基于分层协调优化思想的钢包精炼过程钢水质量的稳定控制方法

Stable control of the steel quality in ladle furnace refining process is one of the important technical problems for producing steel products with high quality and high additional value. The key to obtain stability is to systematize and coordinate the LF process control. However, the existing research is not able to satisfy the practical requirement, for they commonly consider the research in the individual way for investigating the optimal control methods of temperature, chemical composition and purity of steel. In this study, for stable control of the steel quality, research in carried out to develop the multi-objective coordinated optimization control system which integrates stratified coordinated optimization strategy with intensive and flexible technic design as well as the closed-loop control technique. For obtaining cooperation, the existing individual optimization techniques are assembled with the proposed stratified coordinated optimization algorithm. To adapt to the production condition changes caused by multiple steel grades and working conditions, intensive technic process mode and flexible designed technic parameters are proposed. In order to provide basic for process optimization and closed-loop feedback control, high precise models for temperature and chemical composition prediction are developed based on the hybrid modeling methods which are deeply essential amalgamation of data information and mechanism modeling. The proposed methods are validated and improved on the existing experimental platform, and gradually applied to 1~2 LF production processes. It is expected to be a practical technique by the integration of technic and process control.

钢包精炼（LF）过程钢水质量的稳定控制是实现高质量、高附加值钢材产品生产面临的重要技术难题之一，稳定控制的关键在于LF过程控制的系统化和协调化，而目前的研究大多采用单体设计的方式对钢水温度、成分和纯净度的优化控制问题进行研究，不能满足系统性、协调性的实际需求。本项目针对LF钢水质量的稳定控制问题，提出以分层协调优化为核心、以工艺制度集约化与柔性化设计以及过程闭环控制为基础的多目标协调优化控制系统。提出分层结构的协调优化方法将现有各单体优化技术进行有机集成，实现过程操作的统筹调配；提出集约化和柔性化的工艺流程与工艺参数设计方法，以适应多钢种、多工况的生产条件变化。基于数据信息与机理建模深度融合的思想建立钢水温度和成分的高精度预报模型，为过程优化和闭环控制提供基础；将所提方法在已有实验平台上进行验证和完善，并逐步推广应用到1～2个LF生产过程。期望工艺与过程控制的一体化设计，形成实用化技术。

本项目围绕LF精炼过程钢水温度、成分等多质量指标的精准控制问题开展了理论研究与应用研究，为了实现全过程操控环节的系统整合与协调，本项目提出了基于分层协调优化控制思想的解决方案，在LF过程多目标协调优化控制涉及的建模理论、优化控制理论与实用化技术等方面提出了具体的解决方法，并在国内某钢厂进行了试点应用。围绕上述总体目标，结合LF过程生产工艺的实际特点，本项目提出了具有实时预测能力的LF过程钢水温度、成分预测模型建模方法、涵盖多质量指标协同控制的分层协调优化控制方法及面向实际应用的标准化、智能化生产过程监控方法。具体来说：1）为解决操控决策所需的被控量信息反馈问题，提出了以钢水温度、成分监测与预报为目的的LF多质量指标工艺模型建模方法；2）为解决钢种更换、工况变化等带来的控制策略与生产条件适应匹配难题，提出和建立了柔性化的冶炼工艺流程模式和工艺参数设计方法；3）为消除各质量指标单体优化控制系统间的不利干扰和冲突问题，提出了柔性工艺约束下的钢水多质量指标的分层协调优化控制方法。