首页|基于空时模型的航空管路卡箍故障诊断研究

基于空时模型的航空管路卡箍故障诊断研究

扫码查看
针对航空液压管路卡箍振动信号受强噪声干扰,导致航空卡箍故障难以精准识别的问题,提出一种空时模型的航空卡箍故障诊断新方法.建立空间特征提取模型,对航空卡箍的故障特征进行局部融合.在空间模型中引入GRU模块,提取航空卡箍故障信号中的全局特征.结果表明:设计的空时故障诊断模型可实现航空卡箍故障的精准识别.与目前所用的深度卷积神经网络模型、门控循环单元神经网络模型、循环神经网络模型、支持向量机和误差反向传播神经网络模型等5种先进的故障诊断方法进行对比分析,所提方法对航空卡箍故障识别具有优越性.
Research on Fault Diagnosis for Aero-Pipeline Clamp Based on Space-Time Model
Aiming at the problem that the vibration signal of aviation hydraulic pipe clamp is interfered by strong noise,it is diffi-cult to accurately identify aviation clamp fault,a new method of aviation clamp fault diagnosis based on space time model was proposed.A spatial feature extraction model was established to carry out local fusion of fault features of aviation clamp.The GRU module was intro-duced into the spatial model to extract the global features of the aviation clamp fault signal.The results show that:the designed space-time fault diagnosis model can be used to realize accurate identification of aviation clamp faults.It was compared with five advanced fault diagnosis methods currently used,including deep convolutional neural network model,gated recurrent unit neural network model,recur-rent neural network model,support vector machine and error back propagation neural network model.The proposed method has advanta-ges in fault identification of aviation clamp.

fault diagnosisspatial feature extractiontime feature extractionaviation pipeline clamp

王铜宇、袁晟友、李开泰、米承权、林洁如、杨同光

展开 >

东北大学机械工程与自动化学院,辽宁沈阳 110819

临沂大学机械与车辆工程学院,山东临沂 276012

故障诊断 空间特征提取 时间特征提取 航空管路卡箍

2024

机床与液压
中国机械工程学会 广州机械科学研究院有限公司

机床与液压

CSTPCD北大核心
影响因子:0.32
ISSN:1001-3881
年,卷(期):2024.52(7)
  • 20