核技术利用辐射安全监管时空分布模型

环保部门的监管是保障核技术利用设备安全使用的必要环节,现阶段缺乏直观的、可量化的监管手段,导致管理手段低效、监管工作难以合理开展。因此必须通过模式化、指标参考化的监管手段,合理有效安排监管工作,避免监管薄弱环节与难以管理的问题。基于指标建立对检查工作进行多维度分析,并利用卷积神经网络(convolutional neural networks, CNN)对核技术设备持有单位建立风险等级划分的时空分布模型,为核安全监管机关提供可量化的管理依据,并且利用可视化手段辅助管理。结果表明,所建模型指标代表性高,风险等级划分合理,并且能有效提高监管质量与效率。

Establishing a Temporal and Spatial Distribution Model of the Radiation Safety Supervision of Nuclear Technology Utilization

The supervision of the environmental protection department is the necessary link to ensure the safe use of nuclear technology. At present, there is no intuitive and quantifiable regulatory means, which leads to inefficient management methods and difficult to carry out supervision work. Therefore, it is necessary to arrange supervision work reasonably and effectively to avoid regulatory weaknesses and difficult management issues. The study established a multi-dimensional analysis of inspection work based on indicators, and used convolutional neural network(CNN) to analyze the temporal and spatial distribution of risk technology classification of nuclear technology equipment holders. This model provides a quantifiable management basis for nuclear safety regulatory agencies and uses visualization to aid management. The experimental results show that the model can effectively improve the quality and efficiency of supervision, which have high representation of indicators and reasonable division of risk levels.