鲁棒控制器在网络控制中的应用研究——基于改进人工蜂群优化方法设计

The Research of Robust Controller in Network Control——Based on Improved Artificial Bee Colony Optimization

在设计H∞回路成形控制器并将其应用于网络温度控制系统时,权函数参数的选择十分关键。但在寻找满足设计要求的权函数参数过程中,用传统的人工蜂群算法会出现收敛速度过慢、容易陷入局部最优等问题。基于此,本文对传统的人工蜂群算法进行了改进,用改进后的人工蜂群算法进行了控制器的设计。实验结果表明,改进算法之后设计的控制器不仅能够保证网络温度控制系统稳定在平衡位置,而且控制效果更好。

Selecting the appropriate weight function parameters is very important for designing loop shaping controller to apply to network temperature control system. But slow convergence speed and easily plunged into local optimum is disadvantages of traditional artificial bee colony algorithm for looking for the requirement weight function parameters. So traditional artificial bee colony algorithm is improved. The parameters of weight function which is got by improved artificial bee colony algorithm can ensure the better controller performance. The experiment results show that controller is designed by improved artificial bee colony algorithm not only keep network temperature control system stability in the equilibriumposition, but also get better result compare to traditional artificial bee colony.