摘要
针对一类噪声方差未知的随机系统,基于不同加权因子设计多个参数辨识器辨识模型参数,在此基础上,构成多模型自适应控制器.在每个采样时刻基于指标切换函数选择最佳辨识模型,并将基于此最佳模型设计的控制器切换为当前控制器.同时,证明了多个模型控制器之间相互切换时整个闭环系统是全局收敛的.仿真结果表明,同单一自适应模型控制器相比,这种基于多个不同加权因子的多模型自适应控制器在模型参数发生跳变时可很好地改善被控对象的控制品质.
Multiple parameter identifiers based on different weighting factors are set up for a kind of stochastic system with unknown noise variance. A multiple model adaptive controller is formed based on these identified models. At every sample time, an index switching function is used to select the best model identified, and the controller based on this model is switched as the controller of the system. It is proved that the closed-loop system is globally convergent. Simulation result shows that the stochastic system multiple model adaptive controller can improve the control performance greatly compared with the conventional stochastic system adaptive controller, especially for the system with jumping parameters.
出处
《控制与决策》
EI
CSCD
北大核心
2008年第11期1226-1230,共5页
Control and Decision
基金
国家自然科学基金项目(60604002)
北京科技新星计划项目(2006B23)
国家杰出青年基金项目(69825106)
关键词
随机系统
多模型
自适应控制
Stochastic system
Multiple model
Adaptive control
