双通道时延脉冲耦合神经网络的AOV-网拓扑排序

Activity On Vertex Network topological sort using Dual Channels Delayed Pulse Coupled Neural Networks

在时延脉冲耦合神经网络DPCNN的基础上提出了双通道时延脉冲耦合神经网络(DCDPCNN,Dual Channels DPCNN)模型,并提出了利用DCDPCNN来实现AOV-网拓扑排序算法。该算法在深度优先搜索的同时兼顾广度优先搜索,同时忽略节点进栈顺序,在求得的拓扑序列的个数、计算中的临时数据量、有向环判断、计算速度方面,比传统算法有了较大的改进。

This paper presents the DCDPCNN （Dual Channels Delayed Pulse Coupled Neural Networks） model based on the DPCNN,and suggests an approach using DCDPCNN to implement the AOV-Network（Activity On Vertex Network） topological sort algorithm.The algorithm has some advantages in using depth-first search at the same time considering breadth-first search and neglecting the nodes pushing stack order.Compared with the traditional algorithms,the algorithm has plumy improvement in obtaining the number of topological sequences,calculating the quantity of temporal data,judging directional loop,and calculating speed.