报告摘要：For cost and technological reasons, there always exist some variables and parameters which cannot be measured. Moreover, the measurements usually contain noises. Sometime, fast estimations with convergence in finite-time are required in on-line applications. For these reasons, the Modulating Functions Method (MFM) originally introduced for system identification has been applied and extended in the fields of signal processing and automatic control, for parameter estimation, state estimation, and numerical differentiation, etc. This method has the following advantages. The obtained estimators are exactly given by integral formulas of the observation signal. The knowledge of initial conditions is not needed. Thus, fast estimations can be provided using a sliding integration window. Moreover, thanks to the integrals in the formulas, this method is robust with respect to corrupting noises. In this talk, the method will be explained by giving simples examples for different estimation tasks.

报告人简介：刘大研于2011 年获法国里尔大学应用数学专业博士学位。2013 年，他受聘于法国中部卢瓦尔河谷国立应用科学学院，担任终身制副教授，并于 2022 年获得法国博士生导师资格与教授职称。刘教授现任国际自动控制联合会及中国自动化学会相关专业委员会委员，同时担任《International Journal of Robust and Nonlinear Control》等多本 SCI 收录期刊的副主编。近年来，他主持或参与中法美三国科研项目共计 10 余项，已发表学术论文 120 余篇，其中 SCI 期刊论文 60 余篇，成果发表于《Automatica》《IEEE Transactions on Automatic Control》等自动控制领域国际顶级期刊，另获授权发明专利4项。

报告时间：2026年8月4日 16:00-18:00

报告地点：北衡楼1421