Online Prediction Method of Cement Clinker f-Cao Based on K-ELM
Shizeng Lu, Hongliang Yu, Xiaohong Wang, Fangqian Ning, Peirui Zhao, Rongfeng Zhang · 2018
Online prediction of f-Cao content is of great significance for timely control of cement rotary kiln to ensure its stable operation. In this paper, an on-line prediction method of f-Cao content based on kernel extreme learning machine (K-ELM) is proposed. According to the cement process, the input and output of the K-ELM model and its time point matching relationship are determined. On this basis, considering the continuity of cement production, the sliding time window method is combined with the K-ELM prediction method to update the K-ELM model to achieve online prediction of f-Cao content. Finally, taking the production data of a cement company as an example, the online prediction method of f-Cao content was verified. The experimental results showed that the variation trend of the online predicted value of f-Cao was similar to that of the laboratory test value of f-Cao. This paper provided a certain technical basis for in-depth study of f-Cao online prediction to further realize the energy saving and quality improvement of clinker production.