Revisiting robustness of the union-of-subspaces model for data-adaptive learning of nonlinear signal models

2014 ◽  
Author(s):  
Tong Wu ◽  
Waheed U. Bajwa
Keyword(s):  
Adaptive Learning ◽  
Union Of Subspaces ◽  
Data Adaptive ◽  
Nonlinear Signal ◽  
Signal Models

Model Adaptive Learning in Process Control of Strip Cold Rolling

2013 ◽  
Vol 423-426 ◽  
pp. 775-779
Author(s):  
Jin Lan Bai ◽  
Xiang Li ◽  
Jun Sheng Wang
Keyword(s):  
Process Control ◽  
Cold Rolling ◽  
Adaptive Learning ◽  
Rolling Mill ◽  
Calculated Data ◽  
Processing Parameters ◽  
Learning Program ◽  
Cold Rolling Mill ◽  
Line Process ◽  
Data Adaptive

In this paper, mathematic models of processing parameters and their adaptive learning principle in strip cold rolling mill are introduced. Exponential smoothing method is used during model adaptive learning. According to the contrast between actual and calculated data, adaptive learning coefficients in the process control models are modified, thus the precision of presetting model is improved. Based on three kinds of adaptive learning modes, corresponding model adaptive learning program is developed for strip cold rolling. The practical application shows that the accuracy of this method can meet the requirement of on-line process control, and it is suitable for process control in strip cold rolling mill.

Sign in / Sign up

Export Citation Format

Share Document