Title :
Recursive PLS-based soft-sensing for outlet temperature of new dry cement decomposing furnace
Author :
Xu Liu ; Lihui Feng ; Kaifen Wen
Author_Institution :
Fac. of Inf. Eng. & Autom., Kunming Univ. of Sci. & Technol., Kunming, China
Abstract :
There are many factors and hysteresis affecting the outlet temperature of cement decomposing furnace. In order to improve real time control of the outlet temperature, this paper adopted soft-sensing method based on Recursive PLS to predict it. According to historical data in the DCS of a 2000t/d new dry cement process production line, the main auxiliary variables were determined, the soft prediction model of the outlet temperature was established by using MLR, PCA, PLS and Recursive PLS. Compared and verified, the result shows that Recursive PLS-based soft prediction model had better performance.
Keywords :
cements (building materials); furnaces; hysteresis; least squares approximations; prediction theory; recursive estimation; regression analysis; temperature control; temperature sensors; MLR; PCA; auxiliary variable determination; dry cement decomposing furnace; dry cement process production line; hysteresis; multiple linear regression; outlet temperature control; partial least square analysis; principal component analysis; recursive PLS-based soft-sensing method; soft prediction model; Data models; Furnaces; Predictive models; Principal component analysis; Production; Temperature; Temperature control; Cement decomposing furnace; Recursive PLS; Soft-sensing; The outlet temperature;
Conference_Titel :
Control and Decision Conference (CCDC), 2013 25th Chinese
Conference_Location :
Guiyang
Print_ISBN :
978-1-4673-5533-9
DOI :
10.1109/CCDC.2013.6561403
