Aggressive Online Control of a Quadrotor via Deep Network Representations of Optimality Principles

Li, Shuo; Öztürk, Ekin; Wagter, Christophe De; Croon, Guido de; Izzo, Dario

doi:10.1109/icra40945.2020.9197443

Cited by 16 publications

(11 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Please note though that we and others are quickly developing deep learning approaches that can cross the reality gap for performing visual odometry Sanket et al (2020), tracking of predetermined optimal trajectories Kaufmann et al (2020) or even for full optimal control Li et al (2020b). AIRR has already been a driving force to develop AI methods that will successfully bridge the reality gap, even for robots that are difficult to model in detail upfront.…”

Section: Conclusion and Discussionmentioning

confidence: 99%

The Artificial Intelligence behind the winning entry to the 2019 AI Robotic Racing Competition

De Wagter,

Paredes-Vallés,

Sheth

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Robotics is the next frontier in the progress of Artificial Intelligence (AI), as the real world in which robots operate represents an enormous, complex, continuous state space with inherent realtime requirements. One extreme challenge in robotics is currently formed by autonomous drone racing. Human drone racers can fly through complex tracks at speeds of up to 190 km/h. Achieving similar speeds with autonomous drones signifies tackling fundamental problems in AI under extreme restrictions in terms of resources. In this article, we present the winning solution of the first AI Robotic Racing (AIRR) Circuit, a competition consisting of four races in which all participating teams used the same drone, to which they had limited access. The core of our approach is inspired by how human pilots combine noisy observations of the race gates with their mental model of the drone's dynamics to achieve fast control. Our approach has a large focus on gate detection with an efficient deep neural segmentation network and active vision. Further, we make contributions to robust state estimation and risk-based control. This allowed us to reach speeds of 9.2m/s in the last race, unrivaled by previous autonomous drone race competitions. Although our solution was the fastest and most robust, it still lost against one of the best human pilots, Gab707. The presented approach indicates a promising direction to close the gap with human drone pilots, forming an important step in bringing AI to the real world.

show abstract

Section: Conclusion and Discussionmentioning

confidence: 99%

The Artificial Intelligence behind the winning entry to the 2019 AI Robotic Racing Competition

De Wagter,

Paredes-Vallés,

Sheth

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…One significant advantage of direct methods is their ability to easily handle more complicated OCPs, such as those with path constraints. In the context of supervised learning, [33,20] use Hermite-Simpson direct collocation to generate data for finite-horizon OCPs. Alternatively, Radau pseudospectral methods [30,9] are ideal for solving infinite-horizon open loop OCPs, though they have not yet been used for supervised learning.…”

Section: Data Generationmentioning

confidence: 99%

“…This line of work has been futher developed using nonlinear regression with NNs [12,11,24,25,26] and sparse polynomials [3], significantly increasing the maximum feasible problem dimension. A variation of the method in [25] is proposed by [7], in which the value gradient is directly approximated without learning the value function itself, while [31,33,20,12,11] use NNs to directly approximate the control policy without solving the HJB equation.…”

Section: Introductionmentioning

confidence: 99%

Neural network optimal feedback control with enhanced closed loop stability

Nakamura-Zimmerer,

Gong,

Kang

2021

Preprint

View full text Add to dashboard Cite

Recent research has shown that supervised learning can be an effective tool for designing optimal feedback controllers for high-dimensional nonlinear dynamic systems. But the behavior of these neural network (NN) controllers is still not well understood. In this paper we use numerical simulations to demonstrate that typical test accuracy metrics do not effectively capture the ability of an NN controller to stabilize a system. In particular, some NNs with high test accuracy can fail to stabilize the dynamics.To address this we propose two NN architectures which locally approximate a linear quadratic regulator (LQR). Numerical simulations confirm our intuition that the proposed architectures reliably produce stabilizing feedback controllers without sacrificing performance. In addition, we introduce a preliminary theoretical result describing some stability properties of such NN-controlled systems.

show abstract

“…This line of work has been futher developed using nonlinear regression with NNs [14,13,29,30,32] and sparse polynomials [2], significantly increasing the maximum feasible problem dimension. Alternatively, one can directly approximate the value gradient [8,31] or control policy [37,40,25,14,13,31].…”

Section: Introductionmentioning

confidence: 99%

Neural Network Optimal Feedback Control with Guaranteed Local Stability

Nakamura-Zimmerer¹,

Gong²,

Kang³

2022

Preprint

View full text Add to dashboard Cite

Recent research shows that deep learning can be an effective tool for designing optimal feedback controllers for high-dimensional nonlinear dynamic systems. But the behavior of these neural network (NN) controllers is still not well understood. In particular, some NNs with high test accuracy can fail to even locally stabilize the dynamic system. To address this challenge we propose several novel NN architectures, which we show guarantee local stability while retaining the semi-global approximation capacity to learn the optimal feedback policy. The proposed architectures are compared against standard NN feedback controllers through numerical simulations of two highdimensional nonlinear optimal control problems (OCPs): stabilization of an unstable Burgers-type partial differential equation (PDE), and altitude and course tracking for a six degree-of-freedom (6DoF) unmanned aerial vehicle (UAV). The simulations demonstrate that standard NNs can fail to stabilize the dynamics even when trained well, while the proposed architectures are always at least locally stable. Moreover, the proposed controllers are found to be near-optimal in testing.

show abstract

Aggressive Online Control of a Quadrotor via Deep Network Representations of Optimality Principles

Cited by 16 publications

References 15 publications

The Artificial Intelligence behind the winning entry to the 2019 AI Robotic Racing Competition

The Artificial Intelligence behind the winning entry to the 2019 AI Robotic Racing Competition

Neural network optimal feedback control with enhanced closed loop stability

Neural Network Optimal Feedback Control with Guaranteed Local Stability

Contact Info

Product

Resources

About