Title: Automated Generation of Realistic Near-Optimal Aircraft Trajectories
Authors: Janne Karelahti, Kai Virtanen, and John Öström
Status: Journal of Guidance, Control, and Dynamics, to appear.
Keywords: Trajectory optimization, missile avoidance, nonlinear programming, inverse simulation.
A new approach towards the automated solution of realistic near-optimal aircraft trajectories is introduced and implemented in a software named Ace. In the approach, the optimal open-loop trajectory for a three degree-of-freedom aircraft model is first solved by using direct multiple shooting. Then, the obtained trajectory is inverse simulated with a more sophisticated five degree-of-freedom performance model by using an integration inverse method based on Newton's iteration. The trajectories are evaluated visually and by analyzing errors between the trajectories. If the errors remain within a suitable application-specific tolerance, the inverse simulated trajectory can be considered a realistic near-optimal trajectory that could be flown by a real aircraft. Otherwise, the parameters affecting the optimization and inverse simulation are altered and the computations are repeated. The example implementation of the approach, the Ace software, contains a graphical user interface that provides a user-oriented way for analyzing aircraft minimum time and missile avoidance problems. In the software, the computation of the optimal and inverse simulated trajectories is fully automated. The approach is demonstrated with numerical examples by using Ace.