Learning dynamics for improving control of overactuated flying systems

TL;DR

This paper introduces a hybrid data-driven and physics-based control approach for overactuated flying vehicles, significantly improving trajectory tracking accuracy by reducing errors through learned model corrections.

Contribution

It presents a novel combined modeling approach using Gaussian Processes and first-principle models to enhance control of complex overactuated flying systems.

Findings

Attitude trajectory error reduced by 32% on average.

The method effectively handles model uncertainties and interference.

Validation on a tilt-arm vehicle demonstrates improved control performance.

Abstract

Overactuated omnidirectional flying vehicles are capable of generating force and torque in any direction, which is important for applications such as contact-based industrial inspection. This comes at the price of an increase in model complexity. These vehicles usually have non-negligible, repetitive dynamics that are hard to model, such as the aerodynamic interference between the propellers. This makes it difficult for high-performance trajectory tracking using a model-based controller. This paper presents an approach that combines a data-driven and a first-principle model for the system actuation and uses it to improve the controller. In a first step, the first-principle model errors are learned offline using a Gaussian Process (GP) regressor. At runtime, the first-principle model and the GP regressor are used jointly to obtain control commands. This is formulated as an optimization problem, which avoids ambiguous solutions present in a standard inverse model in overactuated systems, by only using forward models. The approach is validated using a tilt-arm overactuated omnidirectional flying vehicle performing attitude trajectory tracking. The results show that with our proposed method, the attitude trajectory error is reduced by 32% on average as compared to a nominal PID controller.