Safety Barrier Certificates for Stochastic Control Systems with Wireless Communication Networks

TL;DR

This paper develops a formal method using control barrier certificates to ensure safety in stochastic control systems with wireless communication, accounting for packet losses and delays, demonstrated on motor control scenarios.

Contribution

It introduces a systematic sum-of-squares and matrix inequalities approach for synthesizing safety certificates considering wireless network effects.

Findings

Successfully applied to a permanent magnet synchronous motor.

Designed a safety controller for automotive electric steering.

Provided probabilistic safety guarantees within finite time horizons.

Abstract

This work is concerned with a formal approach for safety controller synthesis of stochastic control systems with both process and measurement noises while considering wireless communication networks between sensors, controllers, and actuators. The proposed scheme is based on control barrier certificates (CBC), which allows us to provide safety certifications for wirelessly-connected stochastic control systems. Despite the available literature on designing control barrier certificates, there has been unfortunately no consideration of wireless communication networks to capture potential packet losses and end-to-end delays, which is absolutely crucial in safety-critical real-world applications. In our proposed setting, the key objective is to construct a control barrier certificate together with a safety controller while providing a lower bound on the satisfaction probability of the safety property over a finite time horizon. We propose a systematic approach in the form of sum-of-squares optimization and matrix inequalities for the synthesis of CBC and its associated controller. We demonstrate the efficacy of our approach on a permanent magnet synchronous motor. For the application of automotive electric steering under a wireless communication network, we design a CBC together with a safety controller to maintain the electrical current of the motor in a safe set within a finite time horizon while providing a formal probabilistic guarantee.