A robust temperature prediction model of shuttle kiln based on ensemble random vector functional link network

•A novel model is proposed to predict sintering temperature of shuttle kiln.•Mutual information is proposed to select optimal features.•A robust RVFLN method is developed to build prediction model.•Adaboost and offset compensation are utilized to enhance RRVFLN model.•The proposed AB-RRVFLN model pr...

Full description

Saved in:
Bibliographic Details
Published in:Applied thermal engineering 2019-03, Vol.150, p.99-110
Main Authors: Zhang, Lei, Zhang, Xiaogang, Chen, Hua, Tang, Hongzhong
Format: Article
Language:English
Subjects:
AdaBoost algorithm
Kilns
Machine learning
Mathematical models
Offset compensation
Outliers (statistics)
Predictions
Random vector functional link network (RVFLN)
Shuttle kiln
Sintering
Sintering temperature prediction
Temperature
Thermodynamic properties
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•A novel model is proposed to predict sintering temperature of shuttle kiln.•Mutual information is proposed to select optimal features.•A robust RVFLN method is developed to build prediction model.•Adaboost and offset compensation are utilized to enhance RRVFLN model.•The proposed AB-RRVFLN model presents high predictive accuracy. In the sintering process for shuttle kiln, the sintering temperature is the most important thermal parameter, which plays a critical role in stable and efficient operation of the production process. However, due to the complex structure of shuttle kiln, this poses a great challenge for accurate and reliable online prediction of sintering temperature. Therefore, in this paper, a novel prediction model based on ensemble random vector functional link network (AB-RRVFLN) is proposed to predict the temperature. First, the mutual information method is proposed to select the optimal variables for prediction model. Then, considering the outlier influence of industrial data, a robust random vector functional link network (RRVFLN) based on the iteratively reweighted least squares is developed to build prediction model. Moreover, AdaBoost strategy is integrated with RRVFLN model to enhance prediction performance. Finally, the model output is further corrected by offset compensation technique. Based on industrial data from shuttle kiln, the application results on sintering temperature prediction demonstrate that the proposed ensemble method achieves better accuracy and shows stronger robustness than other methods.
ISSN:1359-4311
1873-5606
DOI:10.1016/j.applthermaleng.2018.12.092