摘要
本文提出一种基于约束区域的神经网络模型,构造了它的Liapunov能量函数,运用LaSalle不变性原理证明了它的大范围渐近稳定性,探讨了收敛速率,并给出在求解约束二次规划中的应用。
This paper presents a kind of neural network model based on the constraint domain, defines its Liapunov energy function, shows its global asymptotic stability by using LaSalle invariant principle, and discusses its convergence ratio. It also gives the model's application in quadratic programming problems with constraints.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
1998年第4期474-478,共5页
Pattern Recognition and Artificial Intelligence
关键词
神经网络
收敛速率
二次规划
规划
Neural Network, Global Asymptotic Stability, Convergence Ratio, Quadratic Programming Problems
