather observation data were provided by the Japan Meteorological Agency. The authors received some information on winter traffic in Sapporo from specially appointed assistant professor Dr. Hajime Imura at Hokkaido University, Japan.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Kinoshita, A., Takasu, A., Adachi, J. (2017). Latent Variable Model for Weather-Aware Traffic State Analysis. In: Kotzinos, D., Laurent, D., Petit, JM., Spyratos, N., Tanaka, Y. (eds) Information Search, Integration, and Personlization. ISIP 2016. Communications in Computer and Information Science, vol 760. Springer, Cham. https://doi.org/10.1007/978-3-319-68282-2_4
Download citation
DOI: https://doi.org/10.1007/978-3-319-68282-2_4
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-68281-5
Online ISBN: 978-3-319-68282-2
eBook Packages: Computer ScienceComputer Science (R0)