基于卷积神经网络的发动机气路故障诊断方法

深度学习是一种新的基于特征表示的机器学习方法。深度学习模型包含多个隐藏层,可以通过对输入数据进行自动学习来获取隐藏的功能层中的特征信息。与传统的诊断方法相比,深度学习具备从原始信息中提取更丰富的特征的能力,因此已经成为基于机器学习的故障诊断研究的新方向,为发动机气路等复杂系统故障诊断带来了新思路。结合发动机气路试验数据的特点与深度学习的优势,提出基于卷积神经网络的故障诊断方法,包括预处理、模型训练及优化等过程,并实现了复杂系统故障诊断预测算法平台。经某发动机气路试验仿真数据实例验证,提出的方法具有较好的可行性和效果,能够充分利用深度学习的优点,更准确地识别发动机气路的健康状况。

Aero-engine Gas Path Fault Diagnostic Method based on Convolutional Neural Network

Deep learning is a new machine learning method based on feature representation. The deep learning model consists of multiple hidden layers, and the feature information in the hidden functional layer can be obtained by automatically learning the input data. Compared with traditional diagnostic methods, deep learning has the ability to extract more abundant features from the original information, so it has become a new area of machine learning-based fault diagnosis research. It brings new idea of the complex system fault diagnostic such as aero-engine gas path. Combining the characteristics of complex system test data and the advantages of deep learning, a fault diagnostic method based on convolutional neural network is proposed, including preprocessing, model training and optimization. Then a complex system fault diagnostic algorithm platform based-on deep learning method is realized. The simulation method of an aero-engine gas path test proves that the proposed method has good feasibility and effect, it can make full use of the advantages of deep learning and more accurately identify the health state of the aero-engine gas path.