Adapted neural network of information support subsystem

S. Semenov, О. Lipchanska, M. Lipchanskyi
Science and Technology of the Air Forces of the Forward Forces of Ukraine.— 2019. — № 1(34). – P. 102-106.
Topic of the article: Development of radio engineering, as well as air force communication
UDK 004.415:621.39:510.589
Article language: english
Annotations languages:


Annotation: Safety of human life, the safety of his material values are main priorities in modern society. Objects of critical infrastructure are in a special risk zone. Accident statistics for them has remained high in recent years. Increased risk and a large number of incidents, including abroad, emphasize the relevance of this problem. An adapted neural network has been proposed for monitoring the situation at a railway crossing and informing the train driver of information about unexpected obstacles through the subsystem of information support in order to reduce the likelihood of an accident or reduce the severity of its consequences. Images from a railway crossing video surveillance camera are obtained. The results of neural network training and modeling using image data are given.

References

1. The official site of SAI.GOV.UA (2018), “Statystyka avariynosti v Ukrayini” [Accident Statistics in Ukraine], available at: www.sai.gov.ua/ua/ua/static/21.htm (accessed 27 December 2018).
2. The official site of EC.EUROPA.EU (2018), Rail accident fatalities in the EU, available at: https://ec.europa.eu/eurostat/statistics-explained/index.php/Rail_accident_fatalities_in_the_EU (accessed 27 December 2018).
3. The official site of ORR.GOV.UK (2018), Rail Safety Statistics 2017-18 Annual Statistical Release, available at: www.orr.gov.uk/_data/assets/pdf_file/0016/39103/rail-safety-statistics-2017-18.pdf (accessed 27 December 2018).
4. The official site of STATISTA.COM (2018), Number of rail accidents and incidents in the United States from 2013 to 2017, available at: https://www.statista.com/statistics/204569/rail-accidents-in-the-us (accessed 27 December 2018).
5. The official site of BUDPORT.COM.UA (2018), “V Ukraine khotyat umen'shit' kolichestvo zheleznodorozhnykh per-eyezdov” [In Ukraine, they want to reduce the number of railway crossings], available at: www.budport.com.ua/news/9813-v-ukraine-hotyat-umenshit-kolichestvo-zh-d-pereezdov (accessed 27 December 2018).
6. The official site of GLOBALRAILWAYREVIEW.COM (2018), The rail sector’s efforts in improving level crossing safety, available at: https://www.globalrailwayreview.com/article/73517/level-crossings-improving-safety (accessed 27 December 2018).
7. The official site of UNECE.ORG (2018), Low cost solutions to improve safety at level crossings in Hungary, available at: https://www.unece.org/fileadmin/DAM/trans/doc/2014/wp1/ECE-WP1-GE1-2014-9e.pdf (accessed 27 December 2018).
8. Douglas, L. Reilly (2010), A Neural Network Video Sensor Application for Rail Crossing Safety, IDEA Program, pp. 27-33, available at: www.onlinepubs.trb.org/onlinepubs/archive/studies/idea/finalreports/highspeedrail/hsr-10final_report.pdf (accessed 27 December 2018).
9. Tao, Ye, Baocheng, Wang, Ping, Song and Juan, Li (2018), Automatic Railway Traffic Object Detection System Using Feature Fusion Refine Neural Network under Shunting Mode, Sensors, No. 18(1916), pp. 1-19, available at: https://www.researchgate.net/publication/325730416_Automatic_Railway_Traffic_Object_Detection_System_Using_Feature_Fusion_Refine_Neural_Network_under_Shunting_Mode/ fulltext/5b20963d0f7e9b0e373efd54/325730416_Automatic_Railway_Traffic_Object_Detection_System_Using_Feature_Fusion_Refine_Neural_Network_under_Shunting_Mode.pdf (accessed 27 December 2018).
10. LeCun, Y. and Bengio, Y. (1995), Convolutional Networks for Images, Speech, and Time-Series, The Handbook of Brain Theory and Neural Networks, MIT Press, 14 p.
11. Solla, S. and LeCun, Y. (1991), Constrained Neural Networks for Pattern Recognition, Neural Networks: Concepts, Applications and Implementations, Vol IV, Prentice Hall, pp. 142-161.
12. Nair, V. and Hinton, G.E. (2010), Rectified Linear Units Improve Restricted Boltzmann Machines, Proceedings of the 27Th International Conference on Machine Learning, Haifa, Israel, pp. 1-8.
13. Goodfellow, I., Bengio, Y. and Courville, A. (2017), Deep Learning, 66 p., available at: www.deeplearningbook.org/front_matter.pdf (accessed 27 December 2018).
14. Nielsen, M. (2017), Neural Networks and Deep Learning, available at: www.neuralnetworksanddeeplearning.com/chap2.html (accessed 27 December 2018).

Reference:
 Semenov, S.H., Lipchanska, O.V. and Lipchanskyi, M.V. (2019), Adapted neural network of information support subsystem, Science and Technology of the Air Force of Ukraine, No. 1(34), pp. 102-106. https://doi.org/10.30748/nitps.2019.34.14.