【摘要】 根据快堆裂变产物行为的动力学模型和燃料元件包壳氧化腐蚀的机理，选取燃料初始Ｏ／Ｍ比、燃耗和包壳内壁温度作为自变量，以６个国家的堆内试验参数和８２个ＦＣＣＩ测量值为样品数据，用神经网络扩展δ算法拟合出这些变量与ＦＣＣＩ深度值间的非线性映射关系式。以该方法计算的ＦＣＣＩ深度和相应的堆内ＦＣＣＩ测量值作对比，得出计算值与测量值的相符率、计算值的均方差及其置信度，明显优于现有ＦＣＣＩ经验方程。用该方法分析预测了ＣＥＦＲ设计参数下的ＦＣＣＩ深度。更多还原

【Abstract】 Based on the kinetics mode of fission products behaviour and the oxidation corrosion mechanism of fuel element cladding,the initial O/M ratio of fuel,burn up and temperature at inner wall of cladding are selected as the variables of nerve net extension δ calculation method.In this method,the parameters and FCCI measurement values in pile tests for six countries are used to draw out the nolinear shine upon relationship between these variables and FCCI depth.Comparing the FCCI calculation with its measurement values,the accord ratio between calculation and measurement values,the square average error and the reliability of calculation values are more excellent than that of the existent equations.According to the design parameters of CEFR,the FCCI depth of CEFR with above nerve net extension δ calculation method was evaluated.更多还原