Makine Öğrenmesi Yöntemleri ile Meteorolojik Parametrelere Dayalı Yol Görünürlüğü Tahmini

Öz

Trafik güvenliğini sağlamada en önemli parametrelerden birisi yol görünürlüğüdür. Yol görünürlüğü yolun geçtiği bölgedeki iklim koşullarının yanı sıra, yolun geometric tasarımına ve aydınlatma koşullarına bağlıdır. Görünürlük, sıcaklık, nem, rüzgar hızı, basınç, sis, yağış tipi gibi meteorolojik parametrelere bağlıdır. Bu çalışmada, trafik güvenliğini sağlamak için yol görünürlük tahminlemesi yapılmıştır. Yol görünürlük tahmini için Makine öğrenme metotları kullanılmıştır. Makine öğrenme metotları Random Forest, Extra Tree ve Gradient Boosting yöntemleri ile geliştirilmiştir. Modelde, 2006-2016 yılları arasında Macaristan’ın Szeged şehrindeki sıcaklık, nem, rüzgar hızı, basınç, yağış tipi, görünürlük gibi 96453 meteorolojik data seti kullamılmıştır. Geliştirilen modeller belirtme katsayısı (R2) ve Karesel ortalama hata (KOH) ile değerlendirilmiştir. Değerlendirme sonucunda random forest methodu en iyi sonucu vermiştir.

Anahtar Kelimeler

• Yol güvenliği
• görünürlük
• makine öğrenmesi
• akıllı ulaşım sistemleri
• meteorolojik parametreler

Prediction of Road Visibility Based on Meteorological Parameters by Machine Learning Methods

Abstract

One of the important parameters in ensuring traffic safety is road visibility. Road visibility depends on the geometric design of the road, lighting conditions, as well as the climatic conditions in the area where the road passes. Visibility depends on meteorological parameters such as temperature, humidity, wind speed, pressure, fog, precipitation type. In this study, it is aimed to predict road visibility to ensure traffic safety. Machine learning methods were used for road visibility estimation. Machine learning models were developed with Random Forest, Extra Tree and Gradient Boosting methods. In the models, 96453 meteorological data sets such as temperature, humidity, wind speed, pressure, precipitation types, visibility were used between 2006 and 2016 in Szeged, Hungary. Developed models were evaluated with coefficient of determination (R2) and Root mean squared error (RMSE). As a result of the evaluation, the random forest method gave the best result.

Keywords

• road safety
• visibility
• machine learning
• intelligent transport system
• meteorological parameters

Kaynakça

1. Babari, R., Hautière, N., Dumont, É., Paparoditis, N., & Misener, J. (2012). Visibility monitoring using conventional roadside cameras–Emerging applications. Transportation research part C: emerging technologies, 22, 17-28.
2. Chaabani, H., Werghi, N., Kamoun, F., Taha, B., & Outay, F. (2018). Estimating meteorological visibility range under foggy weather conditions: A deep learning approach. Procedia Computer Science, 141, 478-483.
3. Cornejo-Bueno, S., Casillas-Pérez, D., Cornejo-Bueno, L., Chidean, M. I., Caamaño, A. J., Cerro-Prada, E., ... & Salcedo-Sanz, S. (2021). Statistical Analysis and Machine Learning Prediction of Fog-Caused Low-Visibility Events at A-8 Motor-Road in Spain. Atmosphere, 12(6), 679.
4. Friedman, J. H. (2001). Greedy function approximation: a gradient boosting machine. Annals of statistics, 1189-1232.
5. Geurts, P., Ernst, D., & Wehenkel, L. (2006). Extremely randomized trees. Machine learning, 63(1), 3-42.
6. Hammed, M. M., AlOmar, M. K., Khaleel, F., & Al-Ansari, N. (2021). An Extra Tree Regression Model for Discharge Coefficient Prediction: Novel, Practical Applications in the Hydraulic Sector and Future Research Directions. Mathematical Problems in Engineering, 2021.
7. Ho, T. K. (1995, August). Random decision forests. In Proceedings of 3rd international conference on document analysis and recognition (Vol. 1, pp. 278-282). IEEE.15 John, V., Liu, Z., Guo, C., Mita, S., & Kidono, K. (2015, November). Real-time lane estimation using deep features and extra trees regression. In Image and Video Technology (pp. 721-733). Springer, Cham.
8. Jonnalagadda, J., & Hashemi, M. (2020, August). Forecasting atmospheric visibility using auto regressive recurrent neural network. In 2020 IEEE 21st International Conference on Information Reuse and Integration for Data Science (IRI) (pp. 209-215). IEEE.

9. Kwon, T. M. (2004). Atmospheric visibility measurements using video cameras: Relative visibility.
10. Lakshmi, C. R., Rao, D. T., & Rao, G. S. (2017, September). Fog detection and visibility enhancement under partial machine learning approach. In 2017 IEEE International Conference on Power, Control, Signals and Instrumentation Engineering (ICPCSI) (pp. 1192-1194). IEEE.
11. Negru, M., & Nedevschi, S. (2013, September). Image based fog detection and visibility estimation for driving assistance systems. In 2013 IEEE 9th International Conference on Intelligent Computer Communication and Processing (ICCP) (pp. 163-168). IEEE.
12. Ortega, L., Otero, L. D., & Otero, C. (2019, April). Application of machine learning algorithms for visibility classification. In 2019 IEEE International Systems Conference (SysCon) (pp. 1-5). IEEE.
13. U.S. Department of TransportationFederal Highway Administration, (2022). Low visibility. Available: https://ops.fhwa.dot.gov/weather/weather_events/low_visibility.htm
14. URL-1, (2022). kaggle website. [online]. Available. https://www.kaggle.com/budincsevity/szeged-weather
15. Uyanık, T., Karatuğ, Ç., & Arslanoğlu, Y. (2021). Machine learning based visibility estimation to ensure safer navigation in strait of Istanbul. Applied Ocean Research, 112, 102693.
16. Yang, L., Muresan, R., Al-Dweik, A., & Hadjileontiadis, L. J. (2018). Image-based visibility estimation algorithm for intelligent transportation systems. IEEE Access, 6, 76728-76740.
17. Yufeng, W., Du Jiamin, Y. Z., Yuehui, S., & Dengxin, H. (2022). Atmospheric visibility prediction by using the DBN deep learning model and principal component analysis.
18. Zhao, J. I. N., Kangjun, Q. I. U., & Miaomiao, Z. H. A. N. G. (2021). Investigation of Visibility Estimation Based on BP Neural Network. Journal of Atmospheric and Environmental Optics, 16(5), 415.