湘南学院学报
湘南學院學報
상남학원학보
JOURNAL OF XIANGNAN UNIVERSITY
2005年
5期
1-8
,共8页
Gohen-Grossberg神经网络%渐近稳定性%李雅普诺夫泛函
Gohen-Grossberg神經網絡%漸近穩定性%李雅普諾伕汎函
Gohen-Grossberg신경망락%점근은정성%리아보낙부범함
Cohen-Grossberg neural networks%Asymptotic stability%Lya-punovfunctional
研究多时滞Cohen-Grossberg神经网络的全局渐近稳定性,利用新的不等式技术,同伦映射及李雅普诺夫泛函方法,我们获得了几个新的准则.同已有的结果相比,我们不要求函数有界,可微或严格增加等条件.我们的结果更容易验证且具有较少的限制.此外,给出了一个例子来证明我们的结果的优势.
研究多時滯Cohen-Grossberg神經網絡的全跼漸近穩定性,利用新的不等式技術,同倫映射及李雅普諾伕汎函方法,我們穫得瞭幾箇新的準則.同已有的結果相比,我們不要求函數有界,可微或嚴格增加等條件.我們的結果更容易驗證且具有較少的限製.此外,給齣瞭一箇例子來證明我們的結果的優勢.
연구다시체Cohen-Grossberg신경망락적전국점근은정성,이용신적불등식기술,동륜영사급리아보낙부범함방법,아문획득료궤개신적준칙.동이유적결과상비,아문불요구함수유계,가미혹엄격증가등조건.아문적결과경용역험증차구유교소적한제.차외,급출료일개례자래증명아문적결과적우세.
In this paper, we study the global asymptotic stability of Cohen-Grossbergneural networks involving multiple delays. By using new inequality techniques, homotopic mapping and constructing a new type Lyapunov func tional method, we obtain some new criteria guaranteeing global asymptoticstability. In our results, we do not require activation functions to be bounded,differentiable or strictly increasing. The obtained results are better than those in literatures. The presented results are more easily to verify and turn out tobe less restrictive than those given in the earlier literature. One example is also worked out to demonstrate the advantages of our results.
