Parameter Stress Response Prediction for Vehicle Dust Extraction Fan Impeller Based on Feedback Neural Network
Vehicles exhibit complex failure modes and mechanisms because of their extreme service environments and severe external loads. The increasing level of integration in these vehicles is also driving more stringent reliability requirements, but conventional methods for reliability analysis require sign...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-06-01
|
| Series: | Machines |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-1702/13/6/496 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Vehicles exhibit complex failure modes and mechanisms because of their extreme service environments and severe external loads. The increasing level of integration in these vehicles is also driving more stringent reliability requirements, but conventional methods for reliability analysis require significant calculations, necessitating the use of surrogate models. At present, in the field of the reliability analysis of vehicle dust extraction impellers, although there are various research methods, the research on using surrogate models for relevant analysis is still not perfect. In particular, there are few studies specifically focused on dust extraction impellers. This study established a three-dimensional finite element parametric model of one such fan to simulate the impeller blade stress output for 500 parameter sets. The feedback neural network, backpropagation neural network, and quadratic polynomial response surface were subsequently used as surrogate models to learn the relationship between the parameters and output responses in these data. Comparisons of the results indicated that the feedback neural network exhibited the highest accuracy when predicting the stress responses of the dust extraction fan impeller to changes in parameter values. Through a comparative analysis of multiple surrogate models, this study determined the advantages of the feedback neural network in predicting the impeller stress response. It provides a more efficient and accurate method for reliability analysis in this field and helps to promote the development of reliability research on vehicle filtration systems. |
|---|---|
| ISSN: | 2075-1702 |