| 引用本文: | 李艳姣,张森,尹怡欣,张杰.基于数据驱动的高炉料面优化决策模型研究[J].控制理论与应用,2018,35(3):324~334.[点击复制] |
| LI Yan-jiao,ZHANG Sen,YIN Yi-xin,ZHANG Jie.Research on optimization model of blast furnace burden surface based on data driven[J].Control Theory and Technology,2018,35(3):324~334.[点击复制] |
|
| 基于数据驱动的高炉料面优化决策模型研究 |
| Research on optimization model of blast furnace burden surface based on data driven |
| 摘要点击 2059 全文点击 772 投稿时间：2017-05-07 修订日期：2017-10-15 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 10.7641/CTA.2017.70302 |
| 2018,35(3):324-334 |
| 中文关键词 高炉布料 料面优化 煤气利用率 约束条件 超限学习机 自适应粒子群算法 |
| 英文关键词 blast furnace burden distribution burden surface optimization gas utilization ratio constraint conditions extreme learning machine adaptive particle swarm optimization algorithm |
| 基金项目 国家自然科学基金重点项目,国家自然科学基金 |
|
| 中文摘要 |
| 高炉炼铁是一个典型的高能耗、高排放、高污染的工业环节. 合理的炉料分布能够形成更加合理的煤气流
分布, 使得炉内的化学反应更加充分, 对高炉长期稳顺运行和节能减排具有重要作用. 本文针对基于经验的料面形
状决策不能根据炉况变化做出准确和及时的调控的缺陷, 提出了基于数据驱动的高炉料面形状优化决策模型. 首
先, 基于现场采集的数据, 在考虑高炉生产实际情况约束和变量上下限约束的情况下, 建立了以煤气利用率为评价
函数的料面优化模型. 然后, 为了提高模型的精度和决策性能, 提出了一种误差补偿超限学习机(extreme learning
machine, ELM)方法用于建立料面优化过程模型, 以减少模型与实际生产过程之间的误差. 在此基础上, 采用带有约
束条件的自适应粒子群算法对模型进行求解. 最后, 通过仿真实验验证了所建模型和优化方法的有效性, 实验结果
表明所构建的高炉料面优化决策模型能够及时根据生产情况的变化给出合理的料面形状, 满足现场生产的需求, 使
高炉高效稳定运行. |
| 英文摘要 |
| Blast furnace ironmaking is a typical industrial process with high energy consumption, high emission and
high pollution. The reasonable burden distribution can make the gas flow distribution more reasonable and the chemical
reaction in the furnace more fully. It plays an important role in the long-stable operation and energy saving and emission
reduction of blast furnace. For the difficulty to accurately adjust the burden surface by operators from their experience when
the production situation changes, the optimization model of burden surface profile based on data driven is proposed. Firstly,
taking the gas utilization ratio as the evaluation index, the burden surface optimization model is established based on the
data collected from the blast furnace with considering the constrains of the actual situation and upper and lower bounds of
the variables. Next, in order to improve the model accuracy and decision performance, a novel error compensation extreme
learning machine (ELM) is proposed to establish the process model to reduce the error between the model and actual
production process. On this basis, the adaptive particle swarm optimization (APSO) algorithm with constrains is applied
for relative calculations. Finally, the validity of the model and the optimization method are verified by the simulation
experiments. The simulation results demonstrate that this scheme can give a reasonable burden surface profile according to
the change of production situation to meet the production demand and make the blast furnace work efficient and stable. |
|
|
|
|
|