Ping Zhou
Title: Data-driven Robust Modeling and Learning Modeling of Ironmaking Blast Furnace
Abstract: Blast furnace automation is a hot and difficult research topic in the field of metallurgical engineering and industrial automation. Aimed at the problems of imperfect data quality and nonlinear dynamic time-variation in the blast furnace ironmaking process which is difficult to model with conventional methods, this talk introduces some of the team’s recent work on data-driven blast furnace ironmaking process modeling. My talk focusses on the robust modeling methods and online learning modeling methods for reliable quality prediction, as well as the subspace identification modeling method for quality control.
Biography: Ping Zhou is a professor and doctoral tutor of Northeastern University. He was selected as the top young talents in the National “Ten Thousand People Program”, “Xingliao Talent” and the hundred talent level of “Hundred, Thousand and Ten Thousand Talent Project” in Liaoning Province. He is also a senior member of IEEE, a member of MMM Technical Committee of the International Federation of Automatic Control, a member of the “Random Neural Networks and Learning Systems” Working Group of the Neural Network Technical Committee of IEEE Computational Intelligence Society, a member of the Process Control Committee of the Chinese Association of Automation and other professional committees. He is mainly engaged in the research of industrial process modeling, control and operation optimization. He published more than 100 journal papers, published 2 academic monographs as the first author, and had more than 40 invention patents authorized. Besides, he has presided over more than 10 national and provincial ministerial projects including the major projects of the National Natural Science Foundation. Moreover, he won more than 10 awards such as the Second Prize of Natural Science of the Ministry of Education and the Excellent Doctor of the first Chinese Association of Automation.
Abstract: Blast furnace automation is a hot and difficult research topic in the field of metallurgical engineering and industrial automation. Aimed at the problems of imperfect data quality and nonlinear dynamic time-variation in the blast furnace ironmaking process which is difficult to model with conventional methods, this talk introduces some of the team’s recent work on data-driven blast furnace ironmaking process modeling. My talk focusses on the robust modeling methods and online learning modeling methods for reliable quality prediction, as well as the subspace identification modeling method for quality control.
Biography: Ping Zhou is a professor and doctoral tutor of Northeastern University. He was selected as the top young talents in the National “Ten Thousand People Program”, “Xingliao Talent” and the hundred talent level of “Hundred, Thousand and Ten Thousand Talent Project” in Liaoning Province. He is also a senior member of IEEE, a member of MMM Technical Committee of the International Federation of Automatic Control, a member of the “Random Neural Networks and Learning Systems” Working Group of the Neural Network Technical Committee of IEEE Computational Intelligence Society, a member of the Process Control Committee of the Chinese Association of Automation and other professional committees. He is mainly engaged in the research of industrial process modeling, control and operation optimization. He published more than 100 journal papers, published 2 academic monographs as the first author, and had more than 40 invention patents authorized. Besides, he has presided over more than 10 national and provincial ministerial projects including the major projects of the National Natural Science Foundation. Moreover, he won more than 10 awards such as the Second Prize of Natural Science of the Ministry of Education and the Excellent Doctor of the first Chinese Association of Automation.