In order to solve the problems of fast attenuation, strong interference and ambiguous disturbance characteristics of magnetic signals in the magnetic field proximity detection of underwater targets, a time series disturbance signal detection method for underwater targets with industrial frequency magnetic field based on hybrid neural network and attention mechanism (att-CNN-GRU) is presented. This method combines CNN, GRU neural network and Attention mechanism to fit the signal, and constructs a classification neural network to classify and identify the target signal. The prediction and detection performance of this method is compared with that of CNN-LSTM model without attention mechanism and single CNN and LSTM network model. The measured results show that the fitting effect reduces the error by 36.24%, 14.44%, 4.878% and the target detection accuracy reaches 83.3% compared with the above traditional methods. This method can be widely used in underwater target detection as an auxiliary detection method.