Data-driven multi-time scale model predictive control for blast furnace gas utilization rate
AN Jian-qi
ZHAO Guo-yu
HE Yong
LI Wei-jun
GUO Yun-peng
WU Min
Abstract：In a blast furnace,gas utilization ratio(GUR)is an important indicator for measuring energy consumption and stable operation,which is affected by the operation of burden and blast supply at different time scales.The existing research methods on gas utilization rate are only conducted on a single time scale,ignoring the multi-time scale characteristics,which leads to the limited accuracy of gas utilization rate prediction and control.This paper presents a multi-time-scale gas utilization rate model predictive control method(MTSGURMPC)for blast furnaces based on data-driven.First,the influence of burden and blast supply on gas utilization rate in multi-time scales is analyzed by combining empirical model decomposition and correlation analysis.Then,this paper establishes a long-time-scale model for burden and a short-time-scale model for blast supply,a multi-time-scale model predictive control structure is presented to search for the optimal operating strategy.The presented structure divides the gas utilization rate into different scales for model predictive control,taking into account blast furnace multi-time scales and dynamic optimization characteristics of model predictive control,which leads to continuous feedback optimization to approach the optimal solution.Finally,industrial experiments are conducted based on a blast furnace,and the results show that the method achieves accurate prediction and control.
Keywords：blast furnace gas utilization ratedata-driven modelingmulti-time-scale systemmodel predictive controlempirical mode decomposition
Publication Date：2025-01-27
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：13( 189-201 )