基于多目标拆分优化思维的拥塞网络数值调度方法

针对网络拥塞数值调度中存在的盲目性问题,提出了一种基于多目标拆分优化的网络拥塞数值调度方法.将拥塞网络的数值调度问题进行模型化表示,并将拥塞过程调度的最优问题分解为多个目标同时优化问题:即信道最优任务分配问题和路由拥塞调度问题.根据粒子群算法,对信道分配问题的最优解进行计算,同时设计约束模型并利用遗传算法求解拥塞调度问题,实现了在拥塞状态下的网络数值调度.结果表明,所提出算法获得的拥塞调度方案具有较好的可执行性.

Numerical scheduling method for network congestion based on multi objective resolution optimization

In order to solve the blindness problem existing in the numerical scheduling of network congestion, a numerical scheduling method for network congestion based on multi objective resolution optimization was proposed. The modeling expression for the numerical scheduling problem of network congestion was carried out. In addition, the optimal problem of congestion process scheduling was decomposed into the multi objective optimization problem at the same time, namely the optimal task allocation problem of channel and routing congestion scheduling problem. According to the particle swarm optimization （PSO） algorithm, the calculation for the optimal solution of channel allocation problem was performed. Meanwhile, the constraint model was designed, and the congestion scheduling problem was solved with the genetic algorithm （GA）. Moreover, the numerical scheduling of network at the congestion state was realized. The results show that the congestion scheduling scheme in the proposed algorithm has good enforceability.