TITLE:
Research on Fault Detection and Classification of Industrial Equipment Based on Machine Learning
AUTHORS:
Gangying Cai
KEYWORDS:
Industrial Equipment Fault Detection, Fault Diagnosis, Support Vector Machine, Decision Tree CART, Random Forest
JOURNAL NAME:
Journal of Computer and Communications,
Vol.13 No.4,
April
28,
2025
ABSTRACT: In the context of the rapid development of intelligent manufacturing, the stable operation of mechanical equipment is crucial for maintaining industrial production continuity and achieving economic benefits. Timely identification of potential fault risks based on equipment operation data can facilitate accurate maintenance and enhance production safety and efficiency. This study conducts fault detection and classification modeling using operational data from industrial equipment provided by a manufacturing enterprise. First, the raw data underwent data cleaning, outlier removal, and imputation of missing values. Key influencing factors, including plant temperature, equipment temperature, rotational speed, torque, hours of use, and equipment quality level, were identified through independence tests and ANOVA. For fault detection modeling, support vector machine (SVM) and decision tree (CART) algorithms were employed to determine whether the equipment experienced fault, with model performance evaluated using multiple evaluation metrics. For fault classification, a multi-classification model based on the random forest algorithm was developed to identify specific fault types. Furthermore, feature importance analysis was performed to quantify the impact of different features on various fault types, revealing the potential causes of faults. This study offers practical value for intelligent maintenance and predictive overhaul of manufacturing equipment, providing both data-driven insights and methodological reference for industrial applications.