摘要
本文提出了线性约束条件下任意凸规划的神经网络新模型,与原模型相比,缩小了规模,便于工程实现。网络是全局稳定的,并能收敛到最优点。
This Paper presents new neural network models for convex programming with linear constraints. Compared with original models, the new ones have smaller scale and, so they are easy to be implemented. The works are globally stable and can converge to the optimal point.
出处
《信号处理》
CSCD
2001年第2期104-109,共6页
Journal of Signal Processing
基金
国家自然科学基金
原邮电部预研基金资助课题