Title: Learning and Feedback in the Control of Uncertain Systems
题目：不确定系统控制中的学习与反馈
Abstract: Learning and feedback are complementary mechanisms in dealing with uncertain dynamical systems. Learning plays a basic role in the design of control systems, and feedback makes it possible for a control system to perform well in an open environment with uncertainties. In this talk, some basic results will be presented when online learning is combined with feedback in the control of uncertain dynamical systems. We will first consider the celebrated self-tuning regulators (STR) in adaptive control of uncertain linear stochastic systems, where the STR is designed by combining the recursive least-squares estimator with the minimum variance controller. The convergence of this natural and seemingly simple adaptive system had actually been a longstanding open problem in control theory. Next, we will discuss the rationale and foundation behind the widespread successful applications of the well-known proportional-integral-derivative (PID) control for nonlinear uncertain systems and provide a new online learning-based design method. Finally, we will present some basic results on more fundamental problems concerning the maximum capability and limitations of the feedback mechanism in dealing with uncertain nonlinear systems. These results may offer useful implications for the design and analysis of more complicated control systems where AI is combined with online feedback control.
摘要：学习与反馈是处理不确定动态系统的两种互补机制。学习在控制系统设计中发挥着基础性作用，而反馈则使控制系统能在充满不确定性的开放环境中良好运行。本次报告将阐述当在线学习与反馈控制相结合时，针对不确定动态系统控制所取得的基础性研究成果。我们将首先探讨不确定线性随机系统自适应控制中著名的自校正调节器（STR）——该设计通过将递推最小二乘估计器与最小方差控制器相结合实现。这一自然且看似简单的自适应系统，其收敛性问题实则是控制理论中长期悬而未决的难题。其次，我们将剖析非线性不确定系统中广泛成功的比例-积分-微分（PID）控制背后的理论基础，并提出一种新型的基于在线学习的设计方法。最后，针对反馈机制处理不确定非线性系统的最大能力与根本性局限等更基础的问题，我们将展示若干基本结论。这些成果可为人工智能与在线反馈控制相结合的复杂控制系统设计与分析提供重要启示。
Bio: Lei GUO is a professor at the Academy of Mathematics and Systems Science, Chinese Academy of Sciences (CAS). He is a Fellow of IEEE, Member of CAS, Fellow of the Academy of Sciences for the Developing World (TWAS), Foreign Member of the Royal Swedish Academy of Engineering Sciences, and Fellow of the International Federation of Automatic Control (IFAC). In 2014, he was awarded an honorary doctorate by the Royal Institute of Technology (KTH), Sweden. In 2019, he was awarded the Hendrik W. Bode Lecture Prize by the IEEE Control Systems Society "for fundamental and practical contributions to the field of adaptive control, system identification, adaptive signal processing, stochastic systems, and applied mathematics". His current research interests include adaptive (learning, filtering, control and games) theory of stochastic systems, control of uncertain nonlinear systems, game-based control systems, multi-agent complex systems, and man-machine integration systems, etc.
简介：郭雷，中国科学院系统科学研究所研究员，国际电气与电子工程师学会会士（IEEE Fellow）、中国科学院院士、发展中国家科学院（TWAS）院士、瑞典皇家工程科学院外籍院士、国际自动控制联合会（IFAC）会士。2014年获瑞典皇家理工学院（KTH）荣誉博士学位，2019年因"在自适应控制、系统辨识、自适应信号处理、随机系统及应用数学领域的根本性与实用性贡献"被IEEE控制系统学会授予亨德里克·W·博德讲座奖（Hendrik W. Bode Lecture Prize）。其主要研究方向包括：随机系统的自适应（学习、滤波、控制与博弈）理论、不确定非线性系统控制、基于博弈的控制系统、多智能体复杂系统及人机融合系统等。