作者: Chuandong Li , Xiaofeng Liao , Rong Zhang
DOI: 10.1016/J.CHAOS.2004.09.052
关键词: Mathematics 、 Control theory 、 Interconnection matrix 、 Linear matrix inequality 、 Set (abstract data type) 、 Constant (mathematics) 、 Content-addressable memory 、 Artificial neural network 、 Exponential stability 、 Delay dependent 、 General Mathematics
摘要: Abstract For bi-directional associative memory (BAM) neural networks (NNs) with different constant or time-varying delays, the problems of determining exponential stability and estimating convergence rate are investigated in this paper. An approach combining Lyapunov–Krasovskii functional linear matrix inequality (LMI) is taken to study problems, which provide bounds on interconnection activation functions, so as guarantee system’s stability. Some criteria for stability, give information delay-dependent property, derived. The results obtained paper one more set easily verified guidelines delayed BAM (DBAM) networks, less conservative restrictive than ones reported far literature. typical examples presented show application