S.I. Popkov – Post-graduate Student, Computer Science Faculty, Moscow State University of Psychology and Education, Moscow, Russia

E-mail: rslw25@gmail.com

The project is aimed at solving the actual and unresolved task, which is a creation of fully automated unmanned aerial vehicles (such as drones). A probabilistic model of the applied multi-agent system behaviour, representing the game interaction of a set of agents and the goal, has been developed for such cause. The behaviour of agents is non-deterministic and, therefore, unpredictable for the target. The system should allow both coherent and autonomous behaviour of agents, depending on whether agents receive information about the presence and position of other active agents. The behaviour of the agent is determined by an algorithm that involves identifying the parameters of a probability model using optimized goal-based functions expressing the group and individual probabilities of defeating the target.

Unification of the segmental units of the system into one application package allowing to train the operators and predict the system development according to the aforementioned model. 

The developed model (and appropriate algorithm) allows to control the behaviour of relevant applied multi-agent systems. The model also allows forecasting and analysis of the game actions to design a battle strategy both in the conditions of an autonomous movement of a combat unit and in the collective coherent movement of a combat group. The probabilistic model of the behaviour of the system is supplemented by formulas for the dynamic calculation of the probability distributions for the target defeat by agents (and agent defeat by the target). The applied mathematical methods make it possible to easily adapt the model to different combat situations, including the moving and stationary target representing the enemy, various types of weapons, weather and other conditions affecting the outcome of combat, as well as various types of combat equipment and specific characteristics such as speed movement, the presence or absence of a specific equipment, the type of combat unit and other characteristics. The mathematical method for analysing the general laws of behaviour of the system is presented. Software implementation of the model can be embedded and used for a wide range of battle units, both ground and air, including modern unmanned aerial vehicles and robotic systems, as well as simulators for complex system operators. The basic research methods used in this work are computer and mathematical modeling.

The need of such models for the military vehicles has been revealed through the experienced military conflicts in recent years. These complexes are supposed to replace modern systems, managed solely by operators, to increase their efficiency, in particular, by optimizing their behaviour and increasing the speed of decision-making.

1. Kuravskiy L.S., Popkov S.I. Predstavleniye obshchikh zakonomernostey povedeniya mnogoagentnoy sistemy s pomoshchyu eye makroparametrov. Neyrokompyutery: razrabotka. primeneniye. 2018. № 1. S. 13-25.  (in Russian).
2. Popkov S.I. Metod vneshney optimizatsii dlya identifikatsii markovskikh protsessov. Informatsionnyye Tekhnologii. 2018. T. 24. № 10. S. 633-641. DOI: 10.17587/it.24.633-641 (in Russian).
3. Popkov S.I. Programmnaya realizatsiya mezhyazykovogo vzaimodeystviya na baze dinamicheskikh bibliotek. Neyrokompyutery: razrabotka. primeneniye. 2018. № 3. S. 39-49. (in Russian).
4. Kuravsky L., Yuryev G. On the approaches to assessing the skills of operators of complex technical systems. In: Proc. 15th International Conference on Condition Monitoring & Machinery Failure Prevention Technologies. Nottingham. UK. September 2018. P. 1-25.
5. Kuravskiy L.S., Margolis A.A., Marmalyuk P.A., Panfilova A.S., Yuryev G.A. Matematicheskiye aspekty kontseptsii adaptivnogo trenazhera. Psikhologicheskaya nauka i obrazovaniye. 2016. T. 21. № 2. C. 84–95.  DOI: 10.17759/pse.2016210210 (in Russian).
6. Kuravsky L.S., Margolis A.A., Marmalyuk P.A., Panfilova A.S., Yuryev G.A., Dumin P.N. A probabilistic model of adaptive training. Applied Mathematical Sciences. 2016. V. 10. № 48. P. 2369-2380. [Elektronnyy resurs]. URL: http://dx.doi.org/10.12988/ams.2016.65168 (data obrashcheniya: 30.01.2019 г.).
7. Trenazherno-obuchayushchaya sistema [Elektronnyy resurs]. URL: https://www.youtube.com/watch?v=3CDsMHzh-Co (data obrashcheniya: 30.01.2019). (in Russian).
8. Trenazhernaya sistema (stykovka) [Elektronnyy resurs]. URL: https://www.youtube.com/watch?v=0eVBXK0vivA (data obrashcheniya: 30.01.2019). (in Russian).
9. Kuravsky L.S., Popkov S.I., Artemenkov S.L. An applied multi-agent system within the framework of a player-centered probabilistic computer game. International Journal of Modeling. Simulation. and Scientific Computing. 2018. V. 9. № 1. 17500635. 17p. DOI: 10.1142/S1793962317500635
10. Kuravsky L.S., Popkov S.I. Forecasting macro parameters representing the behavior of an applied multi-agent system. International Journal of Modeling. Simulation. and Scientific Computing. 2018. V. 9. № 6. 1850052. 15p. DOI: 10.1142/S1793962318500526
11. Kuravsky L.S., Popkov S.I. Forecasting behavior of a stochastic multi-agent system. - In: Proc. 15th International Conference on Condition Monitoring & Machinery Failure Prevention Technologies. Nottingham. UK. September 2018. P. 129-137.
12. HTML 5.2. W3C Recommendation. 14 December 2017 [Elektronnyy resurs]. URL: https://www.w3.org/TR/html5/ (data obrashcheniya: 30.01.2019).
13. Business Process Management [Elektronnyy resurs]. URL: https://www.bonitasoft.com/business-process-management-bpm (data obrashcheniya: 30.01.2019).