AUTHOR=Yao Yuantao , Ge Daochuan , Yu Jie , Xie Min 

TITLE=Model-Based Deep Transfer Learning Method to Fault Detection and Diagnosis in Nuclear Power Plants

JOURNAL=Frontiers in Energy Research

VOLUME=Volume 10 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/energy-research/articles/10.3389/fenrg.2022.823395

DOI=10.3389/fenrg.2022.823395

ISSN=2296-598X

ABSTRACT=Deep learning-based nuclear intelligent fault detection and diagnosis (FDD) methods have been widely developed and achieved very competitive results with the progress of artificial intelligence (AI) technology. However, the pre-trained model for diagnosis tasks is hard to achieve good performance when the reactor operation conditions are updated. On the other hand, retraining the model for a new dataset will waste computing resources. This paper proposes an FDD method for cross-condition and facility tasks based on the optimized transferable convolutional neural network (CNN) model. First, by using the pre-trained model's prior knowledge, the model's diagnosis performance to be transferred for source domain datasets is improved. Second, a model-based transfer learning strategy is adopted to freeze the feature extraction layer in a part of the training model. Third, the training data in target domain datasets are used to optimize the model layer by layer to find the optimization model with the transferred layer. Finally, the proposed comprehensive simulation platform provides source and target cross-condition and cross-facility datasets to support case studies. The designed model utilizes the strong nonlinear feature extraction performance of a deep network and applies the prior knowledge of pre-trained models to improve the accuracy and timeliness of training. The results show that the proposed method is superior to achieving good generalization performance at less training epoch than the retraining benchmark deep CNN model.