火箭推进

液体火箭发动机健康检测技术是提高火箭安全性和可靠性的重要技术之一,对其进行研究具有重要的学术和工程应用价值。目前的健康检测方法大多基于特征提取和专家经验,智能检测技术水平急需提高。提出了一种基于卷积自编码器的液体火箭发动机健康状态智能检测方法,对发动机多传感器监测数据进行无监督的特征提取和重构,完成对训练集的学习,并基于隐含层特征和重构误差确定健康状态的中心和监测阈值,然后对测试集各机组状态进行识别,通过对各部件的损失计算可以进一步确定各部件的健康状态。提出方法通过热试车实验数据进行了验证,该方法在测试集上取得了88.9%的准确率,并能够监测整机及各部件的健康状态和退化趋势,结果表明提出方法的有效性和应用潜力。

The health state detection technology of liquid rocket engine is very important to improve the safety and reliability of rockets,which has important academic research and application value.The current research methods are mostly based on feature extraction and expert experience,while the research on intelligent detection technology needs to be improved.In this paper,Multi-component Two-stage Fusion Convolutional Denoising Auto-Encoder(MTAE)for intelligent detection of liquid rocket engine health state was proposed.First,unsupervised feature extraction and reconstruction of multi-sensor monitoring data were performed.After completing the learning of training set with the MTAE,the center of health and the monitoring threshold were determined based on the hidden layer features and reconstruction errors.Then,the state of each group of the testing set would be identified.In addition,the health state of each component could be further determined by calculating the loss of each component.Finally,the proposed method was verified by the data collected from the static firing test process.The method achieves an accuracy of 88.9% on the test set and could monitor the health and degradation trends of the whole engine and each component.The results show the effectiveness and application potential of the proposed method.

引言
1 卷积降噪自编码器
2 基于卷积自编码器的液体火箭发动机健康检测方法
3 某型号发动机试车数据分析
4 结论

图1 提出方法的流程图<br/>Fig.1 Flow chart of the proposed method

图1 提出方法的流程图
Fig.1 Flow chart of the proposed method

图2 MTAE网络结构图<br/>Fig.2 Structure of the MTAE

图2 MTAE网络结构图
Fig.2 Structure of the MTAE

图3 某型号液体火箭发动机的主要结构<br/>Fig.3 The main structures of the liquid rocket engine

图3 某型号液体火箭发动机的主要结构
Fig.3 The main structures of the liquid rocket engine

表1 测试集结果统计和评估<br/>Tab.1 Testing set results statistics and evaluation

表1 测试集结果统计和评估
Tab.1 Testing set results statistics and evaluation

图4 训练集重构误差及统计(置信度96%),健康状态中心及检测阈值。<br/>Fig.4 Reconstruction error of training set(confidence level 96%),health state center and detection threshold

图4 训练集重构误差及统计(置信度96%),健康状态中心及检测阈值。
Fig.4 Reconstruction error of training set(confidence level 96%),health state center and detection threshold

图5 训练集隐含层特征及统计(置信度96%),健康状态中心及检测阈值。<br/>Fig.5 Hidden layer feature of training set(confidence level 96%),health state center and detection threshold

图5 训练集隐含层特征及统计(置信度96%),健康状态中心及检测阈值。
Fig.5 Hidden layer feature of training set(confidence level 96%),health state center and detection threshold

图6 测试集重构误差<br/>Fig.6 Reconstruction error of testing set

图6 测试集重构误差
Fig.6 Reconstruction error of testing set

图7 测试集隐含层特征<br/>Fig.7 Hidden layer feature of testing set

图7 测试集隐含层特征
Fig.7 Hidden layer feature of testing set

表2 测试集结果统计和评估<br/>Tab.2 Testing set results statistics and evaluation

表2 测试集结果统计和评估
Tab.2 Testing set results statistics and evaluation

[1] 张楠,孙慧娟.低温液体火箭发动机重复使用技术分析[J].火箭推进,2020,46(6):1-12. ZHANGN,SUN H J.Analysis on the reusable cryogenic liquid rocket engine technology[J].Journal of Rocket Propulsion,2020,46(6):1-12.
[2] 孙百红,田川.基于特征频段RMS值的发动机故障实时监测方法[J].火箭推进,2019,45(4):74-78. SUN B H,TIAN C.The fault real-time monitoring method for engine based on RMS value of characteristic frequency band[J].Journal of Rocket Propulsion,2019,45(4):74-78.
[3] ZHU X B,CHENG Y Q,WU J J,et al.Steady-state process fault detection for liquid rocket engines based on convolutional auto-encoder and one-class support vector machine[J].IEEE Access,2019,8:3144-3158.
[4] 吴玉洋,李宁宁,薛薇,等.改进PSO优化LSSVM的液体火箭发动机故障检测[J].计算机仿真,2020,37(5):49-54.
[5] 刘付成.人工智能在航天器控制中的应用[J].飞控与探测,2018,1(1):16-25.
[6] 周磊,朱子环,耿卫国,等.美国液体火箭发动机试验中健康管理技术研究进展[J].导弹与航天运载技术,2013(5):20-25.
[7] 李洪.智慧火箭发展路线思考[J].宇航总体技术,2017,1(1):1-7.
[8] 吴建军,程玉强,崔星.液体火箭发动机健康监控技术研究现状[J].上海航天,2020,37(1):1-10.
[9] 宋征宇.运载火箭远程故障诊断技术综述[J].宇航学报,2016,37(2):135-144.
[10] IVERSON D L,MARTIN R,SCHWABACHER M,et al.General purpose data-driven monitoring for space operations[J].Journal of Aerospace Computing,Information,and Communication,2012,9(2):26-44.
[11] 赵汝岩,崔海斌,单崇喆.基于监测信息的固体火箭发动机健康状态评估技术[J].固体火箭技术,2019,42(3):384-388.
[12] 李雷,谢立,张永杰,等.数据挖掘在运载火箭智能测试中的应用[J].航空学报,2018,39(S1):722203.
[13] 孙成志,闫晓东.基于神经网络和证据理论的火箭发动机故障诊断[J].宇航总体技术,2020,4(4):20-30. SUN C Z,YAN X D.Fault diagnosis of rocket engine based on neural network and evidence theory[J].Astronautical Systems Engineering Technology,2020,4(4):20-30.
[14] 张晨曦,唐曙,唐珂.迁移学习下的火箭发动机参数异常检测策略[J].计算机应用,2020,40(9):2774-2780.
[15] 谭永华,李健,贺元军,等.一种针对液体火箭发动机可靠性的双线评估方法[J].推进技术,2021,42(2):421-430.
[16] DU B,XIONG W,WU J,et al.Stacked convolutional denoising auto-encoders for feature representation[J].IEEE Transactions on Cybernetics,2017,47(4):1017-1027.
[17] 张立智,井陆阳,徐卫晓,等.基于卷积降噪自编码器和CNN的滚动轴承故障诊断[J].组合机床与自动化加工技术,2019(6):58-62.
[18] ZONG B,SONG Q.Deep autoencoding Gaussian mixture model for unsupervised anomaly detection[C]//ICLR 2018.Pullman:Washington State University,2018.
[19] 何涛,黄敏超,胡小平,等.某火箭发动机故障检测及诊断算法设计分析[J].南京航空航天大学学报,2019,51(S1):50-55.

备注

引言

1 卷积降噪自编码器

2 基于卷积自编码器的液体火箭发动机健康检测方法

3 某型号发动机试车数据分析

4 结论

期刊信息