%0 Journal Article
%T Kernel learning one-step-ahead predictive control algorithm using Brent optimization<br>采用Brent优化的核学习单步预测控制算法
%A LIU Yi
%A WANG Hai-qing
%A LI Ping
%A <br>刘 毅
%A 王海清
%A 李 平
%J 控制理论与应用
%D 2009
%I 
%X A novel kernel learning one-step-ahead predictive control (KLOPC) algorithm is presented for the general unknown single-input/single-output nonlinear systems. Firstly, a one-step-ahead predictive model is obtained by using the KL identification framework; secondly, a new one-step-ahead weighted predictive control performance index is formulated; thirdly, the control law is computed via Brent optimization method, which is efficient and reliable in one dimension search without knowing any derivative of the KL identification model. This simple KLOPC scheme has few parameters to be chosen, making it very suitable for real-time control. Simulation results of a nonlinear process show that the new KLOPC algorithm is superior to other methods based on KL model and the well tuned PID controller. The proposed KLOPC strategy also exhibits more satisfactory robustness and adaptation to both additive noise and unknown process disturbance.
%K nonlinear system
%K kernel learning
%K one-step-ahead predictive control
%K Brent optimization<br>非线性系统
%K 核学习
%K 单步预测控制
%K Brent优化
%U http://www.alljournals.cn/get_abstract_url.aspx?pcid=5B3AB970F71A803DEACDC0559115BFCF0A068CD97DD29835&cid=8240383F08CE46C8B05036380D75B607&jid=970898A57DFC021F93AB51667BAED7F7&aid=55FED0181615DDFDB485EA198B7A7DF5&yid=DE12191FBD62783C&vid=96C778EE049EE47D&iid=CA4FD0336C81A37A&sid=D767283A3B658885&eid=4DB1E72614E68564&journal_id=1000-8152&journal_name=控制理论与应用&referenced_num=2&reference_num=10