APPLICATION OF PRINCIPAL COMPONENT ANALYSIS METHOD TO DETECTE ABNORMALITY IN THE ZINC ROASTING PROCESS

Abstract

Roasting is the first and most crucial process in the zinc production process. The
zinc production process is becoming increasingly large in scale and highly complex,
involving a series of combined physical and chemical reactions. Normal operating
conditions can transition into abnormal conditions such as excessive decomposition,
poor oxidation, and layer boiling agglomeration due to fluctuations in raw materials
and leakage in the circulating cooling water system. The stable and safe operation of
the roasting process is of great significance in ensuring the quality of the output zinc,
minimizing industrial pollution, and reducing energy consumption. To ensure safe
and stable operation, and to mitigate safety risks and economic losses, it is crucial to
monitor and detect any abnormalities in the process. One proposed method for
monitoring and detecting two typical abnormal conditions in the roasting process is
Principal Component Analysis (PCA). The analysis method utilizes industrial data
collected from real roasting processes.