This dissertation addresses the sampled-data control problem for dynamic nonlinear systems based on two subject.
One is “How to obtain an improved stability results the sampled-data systems”.
By the construction of the functionals such as newly Lypunov-Krasovskii functional, the looped-functionals, and time-dependent Lyapunov-Krasovskii functionals, and the use of the newly integral inequality,
the stability analysis and controller design for three kinds of nonlinear systems were investigated.
Other is “How to design the control system”.
To access the dynamic systems nearer to the real environment, various constraints is considered when designing the sampled-data controller.
By using the proposed methods on the above topics, three dynamic nonlinear systems were applied.
The first one is the synchronization Lur’e system. A synchronization problem for Lur’e system and delay Lur’e systems with sampled-data control are introduced.
At this time, to reflect noises and perturbations of a designed controller gain, Gaussian process, Bernoulli sequence and random variables are applied to reliable control and uncertainties control scheme.
The second is the stability analysis of neural networks.
The author dealt with two types of neural networks such as chaotic neural networks and neural-network-based systems. By applying the sampled-data control with an actuator saturation, the synchronization for chaotic neural networks is studied.
And, a sampled-data control problem for neural-network-based systems with an optimal guaranteed cost is investigated.
Thirdly, this study addresses sampled-data control in the synchronization problem of complex dynamical networks systems with a coupling time-varying delay.
By constructing the augmented Lyapunov-Krasovskii functionals and using some mathematical techniques, sampled-data synchronization criteria are proposed.
The proposed sufficient conditions for dynamic nonlinear systems with various constraints are obtained as the framework of linear matrix inequalities.
To confirm the validity and superiority of the proposed results, various numerical examples are considered.