A feed-forwarded neural network-based variational Bayesian learning approach for forensic analysis of traffic accident
In this work, a variational Bayesian learning-based computation algorithm is developed to “inversely” identify the deformation field of a crashed car and hence their residual strain fields based on its final damaged structural configuration (wreckage), which is important in three-dimensional traffic...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
2023
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/164161 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| id |
sg-ntu-dr.10356-164161 |
|---|---|
| record_format |
dspace |
| spelling |
sg-ntu-dr.10356-1641612023-01-06T07:08:44Z A feed-forwarded neural network-based variational Bayesian learning approach for forensic analysis of traffic accident Xie, Yuxi Wu, C. T. Li, Boyuan Hu, Xuan Li, Shaofan School of Mechanical and Aerospace Engineering Singapore Centre for 3D Printing Engineering::Mechanical engineering Machine Learning Vehicle Collision In this work, a variational Bayesian learning-based computation algorithm is developed to “inversely” identify the deformation field of a crashed car and hence their residual strain fields based on its final damaged structural configuration (wreckage), which is important in three-dimensional traffic collision reconstruction and its forensic analysis. Different from our previous Generalized Bayesian Regularization Network (GBRN) algorithm (Xie et al. [2002] Computational Mechanics 69, 1191–1212), the present method is based on Variational Bayesian Learning theory coupled with a Feed-forward Neural Network architecture, and it provides a higher computation efficiency. This is because it requires less number of iterations and produces more accurate registration results, since the locality of nodes are greatly preserved during registration process. In this work, we have demonstrated that the developed machine learning algorithm has a unique capability to practically identify the deformation field of a real crashed car and to recover its initial pre-crash state based on residual damaged geometric configuration, and it shows great potential in forensic analysis of car crash and vehicle crashworthiness evaluation. 2023-01-06T07:08:44Z 2023-01-06T07:08:44Z 2022 Journal Article Xie, Y., Wu, C. T., Li, B., Hu, X. & Li, S. (2022). A feed-forwarded neural network-based variational Bayesian learning approach for forensic analysis of traffic accident. Computer Methods in Applied Mechanics and Engineering, 397, 115148-. https://dx.doi.org/10.1016/j.cma.2022.115148 0045-7825 https://hdl.handle.net/10356/164161 10.1016/j.cma.2022.115148 2-s2.0-85131406884 397 115148 en Computer Methods in Applied Mechanics and Engineering © 2022 Elsevier B.V. All rights reserved. |
| institution |
Nanyang Technological University |
| building |
NTU Library |
| continent |
Asia |
| country |
Singapore Singapore |
| content_provider |
NTU Library |
| collection |
DR-NTU |
| language |
English |
| topic |
Engineering::Mechanical engineering Machine Learning Vehicle Collision |
| spellingShingle |
Engineering::Mechanical engineering Machine Learning Vehicle Collision Xie, Yuxi Wu, C. T. Li, Boyuan Hu, Xuan Li, Shaofan A feed-forwarded neural network-based variational Bayesian learning approach for forensic analysis of traffic accident |
| description |
In this work, a variational Bayesian learning-based computation algorithm is developed to “inversely” identify the deformation field of a crashed car and hence their residual strain fields based on its final damaged structural configuration (wreckage), which is important in three-dimensional traffic collision reconstruction and its forensic analysis. Different from our previous Generalized Bayesian Regularization Network (GBRN) algorithm (Xie et al. [2002] Computational Mechanics 69, 1191–1212), the present method is based on Variational Bayesian Learning theory coupled with a Feed-forward Neural Network architecture, and it provides a higher computation efficiency. This is because it requires less number of iterations and produces more accurate registration results, since the locality of nodes are greatly preserved during registration process. In this work, we have demonstrated that the developed machine learning algorithm has a unique capability to practically identify the deformation field of a real crashed car and to recover its initial pre-crash state based on residual damaged geometric configuration, and it shows great potential in forensic analysis of car crash and vehicle crashworthiness evaluation. |
| author2 |
School of Mechanical and Aerospace Engineering |
| author_facet |
School of Mechanical and Aerospace Engineering Xie, Yuxi Wu, C. T. Li, Boyuan Hu, Xuan Li, Shaofan |
| format |
Article |
| author |
Xie, Yuxi Wu, C. T. Li, Boyuan Hu, Xuan Li, Shaofan |
| author_sort |
Xie, Yuxi |
| title |
A feed-forwarded neural network-based variational Bayesian learning approach for forensic analysis of traffic accident |
| title_short |
A feed-forwarded neural network-based variational Bayesian learning approach for forensic analysis of traffic accident |
| title_full |
A feed-forwarded neural network-based variational Bayesian learning approach for forensic analysis of traffic accident |
| title_fullStr |
A feed-forwarded neural network-based variational Bayesian learning approach for forensic analysis of traffic accident |
| title_full_unstemmed |
A feed-forwarded neural network-based variational Bayesian learning approach for forensic analysis of traffic accident |
| title_sort |
feed-forwarded neural network-based variational bayesian learning approach for forensic analysis of traffic accident |
| publishDate |
2023 |
| url |
https://hdl.handle.net/10356/164161 |
| _version_ |
1754611288829329408 |