学术¶
时序预测 - Time Series Prediction
Data-driven Remaining Useful Life (RUL) prediction is one of the core technologies of Prognostics and Health Management (PHM). Committed to improving the accuracy of RUL prediction for aero-engines, this paper proposes a model that is entirely based on the attention mechanism. The attention model is divided into the multi-head self-attention and timing feature enhancement attention models. The multi-head self-attention model employs scaled dot-product attention to extract dependencies between time series; the timing feature enhancement attention model is used to accelerate and enhance the feature selection process. This paper utilises Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) turbofan engine simulation data obtained from NASA Ames’ Prognostics Center of Excellence and compares the proposed algorithm to other models. The experiments conducted validate the superiority of our model’s approach.
超分辨重构 - Super Resolution
一种光流法和深度网络结合的视频高时空分辨率信号处理方法 CN201910906175.1
一种光流法和深度网络结合的视频高时空分辨率信号处理方法，采用基于信号与信息处理算法的方式恢复和重构出高空间和高时间分辨视频序列，即一种视频超分辨率重建方法。其步骤为:将视频按顺序取其帧序列;从视频的第3帧开始，每一帧都与其前后两帧做光流法运动估计;将产生的4副运动估计图像与中间帧合成为5副图像的高维图像块;构建OF深度卷积超分辨率网络，浅层网络提取图像信息，最后一层亚像素卷积层进行重建超分辨率图像;将高维图像块送入深度卷积网络进行训练;最终将降质后或者分辨率较低的视频帧送入网络进行重建。本发明实现重建质量好、重建速度快，相比传统视频超分辨率模型重构效果好，能够进行视频实时重建。
目标检测 - Object Detection
风格迁移增强的机场目标检测方法 CN202210332513.7
目标检测 - Object Detection
[风格迁移增强的机场目标检测方法] 计算机应用与软件（未见刊）
自然语言处理 -
[基于ERNIE-SKEP与BiGRU的航空公司旅客评价情感分析] 计算机仿真（未见刊）
目标检测 - Object Detection
Paper: Domain adaptive object detection with generative adversarial network 2020 International Conference on Internet of Things and Intelligent Applications (ITIA)
Domain adaptive object detection has emerged to alleviate the lack of massive amounts of labeled data in the target field. Currently, Generative Adversarial Network (GAN) has been exhibited its benefits to the Faster R-CNN for some specific application. However, it brings up an open question that GAN still has flexibility and robustness in more generic scenarios. Thus, this paper aims to extensively evaluate its efficiency of the GAN on more generic domain-adaptive object detection. More specifically, it presents four generic domain-adaptive object detection with GAN architectures, containing Faster R-CNN with CycleGAN, Faster R-CNN with DualGAN, RetinaNet with CycleGAN, and RetinaNet with DualGAN. Extensive experiment results demonstrate that GAN-derived model makes significant contribution and consistent improvements in two domain shift scenarios for domain-adaptive object detection. Additionally, comparing with the state-of-the-art methods, domain-adaptive object detection with GAN has been validated its ability of generalization and robustness in more challenging scenarios.
目标检测 - Object Detection
Paper: Research on Baggage Tray Recognition in Airports Based on Residual Neural Network 2022 2nd International Conference on Big Data Engineering and Education (BDEE)
In recent years, civil aviation industry of China has made extraordinary achievements, and since the reform and opening up, civil aviation industry of China has entered a new era of rapid development, and the airport construction has also been developed continuously. At present, the wisdom civil aviation program proposed by the state is being implemented gradually. Wisdom civil aviation refers to the application of new generation technologies such as artificial intelligence, Internet of Things, cloud computing, etc. to the civil aviation industry, making travel and logistics to digital transformation, but there is still a lack of further development. Due to the diversified, non-standardized and non-specified characteristics of checked baggage, trays are introduced as the carrier of checked baggage and baggage handling system in airports, but there are problems such as difficulties in tray recovery and low efficiency. To address this problem, this paper proposes the use of ResNet network architecture to identify whether a pal let is used or not and facilitate the systematic handling of pallet re covery work. The algorithm model proposed in this paper can achieve 95.17% accuracy in the problem of intelligent recognition of whether a pallet is used or not, which verifies the effectiveness of the network model for this problem and provides a favorable help for the development of smart civil aviation.
时序预测 - Time Series Prediction
Paper: 多特征注意力的航空发动机剩余寿命预测模型 航空工程进展
航空发动机性能退化趋势复杂，适时地对其进行剩余寿命预测和检修维护十分重要。提出一种基于多特征注意力的膨胀卷积网络模型来预测航空发动机剩余使用寿命，利用膨胀卷积增强提取序列数据时序信息的能力，同时建立残差连接以改善传统卷积网络中的梯度消失问题。首先采用定长滑动时间窗沿时间维度截取数据，对数据进行重构;再对每个特征对应的时间序列单独应用膨胀卷积提取时序信息;引入特征注意力机制计算各特征之间的相对重要性;在公开的航空发动机数据集上进行验证，并对比现有的主流预测方法。结果表明:该模型在时间序列数据预测方面有着更高的精度。