摘要
利用非线性规划的Log-Sigmoid(L-S)型Lagrange函数及其对偶问题的性质,提出一种求解优化问题的神经网络方法,并讨论网络的收敛性和稳定性条件,算例表明该神经网络能有效求解.
A nural network was pretended on the base of L-S Lagrange function of nonlinear programming and its dual problem.The convergence and stability were discussed and the validity of the networks was demonstrated by several numerical examples.
出处
《湖北大学学报（自然科学版）》
CAS
2012年第2期231-234,共4页
Journal of Hubei University：Natural Science
