Data-Driven Control for Nonlinear Networked Systems on Basis of Two Description Coding Scheme

This paper deals with the tracking control problem for a type of nonlinear networked systems by utilizing the data-driven control algorithm. A scalar-uniform-quantization-based two description coding mechanism is employed, with the hope to alleviate communication bandwidth limitation and packet dropouts phenomenon during data transmission. Such a mechanism first encodes the source signal into two descriptions, and then send the coded information to the decoder via independent network channels. The random packet dropouts phenomenon is considered that is described via Bernoulli sequences. A data-driven control scheme is designed, ensuring that the system output can track the desired reference. By resorting to the dynamic linearization method in combination with certain convex optimization technique, the desired control protocol is established. Finally, the usefulness of our designed control approach is demonstrated via two examples. <italic>Note to Practitioners</italic>&#x2014;Networked systems are composed of sensors, actuators and controllers connected through a shared digital communication network, which have advantages in long-distance control operations. It should be noted that, however, during network communication, data packets would probably suffer dropout due to various reasons such as limited bandwidth, environment abrupt change, device aging and so on, which will degrade system performance or even damage the system. However, so far, in the context of data-driven control, where data play an essential role, such a data-lost phenomenon has not been fully examined when designing the control strategies. Therefore, the main motivation of this paper is to study the tracking control problem of networked systems with data packet dropouts. In this work, we provide an efficient control algorithm that enables the system to track the desired target in spite of packet loss. The core of this method is to transmit data through two independent network channels so as to reduce the probability of the transmission information being completely destroyed at the same time. We show the feasibility and efficiency of our data-driven control scheme on basis of two description coding mechanism by simulation. In future research, we will extend the method to target tracking in multi-sensor networked system subject to data loss.