摘要
在网络控制系统的研究中,支持向量机(SVM)在网络广义预测控制中的应用具有良好控制效果和稳定性。为提高网络性能,对网络控制系统进行模型预测,并将SVM作为广义预测控制(GPC)算法中的预测模型,采用支持向量机的广义预测控制算法。进行预估技术和队列机制,对被控对象选择最合适的控制信号,降低了时延对网络控制系统的危害性,并通过Matlab上仿真结果表明,与PID控制相比较,基于SVM的GPC算法在网络控制方面超调量较小,调整时间较短,控制效果更好。
Based on the principle of minimizing the risk of the network of Support Vector Machines （ SVM ）, generalized predictive control in the network application has a good control effect and stability. It is utilized for the model prediction of the NCS. SVM is applied as the prediction model in the GPC algorithm to design the GPC algorithm based on SVM. Due to the use of advanced forecast technology and queuing strategy in choosing the most appropriate control signals for the control target, the harm caused by time - delay to the NCS was greatly reduced. MATLAB simulation showed that compared to PID control, the GPC algorithm based on SVM has smaller overshoot, shorter adjustment time and better controlling effects with respect to network control .
出处
《计算机仿真》
CSCD
北大核心
2010年第6期163-166,共4页
Computer Simulation
关键词
网络控制系统
网络延迟
支持向量机
广义预测控制算法
Networked Control Systems （NCS）
Network delay
Support Vector Machine （ SVM ）
Generalized Predictive Control （GPC）
