Design of Controllers for Finite-Time Robust Stabilization of Inertial Delayed Neural Networks with External Disturbances
Hong, Nan,Zhou, Zichuan,Xiu, Ruihong
DOI: https://doi.org/10.1007/s11063-023-11206-3
IF: 2.565
2023-03-14
Neural Processing Letters
Abstract:This paper studies the problem of finite-time robust stabilization of inertial delayed neural networks with external disturbances. The finite-time stability research of inertial neural networks can be applied to important fields such as secure communication, so it has great significant research value. However, up to now, there are few previous studies on the finite-time stability of inertial neural networks, thus this paper makes up for this gap. Based on the actual communication networks, we improve the model of inertial neural networks, adding uncertainties and external disturbances. Unlike many previous papers based on scalar sign function, this paper introduces vector sign function, combines the constructed Lyapunov function, some inequality conditions, and related lemmas to design two effective controllers composed of and , which can handle uncertainties and external disturbances of the neural networks well, and realize finite-time robust stability of the neural networks with external disturbances. In addition, the theoretical part of this paper estimates an upper bound on the settling time for the system to reach stability. Under the conditions of a certain strategy, we further optimize the extremes of the settling time so that the system reaches stability in a shorter time. Our results improve and extend some recent works. Finally, two examples are given to verify the validity and correctness of the designed controllers by numerical simulations using MATLAB tool.
computer science, artificial intelligence