ℒ<sub>1</sub>-Adaptive MPPI Architecture for Robust and Agile Control of Multirotors

Pravitra, Jintasit; Ackerman, Kasey A.; Cao, Chengyu; Hovakimyan, Naira; Theodorou, Evangelos A.

doi:10.1109/iros45743.2020.9341154

Cited by 33 publications

(25 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The performance benefit from the adaptation law is at least 50% greater than the data driven model augmentation provided by GP-MPC, without the large computational overhead. When implementing L 1 -MPPI, we saw the same high frequency oscillatory content in the body rates that is evident in the videos released with [34]. While MPPI has shown outstanding results on ground vehicles [45,46], we believe the sampling approach suffers from the curse of dimensionality problem induced by the 4D space.…”

Section: Discussionmentioning

confidence: 57%

“…Similarly, in [34] a Model Predictive Path Integral (MPPI) controller was coupled with a nonlinear reference model L 1 adaptive controller for agile quadrotor flight, however the authors have not shown feasibility of the proposed method on real hardware. No analysis of the adaptive control signal is provided, however video footage released of the simulation performance of the proposed L 1 -MPPI architecture indicates highly oscillatory control performance in the first-person camera view 1 .…”

Section: Related Workmentioning

confidence: 99%

“…We implement the L 1 adaptive controller using a nonlinear reference model [40], which estimates both matched and unmatched uncertainties using a piecewise constant adaptation law [41,42]. The derivation is similar to [34], however we account for the uncertainties directly at the rotor thrust level. First, define R I B = e B x , e B y , e B z as the rotation matrix from the body frame to the inertial frame.…”

Section: L 1 -Adaptive Augmentationmentioning

confidence: 99%

“…We begin by implementing several state of the art controllers in simulation including a Single Rotor Thrust NMPC (SRT-NMPC), a data-driven NMPC (GP-MPC) from [9], MPPI with Baseline Control from [34], INDI-NMPC from [16], and our proposed L 1 -NMPC with and without an aerodynamic model included in the underlying NMPC. All baseline controllers have an aerodynamic model enabled by default.…”

Section: A Simulationmentioning

confidence: 99%

See 3 more Smart Citations

Performance, Precision, and Payloads: Adaptive Nonlinear MPC for Quadrotors

Hanover

Foehn

Sun

et al. 2022

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

Agile quadrotor flight in challenging environments has the potential to revolutionize shipping, transportation, and search and rescue applications. Nonlinear model predictive control (NMPC) has recently shown promising results for agile quadrotor control, but relies on highly accurate models for maximum performance. Hence, model uncertainties in the form of unmodeled complex aerodynamic effects, varying payloads and parameter mismatch will degrade overall system performance. In this letter, we propose L1 -NMPC, a novel hybrid adaptive NMPC to learn model uncertainties online and immediately compensate for them, drastically improving performance over the non-adaptive baseline with minimal computational overhead. Our proposed architecture generalizes to many different environments from which we evaluate wind, unknown payloads, and highly agile flight conditions. The proposed method demonstrates immense flexibility and robustness, with more than 90% tracking error reduction over nonadaptive NMPC under large unknown disturbances and without any gain tuning. In addition, the same controller with identical gains can accurately fly highly agile racing trajectories exhibiting top speeds of 70 km/h, offering tracking performance improvements of around 50% relative to the non-adaptive NMPC baseline.

show abstract

Section: Discussionmentioning

confidence: 57%

Section: Related Workmentioning

confidence: 99%

Section: L 1 -Adaptive Augmentationmentioning

confidence: 99%

Section: A Simulationmentioning

confidence: 99%

See 2 more Smart Citations

Performance, Precision, and Payloads: Adaptive Nonlinear MPC for Quadrotors

Hanover

Foehn

Sun

et al. 2022

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

show abstract

“…Additionally, we validate our approach on real hardware, as opposed to merely in numerical simulations in [12], [13]. We note that L 1 adaptive control has been combined with model predictive control (MPC) with application to quadrotors [15], and it has been used for safe learning and motion planning applicable to a broad class of nonlinear systems in robotic applications [16]- [18]. To put things into perspective, this paper is focused on applying the L 1 adaptive control architecture to robustify an RL policy.…”

Section: A Related Workmentioning

confidence: 94%

Improving the Robustness of Reinforcement Learning Policies with $\mathcal{L}_{1}$ Adaptive Control

Cheng,

Zhao,

Wang

et al. 2021

Preprint

View full text Add to dashboard Cite

A reinforcement learning (RL) control policy trained in a nominal environment could fail in a new/perturbed environment due to the existence of dynamic variations. For controlling systems with continuous state and action spaces, we propose an add-on approach to robustifying a pre-trained RL policy by augmenting it with an L1 adaptive controller (L1AC). Leveraging the capability of an L1AC for fast estimation and active compensation of dynamic variations, the proposed approach can improve the robustness of an RL policy which is trained either in a simulator or in the real world without consideration of a broad class of dynamic variations. Numerical and real-world experiments empirically demonstrate the efficacy of the proposed approach in robustifying RL policies trained using both model-free and model-based methods. A video for the experiments on a real Pendubot setup is available at https://youtu.be/xgOB9vpyUgE.

show abstract