Abstract
In this paper, a new method based on Bayesian inference to reconstruct the radiation field on nuclear facilities is proposed. At first, the net function interpolation method is used to reconstruct the radiation field on the basis of the observed data. Considering the effects of radiation shielding, the radiation field near the shields should be reconstructed. According to the measured data, Bayesian inference succeeds in reconstructing a more accurate radiation field. The effectiveness of this method can be verified by two cases. In the first case, there are two calibration areas. Compared with the radiation field simulated by Monte Carlo method, the average relative errors of the net function method are 11.37% and 23.65%, respectively. After the dose rate distribution of the calibration areas is calibrated by Bayesian inference, the average relative errors become 3.70% and 6.67%. The radiation environment of the other case is more complicated and there are six calibration areas. The results have average relative errors of 9.01%, 8.01%, 13.84%, 2.97%, 3.26% and 12.69% respectively before calibration. After calibration, the corresponding relative errors become 3.54%, 5.31%, 4.38%, 2.93%, 3.18% and 5.47%. Therefore, Bayesian inference can improve the accuracy of the reconstructed radiation field.
