Portal de Periódicos da CAPES

A Unified Label Noise-Tolerant Framework of Deep Learning-Based Fault Diagnosis via a Bounded Neural Network

2024; Institute of Electrical and Electronics Engineers; Volume: 73; Linguagem: Inglês

10.1109/tim.2024.3374322

1557-9662

Sudao He, Wai Kei Ao, Yi‐Qing Ni,

Anomaly Detection Techniques and Applications

Supervised fault diagnosis in mechanical systems, particularly in high-speed train components, faces significant challenges due to the presence of label noise in annotating large-scale monitoring data. This label noise introduces strict requirements for label-noise tolerance and the learning capabilities of fault diagnosis algorithms. This paper presents a unified framework for label-noise fault diagnosis in high-speed train components using a bounded neural network (BNN) to address this issue. The proposed framework consists of multiple basic models with shared weights, enabling the learning of global knowledge across sensor nodes and facilitating the estimation of local states to adapt to dynamic measurement networks. The BNN-based basic model incorporates implicit weighted learning and bounded loss mechanisms, which extract valuable insights from mis-annotated data. Additionally, a tighter bound of loss is introduced, providing theoretical proof and enhancing the label-noise tolerance of the BNN. A surrogate training strategy based on an alternative convex search is established to ensure the stability of the BNN model during the early training stage. This strategy mitigates the risk of failure in the initial training phase of the BNN model. The feasibility and effectiveness of the proposed method are demonstrated through a real-field test conducted on a mechanical system of a high-speed train component. The code is released on https://github.com/sudao-he/Bounded Neural_Network.