Studies on synchronization and Stabilization of neural Networks
Loading...
Date
item.page.authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The research learning of thesis investigates the synchronization and stabilization criterion for various types of neural networks, namely, coupled, memristor-based, Markovian jump, Takagi-Sugeno fuzzy, fractional-order and stochastic neural networks. The primary motive of this thesis is to substantiate a suitable controller to sustain the synchronization, stabilization and disturbance rejection criterion for the considered neural networks through state feedback controller or by the approach of observer framework. Particularly, these neural networks are commonly experienced with factors like time delay, uncertainties, external disturbances, input saturation and impulse effects. In this thesis, different control strategies, say reliable control, sampled-data control, dynamic output feedback control and resilient control are examined. Additionally, quantization strategies and equivalent-input-disturbance approaches are also employed to sustain the system stability. To be precise, synchronization criterion is investigated for a class of coupled neural networks subject to semi-Markov inner coupling and Markov outer coupling. In addition, the disturbance rejection approach for neural networks subject to external disturbance has been investigated by employing equivalent-input-disturbance estimator. Next, mixed Hand#8734; and passivity based guaranteed cost control problem for stochastic neural networks is discussed via observer framework with quadratic cost function. Finally, finite-time stabilization problem of memristive neural networks has been scrutinized via quantization based memory feedback controller and sampled-data controller. By endowing Lyapunov stability theory, linear matrix inequality technique and algebraic graph theory, a newest set of necessary conditions is derived to assure the required results. These conditions are determined using some advanced integral inequalities
newline