Method and system for predicting states of infusion tube

The invention discloses a method and system for predicting states of an infusion tube. The method includes: acquiring internal continuous pressure values of the infusion tube; converting the continuous pressure values to obtain vector data; preprocessing the vector data to obtain input data; inputting the input data into a convolutional neural network to obtain convolutional output values; inputting the output values into a recurrent neural network to determine future states of the tube. By preprocessing of internal real-time pressure values of the infusion tube, influences of variable factorssuch as environments, materials and medicines can be reduced, and applicability and robustness of the method are improved. By adoption of the convolutional neural network for extraction of advanced features, stability and accuracy of prediction results are improved. Taking output of the convolutional neural network as input of the recurrent neural network makes full use of time sequence sensitivity of the recurrent neura