自适应终端迭代学习控制的关键问题研究

发布时间：2018-01-04 10:51

本文关键词：自适应终端迭代学习控制的关键问题研究　出处：《青岛科技大学》2017年硕士论文　论文类型：学位论文

【摘要】：本论文主要是对自适应终端迭代学习控制中存在的初始条件变化、参考轨迹变化等问题进行了进一步的研究,提出了一系列自适应终端迭代学习控制的新方法。论文的主要创新点及贡献可总结如下:第一,针对一类非线性单入单出(SISO)离散时间系统,通过在控制器中加入遗忘因子,以达到提高控制性能的目的,并对其进行了严格的收敛性分析,证明了未知参数和输入输出的有界性,以及跟踪误差的渐进收敛性。同时,仿真结果也验证了加入遗忘因子的控制器具有更好的控制性能。第二,针对一般的多入多出(MIMO)线性时变系统,提出了随机高阶内模的方法来处理迭代变化的初始条件。研究中,期望参考点也是随迭代变化的。从而,所提出的方法克服了传统的迭代学习控制中对相同初始条件和目标轨迹严格相同的限制。严格的理论分析证明了方案的可行性,在仿真中与以前基于相同初始条件提出的控制方法进行比较,有效地验证了所提的基于高阶内模的自适应终端迭代学习控制方法的优势。第三,进一步,将基于高阶内模的方法推广到非线性系统终端迭代学习控制中。高阶内模用于对被控非线性系统的间接描述,以此设计自适应迭代学习控制器。该控制器中不包含被控系统模型的信息,而是只利用可测的输入输出数据,是一种数据驱动的控制方法。通过严格的数学分析,在理论上证明了所提方法是稳定的,并且能够使得跟踪误差渐进收敛到零。仿真研究中加入随机扰动,更进一步验证了所提方案的有效性和实用性。
[Abstract]:In this paper, the problems of the change of initial conditions and the change of reference trajectory in adaptive terminal iterative learning control are studied further. A series of new adaptive terminal iterative learning control methods are proposed. The main innovations and contributions of this paper can be summarized as follows: first, for a class of nonlinear discrete time systems with single input and single output (SISO). The forgetting factor is added to the controller to improve the control performance, and the convergence of the controller is analyzed strictly. The boundedness of unknown parameters and input and output is proved. And the asymptotic convergence of tracking error. At the same time, the simulation results show that the controller with forgetting factor has better control performance. Second, for the general multi-input multi-output MIMO-linear time-varying system. A stochastic high-order internal model method is proposed to deal with the initial conditions of iterative variation. In the study, the expected reference points also change with the iteration. The proposed method overcomes the limitation of the same initial condition and target trajectory in the traditional iterative learning control. The rigorous theoretical analysis proves the feasibility of the scheme. Compared with the previous control methods based on the same initial conditions in simulation, the advantages of the proposed adaptive terminal iterative learning control method based on high-order internal model are validated effectively. Thirdly, further. The method based on higher order internal model is extended to the terminal iterative learning control of nonlinear systems. The higher order internal model is used to describe the controlled nonlinear system indirectly. An adaptive iterative learning controller is designed. The controller does not contain the information of the controlled system model, but only uses measurable input and output data. The proposed method is proved to be stable in theory by strict mathematical analysis, and the tracking error can converge gradually to zero. Random perturbation is added to the simulation research. The effectiveness and practicability of the proposed scheme are further verified.
【学位授予单位】：青岛科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP273

【参考文献】