Differential flatness-based optimal air combat maneuver strategy generation

In this study, we present a "flat" optimal-maneuver generation algorithm for one-to-one aerial combat games. Mostly, such algorithms uses iterative approaches with decision trees and certain heuristics at different levels, which is not possible to apply to fully automated systems. Considering the real-time applicability and sudden decision needs, which is highly critical in combat games, we aim to construct an acyclic algorithm. The differential flatness formulation enables to determine the control input and the state sequence with regards to the desired output trajectory and its derivatives, which are parameterized through b-spline. Then the aerial combat system is translated into an optimization problem through a cost function with aspect angle, bearing angle and range. In this manner, we generate an optimal feasible trajectory that meets all the given and dynamical constraints. For the simulation and demonstration purposes, we have applied this methodology to air combat between two UAVs and shown the results for the randomly generated scenarios.