UDC: 656.13
https://doi.org/10.25198/2077-7175-2025-3-107


viewing

DEVELOPMENT OF AN AUTOREGRESSIVE NEURAL NETWORK MODEL FOR PREDICTING ACCIDENTS IN THE KHABAROVSK TERRITORY

E. V. Shokhirev1, P. P. Volodkin2, V. A. Lazarev3
Pacific National University, Khabarovsk, Russia
1 e-mail: shohirev_kna@mail.ru
2 e-mail: 004167@togudv.ru
3 e-mail: 000136@togudv.ru

A. V. Konstantinov
Mining Institute of the Far Eastern Branch of the Russian Academy of Sciences, Khabarovsk, Russia
e-mail: alex-sdt@yandex.ru

Abstract. The relevance of this study is justified by the fact that the problem of accidents on highways is a serious social and economic problem, one of the possible solutions to which is the prevention and prevention of road accidents. The basis for the prevention and subsequent reduction of accidents on the roads can be the analysis of previous road accidents, the results of which can be used later, with the use of modern technology as a basis for modeling of accidents and the possibility of preventing them in the future. The solution to this problem can be the creation of a prediction model based on statistical data with the use of machine learning in the analysis of road accidents, which will be effective in processing data and making informed decisions to improve road safety and can significantly improve the accuracy of forecasts, which can then be aimed at improving the level of safety. The purpose of the study is to compare the actual accident rates in Khabarovsk Krai for the period from 01.01.2015 to 30.11.2023 with the results of the accident prediction model by training and validation based on recurrent neural network using machine learning methods.

The study applies scientific methods of statistical modeling of time series, analysis of the feature space of data. A neural network of recurrent type was used for modeling and subsequent forecasting, using an autoregressive approach.

Scientific novelty in the form of development of autoregressive neural network model of accident rate forecasting to identify dependencies and patterns of data, which can improve the quality of accident rate forecasting, considering also the possibility of processing a large amount of data with information about the factors affecting the accident rate.

Further research will be aimed at improving the forecasting model, using more information in the input data.

Scientific innovation of the presented research is expressed in the form of developed autoregressive neural network model of accident rate forecasting to identify dependencies and patterns of data, which will improve the quality of accident rate forecasting, taking into account the possibility of processing a large amount of data with information about the factors affecting the accident rate. The prediction model presented in the study can be adapted in the future to predict the number and types of crashes, participants and casualties.

The application of the results of the study aims to improve the quality of road accident prevention and crash avoidance using these prediction models.

Key words: road accidents, accident analysis, traffic safety, neural network, modeling, accident prediction.

Cite as: Shokhirev, E. V., Volodkin, P. P., Lazarev, V. A., Konstantinov, A. V. (2025) [Development of an autoregressive neural network model for predicting accidents in the Khabarovsk Territory]. Intellekt. Innovacii. Investicii [Intellect. Innovations. Investments]. Vol. 3, pp. 107–120.– https://doi.org/10.25198/2077-7175-2025-3-107.


References

  1. Velikhov, P. (2016) [Machine Learning for Natural Language Understanding]. Otkrytyye sistemy SUBD [Open DBMS Systems]. Vol. 1, pp. 18–21. (In Russ.).
  2. Gushchina, P. V., Shakirzyanova, I. F., Zyurina, O. A. (2024) [Intellectual transportation systems: elements and advantages]. Logistika – Yevraziyskiy most. Materialy XIX Mezhdunarodnoy nauchno-prakticheskoy konferentsii [Logistics – Eurasian Bridge. Materials of XIX International Scientific and Practical Conference]. Krasnoyarsk, pp. 70– 75. (In Russ.).
  3. Markovnina, A. I., et al. (2020) [Road traffic injuries. Harm caused during road traffic accidents]. Budushcheye tekhnicheskoy nauki. Sbornik materialov XIX Vserossiyskoy molodezhnoy nauchno-tekhnicheskoy konferentsii [Future of technical science. Proceedings of the XIX All-Russian Youth Scientific and Technical Conference]. Nizhny Novgorod, October 08, pp. 190–191. (In Russ.).
  4. Zvyagin, L. S. (2022) [Peculiarities of modeling in different types of neural networks]. Myagkiye izmereniya i vychisleniya [Soft Measurements and Computations]. Vol. 52. No. 3, pp. 43–54. – https://doi.org/10.36871/2618-9976.2022.03.005. (In Russ.).
  5. Kantorovich, G. G. (2002) [Analysis of time series] Ekonomicheskiy zhurnal Vysshey shkoly ekonomiki [Economic Journal of the Higher School of Economics]. Vol. 6, No. 2, pp. 251–273. (In Russ.).
  6. Kushchenko, L. Е. (2023) [Prediction of traffic accidents as one of the ways to reduce mortality]. Mir transporta i tekhnologicheskikh mashin [World of transport and technological machines]. Vol. 2(81), pp. 67–73. – https://doi.org/10.33979/2073-7432-2023-2(81)-67-73. (In Russ.).
  7. Bobintsev, D. O. (2021) [Generalized scheme of data analysis and forecasting of emergency situations on highways]. Yestestvennyye i tekhnicheskiye nauki [Natural and Technical Sciences]. Vol. 12, pp. 285–288. (In Russ.).
  8. Lvovich, I. Y. (2020) [Features of machine learning methods]. Informatsionnyye tekhnologii [Information technologies]. Vol. 26. No. 9, pp. 499–506. – https://doi.org/10.17587/it.26.499-506. (In Russ.).
  9. Rybin, A. L. (2020) [Paradigms of road traffic]. Avtomobil’nyye dorogi [Automobile roads]. Vol. 1069, pp. 133– 137. (In Russ.).
  10. Shpaltakov, V. P. (2021) [Theoretical foundations of the economy of the transport system]. Innovatsionnaya ekonomika i obshchestvo [Innovative economy and society]. Vol. 1 (27), pp. 2–14. (In Russ.).
  11. Alvarez-Rodriguez, U., et al. (2021) Evolutionary dynamics of higher-order interactions in social networks. Nature Human Behaviour, Vol. 5, No. 5, pp. 586–595. – https://doi.org/10.1038/s41562-020-01024-1. (In Eng.).
  12. Franco Basso, et al. (2021) A deep learning approach for real-time crash prediction using vehicle-by-vehicle data. Accident Analysis & Prevention, Vol. 162. – https://doi.org/10.1016/j.aap.2021.106409. (In Eng.).
  13. Lastrico, R., et al. (2024) Estimation of Kidney’s Blood Vessels Deformations for Robot-Assisted Surgery. In: Lee, SG., An, J., Chong, N.Y., Strand, M., Kim, J.H. (eds) Intelligent Autonomous Systems 18. IAS 2023. Lecture Notes in Networks and Systems, Vol. 794. Springer, Cham. – https://doi.org/10.1007/978-3-031-44981-9_35. (In Eng.).
  14. Li, P., Abdel-Aty, M., Yuan, J. (2020) Real-Time Crash Risk Prediction on Arterials Based on LSTM-CNN. Accident Analysis & Prevention, Vol. 135. – https://doi.org/10.1016/j.aap.2019.105371. (In Eng.).
  15. Liu, Q., et al. (2024) Multi-task Support Vector Machine Classifier with Generalized Huber Loss. Journal of Classification – https://doi.org/10.1007/s00357-024-09488-w. (In Eng.).
Error