Learning Control Policies for Stochastic Systems with Reach-avoid Guarantees
This addresses the challenge of ensuring safety and stability in stochastic control systems, offering a foundational approach with broad applications in robotics and autonomous systems.
The paper tackles the problem of learning controllers for stochastic nonlinear systems with formal reach-avoid guarantees, introducing a novel method based on reach-avoid supermartingales (RASMs) that provides guarantees with a tolerable probability threshold over infinite time, validated on 3 stochastic nonlinear reinforcement learning tasks.
We study the problem of learning controllers for discrete-time non-linear stochastic dynamical systems with formal reach-avoid guarantees. This work presents the first method for providing formal reach-avoid guarantees, which combine and generalize stability and safety guarantees, with a tolerable probability threshold $p\in[0,1]$ over the infinite time horizon. Our method leverages advances in machine learning literature and it represents formal certificates as neural networks. In particular, we learn a certificate in the form of a reach-avoid supermartingale (RASM), a novel notion that we introduce in this work. Our RASMs provide reachability and avoidance guarantees by imposing constraints on what can be viewed as a stochastic extension of level sets of Lyapunov functions for deterministic systems. Our approach solves several important problems -- it can be used to learn a control policy from scratch, to verify a reach-avoid specification for a fixed control policy, or to fine-tune a pre-trained policy if it does not satisfy the reach-avoid specification. We validate our approach on $3$ stochastic non-linear reinforcement learning tasks.