作者: Tae H. Lee , Hieu M. Trinh , Ju H. Park
DOI: 10.1109/TNNLS.2017.2760979
关键词: Point (geometry) 、 Computational complexity theory 、 Artificial neural network 、 Applied mathematics 、 Lyapunov exponent 、 Lyapunov equation 、 Stability (learning theory) 、 Computer science 、 Lyapunov redesign 、 Symmetric matrix
摘要: This paper presents two novel Lyapunov functionals for analyzing the stability of neural networks with time-varying delay. Based on our newly proposed and a relaxed Wirtinger-based integral inequality, new criteria are derived in form linear matrix inequalities. A comprehensive comparison results is given to illustrate from both conservative computational complexity point views.