Inverting Learned Dynamics Models for Aggressive Multirotor Control

TL;DR

This paper introduces a control method for multirotors that inverts learned dynamics models to improve accuracy and disturbance rejection, enabling better trajectory tracking in challenging conditions.

Contribution

The paper proposes a novel inverse dynamics control strategy using learned acceleration error models, enhancing disturbance compensation without complex parameter learning.

Findings

Reduced tracking error in simulations and hardware tests

Effective compensation for a wide range of disturbances

Successful control of high-acceleration trajectories over 7 m/s^2

Abstract

We present a control strategy that applies inverse dynamics to a learned acceleration error model for accurate multirotor control input generation. This allows us to retain accurate trajectory and control input generation despite the presence of exogenous disturbances and modeling errors. Although accurate control input generation is traditionally possible when combined with parameter learning-based techniques, we propose a method that can do so while solving the relatively easier non-parametric model learning problem. We show that our technique is able to compensate for a larger class of model disturbances than traditional techniques can and we show reduced tracking error while following trajectories demanding accelerations of more than 7 m/s^2 in multirotor simulation and hardware experiments.