P.A. Panilov1, T.Yu. Tsibizova2, N.A. Malakhov3

1–3 Bauman Moscow State Technical University (National Research-Tel University) (Moscow, Russia)
1 panilovp.a@bmstu.ru, 2 mumc@bmstu.ru, 3 malahovna@bmstu.ru

To solve the problems of ensuring the reliability and safety of autonomous unmanned aerial vehicles (UAVs), it is necessary to develop cognitive control models that allow us to consider autonomous UAVs as mental subjects capable of perceiving the environment, making decisions and acting on the basis of this information.

Cognitive modeling is a technique that is used to describe and understand the cognitive processes that take place in the human brain. In the context of autonomous UAVs, cognitive modeling of the UAV control system allows us to imagine the UAV as a mental entity that has perception, memory, thinking and decision making. These processes are based on the processing of information coming from the UAV sensors and on decision-making algorithms that determine the behavior of the UAV.

The main approaches to the creation of cognitive models of UAV control are proposed: 1) rules and knowledge; 2) machine learning; 3) hybrid approach.

A cognitive model of an autonomous UAV control system is considered, which consists of several modules: 1) perception module; 2) memory module; 3) module of thinking; 4) module of action. The graph of the cognitive model is presented.

The ontology of the UAV autonomous control system was designed, the concepts, their interactions, properties and rules were defined. An ontology is a formal specification of a conceptual model that describes concepts and relationships between them in a particular field of knowledge. The ontology of the UAV autonomous control system defines the concepts and relationships between them that are used in the cognitive model.

Based on this ontology and the proposed cognitive model, a neural network architecture was developed for solving the UAV control problem, consisting of several layers, each of which is responsible for a specific module of the cognitive model. A neural network architecture graph for an autonomous UAV control system has been constructed.

Thus, the developed cognitive model, ontology and architecture of the neural network are important elements in the development of autonomous UAV control systems. The joint use of the cognitive model, ontology and neural network architecture makes it possible to create more efficient and reliable autonomous UAV control systems. They provide a more accurate perception of the environment, faster and more accurate decision-making and more precise control of the actions of the UAV.

Panilov P.A., Tsibizova T.Yu., Malakhov N.A. Cognitive modeling of autonomous control system of unmanned aerial vehicle. Dynamics of complex systems. 2023. V. 17. № 4. P. 5−11. DOI: 10.18127/j19997493-202304-01 (in Russian)

1. Babichenko A.V., Proletarskaya V.A., Selezneva M.S., Neusypin K.A., Proletarskij A.V. Intellektual'nye sistemy upravleniya bespilotnymi letatel'nymi apparatami. Inzhenernaya fizika. 2018. № 5. S. 94–102 (in Russian).
2. Neusypin K.A., Selezneva M.S. Izmeritel'nyj kompleks s intellektual'noj komponentoj dlya letatel'nogo apparata. Avtomatizaciya. Sovremennye tekhnologii. 2016. № 9. S. 27–30 (in Russian).
3. Cibizova T.Yu., P'o S., Selezneva M.S. Matematicheskoe modelirovanie dinamicheskih sistem s ispol'zovaniem parametricheskoj identificiruemosti. Sovremennye naukoemkie tekhnologii. 2018. № 1. S. 54–60 (in Russian).
4. Vostryh A.V. Model' opisaniya elementov informacionnyh sistem, orientirovannyh na cheloveko-mashinnoe vzaimodejstvie. Pribory i sistemy. Upravlenie, kontrol', diagnostika. 2021. № 11. S. 23–30 (in Russian).
5. Proletarskij A.V., Selezneva M.S. Izmeritel'nyj kompleks s intellektual'noj komponentoj dlya letatel'nogo apparata. V sb.: Sovremennye aspekty fundamental'nyh nauk. Trudy vtorogo mezhdunarodnogo simpoziuma. 2015. S. 196–199 (in Russian).
6. Cibizova T.Yu., Panilov P.A., Kocheshkov M.A. Monitoring bezopasnosti sistemy zashchity informacii kriticheskoj informacionnoj infrastruktury na osnove kognitivnogo modelirovaniya. Izv. Tul'skogo gosudarstvennogo universiteta. Ser.: Tekhnicheskie nauki. 2023.
   № 6. S. 33–41 (in Russian).
7. Proletarskij A.V., Neusypin K.A., Selezneva M.S. Sovremennye intellektual'nye sistemy upravleniya: teoriya i realizaciya. Inzhenernaya fizika. 2020. № 1. S. 9–30 (in Russian).
8. Nikolaev E.I., Muhanov A.N. Ispol'zovanie metodov mashinnogo obucheniya v zadachah obnaruzheniya atak na komp'yuternye seti. Komp'yuternye issledovaniya i modelirovanie. 2018. T. 10. № 3. S. 353–367 (in Russian).
9. Proletarskij A.V., Skvorcova M.A., Terekhov V.I. Gibridnaya intellektual'naya sistema ocenki riskov na osnove nestrukturirovannoj informacii. Nejrokomp'yutery: razrabotka, primenenie. 2017. № 1. S. 66–74 (in Russian).
10. Cibizova T.Yu., Slepcova K.A. Avtomatizirovannaya sistema ucheta dannyh vnutrikorporativnoj seti upravleniya informaciej. Sovremennye problemy nauki i obrazovaniya. 2015. № 1–1. S. 429 (in Russian).
11. Suhanov V.A., Cibizova T.Yu. Arhitektura sistemy razvitiya i podderzhki professional'noj intellektual'noj deyatel'nosti. Dinamika slozhnyh sistem – XXI vek. 2020. T. 14. № 1. S. 23–31 DOI: 10.18127/j19997493-202001-02 (in Russian).
12. Panilov P.A., Ivanov A.M. Agentnaya imitacionnaya model' i ontologiya finansovogo vzaimodejstviya fizicheskih i yuridicheskih lic s bankami na primere bank-klienta. Molodezhnyj nauchno-tekhnicheskij vestnik. 2012. № 5. S. 39 (in Russian).
13. Belyaeva O.N., Stepanova E.V. Nejronnye seti v zadachah prognozirovaniya i klassifikacii. Nauchno-tekhnicheskie vedomosti Sankt-Peterburgskogo gosudarstvennogo politekhnicheskogo universiteta. 2018. T. 3. № 275. S. 144–148 (in Russian).