PCA过程监测方法在密闭鼓风炉铅锌熔炼中的应用

Application of PCA-Based Process Monitoring Approach in Pb-Zn Smelting Process

通过分别导出T^2和SPE统计量均值与过程数据统计参数之间的关系,分析了T^2和SPE统计量的变化趋势以及与密闭鼓风炉实际生产状况的对应关系;基于现场采集的长期历史数据,给出了在密闭鼓风炉过程传感器故障检测中的应用实例;试验结果表明,PCA方法可以快速有效地反映生产过程的变化,生产运用效果表明该方法大大提高了对密闭鼓风炉生产工况的实时监测能力,提高了生产效率。

Principal components analysis （PCA） is an effective method for process monitoring and multiple process variables analysis. The expectations of T^2and SPE statistics are studied and their relations to the statistical parameters of process data are presented. These relationships reveal the influence factors of the T^2 and SPE tests and give a definite description of the detection behavior of PCA. Examples of using PCA to fault detection of sensors in the process is given, which is based on real-world historical date sampling from a typical power plant. Applying PCA to statistical process control of Imperial smelting furnace （ISF）, the change trends of expectations of T^2 and SPE statistics and the relationship with ISF manufacture status are analyzed. The experiment results indicate that PCA method improves the real time inspecting ability for ISF smelting and the efficiency of manufacture.