PID sliding mode controller design and application to active queue management
Tianxiang Qi, Hao Wang · 2016
In this paper, we design a sliding mode controller named PID-SMC for active queue management in the congestion control system. When the instantaneous queue length error is greater than a set-point, the PID-SMC controller employs a proportional-derivative (PD) switching function to reduce the queue length overshoot and achieve fast system response; otherwise, the PID-SMC controller uses a proportional-integral-derivative (PID) switching function to tackle the system uncertainty and ensure high control accuracy. We also use an enhanced reaching law to improve the dynamic performance and overcome the chattering problem. We conduct simulation experiments in the Network Simulator, and the results demonstrate that our proposed PID-SMC method is robust against uncertainties and disturbances. PID-SMC outperforms benchmark active queue management algorithms (e.g., PI, PID, and REM) in terms of responsiveness and robustness under time-varying network conditions.