Wake Estimation and Optimal Control for Autonomous Aircraft in Formation Flight

Abstract

The continued development of sophisticated aircraft with high fidelity control systems will enable autonomous execution of challenging tasks such as aerial refueling and close formation flight. In order to achieve such tasks in autonomous flight, an aircraft must sense other aircraft in close proximity and position itself relative to them. For example, aerial refueling requires the follower aircraft to intercept the filling nozzle attached to the leader aircraft; also, formation-flying aircraft must position themselves strategically to realize benefits of aerodynamic efficiency. This paper uses lifting-line theory to represent a two-aircraft formation and presents a grid-based, recursive Bayesian filter for estimating the wake parameters of the leader aircraft using noisy pressure measurements distributed along the trailing aircraft's wing; the estimator also requires a binary, relative-altitude measurement to break the vertical symmetry. Optimal control strategies are presented to steer the follower aircraft to a desired position relative to the leader while simultaneously optimizing the observability of the leader's relative position. The control algorithms guide the follower aircraft along trajectories that maintain adequate observability, thereby guaranteeing estimator performance. Theoretical results are illustrated using numerical examples of two-aircraft formations.

Tip-Vortex Localization for Cross-Stream Position Control of a Multi-Hole Probe Relative to a Stationary Wing in a Free-Jet Wind Tunnel

“…8,9 However, we implement the control algorithms in real time in a wind tunnel, assuming known kinematics of the leader aircraft, using a grid-based Bayesian filter. DeVries and Paley 6 introduced the idea of using a grid-based Bayesian filter for wake estimation, and we demonstrate the viability of this idea in an experimental setting.…”

An Empirical Model of Rotorcrafy UAV Downwash for Disturbance Localization and Avoidance

“…A low-order vortex model was employed by DeVries and Paley [13] to develop a sensing and estimation strategy to detect the wake of a refueling aircraft, providing a fixed-wing counterpart to the research proposed in this paper. Focusing on fundamental downwash characteristics, Khan et al [14] developed a semi-empirical model that describes the wash generated by a single propeller.…”

Analytic Approach to Optimal Routing for Commercial Formation Flight

“…Studies into areas of biomimicry, such as geese flying in a V formation [5,6], have always interested scientists, and the military have long flown in formation for communicative and defensive purposes [7]. More recent studies assessing both the aerodynamic possibility [8][9][10] and the associated control problem [11][12][13][14] of flying in close proximity to reduce drag, coupled with flight tests [15][16][17], shows promise that flying in formation can reduce fuel burn. According to the aerodynamic studies cited, practical formation flight is likely to demand that aircraft fly closer than current separation minima.…”