Networked and distributed predictive control of non-linear systems subject to asynchronous communication

The asynchronous communication always exists in networked control systems that consist of interconnected heterogeneous subsystems. To improve the closed-loop performance, a networked and distributed predictive controller is developed for such a system in this study. At each time instant, a virtual trajectory generator for the interconnected agents with guaranteed error bounds is proposed to implement the receding horizon optimisation. Then, a dual-mode strategy is designed for the asynchronous distributed model predictive control to make effective use of computational resources. Conditions for ensuring the feasibility of the proposed algorithm and the stabilisation of the closed-loop networked control systems are developed, respectively. Finally, a simulation example of a set of coupled Van der Pol oscillators is presented to illustrate the results.