Review on intelligent recognition of blast furnace gas flow distribution through multimodal data fusion

The carbon emissions from the steel industry represent one of the most challenging issues in industrial emission reduction, with the blast furnace(BF)-based integrated process serving as a critical component for energy conservation and emission reduction. Under the dual constraints of the "carbon peak" and "carbon neutrality" strategic objectives and the global climate governance framework, the technological innovation of low-carbon BF smelting has become the core driver for green transformation in the steel industry. Gas flow distribution, as a key control variable for achieving low-carbon emissions in the BF, is difficult to measure directly due to the inherent opaque and enclosed nature of the BF interior. Intelligent gas flow distribution recognition technology enables high-precision quantitative control of gas flow patterns within the BF, which holds strategic significance for achieving carbon reduction targets in BF operations. Based on an investigation of the formation mechanisms, control factors, and quantitative standards of BF gas flow, this paper systematically reviews the cutting-edge advancements in intelligent recognition algo‑rithms, dynamic prediction models, and multimodal fusion modeling techniques for gas flow distribution. Further‑more, it explores the pathways and trends of intelligent control by integrating practical BF regulation strategies(including top and bottom adjustments). Current intelligent recognition technologies for BF gas flow distribution have made significant progress at multiple levels, including technological development, algorithm innovation, model construction, and practical implementation. These advancements provide solid theoretical foundations and practical guidance for steel enterprises to achieve precise identification and optimized control of BF gas flow, thereby facilitating the efficient and low-carbon development of BF ironmaking. In order to further promote the progress of low-carbon ironmaking technology, intelligent recognition of BF gas flow distribution can further promote, the development and application of novel monitoring data, enhanced integration of multiple innovative technologies, correlation analysis of diverse data sources, and deeper convergence between metallurgical principles and operational practices.