Safe Reach Set Computation via Neural Barrier Certificates

TL;DR

This paper introduces a neural network-based method for real-time safety verification of autonomous systems by generating barrier certificates that over-approximate reachable states, ensuring safety or triggering safety measures.

Contribution

It proposes a novel neural barrier certificate approach that enables efficient online safety verification for both linear and nonlinear autonomous systems.

Findings

Effective in online safety verification for autonomous driving scenarios

Generalizes barrier certificates to unseen state space regions

Validated on linear and nonlinear control models

Abstract

We present a novel technique for online safety verification of autonomous systems, which performs reachability analysis efficiently for both bounded and unbounded horizons by employing neural barrier certificates. Our approach uses barrier certificates given by parameterized neural networks that depend on a given initial set, unsafe sets, and time horizon. Such networks are trained efficiently offline using system simulations sampled from regions of the state space. We then employ a meta-neural network to generalize the barrier certificates to state space regions that are outside the training set. These certificates are generated and validated online as sound over-approximations of the reachable states, thus either ensuring system safety or activating appropriate alternative actions in unsafe scenarios. We demonstrate our technique on case studies from linear models to nonlinear control-dependent models for online autonomous driving scenarios.