论文部分内容阅读
为抑制Internet中的拥塞现象,基于终端滑模控制理论和径向基函数(RBF)神经网络提出了一种拥塞控制算法.将网络参数的变化及非传输控制协议(TCP)数据流的影响等效为系统的不确定项,设计了一种新的终端滑模面,使滑动模态具有更短的有限收敛时间,并证明了滑动模态的渐近稳定性.使用RBF神经网络估计了控制器中不确定项的上界,并根据李亚普诺夫稳定性理论得出了神经网络权值的自适应律.与PID控制器和传统的终端滑模控制器进行了相同网络条件下的仿真对比实验,结果表明:所提出的算法具有更好的鲁棒性和更快的收敛速度.
In order to suppress the congestion in the Internet, a congestion control algorithm is proposed based on terminal sliding mode control theory and radial basis function (RBF) neural network. The changes of network parameters and the influence of non-transmission control protocol (TCP) As a system uncertain term, a new terminal sliding mode surface is designed, which makes the sliding mode have a shorter finite convergence time and proves the asymptotic stability of the sliding mode. RBF neural network is used to estimate the control And the upper bound of the uncertain term is derived and the adaptive law of neural network weights is obtained according to the Lyapunov stability theory.Compared with PID controller and traditional terminal sliding mode controller under the same network conditions, Experimental results show that the proposed algorithm has better robustness and faster convergence speed.