In order to enhance the accuracy and reliability of fault diagnosis in the chemical process,this paper pre-sents a method for chemical process fault diagnosis based on an improved ResNet-GRU neural network.Firstly,an im-proved ResNet model is introduced using pre-activation to extract features from the input data,thereby enhancing the model's ability to capture key features.Secondly,the GRU model is employed to perform temporal modeling on the ex-tracted features,allowing for better capturing of the dynamic changes in fault signals.To validate the effectiveness of this method,conducted experiments using the TE chemical process dataset and analyzed the results using a confusion matrix.The improved ResNet-GRU model achieved an average fault diagnosis rate of 95.16%,surpassing other deep learning methods in terms of fault diagnosis rate and reliability.
fault diagnosisresidual neural networkgated recurrent neural networkdeep learningchemical process
维普
万方数据
