Abstract: Converter steelmaking, a mainstream modern process, produces molten steel by blowing oxygen into molten iron to efficiently remove impurities through oxidation reactions. The development of converter automated steelmaking technology is of paramount importance for achieving precise end-point control, enhancing product quality stability, reducing raw material and energy consumption costs, and promoting the intelligent upgrading of steel manufacturing. This paper elaborates on the architecture and control principles of the converter automated steelmaking system, provides an in-depth analysis of the critical role of detection and sensing technologies in automated steelmaking, and discusses the application of key means such as sub-lance, off-gas analysis, and spectral monitoring in real-time process parameter acquisition and dynamic regulation. Furthermore, it systematically sorts out the core status of models and algorithms in automated steelmaking, offering a detailed analysis of the construction principles and application effects of static control models, dynamic control models, as well as end-point carbon and temperature prediction models based on machine learning. The paper emphasizes that the deep integration of mechanism models with data-driven intelligent models is key to improving control accuracy, while also noting that current research still has shortcomings in aspects such as the data silo phenomenon and limited model generalization capability. Converter automated steelmaking is poised to evolve towards a full-process, self-adaptive intelligent control direction underpinned by industrial internet platforms, deeply integrating digital twin and artificial intelligence technologies, ultimately realizing smart steelmaking driven by metallurgical mechanisms and data.