Dual‐Channel Event‐Triggered Model‐Free Adaptive Control for Discrete‐time Nonlinear Networked Control Systems

Abstract

In this article, a novel dual‐channel event‐triggered model‐free adaptive control (DCET‐MFAC) algorithm is investigated to solve the tracking control problem for nonlinear networked control systems (NCSs). Two independent event‐triggered mechanisms (ETMs) are designed within the MFAC framework to conserve communication resources in both the sensor‐controller and controller‐actuator channels. First, by utilizing a new model‐free state predictor, an ETM is designed to conserve the communication resources in the sensor‐controller channels. Second, another ETM is designed combining with the MFAC algorithm to minimize the network communication usage in the controller‐actuator channel, which further conserves the communication resources. Furthermore, the proposed algorithm completely relies on I/O data, avoiding the mathematical modeling of complex nonlinear systems, thereby enhancing its applicability to scenarios that are difficult to model. Finally, the validity of the proposed algorithm is demonstrated through theoretical analysis and two simulation cases, including a numerical example and a steam‐water heat exchanger system.