Speaker
摘要
核设施事故后，放射性物质可能会释放到外部环境中。因此，有必要对放射性核素的空间分布进行快速评估。仅使用大气扩散模型预测的放射性核素浓度分布可能会与实际浓度分布产生偏差。为了使预测分布与实际分布更趋一致，本研究开发了一种基于克里金插值的方法，利用少量的环境气溶胶监测数据对预测的浓度分布进行快速修正。此外，还开发了一种基于差分进化算法的环境气溶胶监测点位优化方法，以指导布置更具代表性的监测点，并提高修正后浓度分布的准确性。结果表明，在构建的案例研究中，利用优化后的监测点获取环境监测数据显著提高了重点区域的预测准确性，使假阴性区域减少了97%。
Abstract
Radioactive materials can be released into the external environment after nuclear facility accidents. Therefore, it is necessary to rapidly assess the spatial distribution of radioactive nuclides. Using only the concentration distribution of radioactive nuclides predicted by the atmospheric dispersion models may result in deviations from the actual concentration distribution. To make the predicted distribution more consistent with the actual distribution, a method based on Kriging interpolation was developed to rapidly correct the predicted concentration distribution using a small amount of environmental aerosol monitoring data. In addition, a method for optimizing environmental aerosol monitoring locations based on a differential evolution algorithm was developed to guide the placement of more representative monitoring locations and improve the accuracy of the corrected concentration distribution. The results indicated that using optimized monitoring locations to obtain environmental monitoring data enhanced the accuracy of prediction in key areas in the constructed case-studies, reducing false-negative areas by 97%.
| 关键词 | 核设施事故，环境监测，监测点位，克里金插值，差分进化算法 |
|---|---|
| Keywords | Nuclear facility accidents, Environmental monitoring, Monitoring locations, Kriging interpolation, Differential-evolutionary algorithm |
