| 引用本文: | 彭开香,李钢,张凯.基于动态全潜结构投影的热连轧厚度监控[J].控制理论与应用,2012,29(11):1446~1451.[点击复制] |
| PENG Kai-xiang,LI Gang,ZHANG Kai.Strip thickness monitoring in hot strip mill processes based on dynamic total projection to latent structures (T–PLS) algorithm[J].Control Theory and Technology,2012,29(11):1446~1451.[点击复制] |
|
| 基于动态全潜结构投影的热连轧厚度监控 |
| Strip thickness monitoring in hot strip mill processes based on dynamic total projection to latent structures (T–PLS) algorithm |
| 摘要点击 2243 全文点击 1593 投稿时间：2012-04-10 修订日期：2012-06-04 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 10.7641/j.issn.1000-8152.2012.11.CCTA120342 |
| 2012,29(11):1446-1451 |
| 中文关键词 过程监控 故障检测 动态T–PLS 热连轧 |
| 英文关键词 process monitoring fault detection dynamic T–PLS hot strip mill |
| 基金项目 国家自然科学基金资助项目(61074085, 61273173); 北京市自然科学基金资助项目(4122029); 北京市重点学科资助项目(XK100080537); 中央高校基本科研业务费资助项目(FRF–AS–11–004B, FRF–SD–12–008B). |
|
| 中文摘要 |
| 本文利用带钢热连轧生产过程的数据, 采用动态全潜结构投影算法(T–PLS), 建立了带钢厚度的动态模型. 该模型对于厚度有良好的预测精度. 利用动态T–PLS的优点, 把过程变量空间分解为4个正交子空间. 在不同的子空间中, 可以对带钢厚度有关的故障进行监测. 通过热连轧机3个典型故障的检测, 充分验证了动态T–PLS在过程质量监控中的优良性能, 加强了带钢热连轧过程的监控. |
| 英文摘要 |
| By using the hot strip rolling process data, we build a dynamic model for strip thickness based on the dynamic total projection to latent structures (T–PLS) algorithm. This model has a high prediction-accuracy for the thickness. Taking the advantage of T–PLS, we decompose the process variable space into four orthogonal subspaces, in which the process fault related to thickness can be detected. Through the detection of three typical faults in hot strip mill processes, the excellent performances of the dynamic T–PLS algorithm in the process monitoring is fully validated, demonstrating the enhancement of the monitoring qualities of the hot strip mill process. |