摘要
随机动态过程随着过程参数的变化而难于控制。由于过程的不确定性,使得动态跟踪控制不能实现。为优化过程控制以及为保证动态过程控制的鲁棒性,有必要实施更为优越的控制算法。以实现对过程变化参数的动态跟踪,达到稳定动态过程输出的目的。该文给出了针对这类不确定过程的一种最优控制算法。通过预测器的简化算法,构建了最优控制的目标函数。应用非参数模型的对称相似结构原理与迭代学习算法相结合,完成了一种递归优化的预测学习控制算法。通过仿真分析验明了所给控制算法是成功的。实际应用结果表明,这种算法适合于一类不确定过程的动态跟踪控制。
An optimal control scheme is proposed for the uncertainty of random dynamic processes to improve the response behaviors of controlled systems,to reach the dynamic following control and to ensure the robustness of the control. An optimal object function is suructrued by simplifying the prosedure determining predictor.The algorithm of the predictive learning control based on recursive optimization is presented by combining the principle of the similar structure of model-free with learning procedure of iterative optimization. Simulating analyses identify that the presented scheme is successful. Practical applied results show that the scheme suits to the dynamic following control of the uncertain processes.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2003年第4期62-66,共5页
Proceedings of the CSEE
关键词
自适应控制
递归优化原则
预测学习控制算法
目标函数
Controlling object function
Prediction information
Recursive optimization
Learning control
