The development and modernization of unmanned aerial vehicles (UAVs), their control systems, safety and regulation, expansion of functionality, miniaturization and increase of sensor performance, improvement of energy capabilities and routing process, increase of autonomy, contributes to the improvement of the performance of tasks. In order to obtain the best performance, customize the process of UAV team self-organization, and reduce risks for operators, the article proposes to consider the proposed method of rules for self-organization of a team of homogeneous UAVs in solving poorly formalized tasks. The basis for writing a scientific paper is the tendency to constantly increase the relevance of UAVs due to their efficiency and feasibility, as well as the constant development of artificial intelligence and information technology. By thoroughly studying, researching and implementing these areas into UAV team management systems, a number of problems related to the construction of UAV teams using existing methods and models are solved. Such as: the vulnerability of the built UAV team, the inability to continue the task due to the destruction or loss of communication with the main UAV or operator (control point), economic inexpediency, reduction of the cognitive load on operators, and others. It should also be noted that in the realities of war, it is necessary to get ahead of the enemy’s scientific research and create an advantage over him in the field of robotic technologies. A model of rules for the self-organization of a team of homogeneous UAVs in solving poorly formalized tasks is proposed, in which a number of functions (route planning, role distribution, determination of optimal actions, obtaining and processing information) assigned to the onboard system of a robotic air complex can be performed by each element of the UAV team system through their self-organization. The practicality of this method lies in the fact that the artificial intelligence of the UAV will constantly self-learn and improve through the use of machine and deep learning. Thus, the results and time required to complete missions will improve significantly, and the number of control operators will decrease. A number of problems and shortcomings related to the organization of the control system, route planning, role assignment, speed and completeness of information receipt, processing and transmission are being addressed, which in turn improves the security and performance of the system.
Read full abstract