State estimation and dissipativity analysis of neural networks with time-varying delays

This thesis is concerned with state estimation and dissipativity analysis of neural networks with time-varying delays. The first objective is to design suitable delay-dependent state estimators to acquire desired neuron states by using the measurements that are contaminated by external disturbances. To ensure satisfactory system performance, another objective is to provide good dissipativity criteria for delayed neural networks to evaluate some performance indices in a unified framework. The results in this thesis are beneficial to the wide applications of neural networks in various fields, like associate memory, optimization, image processing and so on.