2DCNN-Trans Based Fault Diagnosis Method for Air Handling Units

Addressing the issue that traditional air handling unit (AHU) fault diagnosis methods find it difficult to effectively extract fault features, this paper proposes a fault diagnosis model that combines convolutional neural networks (CNN) and Transformers, suitable for the AHU system. Initially, the 2DCNN is used to extract features from samples that have been transformed into two-dimensional data images, and then the feature-extracted data are fed into the encoder part of the Transformer for automatic fault identification and classification. To validate the effectiveness of this method, seven fault cases from the AHU system of the ASHRAE-1312 project were used for diagnosis, achieving an overall accuracy of 99.94%. Finally, to confirm the feasibility of the model, it was compared with several other models, and good results were obtained.