Title: Pattern moving trajectory: a new dynamics description method

Authors: Zheng-guang Xu; Jin-xia Wu

Addresses: Department of Control Science and Engineering, School of Automation and Electrical Engineering, University of Science and Technology Beijing, Beijing, 100083, China; Key Laboratory of the Ministry of Education for Advanced Control of Iron and Steel Process, University of Science and Technology Beijing, Beijing, 100083, China. ' Department of Control Science and Engineering, School of Automation and Electrical Engineering, University of Science and Technology Beijing, Beijing, 100083, China; Key Laboratory of the Ministry of Education for Advanced Control of Iron and Steel Process, University of Science and Technology Beijing, Beijing, 100083, China

Abstract: For a class of complex production process systems, a dynamics description method based on moving pattern is proposed. Different from traditional modelling methods, the definition of moving pattern is presented and the dynamics of system is directly described by the moving law of working condition pattern rather than differential equation or state space model. First, based on the actual run status data, 'pattern moving space' is constructed by a clustering algorithm. Then a variable called pattern class variable (which has statistical characteristic) is defined in this space to describe the moving of pattern. The prediction model based on pattern class variable is also constructed. This prediction model is different from the existing statistical modelling methods though they have the same structure formally. At last, one practical example of sintering process illustrates that the proposed strategy might provide satisfactory results without prior knowledge of the systems.

Keywords: moving patterns; pattern moving space; pattern class variables; pattern moving trajectory; modelling; dynamics description; run status data; clustering algorithms; prediction models; sintering; production processes.

DOI: 10.1504/IJMIC.2012.051088

International Journal of Modelling, Identification and Control, 2012 Vol.17 No.4, pp.370 - 379

Published online: 17 Dec 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article