炼钢-连铸流程生产调度与过程控制协调优化模块化方法及应用

Steelmaking - continuous casting process has such characteristics as production conditions frequent change in production conditions and no accurate process model. According to the real-time production information, production scheduling assigns manufacturing machines and processing time to each production job using algebraic process model equations to achieve the economic performance indexes. Process controllers realize the stable control loops and track the set point values using process differential model equations. Due to the complex production process and multi-level control structures, production scheduling and process controllers are running separately. Dynamic coordinated optimization of scheduling and process controllers are the theoretical foundation to deal with the separation problems, which are multi-objective dynamic optimization problems with hybrid algebraic and differential equation constraints in nature..Our project proposals the coordinated optimization theoretical methods and system of production scheduling and process controllers in steelmaking - continuous casting process, such as modular structure of coordinated optimization system, closed-loop dynamic optimization and performance evaluation, and multi-objective real-time optimization algorithm etc.. The project includes such research contents: (1) Modular structure methodology of coordinated optimization system of production scheduling and process controllers in steelmaking - continuous casting process. (2) Process performance indexes prediction and dynamic regulation methods based on the production flexible capabilities in time, temperature and composition dimension. (3) Feedback-prediction closed loop optimization strategies, high performance algorithms of multi-objective real time optimization, and analysis theory of system performance. Finally, the above theoretical methods and system will evaluated and verified in a steelmaking plant.

炼钢-连铸流程具有生产条件变化频繁、难以用数学模型准确描述等特性。生产调度根据实时信息将设备和处理时间分配给生产任务以实现生产经济指标，其优化模型为代数方程。过程控制实现控制回路稳定和设定值跟踪，其过程模型为微分方程。由于对象复杂、控制层级多，造成了调度与过程控制的脱节。二者动态协调优化是解决脱节问题的理论基础，其本质是代数-微分方程混杂的多目标动态优化控制问题。.本项目提出炼钢-连铸流程生产调度与过程控制协调优化系统的模块化、闭环动态优化和系统性能分析、多目标协调实时优化算法等理论方法。内容包括：（1）炼钢-连铸生产调度与过程控制协调优化控制系统的模块化方法; （2）基于时间、温度、成分等多维柔性尺度的工序模块指标预报与动态调整方法; （3）炼钢-连铸流程动态协调优化控制的反馈-预报的闭环优化策略、高性能多目标实时优化算法和系统性能分析方法。最后在炼钢厂内进行理论方法和系统的应用验证。

炼钢-连铸流程具有生产条件变化频繁、难以用数学模型准确描述等特性。生产调度根据实时生产信息将设备和处理时间分配给生产任务以实现生产经济指标，其优化模型为代数方程。过程控制实现控制回路稳定和设定值跟踪，其过程模型为微分方程。由于对象复杂、控制层级多，造成了调度与过程控制的脱节。二者动态协调优化是解决脱节问题的理论基础，其本质是代数-微分方程混杂的多目标动态优化控制问题。. 本项目针对炼钢-连铸流程生产调度与过程控制协调优化的实际需求，深入开展研究工作，提出了炼钢-连铸流程生产调度与过程控制协调优化系统的模块化、闭环动态优化和高性能实时优化算法等基础理论和关键技术。本项目取得的主要研究进展包括：（1）构建了炼钢-连铸流程生产协调优化控制系统模块化体系和运行机制; （2）建立了炼钢-连铸流程的时间、温度和成分的优化模型和计算方法; （3）提出了炼钢-连铸流程动态协调优化控制的反馈-预报的闭环优化策略、高性能实时优化算法，分析了算法性能；（4）结合炼钢-连铸流程各生产工序的特点，提出了各工序的柔性尺度定义，并基于此设计了各工序的关键工艺参数的协调优化方法。最后通过炼钢-连铸流程工业实际数据，测试了本项目所提出的调度与过程控制协调优化策略、模型和方法，验证了的本项目研究工作的有效性。