A multi-time-scale fusion prediction model for the gas utilization rate in a blast furnace

摘要

The main operations for a blast furnace are hot-blast supply and burden distribution, which adjust the gas utilization rate (GUR) on different time scales. However, most researches only focused on fixed-time scale relationships between the operations and the GUR. This paper presents a multi-time-scale fusion model to predict the GUR. First, this paper analyzes the multi-time-scale characteristics between the operations and the GUR based on the iron-making mechanisms. Then, this paper explains a decomposition method for the GUR based on empirical mode decomposition and a reconstruction method to get a short-time-scale part the GUR (SPGUR) and a long-time-scale part of the GUR (LPGUR). Next, this paper presents a short-time-scale model to describe the relationship between the hot-blast supply and the SPGUR and a long-time-scale model, between the burden distribution and the LPGUR. Finally, this paper fuses the results of the two models to predict the GUR. The analysis based on actual run data shows that the method predicts the GUR more accurately than that predicted based on a fixed time scale.