Neural network solution of the operational planning task for unmanned aerial vehicles route flight and time setting for ground based objects observation employing the fuzzy logic while displaying these results on the computer screen prior to the start

System analysis, control and data processing


Аuthors

Ivashova N. D.1*, Mikhailin D. A.2**, Chernyakova M. E.3***, Shanygin S. V.4****

1. State Research Institute Engineeringpace University, 125, prospekt Mira, Moscow, 129226, Russia
2. Main research and testing center for robotics Ministry of Defense of the Russian , 5, ul. Seregina, Moscow, 125167, Russia
3. Moscow Aviation Institute (National Research University), 4, Volokolamskoe shosse, Moscow, А-80, GSP-3, 125993, Russia
4. Bauman Moscow State Technical University, MSTU, 5, bldg. 1, 2-nd Baumanskaya str., Moscow, 105005, Russia

*e-mail: nati2405@mail.ru
**e-mail: tau_301@mail.ru
***e-mail: kaf301@mai.ru
****e-mail: sg78dec@mail.ru

Abstract

The article considers the possibility of neural network realization of operational planning of the route flight of the unmanned aerial vehicles (UAVs) while ground based objects observation, when the flight task requires correction due to the change of dynamic situation. It happens when either the observation quality of certain object appeared too low, or video information indicates the new objects occurrence, or observation of a number of of objects was untimely.

The number of ground based objects, coordinates of their position and their relative importance as well as the number of UAVs and their current coordinates at the given planning step are regarded as set in the formulated problem statement. The article demonstrates that the operation planning at each step consists of two operations. The first operation consists in determining the initial set of the most important objects of observation, and the second is assigning for each of them their “own” UAV.

The authors suggest the problem solution with the three layer neural networks of sequential distribution with sigmoidal activation function in the first two layers and relay function in the last layer. The article demonstrates that these neural networks training required a small number of examples. Employing special schemes for determining the number of the primary target and the number of the servicing UAV, the onboard program neural network complex was formed. This complex is capable of realizing the process of operational planning in the real time mode.

At the same time, the objects significance estimation representation in the form of the multiplicative convolution of partial criteria reduces the input signals number, and, hence, the neural network structure dimensionality. The structural diagram of the fuzzy logic expert system for the allowed search time determining, and detecting each of the objects of observation with account for the three factors: the growing risk of detection failure at the searching time increase; significant fuel consumption excess, as well as the time reduction to the flight termination.

The adaptive fuzzy logic expert system, determining the moment of observation termination depending on the current parameters of the dynamic situation, was formed to automate the process of continuing managing servicing of the next object of observation, while searching for it in the proper place.

Keywords:

operational planning, unmanned aerial vehicle, route flight, neural networks, fuzzy logic

References

  1. Marc D. Richards, Darrell Whitley, J. Ross Beveridge. Evolving cooperative strategies for UAV teams, GECCO 2005, Washington, D.C. USA. pp. 1 – 8.

  2. He P., Dai S. Stealth Real-time Paths Planning for Heterogeneous UAV Formation Based on Parallel Niche Genetic Algorithm, Journal of Computational Information Systems, 2014, no. 10(15), pp. 6731 – 6740.

  3. Lebedev G., Goncharenko V., Mikhaylin D., Rumakina A. Aircraft group coordinated flight route optimization using branch-and-bound procedure in resolving the problem of environmental monitoring, ITM Web of Conferences 10, 01003 (2017), Seminar on Systems Analysis, 2017, vol. 10, pp. 1 – 3.

  4. Mikhailin D.A., Allilueva N.V., Rudenko E.M. Trudy MAI, 2018, no. 98, available al: http://trudymai.ru/eng/published.php?ID=90386

  5. Gorban’ A.N. Obuchenie neironnykh setei (Training the neural networks), Moscow, ParaGraf, 1991, 160 p.

  6. Gorban’ A.N., Dunin-Barkovskii V.L., Kardin A.N. et al. Neiroinformatika (Neuroinformatics), Novosibirsk, Nauka, 1998, 321 p.

  7. Uosserman F. Neirokomp’yuternaya tekhnika: Teoriya i praktika (Neurocomputer technology: Theory and practice), Moscow, Mir, 1992, 184 p.

  8. Lebedev G.N., Mirzoyan L.A. Aviakosmicheskoe priborostroenie, 2005, no. 12, pp. 51 – 57. 9. Lebedev G.N., Rumakina A.V. Aviakosmicheskoe priborostroenie, 2014, no. 5, pp. 3 – 8.

  9. Kuzin A.V., Kurmakov D.V., Luk’yanov A.V., Mikhailin D.A. Trudy MAI, 2013, no. 70, available at: http://trudymai.ru/eng/published.php?ID=44540

  10. Mikhailin D.A. Nauchnyi vestnik MGTU GA, 2017, vol. 20, no. 4, pp. 18 – 24.

  11. Lebedev G.N., Mirzoyan L.A., Efimov A.V. Mekhatronika, avtomatizatsiya, upravlenie, 2009, no. 11, pp. 60 – 65.

  12. Lebedev G.N., Malygin V.B. Nauchnyi vestnik MGTU GA, 2015, no. 221, pp. 138 – 143.

  13. Lebedev G.N., Goncharenko V.I., Rumakina A.V. Neirokomp’yutery: razrabotka, primenenie, 2017, no. 7, pp. 49 – 57.

  14. Widrow B., Hoff M.E. Adaptive switching circuits, 1960 IRE WESCON Convention Record, New York IRE, 1960, pp. 96 – 104.

  15. Widrow B., Sterns S.D. Adaptive Signal Processing, New York, Prentice Hall, 1985, 247 p.

  16. Haykin S. Neural Networks: A Comprehensive Foundation, New York, Prentice Hall, 1998, 842 p.

  17. Christopher M. Bishop. Neural Networks for Pattern Recognition, Oxford Univ. Press, 1995, 498 p.

  18. Swingler K. Applying Neural Networks. A Practical Guide, Academic Press, 1996, 303 p.

  19. Brian D. Ripley. Pattern recognition and Neural Networks, Cambridge University Press, 1996, 416 p.

  20. Zadeh S.M., Powers D., Sammut K. Optimal Route Planning with Prioritized Task Scheduling for AUV Missions Article, University, Adelaide, SA 5042, Australia, 2016. pp. 1 – 8.

  21. Stefanov V.A., Fedunov B.E. Bortovye operativno-sovetuyushchie ekspertnye sistemy (BOSES) tipovykh situatsii funktsionirovaniya antropotsentricheskikh (tekhnicheskikh) ob"ektov (Onboard operational advisory expert systems for typical situations of the anthropocentric (technical) facilities functioning), Moscow, Izd-vo MAI, 2006, 191 p.

  22. Podlip’yan P.E., Maksimov A.N. Trudy MAI, 2011, no. 43, available at: http://trudymai.ru/eng/published.php?ID=24769


Download

mai.ru — informational site MAI

Copyright © 2000-2024 by MAI

 Вход