Unsupervised Anomaly Detection for Time Series Data of Spacecraft Using Multi-Task Learning

被引:5
|
作者
Yang, Kaifei [1 ]
Wang, Yakun [1 ]
Han, Xiaodong [1 ]
Cheng, Yuehua [1 ,2 ]
Guo, Lifang [1 ]
Gong, Jianglei [1 ,3 ]
机构
[1] China Acad Space Technol, Inst Telecommun & Nav Satellites, Beijing 100094, Peoples R China
[2] Nanjing Univ Aeronaut & Astronaut, Coll Automat Engn, Nanjing 210016, Peoples R China
[3] Xidian Univ, Sch Aerosp Sci & Technol, Xian 710071, Peoples R China
来源
APPLIED SCIENCES-BASEL | 2022年 / 12卷 / 13期
基金
中国国家自然科学基金;
关键词
spacecraft; unsupervised anomaly detection; time series; deep learning; multi-task learning; temporal convolutional network; LSTM; VAE;
D O I
10.3390/app12136296
中图分类号
O6 [化学];
学科分类号
0703 ;
摘要
Although in-orbit anomaly detection is extremely important to ensure spacecraft safety, the complex spatial-temporal correlation and sparsity of anomalies in the data pose significant challenges. This study proposes the new multi-task learning-based time series anomaly detection (MTAD) method, which captures the spatial-temporal correlation of the data to learn the generalized normal patterns and hence facilitates anomaly detection. First, four proxy tasks are implemented for feature extraction through joint learning: (1) Long short-term memory-based data prediction; (2) autoencoder-based latent representation learning and data reconstruction; (3) variational autoencoder-based latent representation learning and data reconstruction; and (4) joint latent representation-based data prediction. Proxy Tasks 1 and 4 capture the temporal correlation of the data by fusing the latent space, whereas Tasks 2 and 3 fully capture the spatial correlation of the data. The isolation forest algorithm then detects anomalies from the extracted features. Application to a real spacecraft dataset reveals the superiority of our method over existing techniques, and further ablation testing for each task proves the effectiveness of fusing multiple tasks. The proposed MTAD method demonstrates promising potential for effective in-orbit anomaly detection for spacecraft.
引用
收藏
页数:17
相关论文
共 50 条
  • [1] Multi-Task Network Anomaly Detection using Federated Learning
    Zhao, Ying
    Chen, Junjun
    Wu, Di
    Teng, Jian
    Yu, Shui
    [J]. SOICT 2019: PROCEEDINGS OF THE TENTH INTERNATIONAL SYMPOSIUM ON INFORMATION AND COMMUNICATION TECHNOLOGY, 2019, : 273 - 279
  • [2] Multi-task learning for video anomaly detection
    Chang, Xingya
    Zhang, Yuxin
    Xue, Dingyu
    Chen, Dongyue
    [J]. Journal of Visual Communication and Image Representation, 2022, 87
  • [3] Multi-task learning for video anomaly detection*
    Chang, Xingya
    Zhang, Yuxin
    Xue, Dingyu
    Chen, Dongyue
    [J]. JOURNAL OF VISUAL COMMUNICATION AND IMAGE REPRESENTATION, 2022, 87
  • [4] Unsupervised anomaly detection by densely contrastive learning for time series data
    Zhu, Wei
    Li, Weijian
    Dorsey, E. Ray
    Luo, Jiebo
    [J]. NEURAL NETWORKS, 2023, 168 : 450 - 458
  • [5] Unsupervised Multi-task Learning with Hierarchical Data Structure
    Cao, Wenming
    Qian, Sheng
    Wu, Si
    Wong, Hau-San
    [J]. PATTERN RECOGNITION, 2019, 86 : 248 - 264
  • [6] MDLR: A Multi-Task Disentangled Learning Representations for unsupervised time series domain adaptation
    Liu, Yu
    Li, Duantengchuan
    Wang, Jian
    Li, Bing
    Hang, Bo
    [J]. INFORMATION PROCESSING & MANAGEMENT, 2024, 61 (03)
  • [7] Spacecraft Time-Series Anomaly Detection Using Transfer Learning
    Baireddy, Sriram
    Desai, Sundip R.
    Mathieson, James L.
    Foster, Richard H.
    Chan, Moses W.
    Comer, Mary L.
    Delp, Edward J.
    [J]. 2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION WORKSHOPS, CVPRW 2021, 2021, : 1951 - 1960
  • [8] Learning Representations from Healthcare Time Series Data for Unsupervised Anomaly Detection
    Pereira, Joao
    Silveira, Margarida
    [J]. 2019 IEEE INTERNATIONAL CONFERENCE ON BIG DATA AND SMART COMPUTING (BIGCOMP), 2019, : 84 - 90
  • [9] Multi-task Contrastive Learning for Anomaly Detection on Attributed Networks
    Zhang, Junjie
    Ding, Yuxin
    [J]. ADVANCES IN KNOWLEDGE DISCOVERY AND DATA MINING, PT I, PAKDD 2024, 2024, 14645 : 15 - 26
  • [10] Spacecraft Time-Series Online Anomaly Detection Using Deep Learning
    Baireddy, Sriram
    Desai, Sundip R.
    Foster, Richard H.
    Chan, Moses W.
    Comer, Mary L.
    Delp, Edward J.
    [J]. 2023 IEEE AEROSPACE CONFERENCE, 2023,