Bifurcation Dynamics Analysis of A Class of Fractional Neural Networks with Mixed Delays

In this paper, a three-dimensional integer order neural network with mixed time delays is extended to fractional order form. By introducing three virtual neurons, the network proposed can be transformed into a six dimensional fractional order system only involving discrete time delays. The critical threshold with respect to the system delay parameter is obtained, by employing the stable principle and the fractional order theory. The correctness of the achievements is verified by numerical simulation. Ultimately, it is observed that the introduction of fractional order can expand the stability region of the original system selecting discrete delays as bifurcation parameter.