Synchronization-based approach for parameters identification in delayed chaotic neural networks

In this paper, an adaptive procedure to the problem of synchronization and parameters identification for chaotic neural networks with time-varying delay is introduced by combining the adaptive control and linear feedback with appropriate update law. Based on the invariance principle of functional differential equations, all the connection weight matrices can be efficiently estimated according to a simple, rigorous, and systematic technique. This approach is also able to track the changes in the operating parameters of the experimental neural networks rapidly. The speed of synchronization and parameters estimation can be adjusted under the adaptive gain properly chosen. In addition, the method is simple to implement in practice, and it is quite robust against the effect of slight noise in the given time series and the estimated value of a parameter fluctuates around the correct value.