Abstract:
This paper presents a path planning concept based on the Manned-Unmanned Teaming (MUM-T) between the helicopter and a drone. The drone flies ahead of the helicopter to detect possible unexpected obstacles in the mission area and sends the data to the helicopter. The path of the helicopter is automatically replanned to avoid the meteorological and physical obstacles detected by the drone. The path planning is based on the Rapidly-exploring Random Tree (RRT) and the Bidirectional Rapidly-exploring Random Tree (BiRRT) algorithms. The reference trajectory is planned by means of the RRT algorithm and the replanning is performed with the BiRRT. The node connection is realized with the Dubins curves, that force the path to comply with the prescribed limitations on the helicopter's roll angle and flight path angle. The Savitzky-Golay filter is used to smooth the trajectory achieving curvature continuity. A closed-loop simulation model containing the dynamics of the pilot is used to evaluate the feasibility of a candidate trajectory basing on the pilot's workload. The pilot's dynamics is modeled through the frequency-based multi-axis Hess model. A workload assessment method is proposed basing on the aggression factor, which depends on the time histories of the control inputs. The workload quantitative assessment is carried out by comparison with the aggression factor on a "simple" manoeuvre, the spiral.