Research On Synchronous And Anti-synchronous Adaptive Control Of Memristor-based Neural Networks

With the development of memristor,people's living standards are constantly improving.Memristor plays an indispensable role in daily circuits.Because of its memory function,it is considered to be the best to simulate human brain synapses circuit components.In recent years,as the advantages of memristor-based neural networks have gradually emerged,they have been highly concerned by scientists.The research field of memristor-based neural networks is different and the need to solve the problem,the memristor-based neural network on the real number domain has been unable to solve some specific problems,such as the detection of symmetry problems and XOR problem,etc.,such problems need to be solved on the complex number domain.Therefore,the memristor-based complex-valued neural networks have been proposed by scholars.Synchronization and anti-synchronization are important dynamic behaviors in memristor-based neural networks,and they have important application prospects in artificial intelligence collaborative control and secure communication.However,in general synchronous system of memristor-based neural networks,it is not possible to make self-regulation to overcome problems such as parameter mismatch,etc.Therefore,we begin to study adaptive synchronization.This paper mainly uses adaptive technology.By constructing suitable Lyapunov functions,a suitable adaptive feedback controller is constructed in the response system of the synchronous system,and the self-adjusting function of the system is used to achieve adaptive synchronization and anti-synchronization.The synchronization and anti-synchronization of memristorbased neural networks can also be applied to the field of information security,for example: image encryption and associative memory.Chapters 2-3 of this article mainly introduce the application of adaptive synchronization and anti-synchronization to memristor-based real-valued neural networks and memristor-based complex-valued neural networks.Due to the random fluctuations caused by neurotransmitter release and other probabilities,synaptic transmission is a noisy process.Therefore,random interference is unavoidable.Chapter 4 adds the random disturbance factor on the basis of the previous article,and gives the algebraic conditions of adaptive synchronization and anti-synchronization in complex domain.