Research on Core Neutronic Parameter Prediction Based on Neural Network Hyperparameter Optimization Method

[Background]:Neural networks, with their powerful fitting capabilities, can learn the relationships between input and output variables based on large amounts of data, often serving as proxy models for physical programs in the field of engineering calculations, including nuclear engineering calculations. Neutron transport calculations, as one of the core links in neutronics simulations, often suffer from lengthy computational times. However, this issue can also be addressed by utilizing neural network models. Nevertheless, neural network models have a series of hyperparameters that need to be set, but manually adjusting these hyperparameters is laborious, repetitive, and reliant only on experience. Moreover, these hyperparameters are not reusable when solving different problems. [Purpose]: This study aims to improve computational efficiency and provide reference for the application of artificial intelligence in core physics calculation theory.[Methods]: The paper proposed the use of the Bayesian optimization algorithm to adjust fully connected neural network hyperparameters, combined with adaptive learning rate decay and loss function optimization methods. In order to verify the generalization and accuracy of this method，the study fitted the critical core parameters: the effective multiplication factor and the regional integrated flux, obtained from the TAKEDA benchmark problem. [Results]:The results show that the average error of the effective multiplication factor is within 150×10-5, and the average error rate of the regional integral flux on the TAKEDA1 dataset is 1.72%, with a maximum error rate of 7.56%. [Conclusions]: This approach can automatically search for the optimal combination of hyperparameters for different datasets to achieve the best performance, demonstrating high flexibility, efficiency, and strong generalization.