Abstract:Thickness accuracy is one of the important indicators of cold-rolled strip quality. For improving strip quality and stability of production, it is of great significance to quickly diagnose the abnormal thickness and locate the root cause of thickness anomaly. The influencing factors of thickness was determined by the rolling mechanism model, and the residual model of thickness increment was constructed by the multivariable linear regression method. Then the kernel density estimation of the residuals was utilized for detecting thickness anomaly. The causal effect of influencing factors was calculated based on causal inference to identify the root cause of thickness anomaly. The results in-practice show that the proposed method can accurately diagnose thickness anomaly and identify the root cause of anomaly, and the method is more effective than conventional method for cold rolling conditions with high correlation between the variables.
周军, 杨荃, 王晓晨. 基于因果推断的冷轧带钢厚度异常根因分析[J]. 中国冶金, 2023, 33(5): 94-101.
ZHOU Jun, YANG Quan, WANG Xiao-chen. Root cause analysis of thickness anomaly for cold rolled strip based on cause inference[J]. China Metallurgy, 2023, 33(5): 94-101.
Erik Vanhatalo,Murat Kulahci,Bjarne Bergquist. On the structure of dynamic principal component analysis used in statistical process monitoring[J]. Chemometrics and Intelligent Laboratory Systems,2017,167:1.
JIANG Q,YAN X,HUANG B. Performance-driven distributed PCA process monitoring based on fault-relevant variable selection and Bayesian inference[J]. IEEE Transactions on Industrial Electronics,2016,63(1):377.
Gharahbagheri H,Imtiaz S A,Khan F. Root cause diagnosis of process fault using KPCA and Bayesian network[J]. Industrial and Engineering Chemistry Research,2017,56(8):2054.
Van Den Kerkhof P,Vanlaer J,Gins G,et al. Contribution plots for statistical process control:Analysis of the smearing-out effect[C]//2013 European Control Conference. Zürich,Switzerland:ECC 2013,2013:428.
Abadie A,Imbens G W,Kennedy J F. Nber working paper series matching on the estimated propensity score[J]. Econometrica,2009,84(2):78.
[21]
Karim M E,Petkau J,Gustafson P,et al. On the application of statistical learning approaches to construct inverse probability weights in marginal structural Cox models:Hedging against weight-model misspecification[J]. Communications in Statistics-Simulation and Computation,2017,46(10):7668.