Lyapunov Stability
Overview
Lyapunov stability asks whether a system’s behavior stays bounded near an equilibrium or converges toward it, a central concern in control, learning, and physical modeling. Researchers construct Lyapunov functions and use tools such as linear matrix inequalities, integral quadratic constraints, and Foster–Lyapunov drift conditions to certify stability in deterministic, stochastic, and learned systems. Neural networks, reinforcement-learning controllers, optimization algorithms, and sampling-based controllers are increasingly designed or analyzed with these certificates, often alongside benchmarks in robotics and nonlinear control. A key challenge is balancing guarantees that are rigorous and computable with ones that are not overly conservative, especially when stability is only local, noise or approximation errors persist, or systems have multiple attractors.
Latest papers
23 papers2 papers in the last four weeks, down 33% on the four weeks before. 0.0% of all new papers.