自适应控制、有限时间与固定时间估计

讲座名称：自适应控制、有限时间与固定时间估计

讲座人：Denis Efimov  主持人：平续斌

讲座时间：5月10日—5月23日

讲座地点：

讲座1：5月10日16:00 Zoom会议直播（https://univ-lille-fr.zoom.us/j/93965412256?pwd=MjVudzFyRWZLNnFjWGpNNCt5dWorZz09）

讲座2：5月12日16:00 Zoom会议直播（https://univ-lille-fr.zoom.us/j/91316864412?pwd=M05JeHAzUWlYRmVXTDdheEhzamQyUT09）

讲座3：5月15日16:00 Zoom会议直播（https://univ-lille-fr.zoom.us/j/96288230095?pwd=TjNPOFd1OFJ5WWIyWi84cEhmMnd4dz09）

讲座4：5月23日16:00 Zoom会议直播（https://univ-lille-fr.zoom.us/j/93452530679?pwd=MldSUXFmNGJxbWJXbTM1cVRSQldNZz09）

讲座人介绍：

Denis Efimov，2001年获得圣彼得堡国立电气工程大学（俄罗斯）自动控制博士学位；2006 年在俄罗斯科学院圣彼得堡机械工程问题研究所获得工程科学博士。从2006年到2011年，在法国苏佩莱克的L2S CNRS实验室、比利时列日大学的蒙特菲奥雷学院研究所和法国波尔多大学的IMS CNRS 实验室工作。2011年入职法国国家信息与自动化研究所（里尔-北欧中心）。从2018年开始为法国国家信息与自动化研究所Valse团队负责人。主要研究领域包括非线性系统控制、齐次系统估计与控制、监督控制系统、区间观测器估计、有限时间估计和稳定性研究等。目前发表150多篇期刊论文，并且多个IFAC TC的成员和IEEE的高级成员。此外担任控制领域知名期刊IEEE Transactions on Automatic Control、IFAC Journal on Nonlinear Analysis: Hybrid Systems 以及Automatica副主编。

讲座内容：

Indirect adaptive control: The basic problem of estimation and parameter identification is introduced for linear and nonlinear regression. The problem of adaptive control is motivated in direct and indirect setting. The issue of online parameter estimation is introduced and several popular solutions (Normalized gradient algorithm, Normalized least-squares algorithm) are surveyed. Conditions of parametric convergence and persistency of excitation are considered. Robustness of adaptation algorithms is investigated. Indirect adaptive control problematics are given, with the model reference and the pole placement control approaches. Adaptive Observers for Nonlinear Systems: Adaptive observers serve to simultaneously estimate the unmeasured state vector and unknown parameters in dynamical systems. Two main approaches are considered to design of adaptive observers for nonlinear systems. Main definitions and basic proofs are given.

Estimation in finite or fixed time: In this presentation, first, the idea of estimation on a finite interval of time is motivated. Second, the definitions of finite-time an fixed-time stability are given, and the theory of homogeneous systems is presented, which is frequently used for design and analysis of finite-time or fixed-time stable systems. Finally, designs of finite-time and fixed-time converging observers are described. Methods of realization of differentiation algorithms are considered.

Interval observer design: The problem of state estimation under conditions of bounded uncertainty is formulated, and its solutions are discussed for different models of the dynamical systems. Application of interval observers for control design is explained.

主办单位：机电工程学院