A Study on PF–IFF-Based Diagnosis Model of Plant Equipment Failure

• Published in: Applied sciences 2022-01, Vol.12 (1), p.347
• Main Authors: Seo, Min-Young, Hwang, Se-Yun, Lee, Jang-Hyun, Kim, Jae-Gon, Jun, Hong-Bae
• Format: Article
• Language: English
• Subjects: Case studies; CBM; Centrifugal pumps; Clustering; Computer applications; Data mining; diagnosis; Failure; Failure modes; Fault diagnosis; Gas turbine engines; Internet of Things; Methods; Neural networks; Nuclear power plants; plant equipment failure; Preventive maintenance; Sensors; Support vector machines; SVM; TF–IDF; Wavelet transforms
• ISSN: 2076-3417; 2076-3417
• DOI: 10.3390/app12010347

There are two types of maintenance policies for equipment: breakdown maintenance and preventive maintenance. In the case of applying preventive maintenance, the maintenance is carried out based on time or the condition of the equipment. However, with the development of Information and Communications Technologies (ICT) and the Internet of Things (IoT) technology, the data collected from equipment has rapidly increased and the use of Condition-Based Maintenance (CBM) to perform appropriate maintenance based on the condition of the equipment is increasing. In this study, based on gathered sensor data, we introduce an approach to diagnosing the condition of the equipment by extracting specific data features related to the types of failures that occur with equipment. To this end, we used the K-means clustering method, support vector machine (SVM) classifier, and Pattern Frequency–Inverse Failure mode Frequency (PF–IFF) method with the Term Frequency–Inverse Document Frequency (TF–IDF) method. As a case study, we applied the proposed approach to a centrifugal pump and carried out computational experiments for assessing the performance and validity of the proposed approach.