Planejamento de rotas para múltiplos VANTs colaborativos

Abstract

This study explores and proposes solutions for assigning tasks in collaborative systems for route planning with multiple Unmanned Aerial Vehicles (UAVs). To achieve this objective, a hybrid algorithm is employed, combining the Multi-Objective Ant Colony System (MOACS) with the implementation of A* algorithms, along with the Dynamic Window Approach (DWA). This combination is specially designed to facilitate obstacle avoidance, which can be either static or dynamic, a critical feature in environments with high variability. Initially, MOACS is parametrized to optimize results, while experiments are conducted through computational simulations, incorporating obstacles as well as energy and capacity constraints for the UAVs. The results obtained with this method show that the proposal allows for efficient route planning. In addition, the research provides valuable guidance on three fundamental aspects: the estimated time for mission execution, the total distance traveled by UAVs during their operations and the computational cost applied to route planning. These insights are essential for optimizing the use of these vehicles in practical applications, thus contributing to advances in the field of autonomous air transport systems.