非常规跟踪目标下随机系统迭代学习控制算法设计与分析

项目名称： 非常规跟踪目标下随机系统迭代学习控制算法设计与分析

项目编号： No.61304085

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 自动化技术、计算机技术

项目作者： 沈栋

作者单位： 北京化工大学

项目金额： 22万元

中文摘要： 本项目开展对非常规跟踪目标下随机系统迭代学习控制算法设计与分析的研究，这一课题是迭代学习控制领域的研究前沿。主要研究内容包括点对点控制、变轨道跟踪及多目标协调跟踪三类非常规跟踪目标问题。其中点对点控制问题指跟踪目标并非完整轨迹，而是部分指定位置的输出；变轨道跟踪问题指跟踪目标并非沿迭代轴固定不变，而是按一定规律变化或任意变化；多目标协调跟踪问题指跟踪目标包含多条独立的轨迹，由相互关联的多个子系统分别完成。本项目的研究对象是随机系统，需要考虑如何削弱随机噪声对控制效果的影响，因此本项目致力于构建基于随机方法的算法设计与分析框架。本项目的研究将有助于理解随机噪声环境下对非常规跟踪目标的迭代学习控制能力，促进迭代学习控制理论的发展。

中文关键词： 迭代学习控制；点对点控制；随机变化运行长度；多目标协调跟踪；不完备数据

英文摘要： This project conducts research on the design and analysis of iterative learning control (ILC) algorithms of stochastic systems for unusual tracking references, which is a frontier of ILC domain. The main research contents of this project include three classes of unusual tracking problems, i.e. point-to-point control problem, varying-references tracking problem, and coordinated tracking problem for multi-objectives. First, by point-to-point control problem we mean the reference is not an integrated trajectory but only the output at some selected positions. Second, by varying-references tracking problem we mean the references are not fixed along iteration axis but varying according to certain rules or arbitrarily. Third, by coordinated tracking problem for multi-objectives we mean there are multiple independent trajectories, accomplished by multiple interconnected subsystems, respectively. Moerover, this project focuses on stochastic systems, which requires consideration on how to reduce the effect on control performance from stochastic noises, thus this project devotes to building a framework for the algorithms design and analysis based on stochastic approaches. The conduct of this project will help to understand the capability of ILC for unusual tracking references under stochastic noises, and promote further de

英文关键词： Iterative Learning Control；Point-to-Point Control；Iteration Varying Lengths；Coordinated Tracking for Multi-Objectives；Incomplete Data