⚡ electrical engineering

Geometric Inverse Flight Dynamics on SO(3) and Application to Tethered Fixed-Wing Aircraft

This paper presents a coordinate-free, geometric formulation of inverse flight dynamics for fixed-wing aircraft on SO(3) that derives closed-form trajectory-to-input maps for coordinated flight, enabling analytic solutions for tethered maneuvers and bridging aeronautical inverse simulation with robotic geometric modeling.

Original authors: Antonio Franchi, Chiara Gabellieri

Published 2026-07-30

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Antonio Franchi, Chiara Gabellieri

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where robots don't just walk or roll, but fly. While we are used to seeing drones that hover like hummingbirds, there is a different kind of flying machine: the fixed-wing aircraft, which looks more like a bird or a plane. These machines are incredibly efficient at covering long distances, but they are tricky to control because they rely on invisible forces like wind and lift to stay aloft. For a long time, engineers have struggled to teach robots how to fly these planes while carrying heavy things, like a backpack or a cable, without crashing. The core problem is figuring out exactly how to tilt the wings, how fast to spin the propeller, and how to angle the nose to follow a specific path. It's like trying to solve a complex puzzle where the pieces are constantly changing shape. This paper dives into that puzzle, using a branch of math called geometry to describe flight without getting bogged down in confusing coordinate systems, aiming to make it easier for robots to fly fixed-wing aircraft on missions that require them to carry loads or stay tethered to a point.

The authors of this paper, Antonio Franchi and Chiara Gabellieri, have developed a new "recipe" for flying fixed-wing aircraft that is both mathematically elegant and practical for robotics. Think of their method as a reverse-engineering tool. Usually, pilots or autopilots decide what the plane should do (like "turn left") and then calculate the forces needed to make that happen. This paper does the opposite: it starts with a desired path (like "fly in a perfect circle while holding a rope") and works backward to tell the computer exactly what the plane's attitude, speed, and engine power need to be at every single moment. They call this "Inverse Flight Dynamics," and they've built it on a solid geometric foundation that avoids the messy local coordinates often used in traditional aviation, making it much cleaner for robots to understand.

The team tested their recipe on a very specific and interesting scenario: a fixed-wing aircraft tethered to the ground by a cable, flying in a circle around a sphere, like a bird on a string. They wanted to see how the plane should bank (tilt its wings) when the cable is pulling it. Their calculations revealed a surprising "sweet spot." Usually, when a plane turns, it has to bank inward to counteract the force that wants to throw it outward (centrifugal force). However, they found that if the cable pulls with just the right amount of tension, it can perfectly cancel out that outward force. In this special case, the plane can fly with its wings perfectly flat—zero bank angle—while still turning. It's as if the rope is doing all the heavy lifting of the turn, allowing the plane to relax its wings.

They also discovered that this "zero-bank" state is a delicate balance. If the plane flies faster, the outward force gets stronger, and it must tilt its wings inward again. If the cable pulls harder, the plane actually has to tilt its wings outward to fight the rope's pull. The authors showed that this behavior depends entirely on the speed, the length of the cable, and the angle of the circle, and it doesn't even matter what kind of plane it is or how big its wings are; the geometry of the situation dictates the tilt.

To prove their ideas work, the researchers ran computer simulations with a 2 kg aircraft, a 20-meter cable, and a flight speed of 11.7 m/s. They found that when the cable tension was set to 16 Newtons, the plane's bank angle dropped to nearly zero, exactly as their math predicted. At this point, the plane needed an angle of attack (the angle the nose points relative to the wind) of about 16 degrees to stay aloft. These results, which are based on simulations rather than physical flight tests, suggest that their geometric method is a powerful tool for designing flight paths for tethered aircraft. It bridges the gap between the abstract math of robotics and the real-world physics of flying, offering a rigorous way to check if a planned flight path is even possible before a robot ever leaves the ground.

Technical Summary: Geometric Inverse Flight Dynamics on SO(3) and Application to Tethered Fixed-Wing Aircraft

Problem Statement
While aerial manipulation using multirotors has advanced significantly, endurance constraints limit long-duration tasks. Recent paradigms propose using multiple fixed-wing carriers to suspend loads perpetually, but existing robotics models often abstract aircraft dynamics as double integrators, neglecting fixed-wing aerodynamics (lift, drag, angle of attack, and coordinated flight constraints). Conversely, existing aeronautical inverse simulation methods rely on local attitude coordinates, and Airborne Wind Energy (AWE) literature focuses on flight mechanics rather than robotics-oriented inverse synthesis for manipulation. There is a lack of a concise, coordinate-free, geometric framework that bridges these domains to generate realizable attitude and control trajectories for fixed-wing aircraft directly from task-space specifications.

Methodology
The authors propose a robotics-oriented, coordinate-free formulation of inverse flight dynamics (IFD) for fixed-wing aircraft defined on the Special Orthogonal group $SO(3)$.

Geometric Modeling: The dynamics are formulated using a global, world-frame ( $W$ ) for translational forces and a body-frame ( $B$ ) for rotational dynamics. The model adopts a Forward-Left-Up (FLU) convention common in robotics, contrasting with the Forward-Right-Down (FRD) convention in aeronautics. Aerodynamic directions (drag, lift, side) are defined geometrically using vectors (flow, chord, span) rather than local angles, ensuring global validity and avoiding singularities.
Inverse Synthesis: Given a desired Center of Mass (CoM) trajectory and external loads, the method constructs a closed-form, pointwise quasi-steady solution.
- Coordinated Flight: The method enforces zero sideslip ( $\beta = 0$ ), geometrically aligning the body's lateral axis with the trajectory binormal.
- Attitude and Velocity: The attitude matrix $R(t)$ and angular velocity $\omega^B(t)$ are derived geometrically from the required net force and the air-relative velocity vector.
- Thrust and Angle of Attack (AoA): The 2D force balance in the plane defined by the air-relative velocity and the required lift direction is solved to find the thrust magnitude $T$ and AoA $\alpha$ . This involves solving implicit equations, which admit closed-form solutions under simplified lift/drag laws (e.g., using Cardano's formula for cubic equations).
- Moment Coefficients: Required aerodynamic moment coefficients ( $C_\ell, C_m, C_n$ ) are recovered component-wise to satisfy rotational dynamics, followed by control allocation to determine physical surface deflections.
Specific Application: The framework is applied to a tethered fixed-wing aircraft flying at a constant speed along a spherical parallel (a circle of latitude) with the tether attached at the CoM and pulling toward the sphere's center.

Key Contributions

Geometric SO(3) Model: A Newton-Euler formulation for fixed-wing aircraft that uses globally valid aerodynamic directions and is compatible with standard stability derivatives, providing a foundation for geometric control and estimation in robotics.
Closed-Form Inverse Flight Dynamics: A method to compute attitude, angular velocity, thrust, and AoA directly from a trajectory and external loads, recovering aerodynamic moment coefficients component-wise.
Analytic Tethered Flight Analysis: Application of the theory to tethered flight on spherical parallels, yielding analytic expressions for the required bank angle.
Zero-Bank Locus Identification: The derivation of a specific condition where tether tension exactly balances centrifugal effects, resulting in a zero-bank (level-wing) attitude. This highlights a decoupling between aerodynamic coordination and the apparent gravity vector.
Minimal-Thrust Solution: Under simple lift/drag laws, the paper provides a closed-form solution for the AoA that minimizes thrust for constant-speed circular motion.

Results

Bank Angle Behavior: The study derives an explicit relationship for the bank angle $\mu$ as a function of dimensionless speed ( $\kappa$ ), normalized tether force ( $\eta$ ), and trajectory colatitude ( $\theta$ ).
Regime Analysis: Three distinct regimes are identified:
- Low Force/High Speed: Centrifugal force dominates, requiring inward banking ( $\mu > 0$ ).
- Zero Bank: Tether tension balances centrifugal force, allowing level wings ( $\mu = 0$ ).
- High Tension/Low Speed: Tether pull exceeds centrifugal force, requiring outward banking ( $\mu < 0$ ) to generate lift opposing the tether.
Sensitivity Analysis: The paper quantifies how the bank angle responds to changes in tether tension, speed, and trajectory geometry. Increasing tether tension or trajectory radius (lowering latitude) promotes outward banking, while increasing speed promotes inward banking.
Numerical Validation: Simulations using realistic parameters for a 2 kg aircraft confirm the analytic predictions. When tether tension balances centrifugal force, the bank angle approaches zero, and the required AoA is approximately $16^\circ$ for the specific test case.

Significance and Claims
The authors claim this work bridges the gap between inverse simulation in aeronautics and geometric modeling in robotics. By providing a rigorous, coordinate-free building block, the framework enables trajectory design and feasibility checks for fixed-wing carriers in manipulation tasks, moving beyond the double-integrator abstractions previously used. The identification of the zero-bank locus offers a specific "sweet spot" for efficiency in tethered flight, where aerodynamic banking is unnecessary. The authors position this work as a foundational step toward realizing non-stop flight paradigms for cable-suspended load manipulation, noting that future work is required to address wind effects, flexible cable dynamics, and experimental validation.

More like this