摘要
基于最优性的充要条件 ,提出了一种解线性约束非线性凸规划的新神经网络 ,构造了恰当的Lyapunov函数 ,证明了其稳定性 .该模型不需要设定网络参数 ,能同时求解原问题与对偶问题 ,并且当目标函数严格单调时 ,它能大范围渐近收敛于原问题的精确解 .模拟实验表明新模型不仅可行 。
This paper presents a new neural network for nonlinear convex programming problems with linear constraints, defines its Lyapunov function, and proves its stability. There need be no parameter in the proposed neural network, which can solve simultaneously the primal and dual problems, and guarantee to asymptotically converge to an exact optimal solution in the large when the objective function is strictly convex. The feasibility and effectiveness of the proposed neural network are supported by the simulation experiments.
出处
《西安电子科技大学学报》
EI
CAS
CSCD
北大核心
2002年第1期52-55,共4页
Journal of Xidian University
基金
陕西师范大学校级重点科研项目
关键词
线性约束
非线性规则
神经网络
稳定性
收剑性
linear constraints
nonlinear programming
neural network
stability
convergence
