The nuclide depletion calculation is of great significance in the design and research of the reactor core scheme.The change of nuclide content can be described by the Bateman equation[1],and the kernel of reactor burn...The nuclide depletion calculation is of great significance in the design and research of the reactor core scheme.The change of nuclide content can be described by the Bateman equation[1],and the kernel of reactor burnup calculation is the solution of Bateman equation.The classic calculation program for nuclide depletion has procedures such as ORIGEN.Since the half-life values of nuclide are widely distributed,the norm of fuel consumption coefficient matrix of the nuclides is very large,so it is difficult to numerically solve the equation.展开更多
学习机器性能是决定智能位移反分析效果的关键,针对现有智能反分析存在的问题,将高斯过程回归(Gaussian Process Regression,简称GPR)引入隧道工程计算模型参数的反演,并采用单一各向同性核函数之和作为GPR的组合核函数以提高其泛化性...学习机器性能是决定智能位移反分析效果的关键,针对现有智能反分析存在的问题,将高斯过程回归(Gaussian Process Regression,简称GPR)引入隧道工程计算模型参数的反演,并采用单一各向同性核函数之和作为GPR的组合核函数以提高其泛化性能。为克服传统共轭梯度法优化求取最优GPR超参数的缺陷,改用十进制遗传算法替代共轭梯度法在训练过程中搜索GPR最优超参数,并编制了相应的计算程序。结合北口隧道施工监测进行了算法程序的应用,并与进化–单一核函数高斯过程回归算法和进化支持向量回归(SVR)算法的应用结果作了对比,结果表明本文提出的进化高斯过程算法显著提高了反演精度,可以应用于岩土工程计算模型参数的反演辨识,并为类似工程提供了借鉴。展开更多
文摘The nuclide depletion calculation is of great significance in the design and research of the reactor core scheme.The change of nuclide content can be described by the Bateman equation[1],and the kernel of reactor burnup calculation is the solution of Bateman equation.The classic calculation program for nuclide depletion has procedures such as ORIGEN.Since the half-life values of nuclide are widely distributed,the norm of fuel consumption coefficient matrix of the nuclides is very large,so it is difficult to numerically solve the equation.
文摘学习机器性能是决定智能位移反分析效果的关键,针对现有智能反分析存在的问题,将高斯过程回归(Gaussian Process Regression,简称GPR)引入隧道工程计算模型参数的反演,并采用单一各向同性核函数之和作为GPR的组合核函数以提高其泛化性能。为克服传统共轭梯度法优化求取最优GPR超参数的缺陷,改用十进制遗传算法替代共轭梯度法在训练过程中搜索GPR最优超参数,并编制了相应的计算程序。结合北口隧道施工监测进行了算法程序的应用,并与进化–单一核函数高斯过程回归算法和进化支持向量回归(SVR)算法的应用结果作了对比,结果表明本文提出的进化高斯过程算法显著提高了反演精度,可以应用于岩土工程计算模型参数的反演辨识,并为类似工程提供了借鉴。
