Traceability method of quality defects causes in continuous casting slab based on correlation and principal component analysis
Online tracing of the causes of quality defects in continuous casting slab plays an important role in improving the efficiency of defect resolution.The traditional offline traceability method has a certain lag and can no longer meet the needs of online monitoring.Taking longitudinal cracks in slab as the research object,the influencing factors of longitudinal cracks were analyzed,and a big data in-tegration platform for continuous casting process was constructed.Correlation analysis and principal component analysis methods were used to establish traceability models for the causes of quality de-fects in slab suitable for different application scenarios.Regarding the longitudinal cracks in Q235B slab produced by a certain steel plant,the correlation analysis results show that the narrow left wide internal heat flux ratio of the mold and the superheat degree of the molten steel have a significant im-pact,with influence weights of-0.28 and 0.15,respectively.The results of principal component analysis indicate that the superheat of molten steel and the ratio of narrow left wide external heat flux in the mold are the main influencing factors on longitudinal cracks.Based on the analysis of the trace-ability results,it was found that the cooling water quality of the mold was not qualified,resulting in uneven cooling and frequent longitudinal cracks.The successful online application of the method for tracing the causes of quality defects in continuous casting slab has reduced the recurrence rate of de-fects,which is of great significance for achieving intelligent continuous casting.
NETL
NSTL
万方数据
