基于模糊免疫PID的时滞网络自适应主动队列管理

Adaptive active queue management of delay network based on fuzzy immune PID

提出了一种基于灰预测和模糊免疫PID控制的时滞网络自适应主动队列管理(AQM)算法FIGAPID,旨在增强AQM算法动态自适应能力,同时补偿网络时滞,综合提高AQM算法性能。该算法借助免疫反馈机理进行PID参数的在线自适应调整,采用模糊非线性逼近的方法进行免疫反馈函数的确定;采用等维新息滚动灰预测实现路由器队列长度的超前预测,补偿AQM控制的反馈滞后。对比传统PID算法,仿真验证了FIGAPID的有效性,表明算法能快速稳定地适应动态时滞网络环境变化,收敛于路由器队列长度期望值,同时具有较小的数据丢包率。

In order toimprove active queue management （AQM） algorithm＇s performance synthetically by enhancing algorithm＇s self-adapting and compensating feedback delay, a novel AQM algorithm for delay network based on fuzzy immune adaptive PID control and gray-prediction （FIGAPID） was proposed. PID parameters＇ online self-adapting was implemented by immune feedback mechanism, and immune feedback function was established by fuzzy nonlinear approximation. Moreover, a gray-prediction algorithm based on consistent dimension innovation was successfully introduced into feedback data＇s advanced prediction to compensate AQM＇s feedback delay. Contrasted with traditional PID algorithm, FIGAPID is validated by simulation results. It can adjust itself to new network conditions rapidly and stably, can converge to queue size-setting value, and can get lesser packets loss rate.