InvSim algorithm for pre-computing airplane flight controls in limited-range autonomous missions, and demonstration via double-roll maneuver of Mirage III fighters

TL;DR

This paper introduces InvSim, an inverse simulation algorithm for pre-computing airplane flight controls in limited-range autonomous missions, demonstrated through a double-roll maneuver of Mirage III fighters, combining mathematical modeling and numerical methods.

Contribution

It develops a mathematical framework and numerical procedure for inverse simulation of flight controls, enabling prediction of control inputs for specified flight trajectories.

Findings

Successfully computed control inputs for a double-roll maneuver

Validated InvSim with Mirage III fighter simulation

Demonstrated accurate trajectory tracking using the algorithm

Abstract

In this work, we start with a generic mathematical framework for the equations of motion (EOM) in flight mechanics with six degrees of freedom (6-DOF) for a general (not necessarily symmetric) fixed-wing aircraft. This mathematical framework incorporates (1) body axes (fixed in the airplane at its center of gravity), (2) inertial axes (fixed in the earth/ground at the take-off point), wind axes (aligned with the flight path/course), (3) spherical flight path angles (azimuth angle measured clockwise from the geographic north, and elevation angle measured above the horizon plane), and (4) spherical flight angles (angle of attack and sideslip angle). We then manipulate these equations of motion to derive a customized version suitable for inverse simulation flight mechanics, where a target flight trajectory is specified while a set of corresponding necessary flight controls to achieve that maneuver are predicted. We then present a numerical procedure for integrating the developed inverse simulation (InvSim) system in time; utilizing (1) symbolic mathematics, (2) explicit fourth-order Runge-Kutta (RK4) numerical integration technique, and (3) expressions based on the finite difference method (FDM); such that the four necessary control variables (engine thrust force, ailerons' deflection angle, elevators' deflection angle, and rudder's deflection angle) are computed as discrete values over the entire maneuver time, and these calculated control values enable the airplane to achieve the desired flight trajectory, which is specified by three inertial Cartesian coordinates of the airplane, in addition to the Euler's roll angle. We finally demonstrate the proposed numerical procedure of flight mechanics inverse simulation (InvSim).