Abstract:Pyrometallurgy of non-ferrous metals is the core technology for the industrial extraction of base metals such as copper, lead, zinc and nickel in China. However, its high-temperature, multiphase, and strongly coupled reaction characteristics lead to long-standing issues such as low energy efficiency, high pollutant emissions, and insufficient process control accuracy. As artificial intelligence (AI) technologies continue to penetrate the metallurgical industry, they demonstrate systematic value in improving efficiency, reducing costs, and enhancing safety and environmental performance. This review, from a full-process perspective of “data acquisition-parameter prediction-process optimization-equipment maintenance,” comprehensively explores the implementation of key technologies, including multimodal sensor network construction, critical parameter prediction in high-temperature multiphase systems, multi-objective dynamic optimization control, and intelligent fault diagnosis. Studies show that AI-based methods significantly enhance process energy efficiency and environmental performance by improving the accuracy of online parameter monitoring, strengthening multivariable coordinated control, and enabling full-lifecycle management of critical equipment. Nonetheless, challenges remain, such as fluctuating data quality under extreme thermal conditions, limited model generalization in complex scenarios, and insufficient understanding of multi-scale spatiotemporal coupling mechanisms. Future research should focus on developing physics-informed modeling approaches, cross-scale dynamic optimization algorithms, and industrial-grade intelligent decision systems . Through the deep integration of metallurgical thermodynamics, process systems engineering, and information science, a smart pyrometallurgical technology framework aligned with green and low-carbon goals can be established.
