Multi-Granularity Modeling and Analysis for Complex Systems Based on Data Mining
Lingli Guo, Minfang Mao · 2025
This paper proposes a modeling framework for a kind of complex systems. Firstly, an iterative self-organizing data analysis techniques algorithm (ISODATA) of data mining is used to divide process data into a number of clusters. Based on this division, a quotient space in which each cluster is taken as an granular is established. Secondly, a time series granular model is established quotient space-based and predictive model is built via cell mapping theory. Furthermore, by the state space method, the pattern moving state space model is also formulated and a controller is designed. Thirdly, consider the clustering parameter change, the impaction of which on the control system is studied. Finally, a simulation of a sintering process is given to demonstrate the feasibility of the proposed approach.