Data-Driven Safety Controller Synthesis for Unknown Systems with Wireless Communication Networks

This paper offers a formal data-driven scheme for constructing control barrier certificates (CBC) and synthesizing safety controllers for discrete-time control systems. Our framework accommodates scenarios where the mathematical model is unknown, while also considering the presence of wireless communication networks between sensor-controller and controller-actuator links. While existing literature extensively addresses the design of CBC, there has been a notable lack of attention in incorporating wireless communication networks to tackle potential packet losses. This gap poses a greater challenge when considering the absence of knowledge about the system’s model, a crucial aspect in real-world applications. Given a particular rank condition for unknown wirelessly-connected systems, our method provides a linear matrix inequality, constructed based on two input-output trajectories of the system, offering a probabilistic safety assurance across an infinite time horizon. We showcase the efficacy of our data-driven approach over a wirelessly-connected synchronous motor with an unknown model.