Domain Generalization Fault Diagnosis of Rotating Machinery Based on Multimodal Ensemble Learning
Abstrak
Rotating machinery is one of the most common and critical types of machinery in engineering. This type of machinery often operates under harsh conditions for extended periods, which makes it prone to failure during use. Therefore, condition monitoring and fault diagnosis can help address issues promptly and reduce the risk of accidents in rotating machinery. Currently, vibration signals are the predominant single modality used for mechanical fault diagnosis. However, unimodal approaches often lack robustness and struggle to provide reliable diagnostics in practical industrial environments. Complementary data sources can provide unique insights into different aspects of physical degradation. To address these limitations, this paper proposes a multimodal domain generalization diagnostic method based on vibration and acoustic signals. First, the method extracts and fuses features using a feature extraction and fusion model. Then, it employs ensemble learning to achieve transfer diagnosis across different operating conditions. This approach effectively resolves issues of insufficient robustness in single-modal diagnosis and diagnosis under varying operating conditions.
Penulis (8)
Hongpeng Xiao
Zhe Cheng
Zhitao Xing
Niaoqing Hu
Guoji Shen
Yi Yang
Ruizhi Li
Duofu Wang
Akses Cepat
PDF tidak tersedia langsung
Cek di sumber asli →- Tahun Terbit
- 2025
- Bahasa
- en
- Sumber Database
- Semantic Scholar
- DOI
- 10.1109/PHM-Xian66756.2025.11427594
- Akses
- Open Access ✓