Mining of expert rules for blast furnace air volume regulation

Air volume regulation is a crucial means for adjusting the blast furnace burden condition, which directly influences the reaction process inside the furnace and is closely related to the quality of hot metal and production efficiency. However, due to the complex characteristics of the blast furnace smelting process, such as large time delay, strong coupling, and nonlinearity, achieving precise regulation of the air volume poses significant challenges. In this paper, based on theoretical knowledge and process logic, we screen the monitoring parameters that affect air volume regulation. These parameters include charge velocity, static pressure, full differential pressure, utilization rate of coal gas, permeability, and furnace temperature. By comparing various encoding methods, an effective discretization encoding of continuous monitoring data is realized through a combination of normal distribution division and empirical correction. The FP-Growth algorithm is applied to mine the correlation between air volume regulation and monitoring parameters, enabling the development of intelligent expert rules for increasing or decreasing blast furnace air volume and thereby rationally optimizing blast furnace air volume regulation. These rules have been put into application in a 1750 m³ blast furnace in actual production. The application results demonstrate that this system can accurately generate regulation strategies, and the analysis of the rule triggering frequency further validates its effectiveness. This research provides a solid basis for optimizing regulation strategies, significantly enhancing the intelligent level and regulation accuracy of blast furnace smelting, and thus playing a crucial role in promoting the intelligent transformation of modern blast furnace ironmaking.