Practical and Performance Oriented Nonlinear Adaptive Robust Control -- Theory and Applications to Precision Control of Modern Mechanical Systems

项目负责人：姚斌单位：普渡大学项目编号：0220179题目：面向实用与性能的非线性自适应鲁棒控制——现代机械系统精密控制理论与应用该项目的目标是(i)为设计实用和性能导向的非线性自适应鲁棒控制器（PARC）开发一个通用框架，该框架明确考虑到典型实现约束的影响，同时不会牺牲可实现的实现性能，并且结构简单，以促进设计和实现；（ii）确定有实际意义的标准和适当的优化算法，可用于系统地确定PARC参数，以优化控制性能；（iii）将PARC集成设计应用于由电磁马达或电液执行器驱动的高速机械系统的精确控制，以及高速机器人操纵器的协调控制。采用了一种完全不同的新设计理念来合成非线性鲁棒反馈，以解决现有自适应鲁棒控制器的结构复杂性——采用适当的具有固定结构的非线性鲁棒反馈，从而使典型的实现约束(如控制输入饱和、测量噪声、通过全局非线性综合和局部拟线性分析相结合的方法，可以明确地分析和量化被忽略的高频动力学。研究既强调一般理论的发展，又强调所提出的PARC与具体应用的结合。教育计划的重点是深入重新审视当前的本科生和研究生控制课程，为学生在控制教育的早期阶段设定正确的控制设计理念，并开发以性能为导向的非线性自适应和鲁棒控制的创新课程。由于工程控制课程中缺乏先进且实用的非线性自适应鲁棒控制设计，以满足工业对高生产率和高质量产品的需求，因此提出了这项研究的动机。一旦成功完成，结果将对未来的控制课程，在本科和研究生水平产生持久的影响。研究结果也对工业产生了直接影响，因为它们直接解决了实践工程师在应用先进非线性控制方面的共同问题。

Principle Investigator: Bin YaoInstitution: Purdue UniversityProposal Number: 0220179Title:Practical and Performance Oriented Nonlinear Adaptive Robust Control -- Theory and Applications to Precision Control of Modern Mechanical SystemsAbstract: The goals of this project are to (i) develop a general framework for the design of practical and performance oriented nonlinear adaptive robust controllers (PARC) that explicitly take into account the effects of typical implementation constraints while without sacrificing achievable implementation performance much and being structurally simple to facilitate design and implementation, (ii) determine practically meaningful criteria and appropriate optimization algorithms that can be used to systematically determine PARC parameters for an optimized control performance, and (iii) apply the integrated PARC design to the precision control of high-speed mechanical systems driven by either electro-magnetic motors or electro-hydraulic actuators as well as the coordinated control of high-speed robot manipulators. A fundamentally different new design philosophy on synthesizing nonlinear robust feedback is adopted to solve the structural complexity of existing adaptive robust controllers-proper nonlinear robust feedback with fixed structures is used so that the effects of typical implementation constraints such as the control input saturation, measurement noise, and neglected high frequency dynamics can be explicitly analyzed and quantified through a combination of global nonlinear synthesis and local quasi-linear analysis. The research emphasizes both general theory development and the integration of the proposed PARC with the particular characteristics of the application. The education plan focuses on the in-depth re-examination of current undergraduate and graduate control curriculum to set the right control design philosophy for students during their early stage of control education, as well as developing innovative course on performance-oriented nonlinear adaptive and robust controls. The proposed research is motivated by the lack of advanced and yet practical performance-oriented nonlinear adaptive robust control designs that are needed in engineering control curriculum to meet industry needs for high productivity and quality products. Once successfully completed, the results will have a lasting impact on future control curriculum, at both the undergraduate and the graduate level. The results also have a direct impact on industry as they directly address the common concerns of practicing engineers in applying advanced nonlinear controls.