Multiple Mittag-Leffler stability of fractional-order Cohen-Grossberg neural networks with non-monotonic piecewise linear activation functions

摘要

This paper take into account multiple Mittag-Leffler stability of neural networks with non-monotonic piecewise linear activation functions. We have set up multiple sufficient conditions in order to find out 5ⁿ, and 3ⁿequilibrium points of which is local Mittag-Leffler stability by making use of the related knowledge of fractional-order ordinary differential equations and fixed point theorem, non-smooth analysis knowledge. Fractional-order Cohen-Grossberg neural networks with non-monotonic linear activation functions possesses greater capacity than the ones with Mexican-hat-type activation function. A numerical example is manufactured to test and verify the authenticity of the theoretical results.