随着工业4.0时代的来临，制造技术正逐步从自动化、数字化、网络化向智能化方向发展。作为工业4.0的重要策略—信息物理系统(Cyber-physical system ，CPS)，由于其具有自主判断、自主决策、自主调控能力，如何将CPS技术应用于流程工业的智能化制造引起了广泛关注。根据钢铁制造过程的特点，提出在各工序中嵌入CPS质量管控模块的过程控制系统，实现全流程产品质量在线动态管控与优化。重点研究采用嵌入式CPS方法的产品质量在线管控系统架构和产品质量管控模型。在这些模型中，提出一种基于数据驱动的质量异常判定模型，提出基于贡献图的质量异常原因分析方法，讨论动态的工艺参数控制策略。嵌入式CPS产品质量监控系统可以根据生产过程中各个工序的产品质量偏离情况，及时调整后工序的工艺参数，实现产品质量的在线优化。通过汽车钢的应用实例，证实了基于嵌入式CPS模型的动态产品质量管控方法是提高产品质量的稳定性和实现柔性智能化制造的有效途径。

Along with coming into Industrie 4.0 Era, manufacture technologies are developing in smart manufacture system from automatization, digital and networked manufacture system. As an important strategy in Industrie 4.0, how the CPS (Cyber- Physical System) is used on smart manufacture in process industry is attracting widely attention, because it possesses self-discriminating, self-decision and self-control power. Facing to characteristics of steel production, this paper proposes that the CPS models are embedded into process control systems of each process, to achieve online monitoring, control and optimization of product quality. Research focuses on the system framework of online monitoring and control of product quality used CPS, as well as relative models and algorithms. In those models, the anomaly detection method of product quality based on data-driven, and the analysis method of causes resulted in quality outliers based on contribution chart are proposed. Finally, the dynamic control strategy of process parameters is discussed. According as abnormal status of product quality outlier in each process, the proposed quality monitoring and control system embedded CPS models can just in time adjust latter process parameters to achieve online optimization of product quality. The results in practical production for automobile steel show that those methods of dynamic monitoring and control of product quality based on embedded CPS are an effective way to enhance stability of product quality and to realize flexible and smart manufacture.