Multi-output Classification using a Cross-talk Architecture for Compound Fault Diagnosis of Motors in Partially Labeled Condition

TL;DR

This paper introduces a novel cross-talk neural architecture for multi-output fault classification in motors, improving domain adaptation and compound fault diagnosis with superior performance on vibration data.

Contribution

The study proposes a new residual neural dimension reductor (RNDR) architecture with cross-talk design for enhanced multi-output fault diagnosis under partially labeled conditions.

Findings

RNDR improves macro F1 scores over baseline models

Frequency-layer normalization enhances domain adaptation

RNDR's structural benefits are more evident in compound fault scenarios

Abstract

The increasing complexity of rotating machinery and the diversity of operating conditions, such as rotating speed and varying torques, have amplified the challenges in fault diagnosis in scenarios requiring domain adaptation, particularly involving compound faults. This study addresses these challenges by introducing a novel multi-output classification (MOC) framework tailored for domain adaptation in partially labeled target datasets. Unlike conventional multi-class classification (MCC) approaches, the MOC framework classifies the severity levels of compound faults simultaneously. Furthermore, we explore various single-task and multi-task architectures applicable to the MOC formulation-including shared trunk and cross-talk-based designs-for compound fault diagnosis under partially labeled conditions. Based on this investigation, we propose a novel cross-talk architecture, residual neural dimension reductor (RNDR), that enables selective information sharing across diagnostic tasks, effectively enhancing classification performance in compound fault scenarios. In addition, frequency-layer normalization was incorporated to improve domain adaptation performance on motor vibration data. Compound fault conditions were implemented using a motor-based test setup and evaluated across six domain adaptation scenarios. The experimental results demonstrate its superior macro F1 performance compared to baseline models. We further showed that the structural advantage of RNDR is more pronounced in compound fault settings through a single-fault comparison. We also found that frequency-layer normalization fits the fault diagnosis task better than conventional methods. Lastly, we analyzed the RNDR with various conditions, other models with increased number of parameters, and compared with the ablated RNDR structure.