Maneuvering Planning for UAVs in Forest Surveillance and Fire Detection Missions with Kinematic Uncertainties

Unmanned aerial vehicles (UAVs) have brought increasing spotlights to the area of forest fire detection, monitoring, and tracking in latest years because of their high flexibility, efficiency, lower cost, and less risk without the need for pilot onboard. Commonly, heavier and more complex tasks can be accomplished by employing multiple UAVs in a formation instead of a single UAV. However, one potential trouble is that the power could constantly exhaust or faults could happen on UAVs in the formation in the field, therefore the weakened UAVs must be supplanted by new vehicles from the UAV base. Thus, the problem is to find an optimal way to navigate the recent UAVs to join the team and to maintain the formation for the remaining task. Additional difficulties arise when uncertainties of motions are encountered over the flight. To overcome these challenges, an uncertainty-embedded maneuvering planning strategy based on the Star-Minimax algorithm is developed. Simulation results prove that a newly assigned UAV can be effectively navigated to join the UAV team, and the formation can be sustained during the flight even in the presence of kinematic uncertainties. Flight experiments, operated in the NAV Lab at Concordia University, further validate the functional performance of the proposed maneuvering planning strategy in real-time.