Comparative Analysis of Mission Simulation Methods for Evaluating Mission Performance of Unmanned Aerial Vehicles

摘要

Typical missions of civil unmanned aerial vehicles (UAVs) comprise inspecting a designated area by means of onboard sensors as for instance in crop monitoring or search and track missions. The mission performance of a UAV considerably depends on the quality of the sensor data obtained during the mission. For an effective optimization of UAV configurations, the interaction between environment and onboard sensor equipment should be considered already in the conceptual design phase. For this purpose, a Simulink flight dynamic mission simulation model has been developed which generates the trajectory of a designated UAV along prescribed way points. The model features a flight controller unit which computes the required control parameters to guide the aircraft along the waypoints. Furthermore, terrain data is available which allows for the evaluation of the sensor footprint at the ground at any time during the mission simulation. Aiming for a computationally less expensive yet sufficiently accurate mission simulation method, a quasi-stationary model is introduced which is not reliant on a flight control unit. Another benefit of forgoing a flight control unit is the absence of the need of controller gains to be tuned for each individual aircraft configuration. This paper compares both simulation models regarding their generated trajectories. Discrepancies regarding the obtained sensor data are investigated in order to evaluate the suitability of a quasi-stationary mission simulation model without flight control unit to replace the currently used Simulink model. The results of both simulation models show very good accordance. By using the quasi-stationary model, a significant decrease in computation time is achieved.