Guidance laws for autonomous formation flight

Abstract

The objective of this thesis is to find guidance laws for autonomous formation flight. Two different approaches are introduced for autonomous flight. The first approach is a boid-based formation guidance which is derived from swarm intelligence. The boid-based formation guidance "flocks" the unmanned aerial vehicles through simple steering behaviors. Collision avoidance, velocity matching, flock centering is applied and modified for general aircraft. Flock joining is proposed to substitute the flock centering behavior for a more stable formation. Flock joining behavior matches the formation bubble distance to the two closest neighbors. The seek maneuver is applied to orient the flock into a specific direction. The behaviors are studied through simple point mass model simulation. Obstacle avoidance and path following are also proposed in the thesis. The boid-based formation guidance allows for a leader-free decentralized formation which can adapt to the given environment. The second approach is two-dimensional formation guidance laws for formation flight using only line-of-sight angle information. The main idea is to use the line of sight angles to two nearby vehicles to maintain the formation position of the current vehicle. Such a formation guidance law is useful because measurement of the line of sight angles does not require data communication between the vehicles. Two methods of using the line of sight angle information for formation guidance are proposed. The angle information is used to control the flight-path angle and the velocity of the formation vehicle. Approach guidance and formation guidance of the two leading vehicles are also proposed. Stability of the proposed formation guidance laws is analyzed by using the Jacobian at the equilibrium point. Multiple vehicle formation flight is simulated to verify the guidance laws proposed.