A FORMALIZED MODEL FOR ASSESSING THE DEPENDABILITY OF UNMANNED AERIAL VEHICLE CONTROL SYSTEMS BASED ON MULTI-CRITERIA OPTIMIZATION

The intensive deployment of unmanned aerial vehicles in military, civilian, monitoring, and logistics applications increases the requirements for resilience and operational continuity of their control systems. Such systems represent distributed cyber-physical structures that integrate telemetry subsystems, communication channels, data processing modules, decision-making algorithms, and flight control actuators. Under real operating conditions, they are exposed to stochastic failures, communication degradation, software faults, computational resource constraints, and targeted information attacks, which complicates the assurance of dependability. The aim of this study is to develop a formalized integrated model for evaluating the dependability of UAV control systems that harmonizes heterogeneous operational criteria and enables multi-criteria configuration selection under temporal and resource constraints. The model is implemented as a formalized computational framework that includes state vector formation, indicator normalization, weighted aggregation, constraint verification, and decision-making regarding configuration admissibility. An algorithm for computing the integrated dependability indicator with linear complexity relative to the number of system modules and considered risk factors is developed, ensuring the suitability of the approach for embedded and edge computing in real-time systems. A sensitivity analysis method is proposed to identify critical degradation factors. Scenario-based validation is conducted for typical operational modes, including nominal operation, communication degradation, increased cyber risk, and their combined impact. The results confirm the adequacy of the model and its practical applicability for monitoring and adaptive reconfiguration of UAV control systems.