The Stability Analysis Of Uncertain Stochastic Neural Networks With Time-varying Mixed Delays

Neural networks,which simulates human nervous system thinking ways and is equipped with self-learning,association memory and robustness and so on,is a kind of nonlinear dynamical system in order to solve our production requirement in reality.Neural networks have been so widely applied to many important fields such as artificial intelligence,pattern recognition,optimization problems and so on due to neural networks is provided with self-learning,robustness and so on.And then the study of neural networks can help us learn ourselves.Hence the study of stability of neural networks has not only very high value of application,but also has cognitive value.In this thesis,we investigate the stability problem of uncertain stochastic neural networks with discrete and distributed time-delays,and obtain results of the system stability.By improving existent model and assumption condition of neural networks,in this thesis,the delay-dependent mean square exponential stability of uncertain stochastic neural networks with discrete and distributed time-delays is investigated.Then,by analyzing existent research findings,the almost sure exponential stability of uncertain stochastic neural networks with discrete delays is further studied.In this thesis,by constructing mainly a suitable Lyapunov function,and utilizing Lyapunov function method,stochastic analysis method,matrix inequalities and fundamental inequality,the criterion of stability of uncertain stochastic neural networks with time-delays are obtained.The conditions obtained here are expressed in terms of LMIs whose feasibility can be checked easily by Matlab toolbox.