We consider the problem of learning maximal Lyapunov functions to compute the maximal region of attraction of nonlinear dynamical systems using neural networks. Two obstacles limit existing neural-network approaches: 1) finite-width approximations can fail to preserve the strict Lyapunov inequalities near the equilibrium, where they become arbitrarily weak, and 2) training is necessarily performed on a bounded computational domain for which the classical Lyapunov and Zubov theorems do not directly apply. To address these issues, first we derive Lyapunov and Zubov theorems directly on a prescribed compact computational domain, characterizing the maximal region of attraction within that domain as the solution of a boundary-value PDE. Second, we introduce the Taylor Lyapunov network, an architecture that prescribes the quadratic leading-order term via the system's linearization and represents the higher-order remainder through a structured cubic factorization, guaranteeing that any network in this class is a valid local Lyapunov function by construction. Finally, we prove that Taylor Lyapunov networks are universal approximators of maximal Lyapunov functions on compact domains, and combine these results into an unsupervised, physics-informed training procedure. Numerical experiments on systems of varying dimensions and stability types show that the proposed architecture is efficient at estimating the maximum region of attraction.
Figures & tables
Figure 1: Taylor Lyapunov network function for globally stable system ( 27 ). The color scale represents the level sets of the Lyapunov function. The estimated maximal RoA is colored. Arrows indicate the flow of the original system.
Figure 2: Taylor Lyapunov network function for locally stable system ( 28 ). The color scale represents the sub-level sets of the Lyapunov function. The estimated RoA is the colored area. Arrows indicate the flow of the original system.
Figure 3: Taylor Lyapunov network function for the Van der Pol oscillator with μ=1 . The color scale represents the sub-level sets of the Lyapunov function. The estimated RoA is the colored area. Arrows indicate the flow of the original system. The dashed line is the estimated RoA obtained using SOS.
Architecture
Training
( 27 )
( 28 )
( 29 ), μ=1
Lyapunov-Net
LyZNet
0.1%
≃0%
≃0%
Adv.
0.2%
0.11%
0.04%
TNLN
LyZNet
76.1%
79.1%
≃0%
Adv.
83.8%
81.3%
98.6%
Table 1: Ablation study and comparison with other algorithms (Lyapunov-Net: Gaby et al. (2022) , TNLN: Taylor Neural Lyapunov Network, LyZNet: training of Liu et al. (2024) , Adv.: adversarial training). Entries are the RoA coverage for each system.
Dynamical systems studies how states evolve over time, including stability, feedback, chaos, and transitions in physical, biological, and engineered systems.
Constructing Lyapunov functions for nonlinear dynamical systems is a central problem in stability analysis, yet remains challenging. Lyapunov functions are commonly characterized as solutions to first-order partial differential equations (PDEs), but these solutions are typically obtained for single systems, limiting their reuse across systems. In this paper, we study the Lyapunov solution operator that maps a vector field to the corresponding Lyapunov function defined by a dissipation-based Lyapunov PDE. We establish that, on compact subsets of the domain of attraction and under exponential stability assumptions, this operator is well-defined, unique, and continuous with respect to perturbations of both the vector field and the dissipation function. These results provide a theoretical foundation for approximating Lyapunov functions uniformly over families of nonlinear systems. Building on these theoretical foundations, we employ Fourier Neural Operators (FNOs) as a data-driven approximation of the Lyapunov solution operator. Numerical experiments demonstrate that a single trained operator can accurately approximate the numerical Lyapunov functions across parameterized families of dynamics. This illustrates the potential of neural operators for approximating Lyapunov functions.
Amartya Mukherjee, Maxwell Fitzsimmons, David C. Del Rey Fernández +1
Department of Applied Mathematics, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1
We introduce Hyperlyve (Hyperplane Partitioned Lyapunov Verification and Enumeration), a framework for proving the asymptotic stability of nonlinear dynamical systems, including those with neural network controllers. Hyperlyve learns and formally certifies continuous piecewise-affine (CPA) neural Lyapunov functions within a counterexample-guided inductive synthesis (CEGIS) loop. Hyperlyve's key innovation is a parallel algorithm that enumerates the hyperplane arrangement formed by the ReLU activations of deep networks and certifies that Lyapunov conditions hold within each polyhedral region. We derive sound and complete regional tests to verify the asymptotic stability of the exact smooth dynamics (rather than a piecewise-linear approximation) within a compact domain of interest. Furthermore, we formally characterize the computational complexity of the CPA neural Lyapunov verification problem as co-NP-complete, filling a key gap in the literature. Empirical results demonstrate that Hyperlyve enumerates ReLU networks over 150-times faster than competing tools, which contributes to Hyperlyve proving the asymptotic stability of systems that are currently intractable for state-of-the-art verifiers.
Brian Yarbrough, Jesse Wayment, Jingbo Wang +2
The authors are with the Elmore Family School of Electrical Engineering at Purdue University, West Lafayette, IN 47906, USA.
This work investigates the joint learning of neural controllers and control Lyapunov functions (CLFs) for nonlinear discrete-time systems. We exploit the structural properties of Input Convex Neural Networks and Polyak-Łojasiewicz Networks to guarantee positive definiteness of CLF candidates by construction, reducing the conditions requiring formal verification. We develop a mixed-integer programming framework to formally verify the remaining stability conditions and introduce an iterative region of attraction (ROA) expansion procedure to enlarge the certified ROA without retraining. Experiments on an inverted pendulum and a path-tracking system demonstrate the advantages of structured neural CLFs, particularly PLNets, and illustrate how their structural bounds influence the geometry of the certified regions. Our code is available at: https://github.com/JLedesMart/Structured-Neural-Control-Lyapunov-Functions.
Jose Luis Ledesma Martin, Patricia Pauli
Department of Mechanical Engineering, Delft University of Technology, Delft, 2628 CD Delft, The Netherlands · Department of Mechanical Engineering, Eindhoven University of Technology, 5612 AZ Eindhoven, The Netherlands