摘要
自组织特征映射神经网络中所有神经元通过相互竞争和自适应学习而形成空间上的有序结构，这种有序结构能反应出输入矢量集合的内在特征。这一特性可成功地应用于解决组合优化问题。本文根据栅阵列排序问题的性质来设定网络中各种经元的相互竞争原则，建立了一个栅阵列排序算法，并取得了非常好的布图结果。本文算法的时间复杂度为O（n×P2lnp），n为线网个数，p主栅列数。
By means of all the neurons' competition in their activities and adaptive learning, the Self-Organization Feature Mapping (SOFM) neural network can effectively create the spatially organized maps to represent the various features of input signals and their interior relationships. This characteristic can be successfully used to solve combination optimization problems. in this paper, A SOFM algorithm for gate matrix layout is put forward by building up the competitive rules based on the nature of layout problem, and very satisfied results are achieved. The time complexity of this algorithm isO (n×p2lnp),here n and p are the number of the nets and dominant gates, respectively.
出处
《电路与系统学报》
CSCD
1997年第1期40-43,共4页
Journal of Circuits and Systems
关键词
自组织特征映射
栅阵列排序
神经网络
Self-Organization Feature Mapping, Gate matrix layout
Optimization algorithm
