细胞神经网络求解点堆的动力学方程  

Solution of point reactor neutron kinetic equations using cellular neural network

作　　者：李忠勤 Li Zhongqin(School of Electrical&Control Engineering,Heilongjiang University of Science&Technology,Harbin 150022,China)

机构地区：[1]黑龙江科技大学电气与控制工程学院,哈尔滨150022

出　　处：《黑龙江科技大学学报》2018年第6期712-717,共6页Journal of Heilongjiang University of Science And Technology

基　　金：国家自然科学基金项目(51674109);哈尔滨市科技创新人才研究专项资金项目(2017RAQXJ114)

摘　　要：为有效克服点堆中子动力学方程的刚性,准确、快速、稳定地求解方程,同时考虑到缓发中子对堆内中子的影响,采用Simulink建立了三层12×12的细胞神经网络模型,模拟实现1/4均匀裸堆在稳态和插入控制棒两种情况下的中子通量分布,利用二分法求解堆内的有效增殖系数。结果表明,细胞神经网络计算结果正确,能比较准确地预测反应堆中子通量密度。细胞神经网络可用于模拟反应堆堆内中子的动态变化。This paper is an attempt to effectively overcome the stiffness inherent in the point reactor neutron kinetic equations and come up with a more accurate,quick and steady solution of the equation,with a simultaneous consideration of the effect of delayed neutrons on neutrons in the reactor.The targeted study involves establishing a cellular neural network model with the three layers of 12×12 using Simulink;thereby simulating the neutron flux distribution of 1/4 bare homogeneous reactor in the two cases:steady state and insertion of control rods;and obtaining the solution to the effective proliferation coefficient in reactor using dichotomy.The results demonstrate that the cellular neural network could enable a correct calculation and a more accurate prediction of the reactor neutron flux density.The cellular neural network could work better for the simulation of the dynamic changes of neutrons in a reactor.

关 键 词：细胞神经网络 点堆动力学方程 SIMULINK 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]