Akıllı Kavşaklarda Trafik Hacmi Tahmini İçin Makine Öğrenmesi Modellerinin Karşılaştırılması

Abstract

Modern şehirlerin artan nüfusu ve hareketliliği, ulaşım sistemlerinin verimli yönetimini gittikçe daha kritik hale getirmiştir. Bu bağlamda, dijital ikiz kavramı, fiziksel dünyanın dijital bir yansımasını oluşturarak karmaşık sistemleri anlamak, yönetmek ve optimize etmek için güçlü bir araç haline gelmiştir. Ulaşım da bu alanlardan biridir ki akıllı kavşaklar gibi teknolojik altyapılar sayesinde büyük miktarda veri toplanmakta ve bu verilerle dijital ikiz mimarisinin önemli bir unsuru olan tahmin modelleri oluşturulmaktadır. Bu çalışma akıllı kavşaklarda trafik hacmi tahmini uygulamalarında daha etkili ve doğru sonuçlar elde etmek için hangi modelin seçilmesi gerektiğine dair akademik bir anlayış sunmayı amaçlamaktadır. Bunun için Antalya’da bulunan seçilmiş iki akıllı kavşağın her bir geliş kolu için saatlik araç sayım verileri kullanılmıştır. Bu veriler eğitim ve test verisi olarak ayrılmış olup Lineer Regresyon, Polinomal Regresyon, Destek Vektör Regresyonu (SVR) ve Rastgele Orman Regresyonu tabanlı makine öğrenmesi modelleri oluşturulmuştur. Böylece akıllı kavşaklarda her bir kol için trafik hacmi tahmini yapılmıştır. Ortalama Mutlak Hata (MAE) ve En Küçük Kareler (R Kare) yöntemleri ile tahmin modellerinin performans karşılaştırması yapılmıştır. Buna göre Rastgele Orman modelinin diğer önerilen modellere göre daha başarılı performans göstermiş olduğu görülmüştür.

Keywords

• Akıllı Kavşak
• Dijital İkiz
• Büyük Veri
• Makine Öğrenmesi
• Trafik Hacmi Tahmini

Comparison of Machine Learning Models For Traffic Volume Estimation at Smart Intesections

Abstract

The population growth and mobility of modern cities have made efficient management of the transportation systems increasingly critical. According to this, the digital twin concept has become a powerful tool for understanding, managing, and optimizing complex systems by creating a digital reflection of the physical world. Transportation is one of these areas, moreover, with the help of such smart intersections and technological infrastructure, large amounts of data are made up of estimation models which is one of the important elements of digital twins. This study aims to provide an academic understanding of which model should be chosen to have accurate and more effective results during the application of traffic value prediction in smart intersections. Hourly vehicle count data from the arrival branches of two smart intersections in Antalya were utilized for this. This data is separated as training and test data created based on learning machine models including Linear Regression Polynomial Regression, Support Vector Regression (SVR), and Random Forest Regression. Thus, traffic volume estimation was conducted for each branch at smart intersections. Forecasting models were evaluated using Mean Absolute Error (MAE) and Least Squares (𝑅) methods. Accordingly, it was seen that the Random Forest model performed better than the other proposed models.

Keywords

• Smart Intersection
• Digital Twin
• Big Data
• Machine Learning
• Traffic Volume Estimation

References

1. [1] İ. Aynacı, “Dijital ikiz ve sağlık uygulamaları,” İzmir Kâtip Çelebi Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi, vol. 3, no. 1, pp. 70–79, 2020.
2. [2] Z. Wang, K. Han and P. Tiwari, “Digital twin-assisted cooperative driving at non-signalized intersections,” IEEE Transactions on Intelligent Vehicles, vol. 7, no. 2, pp. 198–209, Jun. 2022, doi: 10.1109/TIV.2021.3100465.
3. [3] E. Tomanik, A. J. J. Reyes, V. Tomanik and B. Tormos, “Machine-learning-based digital twins for transient vehicle cycles and their potential for predicting fuel consumption,” Vehicles, vol. 5, no. 2, pp. 583–604, May 2023, doi: 10.3390/vehicles5020032.
4. [4] A. Fuller, Z. Fan, C. Day and C. Barlow, “Digital twin: Enabling technologies, challenges and open research,” IEEE access, vol. 8, pp. 108952–108971, May 2020, doi: 10.1109/ACCESS.2020.2998358.
5. [5] T. Y. Fujii, V. T. Hayashi, R. Arakaki, W. V. Ruggiero, R. Bulla, F. H. Hayashi and K. A. Khalil, “A digital twin architecture model applied with MLOps techniques to improve short-term energy consumption prediction,” Machines, vol. 10, no. 1, Jan. 2022, doi: 10.3390/machines10010023.
6. [6] H. Xu, A. Berres, S. B. Yoginath, H. Sorensen, P. J. Nugent, J. Severino, S. A. Tennille, A. Moore, W. Jones and J. Sany, "Smart mobility in the cloud: enabling real-time situational awareness and cyber-physical control through a digital twin for traffic," IEEE Transactions on Intelligent Transportation Systems, vol. 24, no. 3, pp. 3145-3156, March 2023, doi: 10.1109/TITS.2022.3226746.
7. [7] J. Lai, Z. Chen, J. Zhu, W. Ma, L. Gan, S. Xie and G. Li, “Deep learning based traffic prediction method for digital twin network," Cognitive Computation, vol. 15, no. 5, pp. 1748–1766, September 2023, doi: 10.1007/s12559-023-10136-5
8. [8] “Real-time transportation management’’ aimsun.com, [Online]. Available: https://www.aimsun.com/real-time-transportation-management/. [Accessed: Mar. 27, 2023.]

9. [9] S. Janković, A. Uzelac, S. Zdravković, D. Mladenović, S. Mladenović and I. Andrijanić, “Traffıc volumes predıctıon usıng bıg data analytıcs methods,” International Journal for Traffic and Transport Engineering, vol. 2, no. 0, pp. 184–198, 2021, doi: 10.7708/ijtte2021.11(2).01.
10. [10] C. Bratsas, K. Koupidis, J. M. Salanova, K. Giannakopoulos, A. Kaloudis and G. Aifadopoulou, “A comparison of machine learning methods for the prediction of traffic speed in urban places,” Sustainability (Switzerland), vol. 12, no. 1, Jan. 2020, doi: 10.3390/SU12010142.
11. [11] F. M. N. Ali and A. A. M. Hamed, “Usage apriori and clustering algorithms in WEKA tools to mining dataset of traffic accidents,” Journal of Information and Telecommunication, vol. 2, no. 3, pp. 231–245, Jul. 2018, doi: 10.1080/24751839.2018.1448205.
12. [12] A. N. Espinoza, O. R. L. Bonilla, E. E. G. Guerrero, E. T. Cuautle, D. L. Mancilla, C. H. Mejia and E. I. Gonzalez, “Traffic flow prediction for smart traffic lights using machine learning algorithms,” Technologies (Basel), vol. 10, no. 1, p. 5, January 2022. https://doi.org/10.3390/technologies10010005
13. [13] Y. Hou, Z. Deng and H. Cui, “Short-term traffic flow prediction with weather conditions: based on deep learning algorithms and data fusion,” Complexity, vol. 2021, doi: 10.1155/2021/6662959.
14. [14] W. Alajali, W. Zhou and S. Wen, “Traffic flow prediction for road intersection safety,” in 2018 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI), 812–820. October 2018, Guangzhou, China. Available: IEEE Xplore, https://ieeexplore.ieee.org/document/8560129 [Accessed: 20 Sept. 2023].
15. [15] “Strateji ve Eylem Planları - Haberleşme Genel Müdürlüğü,” uab.gov.tr, Available: https://hgm.uab.gov.tr/strateji-eylem-planlari. [Accessed: Sept. 15, 2023].
16. [16] “MAtchUP Antalya - Antalya Büyükşehir Belediyesi,” matchupantalya.org, [Online]. Available: http://www.matchupantalya.org/ [Accessed: Nov. 1, 2023].
17. [17] A. Kemeç and H. Gül, “Antalya Büyükşehir Belediyesi örneğinde akilli kent uygulamalari,” Kamu Yönetimi Ve Politikaları Dergisi, 2(3), pp. 355-382.
18. [18] M. Akın and Ş. Sağıroğlu, “Büyük veri perspektifinden trafik tahmini traffic prediction based on big data perspective,” in 2019 1st International Informatics and Software Engineering Conference (UBMYK), 2019, Ankara, Turkey, Available: IEEE Xplore https://ieeexplore.ieee.org/document/8965545. [Accessed: 02 Dec. 2023].
19. [19] A. C. Müller and S. Guido, Introduction to machine learning with Python, O'Reilly Media, 2022.
20. [20] J. Brownlee, Deep learning with Python: develop deep learning models on Theano and TensorFlow using Keras, Machine Learning Mastery, 2016.
21. [21] A. Kowalczyk, Support vector machines succinctly,” Syncfusion Inc, 2017.
22. [22] A. Boukerche and J. Wang, “Machine learning-based traffic prediction models for intelligent transportation systems,” Computer Networks, vol. 181, Nov. 2020. doi: 10.1016/j.comnet.2020.107530.
23. [23] “R Kare ve Düzeltilmiş R Kare - Veri bilimi okulu,” veribilimiokulu.com, Available: https://www.veribilimiokulu.com/r-kare-ve-duzeltilmis-r-kare/, [Accessed: February. 16, 2023].
24. [24] “Regresyon analizi,” wikipedia.org, Available: https://tr.wikipedia.org/wiki/Regresyon_analizi, [Accessed: May. 26, 2023].