文献类型：期刊文献

中文题名：过程测量数据校正技术在甲醇工厂的在线应用

英文题名：Online Application of Process Measurement Data Reconciliation in Methanol Plant

作者：刘宝卫[1];赵霞[1];吴胜昔[2];潘兆鸿[3]

机构：[1]上海交通大学电子信息与电气工程学院,上海200240;[2]华东理工大学信息学院,上海200237;[3]上海慧铭自动化信息产业有限公司,上海200233

年份：2008

卷号：29

期号：2

起止页码：20

中文期刊名：自动化仪表

外文期刊名：Process Automation Instrumentation

收录：CSTPCD

基金："十一五"国家"八六三"基金项目(编号:2006AA042171)

语种：中文

中文关键词：制造执行系统;数据校正;随机误差;最小二乘法;序列补偿法

外文关键词：Manufacturing execution system （MES） Data reconciliation Random error Least square method Serial compensation method

摘要：过程数据的质量严重影响了MES中的过程监控、实时优化和控制及其它模块的性能和效益。为获得平衡和准确的过程数据,研发了一种组合了测量数据检验法和GLR检验法的改进型序列补偿法,并将其应用于数据校正中显著误差的侦破和处理。基于该方法并以甲醇工艺过程为背景的一套实时数据校正软件在某大型甲醇工厂的MES中投入在线运行,并获得了满意的效果。该方法经过适当改造后可以应用于某些过程工业的在线数据校正。
The quality of process data significantly affects the performance and profit gained from process monitoring, real-time optimization and control and other applications in MES. To get balanced and correct process data, the modified serial compensation method integrating measurement data test with GLR test was developed to detect and deal with the gross errors. Based on the method, and with the methanol process as background, a data-reconciliation software package is put into on-line service in MES of a large-scale methanol plant. The reconciled data bring about the desired effect. The method can be applied in certain process industries for on-line data reconciliation with slight modification.