计算机应用与软件2012，Vol.29Issue(8)：101-105,5.
基于SVM整定PID控制参数的网络拥塞控制AQM算法
NETWORK CONGESTION CONTROL AQM ALGORITHM BASED ON SVM TUNED PID CONTROL
摘要
Abstract
Focus on the disadvantages of PID active queue management (AQM) in self-tuning, cannot adapt to complex nonlinear network environment, etc. , this paper puts forward an intelligence AQM algorithm-LSPID, which is based on the combination of least squares support vector machine (LS-SVM) and PID control. Simulation experiments show that the LSPID algorithm has better convergence, it can more quickly converge the queue length to approximate expectations, can maintain very good stability and robustness in network conditions with dynamic topology, link capacity change and sudden flow, and the control effect is significantly better than two other AQM algorithms, including the expert intelligence PID and the neural network PID.关键词
主动队列管理/专家智能/神经网络/最小二乘支持向量机/PIDKey words
AQM/Experts intelligence/Neural network/LS-SVM/PID分类
信息技术与安全科学引用本文复制引用
胡鹏莎,张浩平,丁文,徐波..基于SVM整定PID控制参数的网络拥塞控制AQM算法[J].计算机应用与软件,2012,29(8):101-105,5.基金项目
教育科研重大应用示范项目(8509000030) （8509000030）
南京林业大学创新基金项目(X2011-11). （X2011-11）
