Cargando…
Research on Multi-Fault Diagnosis Method Based on Time Domain Features of Vibration Signals
The normal operation of the engine is of great importance for the safety of life and property, so we need to monitor and analyze the state of the engine. Most of the existing methods only diagnose the type of engine fault without further analysis of the severity of the engine fault. Additionally, th...
Autores principales: | , , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
MDPI
2022
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9655301/ https://www.ncbi.nlm.nih.gov/pubmed/36365866 http://dx.doi.org/10.3390/s22218164 |
Sumario: | The normal operation of the engine is of great importance for the safety of life and property, so we need to monitor and analyze the state of the engine. Most of the existing methods only diagnose the type of engine fault without further analysis of the severity of the engine fault. Additionally, the features used for fault diagnosis are not selected according to faults and do not necessarily contain more fault information. In the paper, we propose using Pearson correlation coefficients in combination with faults selects sensors and the corresponding features, and then single-fault diagnosis combined with GRU (gating recurrent unit) is performed by using the selected sensors and features. Since multi-fault diagnosis is more difficult than single-fault diagnosis, more state information is required. Therefore, the multi-fault diagnosis will directly extract the time domain features screened above from all vibration signals, stack them and send them to GRU for multi-fault diagnosis. From the experimental results we can conclude that the feature selection method combining Pearson correlation coefficient and fault state can extract effective features to diagnose the fault type and its severity. Finally, the influence factors of the model are analyzed through comparative experiments, and the results show the effectiveness of the method and the selected model parameters. |
---|