Pith. sign in

REVIEW 8 cited by

Lyapunov-stable neural-network control

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2109.14152 v1 pith:IN47SP74 submitted 2021-09-29 cs.RO cs.SYeess.SY

Lyapunov-stable neural-network control

classification cs.RO cs.SYeess.SY
keywords controllerlyapunovneural-networkcontrollersapproachclosed-loopconditioncontrol
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Deep learning has had a far reaching impact in robotics. Specifically, deep reinforcement learning algorithms have been highly effective in synthesizing neural-network controllers for a wide range of tasks. However, despite this empirical success, these controllers still lack theoretical guarantees on their performance, such as Lyapunov stability (i.e., all trajectories of the closed-loop system are guaranteed to converge to a goal state under the control policy). This is in stark contrast to traditional model-based controller design, where principled approaches (like LQR) can synthesize stable controllers with provable guarantees. To address this gap, we propose a generic method to synthesize a Lyapunov-stable neural-network controller, together with a neural-network Lyapunov function to simultaneously certify its stability. Our approach formulates the Lyapunov condition verification as a mixed-integer linear program (MIP). Our MIP verifier either certifies the Lyapunov condition, or generates counter examples that can help improve the candidate controller and the Lyapunov function. We also present an optimization program to compute an inner approximation of the region of attraction for the closed-loop system. We apply our approach to robots including an inverted pendulum, a 2D and a 3D quadrotor, and showcase that our neural-network controller outperforms a baseline LQR controller. The code is open sourced at \url{https://github.com/StanfordASL/neural-network-lyapunov}.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 8 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. ShardNet: Training Neural Controllers with Hard, Non-Convex Constraints

    eess.SY 2026-06 unverdicted novelty 8.0

    ShardNet enforces non-convex polyhedral safety constraints in neural controllers by construction via a differentiable projection layer, achieving 100% verified safety and over 3x larger safe sets than prior methods on...

  2. Co-Learning Port-Hamiltonian Systems and Optimal Energy-Shaping Control

    eess.SY 2026-04 unverdicted novelty 7.0

    A physics-informed co-learning method jointly learns port-Hamiltonian dynamics and optimal energy-shaping controllers from data, guaranteeing closed-loop passivity and stability for pendulum regulation and swing-up tasks.

  3. Co-Learning Port-Hamiltonian Systems and Optimal Energy-Shaping Control

    eess.SY 2026-04 unverdicted novelty 7.0

    A co-learning approach jointly optimizes a port-Hamiltonian system model and an energy-balancing passivity-based controller from data via alternating optimization with neural networks that embed structure for guarante...

  4. Certified Training with Branch-and-Bound for Lyapunov-stable Neural Control

    cs.LG 2024-11 unverdicted novelty 7.0

    CT-BaB integrates branch-and-bound during training to tighten certified Lyapunov bounds, yielding neural controllers with 164X larger verifiable ROA and 11X faster verification than CEGIS on a 2D quadrotor.

  5. Lyapunov-Guided Training for Hardware-Safe Neural Networks Under Fixed-Point Arithmetic

    cs.LG 2026-07 conditional novelty 6.0

    Monotone Lyapunov projection of layerwise hidden-state energy suppresses two's-complement overflow under wrapping fixed-point QAT/PTQ, recovering 86.55% MNIST accuracy where unconstrained models collapse to chance.

  6. Set-Based Value Function Characterization and Neural Approximation of Stabilization Domains for Input-Constrained Discrete-Time Systems

    eess.SY 2026-03 conditional novelty 6.0

    Set-based value functions on compact sets characterize domains of stabilization for input-constrained discrete-time systems and are learned via physics-informed NNs to estimate DOS and synthesize controllers.

  7. Adaptive Data Harvesting for Efficient Neural Network Learning with Universal Constraints

    cs.LG 2026-05 unverdicted novelty 4.0

    A reinforcement learning policy learns to adaptively harvest data samples, improving empirical constraint satisfaction and training efficiency for Lyapunov NNs and PINNs.

  8. Bridging Control with Neural Network Verifier alpha-beta-CROWN: A Tutorial

    eess.SY 2026-05 unverdicted novelty 3.0

    Tutorial introducing applications of the existing α,β-CROWN verifier to scalable formal verification of neural network controllers via bound computation and domain partitioning.