基于时延脉冲耦合神经网络的AOE-网问题求解算法

Solution algorithm of AOE-Nets problem using delayed pulse coupled neutral networks

摘要: 在时延脉冲耦合神经网络(DPCNN-Delay Pulse Coupled Neural Network)的基础上,提出了mDPC-NN(modified DPCNN)模型.mDPCNN能方便地用于有向网的最大代价路径和最小代价路径的求解,并将其成功地运用其于求解AOE-网(Activity On Edge)有关问题.该方法通过快速并行计算,一次正向计算即可求解AOE-网大部分问题,与传统方法相比体现了较好的优势.

Abstract: It is presented mDPCNN(modified DPCNN) model based on the DPCNN(Delay Pulse Coupled Neutral Network).It is easy to solve the problem of the path with maximum cost and with minimum cost in directional nets,and mDPCNN is used to solve the Activity On Edge(AOE) nets problems successfully.Compared with traditional method in the AOE nets problems,the approach proposed has better advantages in fast parallel calculation,and solves most of the AOE nets problems by using fast parallel calculation and a positive direction calculation.