Industrial control network intrusion detection method based on 1D CNN-BiSRU

The invention provides an industrial control network intrusion detection method based on a 1D CNN-BiSRU, and the method comprises the steps: firstly, carrying out the preprocessing of flow data in a Gas Pipeline data set, obtaining input data, and carrying out the marking of the input data; secondly, inputting the input data into the 1D CNN model to obtain a feature vector; inputting the feature vector into a BiSRU neural network for prediction, and adjusting parameters of the BiSRU neural network by comparing a prediction result with the loss of an actual label value to obtain a detection model; and finally, inputting the flow data in the Gas Pipeline data set into the detection model, and outputting a classification result. According to the method, the spatial features of the network traffic data of the ICS are learned through the one-dimensional CNN, the BiSRU can learn the bidirectional structural features of the network traffic data of the ICS through forward and backward input, and accurate detection is