SCENARIO-PRECEDENT APPROACHES TO THE COORDINATION OF HETEROGENEOUS UNMANNED AERIAL VEHICLE ENSEMBLES

Abstract

The article addresses the pressing issue of coordinating heterogeneous ensembles of unmanned aerial vehicles (UAVs) under conditions of high dynamism and uncertainty. A coordination approach based on a scenario-precedent model is proposed, which integrates elements of centralized, decentralized, and hybrid control. Particular attention is paid to the effective distribution of tasks among different types of UAVs, considering their unique capabilities and limitations.The objective of the study is to develop a flexible and adaptive approach capable of ensuring effective interaction among agents with diverse characteristics.The research envisions the creation of a precedent database that will enable the use of accumulated experience to select relevant mission execution scenarios. An adaptive coordination algorithm is proposed, based on a combination of machine learning methods (neural networks) and reinforcement learning, to implement real-time dynamic adaptation.The proposed architecture also includes mechanisms for communication and information exchange between individual UAVs to ensure coherence of actions. The influence of the quality and volume of the precedent database on coordination efficiency is separately analyzed. The algorithm also considers the possibility of unforeseen situations and the need for rapid replanning of actions. The planned simulation allows for a comparative analysis of the effectiveness of the developed approach against existing methods, particularly in terms of decision-making speed and fault tolerance.The results of the study demonstrate that the use of the scenario-precedent approach contributes to reducing decision- making time, increasing system flexibility, and enhancing resilience in a dynamic environment. The findings confirm the relevance of further development of the proposed model, particularly by expanding the precedent database and improving the adaptation algorithms.