Aeroengine fault risk early warning model based on improved DRSN
-
摘要:
航空发动机属于多发性故障机械，运用先进的计算训练方法可有效地实现准确的风险预警分析，为发动机的运维指导提供参考。在发动机故障风险预警征兆数据集中提取多变量时间序列样本，将样本矩阵化，转换为灰度图样本。预处理并增强图像数据样本，热编码化序列样本标签。深度残差收缩网络（deep residual shrinkage network，DRSN）中融入深度注意力机制与带有阈值的残差收缩块，获取高判别性特征，实现软阈值化。结合长短时记忆神经网络层与多个隐层，改进DRSN模型，使用主成分分析重构特征与主元提取，累积可解释方差贡献率为93.7%。对潜在20种故障征兆识别、分类并预警，训练精确度为96.1%。提出了改进DRSN航空发动机故障风险预警模型，与其他算法相比有较强的鲁棒性，预警正确率至少提高4.4%。
Abstract:Aero-engine is a kind of mechanical equipment with possible multi-fault risk. The application of advanced computing training method can effectively realize accurate risk early warning analysis, and provide reference for the guidance of engine operation and maintenance. Multivariable time series samples were extracted from the early warning symptom data set of engine failure risk, and the samples were matrix-transformed into gray scale samples. Image samples were preprocessed and enhanced, and sequence sample tags were thermally encoded. Deep attention mechanism and residual shrinkage block with threshold were integrated into the deep residual shrinkage network (DRSN), so as to obtain high discriminant features and realize soft thresholding. Combining long short term memory layers with multiple hidden layers, DRSN model was improved, and principal component analysis was made to reconstruct features and extract principal components. The cumulative interpretable variance contribution rate was 93.7%. The training accuracy for identifying, classifying, and warning 20 potential fault symptoms was 96.1%. An improved early warning DRSN model of engine fault risk was proposed. Compared with other algorithms, this model of strong robustness improved the accuracy by at least 4.4%.
-
图 4 基于改进DRSN的故障风险预警网络结构
Figure 4. Network structure of fault risk early warning based on improved DRSN
图 5 基于改进DRSN的故障风险预警流程
Figure 5. Fault risk early warning process based on improved DRSN
图 6 噪声预警征兆样本与原样本对比
Figure 6. Comparison between noise warning symptom samples and original samples
表 1 故障预警征兆样本分类
Table 1. Classification of fault warning symptom samples
故障类型 故障现象 预警征兆变化量/% F1 风扇叶片结垢 风扇折合流量下降7 风扇效率下降2 F2 风扇叶片顶端间隙变大（磨损） 风扇折合流量下降4 F3 风扇叶片受外物损伤 风扇效率下降5 F4 风扇叶片腐蚀 风扇效率下降2 LC1 增压级叶片结垢 增压级折合流量下降7 增压级效率下降2 LC2 增压级叶片顶端间隙变大（磨损） 增压级折合流量下降4 LC3 增压级叶片受外物损伤 增压级效率下降5 LC4 增压级叶片腐蚀 增压级效率下降2 HC1 压气机叶片结垢 压气机折合流量下降7 压气机效率下降2 HC2 压气机叶片顶端间隙变大（磨损） 压气机折合流量下降4 HC3 压气机叶片受外物损伤 压气机效率下降5 HC4 压气机叶片腐蚀 压气机效率下降2 HT1 高压涡轮叶片结垢 高压涡轮折合流量下降6 高压涡轮效率下降2 HT2 高压涡轮喷嘴腐蚀 高压涡轮折合流量增加6 HT3 高压涡轮叶片磨损 高压涡轮折合流量增加6 高压涡轮效率下降2 HT4 高压涡轮叶片机械损伤 高压涡轮效率下降5 LT1 低压涡轮叶片结垢 低压涡轮折合流量下降6 低压涡轮效率下降2 LT2 低压涡轮喷嘴腐蚀 低压涡轮折合流量增加6 LT3 低压涡轮叶片磨损 低压涡轮折合流量增加6 低压涡轮效率下降2 LT4 低压涡轮叶片机械损伤 低压涡轮效率下降5 
下载: 导出CSV
表 2 故障风险预警征兆预测
Table 2. Prediction of fault risk early warning signs
故障类型 $ \Delta {N_1}{\text{/\% }} $ $ \Delta {N_2}{\text{/\% }} $ $ \Delta {\pi _{\rm{f}}}{\text{/\% }} $ $ \Delta {\pi _{{\rm{lc}}}}{\text{/\% }} $ $ \Delta {\pi _{{\rm{hc}}}}{\text{/\% }} $ $ \Delta {T_{25}}{\text{/\% }} $ $ \Delta {T_5}{\text{/\% }} $ $ \Delta {W_{\rm{f}}}{\text{/\% }} $ 预警征兆 F1 3.5886 3.0583 −1.2689 0.2669 4.7768 1.9338 4.0346 7.1595 $ {y_0} $ LC1 3.5792 −0.0183 0.5637 −8.9666 7.1416 −0.4402 3.4532 1.6750 $ {y_4} $ HC1 1.1138 0.0021 0.2627 3.6565 −4.2153 0.2970 1.6303 1.1169 $ {y_8} $ HT1 1.8460 −0.0288 −0.0746 −0.8456 7.5887 0.0265 −0.1475 0.0834 $ {y_{12}} $ LT1 −4.6522 −0.0621 0.2429 5.6625 −7.2420 0.5248 2.2697 1.2701 $ {y_{16}} $
下载: 导出CSV
表 3 不同模型准确率对比
Table 3. Comparison of model accuracy rates
模型 训练维度 原始数据集/% 添加噪声数据集/% 平均准确率/% 改进DRSN 16×16 99.2 94.9 97.1 DRSN 16×16 98.6 80.3 89.5 Gauss 4 98.2 75.0 86.6 MOG 4 93.2 85.4 92.0 KNN 4 95.3 69.4 92.3 K-means 4 95.5 72.2 92.7 SVM 4 93.2 91.3 92.3 BP 4 90.6 63.4 77.0
下载: 导出CSV
-
[1] 皮骏,马圣,张奇奇,等. 基于改进果蝇算法优化的GRNN航空发动机排气温度预测模型[J]. 航空动力学报,2019,34(1): 8-17. PI Jun,MA Sheng,ZHANG Qiqi,et al. Aero-engine exhaust gas temperature prediction model based on IFOA-GRNN[J]. Journal of Aerospace Power,2019,34(1): 8-17. (in Chinese PI Jun, MA Sheng, ZHANG Qiqi, et al . Aero-engine exhaust gas temperature prediction model based on IFOA-GRNN[J]. Journal of Aerospace Power,2019 ,34 (1 ):8 -17 . (in Chinese)[2] DING Shuiting,ZHOU Huimin,PAN Bochao,et al. An experimental method to obtain the hard alpha anomaly distribution for titanium alloy aeroengine disks[J]. Chinese Journal of Aeronautics,2021,34(4): 67-82. doi: 10.1016/j.cja.2020.08.024 [3] 丁水汀,张弓,蔚夺魁,等. 航空发动机适航概率风险评估方法研究综述[J]. 航空动力学报,2011,26(7): 1441-1451. DING Shuiting,ZHANG Gong,YU Duokui,et al. Review of probabilistic risk assessment on aero-engine airworthiness[J]. Journal of Aerospace Power,2011,26(7): 1441-1451. (in Chinese DING Shuiting, ZHANG Gong, YU Duokui, et al . Review of probabilistic risk assessment on aero-engine airworthiness[J]. Journal of Aerospace Power,2011 ,26 (7 ):1441 -1451 . (in Chinese)[4] WU Yuting,YUAN Mei,DONG Shaopeng,et al. Remaining useful life estimation of engineered systems using vanilla LSTM neural networks[J]. Neurocomputing,2018,275: 167-179. doi: 10.1016/j.neucom.2017.05.063 [5] YANG Shuming,QIU Jing,LIU Guanjun. Hierarchical model-based approach to testability modeling and analysis for PHM of aerospace systems[J]. Journal of Aerospace Engineering,2014,27(1): 131-139. doi: 10.1061/(ASCE)AS.1943-5525.0000203 [6] 王奕惟,莫李平,王奕首,等. 基于全航段QAR数据和卷积神经网络的航空发动机状态辨识[J]. 航空动力学报,2021,36(7): 1556-1563. WANG Yiwei,MO Liping,WANG Yishou,et al. Aero-engine status identification based on full-segment QAR data and convolutional neural network[J]. Journal of Aerospace Power,2021,36(7): 1556-1563. (in Chinese doi: 10.13224/j.cnki.jasp.20200431 WANG Yiwei, MO Liping, WANG Yishou, et al . Aero-engine status identification based on full-segment QAR data and convolutional neural network[J]. Journal of Aerospace Power,2021 ,36 (7 ):1556 -1563 . (in Chinese) doi: 10.13224/j.cnki.jasp.20200431[7] 李娟,周东华,司小胜,等. 微小故障诊断方法综述[J]. 控制理论与应用,2012,29(12): 1517-1529. LI Juan,ZHOU Donghua,SI Xiaosheng,et al. Review of incipient fault diagnosis methods[J]. Control Theory & Applications,2012,29(12): 1517-1529. (in Chinese LI Juan, ZHOU Donghua, SI Xiaosheng, et al . Review of incipient fault diagnosis methods[J]. Control Theory & Applications,2012 ,29 (12 ):1517 -1529 . (in Chinese)[8] 肖乾浩. 基于机器学习理论的机械故障诊断方法综述[J]. 现代制造工程,2021(7): 148-161. XIAO Qianhao. Review on mechanical fault diagnosis methods based on machine learning theories[J]. Modern Manufacturing Engineering,2021(7): 148-161. (in Chinese XIAO Qianhao . Review on mechanical fault diagnosis methods based on machine learning theories[J]. Modern Manufacturing Engineering,2021 (7 ):148 -161 . (in Chinese)[9] 尹宝才,王文通,王立春. 深度学习研究综述[J]. 北京工业大学学报,2015,41(1): 48-59. YIN Baocai,WANG Wentong,WANG Lichun. Review of deep learning[J]. Journal of Beijing University of Technology,2015,41(1): 48-59. (in Chinese YIN Baocai, WANG Wentong, WANG Lichun . Review of deep learning[J]. Journal of Beijing University of Technology,2015 ,41 (1 ):48 -59 . (in Chinese)[10] 马波,赵祎,齐良才. 变分自编码器在机械故障预警中的应用[J]. 计算机工程与应用,2019,55(12): 245-249,264. MA Bo,ZHAO Yi,QI Liangcai. Application of variational auto-encoder in mechanical fault early warning[J]. Computer Engineering and Applications,2019,55(12): 245-249,264. (in Chinese MA Bo, ZHAO Yi, QI Liangcai . Application of variational auto-encoder in mechanical fault early warning[J]. Computer Engineering and Applications,2019 ,55 (12 ):245 -249, 264 . (in Chinese)[11] LI Yingshun,DONG Wan,YI Xiaojian,et al. Engine fault prediction and health evaluation based on oil sensor[C]//2020 Prognostics and Health Management Conference. Piscataway,US: IEEE,2020: 223-228. [12] ZHANG Zhijin,LI He,CHEN Lei. Deep residual shrinkage networks with self-adaptive slope thresholding for fault diagnosis[C]// Proceedings of 2021 7th International Conference on Condition Monitoring of Machinery in Non-Stationary Operations. Guangzhou: IEEE,2021: 236-239. [13] ZHAO Minghang,ZHONG Shisheng,FU Xuyun,et al. Deep residual shrinkage networks for fault diagnosis[J]. IEEE Transactions on Industrial Informatics,2020,16(7): 4681-4690. doi: 10.1109/TII.2019.2943898 [14] LIU Jiuzhou,ZHOU Qianwen,ZHANG Biyin. Deep residual shrinkage network for few-shot learning[C]// Proceedings of 2021 3rd International Conference on Intelligent Control,Measurement and Signal Processing and Intelligent Oil Field. Xi'an: IEEE,2021: 120-124. [15] CHE Changchang,WANG Huawei,NI Xiaoping,et al. Fault diagnosis of rolling bearing based on deep residual shrinkage network[J]. Journal of Beijing University of Aeronautics and Astronautics,2021,47(7): 1399-1406. [16] VIANNA M R,IZQUIERDO L A,BARROS D M,et al. Short- and long-term memory: differential involvement of neurotransmitter systems and signal transduction cascades[J]. Anais Da Academia Brasileira De Ciencias,2000,72(3): 353-364. doi: 10.1590/S0001-37652000000300009 [17] ZOU Hui,XUE Lingzhou. A selective overview of sparse principal component analysis[J]. Proceedings of the IEEE,2018,106(8): 1311-1320. doi: 10.1109/JPROC.2018.2846588 [18] HE Kaiming,ZHANG Xiangyu,REN Shaoqing,et al. Deep residual learning for image recognition[C]// Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas,US: IEEE,2016: 770-778. [19] 刘元芳. 航空发动机气路故障的智能诊断方法研究[D]. 厦门: 厦门大学,2018. LIU Yuanfang. Research on intelligent diagnosis method for aircraft engine gas path fault[D]. Xiamen: Xiamen University,2018. (in ChineseLIU Yuanfang. Research on intelligent diagnosis method for aircraft engine gas path fault[D]. Xiamen: Xiamen University, 2018. (in Chinese) [20] 郝英. 基于智能技术的民航发动机故障诊断和寿命预测研究[D]. 南京: 南京航空航天大学,2006. HAO Ying. Research on civil aviation engine fault diagnosis and life predication based on intelligent technologies[D]. Nanjing: Nanjing University of Aeronautics and Astronautics,2006. (in ChineseHAO Ying. Research on civil aviation engine fault diagnosis and life predication based on intelligent technologies[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2006. (in Chinese) [21] 瞿红春. 民用涡扇发动机健康智能监控技术的研究[D]. 天津: 天津大学,2010. QU Hongchun. Study on civil turbofan engine health intelligent monitoring technologies[D]. Tianjin: Tianjin University,2010. (in ChineseQU Hongchun. Study on civil turbofan engine health intelligent monitoring technologies[D]. Tianjin: Tianjin University, 2010. (in Chinese) -
点击查看大图
计量
- 文章访问数: 212
- HTML浏览量: 90
- PDF量: 47
- 被引次数: 0