Memory-augmented skip-connected autoencoder for unsupervised anomaly detection of rocket engines with multi-source fusion
ISA Transactions · 2023-02 · 133 · 53-65
DOI: 10.1016/j.isatra.2022.07.014 ↗Kurz gefasst
Fachartikel in der Zeitschrift ISA Transactions, 2023-02. Verfasst von Haodong Yan und 4 weiteren Autor:innen. Beteiligte Einrichtung: State Key Laboratory for Manufacturing and Systems Engineering, Xi'an Jiaotong University, Xi'an 710049, PR China. Hauptthemen laut MEDLINE: Algorithms, Generalization, Psychological. Die Arbeit wurde bisher 55-mal zitiert (Crossref) und nennt 46 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Fundamental Research Funds for the Central Universities, Science and Technology Major Project of Guangxi.
Abstract (Originalsprache)
To ensure the safety and stability of the rocket, it is essential to implement accurate anomaly detection on key parts such as the liquid rocket engine (LRE). However, due to the indistinct features of signals under the interference of extreme conditions and the weak distinguishing ability to exist unsupervised methods, it is difficult to distinguish abnormal samples from normal samples, which leads to the failure of reliable anomaly detection. Aiming at this problem, this paper proposed an unsupervised learning algorithm named Memory-augmented skip-connected deep autoencoder (Mem-SkipAE) for anomaly detection of rocket engines with multi-source data fusion. Unlike traditional autoencoders, the input embedding for the decoder is not generated by an encoder but by a combination of memory items that record prototypical patterns of normal samples. Besides, each layer of the encoder and decoder has a skip connection to fully extract the multi-scale features of the normal sample in multi-dimensional space and suppress over-fitting caused by the memory-augmented network. Compared with existing methods and ablation control groups, experiments on four test sets prove the excellent generalization and satisfactory performances of the proposed Mem-SkipAE. Furthermore, the comparison of the single-source model and multi-source model verifies the effectiveness of multi-source fusion.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- State Key Laboratory for Manufacturing and Systems Engineering, Xi'an Jiaotong University, Xi'an 710049, PR China.
- Science and Technology on Liquid Rocket Engine Laboratory, Xi'an Aerospace Propulsion Institute, Xi'an 710100, PR China.
- State Key Laboratory for Manufacturing and Systems Engineering, Xi'an Jiaotong University, Xi'an 710049, PR China. Electronic address: jlstrive2008@mail.xjtu.edu.cn.
Förderung
- National Natural Science Foundation of China
- Fundamental Research Funds for the Central Universities
- Science and Technology Major Project of Guangxi
- National Key Research and Development Program of China
Zitieren
APA
Yan, H., Liu, Z., Chen, J., Feng, Y., & Wang, J. (2023). Memory-augmented skip-connected autoencoder for unsupervised anomaly detection of rocket engines with multi-source fusion. ISA Transactions, 133, 53-65. https://doi.org/10.1016/j.isatra.2022.07.014
Vancouver
Yan H, Liu Z, Chen J, Feng Y, Wang J. Memory-augmented skip-connected autoencoder for unsupervised anomaly detection of rocket engines with multi-source fusion. ISA Transactions. 2023;133:53-65. doi:10.1016/j.isatra.2022.07.014
BibTeX
@article{Yan2023,
author = {Yan, Haodong and Liu, Zijun and Chen, Jinglong and Feng, Yong and Wang, Jun},
title = {Memory-augmented skip-connected autoencoder for unsupervised anomaly detection of rocket engines with multi-source fusion},
journal = {ISA Transactions},
year = {2023},
volume = {133},
pages = {53--65},
doi = {10.1016/j.isatra.2022.07.014}
}
EUVIMED