摘要
质量异常检测技术是保障冶金工业过程产品质量安全性和稳定性的重要手段,是当前过程监控领域的热点研究方向。针对传统的偏最小二乘(PLS)、最小二乘(LS)方法获取的主元空间中可能含有和质量正交的成分及检测实时性不强等问题,提出了分布式最小二乘(DLS)方法,设计了监测统计量,实现了质量异常检测。通过典型的冶金工业流程——热轧过程现场数据进行仿真验证,并与传统方法对比,验证了新算法的有效性。所提出的质量异常检测方法,将为以热轧过程为代表的冶金工业流程的安全监控及自主保障提供理论支撑与技术保证。
Quality abnormality detection techniques are an important means to ensure safety and stability of metallurgical industrial product quality,which,thus,have recently become hotspots in the process monitoring domain.Distributed least squares(DLS)-based quality abnormality detection method is proposed,and the associated statistics are designed,which aim at addressing the issues on orthogonal components in principal component space related to quality and poor real-time performance for the traditional partial least squares(PLS)and least squares(LS).Moreover,a case study on a typical metallurgical industrial process,hot rolling process,is finally given to compare with other methods to demonstrate the advantages of the new approach.The proposed quality abnormality detection framework will provide theoretical basis and technical support for safety monitoring and independent security of hot rolling process as the representative of metallurgical process industries.
作者
姚林
张岩
YAO Lin;ZHANG Yan(Ansteel Group Co. ,Ltd. ,Anshan 114021,China;Beijing Research Institute of Ansteel Co. ,Ltd. ,Beijing 102200,China)
出处
《自动化仪表》
CAS
2020年第8期10-14,19,共6页
Process Automation Instrumentation
关键词
质量异常
最小二乘
故障检测
实时性
分布式优化
正交子空间
监测统计量
热轧过程
Quality abnormality
Least squares(LS)
Fault detection
Real-time performance
Distributed optimization
Orthogonal subspace
Monitoring statistics
Hot rolling process
作者简介
姚林(1965—),男,博士,高级工程师,主要从事轧钢过程质量监测与优化等研究工作,E-mail:ylinresearch@163.com;通讯作者:张岩,男,博士,教授级高级工程师,主要从事轧制过程控制及工业人工智能等研究工作,E-mail:wu_kunkui@163.com。
