FORMAL MODEL OF SCENARIO-PRECEDENT COORDINATION OF HETEROGENEOUS UNMANNED AERIAL VEHICLE ENSEMBLES AND ITS ADAPTATION ALGORITHM IN DYNAMIC ENVIRONMENTS

Abstract

This article addresses the problem of coordinating heterogeneous ensembles of unmanned aerial vehicles (UAVs) in dynamic and uncertain environments using a scenario-precedent approach. Unlike previous studies, which primarily focused on the conceptual system model and general architectural principles, this work emphasizes the formal mathematical specification of the key components of the approach. In particular, a formal model of a mission execution scenario, a precedent structure, and the organization of a precedent base are proposed, enabling the systematic representation and reuse of accumulated experience in typical UAV group missions. The study develops a model oriented toward three basic types of UAVs – micro quadcopters, medium multirotor UAVs, and fixed-wing aircraft – each characterized by distinct technical capabilities, environmental resilience, and functional roles. The model aims to support effective coordination during reconnaissance, observation, and territory monitoring missions, which require synchronized role distribution, rapid trajectory adjustment, and adaptive agent behavior. The article introduces a similarity metric that allows estimating the correspondence between the current environmental state, UAV parameters, and mission execution conditions recorded in the precedent base. Based on this metric, an algorithm for selecting a relevant scenario from the precedent base is proposed, along with a real-time scenario adaptation mechanism leveraging reinforcement learning principles. This enables flexible scenario correction in response to changes in external conditions and individual agent behavior. The results of the study demonstrate that the formalized scenario-precedent model provides a foundation for building robust, adaptive, and scalable coordination systems for heterogeneous UAV ensembles. The proposed solutions form the theoretical basis for further simulation studies, experimental validation, and comparative analysis with existing coordination methods in complex, dynamic, and uncertain environments.