Aiming at the problem that the vibration signal of aviation hydraulic pipe clamp is interfered by strong noise,it is difficult 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 introduced 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,recurrent neural network model,support vector machine and error back propagation neural network model.The proposed method has advantages in fault identification of aviation clamp.