一类广义Markov跳变随机反应扩散神经网络的稳定性研究

Research on Stability for a Class of Generalized Markovian Jumping Reaction-diffusion Neural Networks

通过构造合适的Lyapunov泛函,并根据It?公式沿该神经网络求Lyapunov泛函的随机导数,在Dirichlet边界条件下讨论一类具有Markov跳变参数、随机干扰和混合时滞的广义反应扩散神经网络的全局均方鲁棒渐进稳定性,得到依赖于时滞、扩散系数和扩散空间的稳定性判据。最后,通过数值模拟说明所得结论判据的有效性。

By constructing some suitable Lyapunov functionals and calculating the stochastic derivative of these Lyapunov functionals along neural networks using differential formula,the global mean square and asymptotic robust stability for a class of generalized Markovian jumping stochastic reaction-diffusion neural networks with mixed delays is discussed under Dirichlet boundary conditions. The obtained criteria not only depend on delays but also depend on diffusion coefficients and diffusion spaces. Finally,a numerical example is provided to show the validity of the obtained results criteria.