Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
166 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
42 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Physics-Informed Neural Network Lyapunov Functions: PDE Characterization, Learning, and Verification (2312.09131v3)

Published 14 Dec 2023 in math.OC, cs.LG, cs.SY, and eess.SY

Abstract: We provide a systematic investigation of using physics-informed neural networks to compute Lyapunov functions. We encode Lyapunov conditions as a partial differential equation (PDE) and use this for training neural network Lyapunov functions. We analyze the analytical properties of the solutions to the Lyapunov and Zubov PDEs. In particular, we show that employing the Zubov equation in training neural Lyapunov functions can lead to approximate regions of attraction close to the true domain of attraction. We also examine approximation errors and the convergence of neural approximations to the unique solution of Zubov's equation. We then provide sufficient conditions for the learned neural Lyapunov functions that can be readily verified by satisfiability modulo theories (SMT) solvers, enabling formal verification of both local stability analysis and region-of-attraction estimates in the large. Through a number of nonlinear examples, ranging from low to high dimensions, we demonstrate that the proposed framework can outperform traditional sums-of-squares (SOS) Lyapunov functions obtained using semidefinite programming (SDP).

Definition Search Book Streamline Icon: https://streamlinehq.com
References (45)
  1. FOSSIL: a software tool for the formal synthesis of Lyapunov functions and barrier certificates using neural networks. In Proc. of HSCC, pages 1–11, 2021.
  2. Formal synthesis of Lyapunov neural networks. IEEE Control Systems Letters, 5(3):773–778, 2020.
  3. Automated and sound synthesis of Lyapunov functions with smt solvers. In Proc. of TACAS, pages 97–114. Springer, 2020.
  4. Optimal control and viscosity solutions of Hamilton-Jacobi-Bellman equations, volume 12. Springer, 1997.
  5. Dynamical Systems: Stability Theory and Applications, volume 35. Springer, 1967.
  6. A generalization of zubov’s method to perturbed systems. SIAM Journal on Control and Optimization, 40(2):496–515, 2001.
  7. Neural Lyapunov control. Advances in Neural Information Processing Systems, 32, 2019.
  8. Learning lyapunov functions for hybrid systems. In Proc. of HSCC, pages 1–11, 2021.
  9. Learning region of attraction for nonlinear systems. In Proc. of CDC, pages 6477–6484. IEEE, 2021.
  10. Viscosity solutions of hamilton-jacobi equations. Transactions of the American mathematical society, 277(1):1–42, 1983.
  11. Counter-example guided synthesis of neural network lyapunov functions for piecewise linear systems. In Proc. of CDC, pages 1274–1281, 2020.
  12. Lyapunov-stable neural-network control. In Proc. of RSS, 2021.
  13. Safe control with learned certificates: A survey of neural Lyapunov, barrier, and contraction methods for robotics and control. IEEE Transactions on Robotics, 2023.
  14. Z3: An efficient smt solver. In Proc. of TACAS, pages 337–340. Springer, 2008.
  15. Lawrence C Evans. Partial Differential Equations, volume 19. American Mathematical Society, 2010.
  16. Lyapunov-net: A deep neural network architecture for lyapunov function approximation. In 2022 IEEE 61st Conference on Decision and Control (CDC), pages 2091–2096. IEEE, 2022.
  17. dreal: an SMT solver for nonlinear theories over the reals. In Proc. of CADE, pages 208–214, 2013.
  18. Peter Giesl. Construction of Global Lyapunov Functions Using Radial Basis Functions, volume 1904. Springer, 2007.
  19. Review on computational methods for Lyapunov functions. Discrete & Continuous Dynamical Systems-B, 20(8):2291, 2015.
  20. Lars Grüne. Computing Lyapunov functions using deep neural networks. Journal of Computational Dynamics, 8(2), 2021.
  21. Nonlinear Dynamical Systems and Control: A Lyapunov-based Approach. Princeton University Press, 2008.
  22. Extreme learning machine: theory and applications. Neurocomputing, 70(1-3):489–501, 2006.
  23. Converse Lyapunov functions and converging inner approximations to maximal regions of attraction of nonlinear systems. In Proc. of CDC, pages 5312–5319. IEEE, 2021.
  24. Data-driven computational methods for the domain of attraction and Zubov’s equation. arXiv preprint arXiv:2112.14415, 2021.
  25. Simulation-guided lyapunov analysis for hybrid dynamical systems. In Proc. of HSCC, pages 133–142, 2014.
  26. Estimation of region of attraction for polynomial nonlinear systems: A numerical method. ISA Transactions, 53(1):25–32, 2014.
  27. A sum-of-squares approach to the stability and control of interconnected systems using vector Lyapunov functions. In Proc. of ACC, pages 5022–5028. IEEE, 2015.
  28. Artificial neural networks for solving ordinary and partial differential equations. IEEE Transactions on Neural Networks, 9(5):987–1000, 1998.
  29. Towards learning and verifying maximal neural lyapunov functions. In Proc. of CDC, 2023.
  30. Compositionally verifiable vector neural Lyapunov functions for stability analysis of interconnected nonlinear systems. In Proc. of ACC, submitted, 2024.
  31. Y. Long and M.M. Bayoumi. Feedback stabilization: control lyapunov functions modelled by neural networks. In Proc. of CDC, pages 2812–2814 vol.3, 1993.
  32. Aleksandr Mikhailovich Lyapunov. The general problem of the stability of motion. International Journal of Control, 55(3):531–534, 1992.
  33. Smooth converse lyapunov-barrier theorems for asymptotic stability with safety constraints and reach-avoid-stay specifications. Automatica, 144:110478, 2022.
  34. Help on SOS. IEEE Control Systems Magazine, 30(4):18–23, 2010.
  35. On the construction of Lyapunov functions using the sum of squares decomposition. In Proc. of CDC, volume 3, pages 3482–3487. IEEE, 2002.
  36. A tutorial on sum of squares techniques for systems analysis. In Proc. of ACC, pages 2686–2700. IEEE, 2005.
  37. Danil V Prokhorov. A lyapunov machine for stability analysis of nonlinear systems. In Proc. of ICNN, volume 2, pages 1028–1031. IEEE, 1994.
  38. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational physics, 378:686–707, 2019.
  39. Constructive Nonlinear Control. Springer Science & Business Media, 2012.
  40. Pierpaolo Soravia. Optimality principles and representation formulas for viscosity solutions of hamilton-jacobi equations i. equations of unbounded and degenerate control problems without uniqueness. Advances in Differential Equations, 4(2):275–296, 1999.
  41. Stability region analysis using polynomial and composite polynomial Lyapunov functions and sum-of-squares programming. IEEE Transactions on Automatic Control, 53(2):565–571, 2008.
  42. Local stability analysis using simulations and sum-of-squares programming. Automatica, 44(10):2669–2675, 2008.
  43. Maximal Lyapunov functions and domains of attraction for autonomous nonlinear systems. Automatica, 21(1):69–80, 1985.
  44. Neural Lyapunov control of unknown nonlinear systems with stability guarantees. In Advances in Neural Information Processing Systems, 2022.
  45. V. I. Zubov. Methods of A. M. Lyapunov and Their Application. Noordhoff, 1964.
Citations (15)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now