摘要
研究时变时滞神经网络的鲁棒稳定性和耗散性问题.充分利用积分项的时滞信息和激励函数条件构造一个合适的增广LK泛函;利用自由矩阵积分不等式处理LK泛函的导数,得到一个低保守性的时滞相关稳定判据;将所获得的结论延伸至神经网络的耗散性分析,并推导出一个确保神经网络严格(X,Y,Z)-γ-耗散的充分条件.最后通过3个数值算例验证了所提出方法的可行性和优越性.
The problems of robust delay-dependent stability and dissipativity for neural networks(NNs) with time-varying delay is investigated. A proper augmented Lyapunov-Krasovskii functional(LKF) is constructed, which fully utilizes the information of time-delay in integral term and the neuron activation function conditions. Then, by employing the free- matrix-based integral inequality to handle the derivative of the LKF, a less conservative delay-dependent stability criteria is obtained. The acquired conclusion is extended to the analysis of dissipativity for delayed NNs, and a suficient condition is derived to guarantee the NNs strictly(X, y, Z)--y-dissipative. The superiority and feasibility of presented approaches are verified via three numerical examples.
作者
肖伸平
练红海
陈刚
冯磊
XIAO Shen-pingt LIAN Hong-hai CHEN Gang FENG Lei(School of Electrical and Information Engineering, Hu'nan University of Technology, Zhuzhou 412007, China Key Laboratory for Electric Drive Control and Intelligent Equipment of Hu'nan Province, Zhuzhou 412007, China)
出处
《控制与决策》
EI
CSCD
北大核心
2017年第6期1084-1090,共7页
Control and Decision
基金
国家自然科学基金项目(61672225
61304064)
国家火炬计划项目(2015GH712901)
湖南省自然科学基金项目(2015JJ3064
2015JJ5021)
湖南省教育厅科研基金项目(15B067)
广东省特种光纤材料与器件工程技术研究开发中心开放基金项目
关键词
神经网络
稳定性
耗散性
自由矩阵积分不等式
时变时滞
neural network
stability
dissipativity
free-matrix-based integral inequality
time-varying delay