The first order gradient optimization algorithm is employed to train the network. The training samples stem from the neutronkinetics of the point-reactor.
神经网络的训练采用一阶梯度优化算法，利用点堆中子动力学模型产生训练样本。
2
A kinetics nodal method for the solution of transient two-group multi - dimensional neutron diffusion equation is described.
介绍一种求解瞬态两群多维中子扩散方程的动力学节块方法。
3
The terms for the effects of the neutrons reflected back from outside the reactor are added as extra delayed neutron terms to the reactor kinetics equation.