Unsupervised Anomaly Detection for Liquid Rocket Engines with Multi-source Data Density Estimation Autoencoder

doi:10.3901/JME.2025.02.036

Abstract

Abstract: Liquid rocket engine hot fire tests represent a pivotal phase in the engine manufacturing and delivery process. Conducting high-accuracy anomaly detection during hot fire tests is a crucial method to ensure the secure and dependable operation of the engine, as well as the successful execution of manned lunar missions. To address the technical challenges associated with the high false positive and false negative rates in anomaly detection caused by the interference of multiple component operational responses and the nonlinear system with differential degradation trends among engine components, an unsupervised learning method of multi-source data density estimation autoencoder is proposed based on liquid rocket engine health operation status monitoring data. Fine-grained feature representations are learned using micro patch feature embedding with a global structured information modeling encoder, and then reconstructed by a decoder to learn health state data distributions. Meanwhile, utilizing gaussian mixture density estimation network constraints enables the autoencoder to achieve improved feature representation and information reconstruction. Finally, the anomaly detection evaluation score based on the dual scales of data feature space representation and data distribution is constructed, the normal situation constraint boundaries under the healthy state operation of the liquid rocket engine are defined, and the intelligent detection of anomalies at the engine system level under the multi-source monitoring data is finally realized. The results of the multi-source data anomaly detection task for the ground thermal test of liquid rocket motors under the same type and across types prove the effectiveness and superiority of the proposed method.

Key words: anomaly detection, liquid rocket engines, unsupervised learning, autoencoder

CLC Number:

TG156

LIU Shen, WANG Jun, CHEN Jinglong, LIU Zijun. Unsupervised Anomaly Detection for Liquid Rocket Engines with Multi-source Data Density Estimation Autoencoder[J]. Journal of Mechanical Engineering, 2025, 61(2): 36-45.

References

[1] 李京浩. 基于数据挖掘技术的液体火箭发动机故障检测和诊断研究[D]. 长沙：国防科学技术大学，2007. LI Jinghao. Fault detection and diagnosis of liquid rocket engine based on data mining[D]. Changsha：National University of Defense Technology，2007.
[2] 房芳，郑辉，汪玉，等. 机械结构健康监测综述[J]. 机械工程学报，2021，57(16)：269-292. FANG Fang，ZHEN Hui，WANG Yu，et al. Mechanical structural health monitoring：A review[J]. Journal of Mechanical Engineering，2021，57(16)：269-292.
[3] 王艳梅. 液体火箭发动机故障检测和诊断中数据挖掘技术的应用研究[D]. 长沙：国防科学技术大学，2005. WANG Yanmei. Studies on the application of data mining for fault detection and diagnosis of liquid rocket engine[D]. Changsha：National University of Defense Technology，2005.
[4] XIE Q，TAO G，XIE C，et al. Abnormal data detection based on adaptive sliding window and weighted multiscale local outlier factor for machinery health monitoring[J]. IEEE Transactions on Industrial Electronics，2023，70(11)：11725-11734.
[5] 刘子俊，冯勇，陈景龙，等. 基于多源数据的液体火箭发动机智能异常检测[J]. 火箭推进，2022，48(3)：79-86. LIU Zijun，FENG Yong，CHEN Jinglong，et al. Intelligent anomaly detection of liquid rocket engine with multi-source data[J]. Journal of Rocket Propulsion，2022，48(3)：79-86.
[6] 李栋，孙欣，刘树林. 基于变径边界样本界面检测器的异常度检测方法[J].机械工程学报，2020，56(10)：57-63. LI Dong，SUN Xin，LIU Shulin. Equipment abnormal degree detection approach based on interface detector with variable-sized boundary samples[J]. Journal of Mechanical Engineering，2020，56(10)：57-63.
[7] 胡雷，胡茑庆，秦国军. 双阈值单类支持矢量机在线故障检测算法及应用[J]. 机械工程学报，2009，45(3)：169-173. HU Lei，HU Niaoqing，QIN Guojun. Online fault detection algorithm based on double-threshold OCSVM and its application[J]. Journal of Mechanical Engineering，2009，45(3)：169-173.
[8] 张振臻，陈晖，高玉闪，等. 液体火箭发动机故障诊断技术综述[J]. 推进技术，2022，43(6)：20-38. ZHANG Zhenzhen，CHEN Hui，GAO Yushan，et al. Review on fault diagnosis technology of liquid rocket engine[J]. Journal of Propulsion Technology，2022，43(6)：20-38.
[9] 朱明清. 基于Modelica的液体火箭推进系统故障仿真及系统实现[D]. 武汉：华中科技大学，2016. ZHU Mingqing. Fault simulation and software development of liquid rocket propulsion system based on modelica[D]. Wuhan：Huazhong University of Science and Technology，2016.
[10] 薛薇，武小平，胡慧，等. 基于可视化界面的液氧甲烷发动机故障建模设计与实现[J]. 计算机测量与控制，2018，26(9)：1-4. XUE Wei，WU Xiaoping，HU Hui，et al. Design and implementation of liquid oxygen/methane engine fault model based on visual interface[J]. Computer Measurement ＆ Control，2018，26(9)：1-4.
[11] CHA J，KO S，PARK S Y，et al. Fault detection and diagnosis algorithms for transient state of an open-cycle liquid rocket engine using nonlinear Kalman filter methods[J]. Acta Astronautica，2019，163：147-156.
[12] ALI M，GUPTA U K. An expert system for fault diagnosis in a Space Shuttle main engine[C/CD]// 26th Joint Propulsion Conference，1990.
[13] GUPTA U K，ALI M. LEADER-an integrated engine behavior and design analyses based real-time fault diagnostic expert system for space shuttle main engine (SSME)[C]// Proceedings of the 2nd International Conference on Industrial and Engineering Applications of Artificial Intelligence and Expert Systems，1989，1：135-145.
[14] 张纯良，张振鹏，杨尔辅，等. 空间推进系统故障诊断专家系统研究[J]. 航空动力学报，2004，19(5)：718-723. ZHANG Chunliang，ZHANG Zhenpeng，YANG Erfu，et al. Research on fault diagnoses expert system of space propulsion system[J]. Journal of Aerospace Power，2004，19(5)：718-723.
[15] 汪俊亮，高鹏捷，张洁，等. 制造大数据分析综述：内涵、方法、应用和趋势[J]. 机械工程学报，2023，59(12)：1-16. WANG Junliang，GAO Pengjie，ZHANG Jie，et al. A review of manufacturing big data：connotation，methodology，application and trends[J]. Journal of Mechanical Engineering，2023，59(12)：1-16.
[16] LIU F T，TING K M，ZHOU Z H，et al. Isolation forest[C] // Proceedings of the 2008 Eighth IEEE International Conference on Data Mining，2008：413-422.
[17] PEVNY T. Loda: Lightweight on-line detector of anomalies[J]. Machine Learning，2016，102(2)：275-304.
[18] 赵松波. 基于改进优化算法的液体火箭发动机故障检测与诊断研究[D]. 天津：天津理工大学，2019. ZHAO Songbo. Research on fault detection and diagnosis of liquid propellant rocket engine based on improved optimization algorithm[D]. Tianjin：Tianjin University of Technology，2019.
[19] 张万旋，薛薇，张楠. 基于增量式孤立森林的液体火箭发动机异常检测方法[J]. 航空动力学报，2022，37(8)：1674-1682. ZHANG Wanxuan，XUE Wei，ZHANG Nan. Anomaly detection method of liquid rocket engine based on incremental isolation forest[J]. Journal of Aerospace Power，2022，37(8)：1674-1682.
[20] HAN D，WANG Z，CHEN W，et al. DeepAID: Interpreting and improving deep learning-based anomaly detection in security applications[C]// Proceedings of the 2021 ACM SIGSAC Conference on Computer and Communications Security，2021：3197-3217.
[21] ARELLANO-ESPITIA F，DELGADO-PRIETO M，MARTINEZ-VIOL V，et al. Anomaly Detection in Electromechanical Systems by means of Deep-Autoencoder[C]// 26th IEEE International Conference on Emerging Technologies and Factory Automation (ETFA )，2021：1-6.
[22] 朱晓彬. 基于卷积神经网络的液体火箭发动机故障检测与诊断方法研究[D]. 长沙：国防科学技术大学，2020. ZHU Xiaobin. Research on fault detection and diagnosis methods of liquid-propellant rocket engine based on convolutional neural network[D]. Changsha：National University of Defense Technology，2020.
[23] DOSOVITSKIT A，BEYER L，KOLESNIKOV A，et al. An Image is worth 16x16 words: Transformers for image recognition at scale[C/CD]// Proceedings of the 9th International Conference on Learning Representations，2021.
[24] LEE Y，KANG A P. AnoViT: Unsupervised anomaly detection and localization with vision Transformer-based encoder-decoder[J]. IEEE Access，2022，10：46717-46724.
[25] YOU Z，CUI L，SHEN Y，et al. A unified model for multi-class anomaly detection[C/CD]// Proceedings of the 36th Conference on Neural Information Processing Systems，2022.