Russian Federation
Russian Federation
Russian Federation
UDK 004.93 Распознавание и преобразование образов
the article analyzes the research and works on processing Sensor data processing and navigation for mobile objects, including those with automatic control. automatic control. It is pointed out the necessity of integration of solutions based on vision and neural networks, taking into account the state of the environment. on the basis of vision and neural networks, taking into account the state of the environment. environment. The description of the algorithm and methods for complexing data received from sensor sensors of the onboard vision system of the rolling stock. Within the framework of the algorithm work the trajectory data\ processing: confirmation (removal) of the observed objects, construction of their trajectories, estimation of velocities and coordinates, formation of the list of global objects. To calculate the object motion model, an extended Kalman filter is applied Kalman filter, Hungarian algorithm and calculation of the Mahalanobis distance. Detailed The stages of detection, tracking and identification of obstacle objects are described in detail, as well as determination of their parameters (class, coordinates and velocity). The Experimental indicators used in this study. Relevance relevance and shown practical applicability of the described approach for the tasks of automatic control of rolling stock in the conditions of a digital railroad
vision system, sensor fusion, object identification, object tracking, covariance matrix, clustering, extended Kalman filter, Hungarian algorithm, Mahalanobis distance
1. Rozenberg E. N., Dzyuba Yu. V., Batraev V. V. O napravleniyah razvitiya cifrovoy zheleznoy dorogi // Avtomatika, svyaz', informatika. 2018. № 1. S. 9–13.
2. Ohotnikov A. L., Popov P. A. Bespilotnoe upravlenie lokomotivom: vchera, segodnya i zavtra // Avtomatika, svyaz', informatika. 2019. № 8. S. 12–17. DOI:https://doi.org/10.34649/AT.2019.8.8.002
3. Otte M. W. A Survey of Machine Learning Approaches to Robotic Path-Planning. University of Colorado at Boulder: Boulder, CO, USA, 2009.
4. Dulac-Arnold G., Mankowitz D., Hester T. Challenges of Real-World Reinforcement Learning // arXiv. 2019. URL: arXiv:1904.12901
5. Zhao Y., Zhang Y., Wang S. A Review of Mobile Robot Path Planning Based on Deep Reinforcement Learning Algorithm // J. Phys. Conf. Ser. 2021. Vol. 2138. P. 012011.
6. Deshpande S., Kashyap A. K., Patle B. K. A review on path planning AI techniques for mobile robots // Robot. Syst. Appl. 2023. Vol. 3. P. 27–46.
7. Janji S., Kliks A. Neural Networks for Path Planning // arXiv. 2022. URL: arXiv:2207.00874
8. Bharadwaj H., Kumar E. V. Comparative study of neural networks in path planning for catering robots // Procedia Comput. Sci. 2018. Vol. 133. P. 417–423.
9. Obstacle detection in a field environment based on a convolutional neural network security / T. Li [et al.] // Enterp. Inf. Syst. 2022. Vol. 16. P. 472–493.
10. Nowakowski M., Kurylo J. Usability of Perception Sensors to Determine the Obstacles of Unmanned Ground Vehicles Operating in Off-Road Environments // Appl. Sci. 2023. Vol. 13. P. 4892.
11. Deep Multi-modal Object Detection for Autonomous Driving / A. Ennajar [et al.] // Proceedings of the 2021 18th International Multi-Conference on Systems, Signals & Devices (SSD) (Monastir, Tunisia, 22–25 March 2021). P. 10.
12. Reinoso O., Payá L. Special Issue on Visual Sensors // Sensors. 2020. Vol. 20. P. 910.
13. 3D Recognition Based on Sensor Modalities for Robotic Systems: A Survey / S. Manzoor [et al.] // Sensors. 2021. Vol. 21. P. 7120.
14. Xie D., Xu Y., Wang R. Obstacle detection and tracking method for autonomous vehicl based on threedimensional LiDAR // Int. J. Adv. Robot. Syst. 2019. Vol. 16. P. 1729881419831587.
15. Sravnenie ocenok kachestva kompleksirovaniya raznospektral'nyh izobrazheniy i algoritmy uluchsheniya kachestva nizkokontrastnyh izobrazheniy dlya sistem videonablyudeniya / S. N. Kirillov [i dr.] // Cifrovaya obrabotka signalov i ee primenenie — DSPA‑2019: doklady 21‑y Mezhdunarodnoy konferencii. 2019. S. 403–408.
16. Zhuravlev Yu. I., Ryazanov V. V., Sen'ko O. V. «Raspoznavanie». Matematicheskie metody. Programmnaya sistema. Prakticheskie primeneniya. M.: FAZIS, 2006. 176 s.
17. Bar-Shalom Y., Li X.-R. Multitarget-multisensor tracking: principles and techniques. Vol. 19. Storrs, CT: YBs, 1995.
18. Eddous M., Stensfild R. Metody prinyatiya resheniy / per. s angl. pod red. chlen-korr. RAN I. I. Eliseevoy. M.: Audit, YuNITI, 1997. 590 s.
19. Ohotnikov A. L. Situacionnoe bespilotnoe upravlenie // ITNOU: Informacionnye tehnologii v nauke, obrazovanii i upravlenii. 2019. № 3(13). S. 96–107.
20. Gizzatov A. S., Ivanov V. F. Privyazka koordinat podvizhnogo sostava k cifrovoy modeli puti // Navigaciya i upravlenie dvizheniem: materialy HXV Yubileynoy konferencii molodyh uchenyh (Sankt-Peterburg, 21–24 marta 2023 goda). SPb.: Koncern «Central'nyy nauchno-issledovatel'skiy institut «Elektropribor», 2023. S. 32–34.
21. ISO 8855:2011 — Road vehicles — Vehicle dynamics and road-holding ability — Vocabulary. URL:/ https://www.iso.org/standard/51180.html
22. Sposob avtomaticheskoy kalibrovki datchikov mashinnogo zreniya rel'sovogo transportnogo sredstva: patent № 2811766 C1 Rossiyskaya Federaciya, MPK G06T 7/80 / I. A. Deylid [i dr.]; zayavl. 20.04.2023; opubl. 17.01.2024; zayavitel' OAO «RZhD».
23. Ustroystvo kompleksirovaniya dannyh ot obnaruzhennyh prepyatstviy dlya skorostnogo elektropoezda: patent № 042050 / V. F. Ivanov [i dr.]; zayavl. 04.02.2021; opubl. 29.12.2022; zayavitel' OAO «RZhD».
