基于堆叠稀疏自编码神经网络的航空发动机剩余寿命预测方法研究

ResearchonPredictionMethodofAeroengineResidualLifeBasedonStackedSparseAutomaticEncoder

摘要:

航空发动机是飞行器的核心动力系统,工作环境恶劣,对其进行状态监测和寿命预测是保障飞行器安全可靠运行的重要技术手段。本文研究一种基于堆叠稀疏自编码神经网络的航空发动机剩余寿命预测方法,首先将多个自编码网络连接构成深度堆叠自编码网络,选取发动机的状态数据作为网络的训练输入,使网络逐层智能提取数据间的分布式规则,从而构建发动机退化的堆叠自编码学习模型。通过采用BP神经网络对发动机剩余寿命区间进行分类,作为发动机剩余寿命预测的结果。通过使用PHM2008挑战赛中发动机退化数据对本文研究方法进行了验证,结果验证了堆叠自编码网络深度学习方法对航空发动机剩余寿命预测的有效性。

Abstract:

Aeroengine is the core power system of the aircraft. The working environment is harsh. The state monitoring and life prediction are important technical means to ensure the safe and reliable operation of the aircraft. This paper proposes a method for predicting the remaining life of aeroengine based on stacked sparse autoencoder. Firstly, multiple self-encoding networks are connected to form a deep stack self-encoding network, and the state data of the engine is selected as the training input of the network to make the network layer-by-layer intelligent extraction. Distributed rules between data to build an engine-degraded stacked self-encoding learning model. The BP residual neural network is used to classify the remaining life of the engine as a result of the prediction of the remaining life of the engine. The proposed method is validated by using the engine degradation data in the PHM2008 Challenge. The results verify the effectiveness of the stacked self-encoding network deep learning method for the prediction of remaining life of aeroengine.