基于全局经济效益最优的铁前大系统建模与优化

Parsing and optimizing the “cost per ton pig iron” in ironmaking process via system modeling is the realistic need for production management and decision making in steel enterprises. This project aims at constructing a distributed hierarchical intelligent control (decision) model in iromaking process to maximize the global economic profit according to the supply-and-demand situations of crude ore and fuel in the market and optimize a series of procedures including procurement, transportation, coking, sintering, ironmaking etc.. Firstly, a long process static game coking and sintering minimum cost burden model is constructed by introducing the concept of “virtual inventory”. The idea of cost tracing is used to build the organization level to achieve the purpose of “lowest cost per ton pig iron”. Secondly, the coordination level will be established by co-integrating between information fusion and data synchronization. In the coordination level, several tasks are implemented, including development of coordination rules, completion of the receiving organizational level instruction and each subtask process execution feedback, and optimization of the indicators and parameters. After that, we construct production process executive level according to the output sub goals of coordination level, whose core is to analyze relationship between linear and nonlinear of furnace temperature index [Si] by using additive nonlinear systems (ANLS), so the blast furnace temperature prediction and control model adaptive to different crude ores and fuels can be established. Finally, we will explore the effect of market volatility and production situation on the new ironmaking system model.

基于铁前大系统视角建模以解析并优化“吨铁成本构成”是钢铁企业生产管理和决策的现实需要。本项目拟结合钢铁行业原燃料和产品市场供需波动现状，以全局经济效益最优为总目标，以高炉冶炼过程为控制核心，建立包括采购、运输、焦化、烧结以及炼铁等诸多子工序在内的铁前大系统集散递阶智能控制（决策）模型。. 首先，引入“虚拟库存”构建长流程静态博弈焦化和烧结最小成本配料模型，利用成本倒追溯的思路，构建服务于“吨铁成本最低”的组织级；其次，通过信息融合和数据同步协整，研制协调规则，完成接受组织级指令和每一子流程执行反馈信息并优化执行级的指标和参数，构建模型的协调级；然后，根据协调级输出的子目标构建生产过程执行级：核心环节是利用可加非线性系统（ANLS）解析炉温[Si]的线性与非线性影响关联，建立起能适应不同原燃料方案的高炉炉温预测控制模型；最后，探求市场波动及生产状况变化对新建铁前大系统模型性能的影响。

基于铁前大系统视角建模以解析并优化“吨铁成本构成”是钢铁企业生产管理和决策的现实需要。本项目结合钢铁行业原燃料和产品市场供需波动现状，以全局经济效益最优为总目标，以高炉冶炼过程为控制核心，建立并优化包括采购、运输、焦化、烧结以及炼铁等诸多子工序在内的铁前大系统集散递阶智能控制（决策）模型。. 首先，引入“虚拟库存”构建长流程静态博弈焦化和烧结最小成本配料模型，利用成本倒追溯的思路，构建服务于“吨铁成本最低”的组织级；其次，通过信息融合和数据同步协整，研制协调规则，完成接受组织级指令和每一子任务流程执行反馈信息并优化执行级的指标和参数，构建模型的协调级；然后，根据协调级输出的子目标构建生产过程执行级：核心环节是利用可加非线性系统（ANLS）解析炉温[Si]的线性与非线性影响关联，建立起能适应不同原燃料方案的高炉炉温预测控制模型；最后，探求市场波动及生产状况变化对新建铁前大系统模型性能的影响。. 就上述内容，本项目取得了一系列重要成果，共发表学术论文15篇，其中10篇被SCI 收录，3篇被EI 收录。