An intelligent moving window sparse principal component analysis-based case based reasoning for fault diagnosis: Case of the drilling process

The drilling process is an important step in petrochemical industries, but the drilling process is risky and costly. In order to improve the safety and cost the impact of faults in the drilling process, this paper proposes intelligent moving window based sparse principal component analysis (MWSPCA) integrating case-based reasoning (CBR) (MWSPCA-CBR) in the fault diagnosis of the drilling process in the petrochemical industry. Through introducing sparsity into the PCA model, the Lasso constraint function of the MWSPCA method is used to optimize the sparse principals. The corresponding T-2 and Q statistics calculated by the selected sparse principals decide whether the faults have occurred, and the occurrence time of the anomaly is quickly located based on the MWSPCA method. Then the CBR method is used to analyze the anomaly data to identify the possible fault types, and provide the relational handling methods for real-time monitoring experts. Finally, the MWSPCA method is verified based on the intelligent diagnosis of the Tennessee Eastman (TE) process, reducing false negatives and false positives and improving the accuracy rate and the diagnosis speed. Furthermore, the proposed method is applied to analyze the data of the drilling process. The experimental results demonstrate that the proposed method can effectively diagnosis faults in the drilling process and reduce risks and costs in the petrochemical industry. (C) 2021 ISA. Published by Elsevier Ltd. All rights reserved.