Answers are grounded in this paper and cite their sources.
To achieve the accurate measurement of the activity size distribution of radon progeny and identify the distribution mode automatically, a measurement method based on the screen diffusion battery (SDB) and the Multilayer Perceptron (MLP) machine learning algorithm was developed in this study. Comparison experiments and field measurements were also carried out. Results show that the MLP algorithm presents superior performance in accurately identifying the unimodal or bimodal distribution and avoiding unambiguous solutions compared with the traditional ex-pectation maximization (EM) and Twomey inversion algorithms. Meanwhile, the SDB+MLP method proposed in this study has principle-based advantages over the conventional impactor method and SMPS indirect measurement method for solving parameters and the fraction of small-sized nucleation mode radon progeny. Field measurement results show that the SDB+MLP system can be applied to the activity size distribution measurement of the attached radon progeny. And the activity size distribution of the attached radon progeny in the outdoor environment all exhibits a bimodal distribution with the average diameter of the nucleation mode is 22 nm and that of the accumulation mode is 229 nm, and only a small portion of those show a unimodal distribution in the indoor environment, while the average diameters of the nucleation mode and accumulation mode separately are 25 nm and 209 nm. For the average fraction of the accumulation mode, a little difference of 66% and 69% for indoors and outdoors is shown, and the range is from 52% to 84% for outdoors and 49% to 87% for indoors.
No takes yet. Share an insight, caveat, or question.
Ma et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: