Research on Inverse Reconstruction Method of Radiation Dose Distribution Based on RBF Interpolation

Jifeng Hu,Yongkuo Liu,Zhitao Chen,Xinrong Qian
DOI: https://doi.org/10.1115/icone31-134809
2024-01-01
Abstract:Abstract The precise reconstruction of radiation dose fields holds paramount significance in radiation protection during the decommissioning of nuclear facilities. In this study, a radiation field reconstruction method based on sparse scattered monitoring data was employed, utilizing Radial Basis Function (RBF) interpolation. We conducted a detailed analysis of both RBF interpolation and its compact support variant and validated their reconstruction effectiveness through four sets of simulation experiments. The first two experiments simulated unshielded environments with a single radiation source and multiple radiation sources, respectively. The subsequent two experiments simulated shielded environments with a single radiation source and multiple radiation sources, respectively. The reconstructed results were compared with those obtained using the Kriging method, with 2.5% of all nodes utilized as measurement data in the experiments. The four sets of experiments utilizing RBF interpolation and compact support RBF interpolation exhibited commendable reconstruction performance, with average relative errors of 1.06%/3.52%, 1.13%/2.56%, 8.38%/11.59%, and 3.98%/5.65%, respectively. In comparison to the Kriging method, RBF interpolation consistently outperformed in all experiments. It is worth noting that although the overall performance of compact support RBF is relatively inferior to that of the Kriging method, its reconstruction accuracy at the locations of radiation sources surpasses that of the Kriging method. These research findings provide preliminary evidence of the potential application prospects of RBF interpolation in radiation field reconstruction.
What problem does this paper attempt to address?