Run-Time Safety Monitoring of Neural-Network-Enabled Dynamical Systems

Abstract

Complex dynamical systems rely on the correct deployment and operation of numerous components, with stateof-the-art methods relying on learning-enabled components in various stages of modeling, sensing, and control at both offline and online levels. This paper addresses the run-time safety monitoring problem of dynamical systems embedded with neural network components. A run-time safety state estimator in the form of an interval observer is developed to construct lowerbound and upper-bound of system state trajectories in run time. The developed run-time safety state estimator consists of two auxiliary neural networks derived from the neural network embedded in dynamical systems, and observer gains to ensure the positivity, namely the ability of estimator to bound the system state in run time, and the convergence of the corresponding error dynamics. The design procedure is formulated in terms of a family of linear programming feasibility problems. The developed method is illustrated by a numerical example and is validated with evaluations on an adaptive cruise control system.

Monitoring Object Detection Abnormalities via Data-Label and Post-Algorithm Abstractions

“…In this work, we focus on monitoring methods without exploiting sensor modalities and diversities, where there exist also monitors equipped on the system level for fault checking and recovery [5], [6]. Our attention is further restricted to monitoring modules implemented using learning-based approaches.…”