Planejamento de caminhos tridimensionais para veículos aéreos não tripulados

Abstract

Unmanned aerial vehicles (UAV’s) have been the target of research in many applications where there are hazardous conditions to the integrity of the crew. A great effort has been made to increase the autonomy of UAV’s to perform these missions. In the military domain these studies have shown good results. On the other hand, in civil area, researches are still a challenge, due to the lack of investments for the development and absence of specific rules. However, in some situations, the lack of pilots onboard becomes mission execution a nontrivial task. An example is missions on dense forest environments with, specific feature of the Amazon region. This scenario presents a critical role both in autonomous navigation mode and the remotely operated mode, due to the difficulty of navigation in such environments. The use of planners has enabled obtaining paths or trajectories that satisfy criteria navigation performance meeting the requirements of the mission. In general, a set of points is defined in the vehicle navigation space and the planner finds a path that passes through all of these points, taking into account the constraints of motion, time and / or energy consumption of the vehicle. This work addresses the construction of a path planner for fixed wing UAV navigation in three dimensional spaces about sparse forest regions with restriction of flight due to weather and environmental conditions. Unlike the traditional way, the methodology developed here gets specific points in three-dimensional space regions that favor the criteria for flight safety and maneuverability of the aircraft and also meet the requirements of the mission, paving the way from the interpolation of these points with points crossing the coming mission. Experimental tests by simulation in different scenarios show the applicability of UAVpath planner in actual service security and improving mission performance.