An Efficient Multi-PCA Based On-line Monitoring Scheme for Multi-Stages Imperial Smelting Process

作者：Hu, Zhi-kun^*; Chen, Zhi-wen; Gui, Wei-hua; Yang, Chun-hua; Yin, Lin-zi; Peng, Xiao-qi

来源：International Journal of Control Automation and Systems, 2013, 11(2): 317-324.

DOI：10.1007/s12555-011-9105-x

• Dynamic PCA
• FOP based recursive DPCA
• imperial smelting process
• process monitoring and fault diagnosis

Imperial Smelting Process (ISP) is one of the main methods for Zinc and Lead smelting. In this paper, we firstly focus on its unique characteristics that ISP consists of many feeding cycles repeated in a similar way, and the normal process changes and multiple operating conditions would be frequently found in the process which would always lead to false alarms. To this end, we classify each feeding cycles of ISP into initial stage, transition stage, and steady stage, and use standard PCA, first order perturbation (FOP) based recursive dynamic PCA and multi-mode approach to deal with them respectively. The monitoring results using real data show that the proposed scheme can be able to eliminate false alarms and detect faults efficiently.

• 出版日期2013-4
• 单位中南大学