Using 2D CNN models with mid-level fusion-based approach for multi-classification of bearing faults
Master thesis
Permanent lenke
https://hdl.handle.net/11250/3107216Utgivelsesdato
2023Metadata
Vis full innførselSamlinger
Sammendrag
Bearings play a pivotal role in providing an efficient life cycle for industrial machinery. They are a key component in facilitating horizontal and rotational movements within the equipment, and the continuous uptime of the bearing ensures the efficiency, safety, and reliability of the equipment. Despite the simplistic architecture of this element, any defect compromises its efficiency and can result in a full breakdown. The criticality of ensuring optimal functionality underscores the significance of our research. Thereby, we evaluate and develop a fusion architecture by employing varied models and features to classify the defects within the bearing