基于非线性干扰观测器的主动队列管理

针对主动队列管理系统存在UDP流干扰的拥塞问题,本文基于非线性干扰观测器提出了一种主动队列管理算法。采用非线性干扰观测器对UDP流干扰进行估计,从而减小了主动队列管理系统的干扰。然后利用设计的非线性干扰观测器,使用反步滑模方法设计了一种主动队列管理算法。由于观测误差的界未知,设计了一种自适应律来避免必须事先知道观测误差的界。仿真结果表明,该算法对UDP流干扰具有较强的鲁棒性,能使队列长度较好地稳定在期望的队列长度附近。

Active queue management based on nonlinear disturbance observer

For the congestion problem of active queue management systems in the presence of UDP flow disturbances, an active queue management algorithm is proposed based on nonlinear disturbance observer. A nonlinear disturbance observer is used to estimate the UDP flow disturbances so that the disturbances of the active queue management systems are reduced. An active queue management algorithms is designed based on the proposed nonlinear disturbance observer with the backstepping sliding mode method. Since the bound of the observation error is unknown, an adaptive law is designed so that the bound of the observation error is not required to be known in advance. Numerical simulations demonstrate that the algorithm has the better robustness against the UDP flow disturbances and can stabilize the queue length at the desired queue size.