【作者】 刘明； 朱绍文；

【Author】 Liu Ming Zhu Shaowen Department of Computer Science , Central China Normal University , Wuhan,Hubei 430079

【机构】 华中师范大学；

【摘要】 对于计算机高速网络中能控交通流的调节,PI控制器的参数优化直接影响控制系统的性能。本文针对单个节点的交通流网络模型,提出将遗传算法应用于一个具有比例加积分(PI)结构的拥塞控制器的参数整定,给出了构造控制器反馈系数的遗传算法辨识过程。仿真结果表明,遗传算法可以较准确地获取控制器的反馈系数,也为计算机网络的性能分析和控制研究提供了方法性依据。更多还原

【Abstract】 For the flow regulation of the controllable traffic in high-speed computer networks, optimization of control parameters directly influence on a PI controller. With regard to a switching network model of a single node, the paper proposes the method of applying genetic algorithm to get the solution to the optimal identification of a proportional-plus-integrative (PI) controller for congestion controlling, and gives the progress of parameter identification. The simulation shows that genetic algorithm is an easy and effective method to obtain the feedback coefficients, and the method has a great meaning in performance analysis and control problem of computer networks.更多还原