IJCAS

In this paper, the problem on master-slave synchronization is investigated for a class of delayed neural networks (DNNs) subject to multiple variable disturbances, where the synchronization can be widely utilized in many engineering fields such as teleoperation control, secure communication, and so on. Initially, the unknown disturbances are assumed to be generated by two external systems, in which the additional ones are required to belong to the space of L2[0,+∞). Then based on disturbance-observer-based-control method, H∞ control technique, and composite controller, a sufficient condition on designing the observer gains and controller ones are established to ensure that the overall closed-loop system is asymptotically stable with H∞ control performance. Furthermore, by using matrix transformation technique, the derived condition is converted into the linear matrix inequality (LMI) forms, which can be easily checked and present much less conservatism. Finally, some simulations and comparisons in an example are given to illustrate the effectiveness of our control methods.

Download: http://link.springer.com/article/10.1007/s12555-019-0783-0

Keywords: Disturbance observer-based control, master-slave synchronization, multiple unknown disturbances, neural networks, time-varying delay.