In the wake of the era of big data, the techniques of deep learning have become an essential research direction in the machine learning field and are beginning to be applied in the steel industry. The sintering process is an extremely complex industrial scene. As the main process of the blast furnace ironmaking industry, it has great economic value and environmental protection significance for iron and steel enterprises. It is also one of the fields where deep learning is still in the exploration stage. In order to explore the application prospects of deep learning techniques in iron ore sintering, a comprehensive summary and conclusion of deep learning models for intelligent sintering were presented after reviewing the sintering process and deep learning models in a large number of research literatures. Firstly, the mechanisms and characteristics of parameters in sintering processes were introduced and analysed in detail, and then, the development of iron ore sintering simulation techniques was introduced. Secondly, deep learning techniques were introduced, including commonly used models of deep learning and their applications. Thirdly, the current status of applications of various types of deep learning models in sintering processes was elaborated in detail from the aspects of prediction, controlling, and optimisation of key parameters. Generally speaking, deep learning models that could be more effectively implemented in more situations of the sintering and even steel industry chain will promote the intelligent development of the metallurgical industry.

大数据时代到来，深度学习技术已成为机器学习领域的重要研究方向，并开始在钢铁行业得到应用。烧结过程是一个极其复杂的工业场景。作为高炉炼铁工业的主要工艺，对于钢铁企业具有巨大的经济价值和环保意义。这也是深度学习尚处于探索阶段的领域之一。为了探讨深度学习技术在铁矿石烧结中的应用前景，在回顾大量研究文献中的烧结过程和深度学习模型的基础上，对智能烧结的深度学习模型进行了全面的总结和总结。首先详细介绍和分析了烧结过程参数的机理和特点，然后介绍了铁矿石烧结模拟技术的发展。其次，介绍了深度学习技术，包括深度学习的常用模型及其应用。第三，从关键参数的预测、控制和优化等方面详细阐述了各类深度学习模型在烧结过程中的应用现状。总体而言，深度学习模型能够在烧结乃至钢铁产业链的更多场景中得到更有效的落地，将推动冶金行业的智能化发展。