由于北斗卫星短报文通信技术存在数据包丢失，在对船舶燃气轮机远程态势感知时，面临信息不完整、数据挖掘困难等问题。针对统计学丢失包插补存在的缺陷，提出了一种多元变量两阶段燃气轮机丢失包智能插补算法。在时间尺度上设计了具有交叉索引机制的双向门控循环单元，通过交叉索引来有效规避已丢失的数据包，同时在属性尺度上建立了高斯聚合去相关自编码器，利用燃气轮机运行过程中的多元变量高斯特性，引入高斯聚合因子来优化去相关自编码器的重构误差，实现了北斗卫星通信链路波动下的远程燃气轮机数据包丢失插补。仿真实验结果验证了所提插补算法应用于燃气轮机远程态势感知的有效性。
Due to packet loss in BeiDou satellite short message communication technology, remote situational awareness of gas ship turbines faces issues such as incomplete information and difficulty in data mining. A multi variable two-stage intelligent imputation algorithm for gas turbine missing packets is proposed to address the shortcomings of statistical missing packet imputation. A BiGRU with cross indexing mechanism was designed on the time scale to effectively avoid lost data packets. Meanwhile, a Gaussian Polymerization TRAE was established on the attribute scale, effectively utilizing the multivariate Gaussian characteristics during the operation of gas turbines. The Gaussian Polymerization factor was introduced to optimize the reconstruction error of the decorrelation autoencoder, achieving remote gas turbine data packet loss imputation under fluctuations in the Bei Dou satellite communication link. The simulation experimental results validate the effectiveness of the proposed imputation algorithm applied to remote situational awareness of gas turbines.
2025,47(20): 132-139 收稿日期：2025-1-20
DOI：10.3404/j.issn.1672-7649.2025.20.020
分类号：U664.1；TK477
基金项目：国家自然科学基金资助项目（51272066）
作者简介：吴迪(1994-),男,博士研究生,研究方向为数据驱动的燃气轮机智能控制与优化
参考文献：
[1] 柯涛, 宋佳, 吴俭, 等. 基于多元信息融合的无人艇自主航行环境态势智能感知方法及验证[J]. 舰船科学技术, 2022, 44(24): 76-80.
KE T, SONG J, WU J, et al. Intelligent perception method and verification of the environment situation of autonomous navigation of unmanned vehicles based on multiple information fusion[J]. Ship Science and Technology, 2022, 44(24): 76-80.
[2] LIU S, WU D, SUN H, et al. A novel beidou satellite transmission framework with missing package imputation applied to smart Ships[J]. IEEE Sensors Journal, 2022, 22(13): 13162-13176.
[3] 周春荣, 蒋郑红. 考虑随机因素的舰船网络态势预测模型[J]. 舰船科学技术, 2020, 42(18): 130-132.
ZHOU C R, JIANG Z H. A ship network situation prediction model considering stochastic factors[J]. Ship Science and Technology, 2020, 42(18): 130-132.
[4] CHEN B, YE W. Low-error data recovery based on collaborative filtering with nonlinear inequality constraints for manufacturing processes[J]. IEEE Transactions on Automation Science and Engineering, 2020, 18(4): 1602-1614.
[5] WANG L, WU S, WU T, et al. HKMF-T: Recover from blackouts in tagged time series with hankel matrix factorization[J]. IEEE Transactions on Knowledge and Data Engineering, 2020, 33(11): 3582-3593.
[6] HEMANTH G, RAJA C. Proposing suitable data imputation methods by adopting a stage wise approach for various classes of smart meters missing data–practical approach[J]. Expert Systems with Applications, 2022, 187: 115911.
[7] YIN M, CHEN R, LIN R, et al. Multi-SSALvcAE: Self-supervised adversarial learning-based view-common latent auto encoders for multiview clustering[J]. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2024, 54(9): 5456-5467.
[8] YUAN Z, LIU J, LIU Y, et al. A multi-task analysis and modelling paradigm using LSTM for multi-source monitoring data of inland vessels[J]. Ocean Engineering, 2020, 213: 107604.
[9] HANG H, LIN L, ZHAO L, et al. TSHNN: Temporal-spatial hybrid neural network for cognitive wireless human activity recognition[J]. IEEE Transactions on Cognitive Communications and Networking, 2024, 10(6): 2088-2101.
[10] ZHANG Y, CHEN J, LIU L, et al. Generating imperceptible and cross-resolution remote sensing adversarial examples based on implicit neural representations[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62.
[11] LUO X, WANG A, ZHAN C, et al. Latraivf: An infrared and visible image fusion method based on latent regression and adversarial training[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 1-16.
[12] ZHANG Y, THORBURN P, XIANG W, et al. SSIM—A deep learning approach for recovering missing time series sensor data[J]. IEEE Internet of Things Journal, 2019, 6(4): 6618-6628.
[13] AWAN S, BENNAMOUN M, SOHEL F, et al. Imputation of missing data with class imbalance using conditional generative adversarial networks[J]. Neurocomputing, 2021, 453: 164-171.
[14] WANG Y, LI D, LI X, et al. PC-GAIN: Pseudo-label conditional generative adversarial imputation networks for incomplete data[J]. Neural Networks, 2021, 141: 395-403.
[15] ZHOU X, LIU X, LAN G, et al. Federated conditional generative adversarial nets imputation method for air quality missing data[J]. Knowledge-Based Systems, 2021, 228: 107261.
[16] NGUYEN B L, DOAN A D, CHIN T J, et al. Sensor allocation and online-learning-based path planning for maritime situational awareness enhancement: a multi-agent approach[J]. IEEE Transactions on Intelligent Transportation Systems, 2024, 25(9): 11635-11647.
[17] ZHAO Q, ZHANG Y, ZHAO Z, et al. A joint inversion approach of electromagnetic and acoustic data based on pearson correlation coefficient[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 25: 11635-11647.
[18] 徐杨, 廖小琴. 融合双向门控循环单元和卷积神经网络的歧视性言论鉴别[J]. 武汉大学学报(理学版), 2020, 66(2): 111-116.
[19] LAI X, WU X, ZHANG L, et al. Imputations of missing values using a tracking-removed autoencoder trained with incomplete data[J]. Neurocomputing, 2019, 366: 54-65.
[20] 彭维仕. 高超声速飞行器命中精度高准确度评估[J]. 北京航空航天大学学报, 2022, 48(11): 2130-2137.
[21] CAO W, WANG D, KI J, et al. Brits: bidirectional recurrent imputation for time series[J]. Advances in Neural Information Processing Systems, 2018, 31.
[22] ZHU B, LAM J, SHEN J, et al. Input-output gain analysis of linear discrete-time systems with cone invariance[J]. IEEE Transactions on Automatic Control, 2024, 69(12): 8751-8757.
