论文部分内容阅读
从堆芯燃料管理装载模式 (L P)优化中的非确定多项式 (NP)特征和全局寻优要求 ,探讨了遗传算法在 L P优化中的应用。结合 L P优化和遗传算法特点 ,建立了 L P优化的数学模型的矩阵表示。为了解决组件矩阵的有效性问题 ,对组件矩阵进行基于位置的遗传操作 ;考虑到可燃毒物数目的独立性 ,对可燃毒物矩阵进行随机遗传操作 ;对旋向矩阵进行经典遗传操作。类比旅行商问题 ,完成了 L P优化遗传算法编码和解码。并可与相应的堆物理计算程序构成一个完整的堆芯燃料管理程序。
From the non-deterministic polynomial (NP) characteristics and the global optimization requirements in the LMP optimization of core fuel management, the application of genetic algorithm in LP optimization is discussed. Combined with the characteristics of L P optimization and genetic algorithm, a matrix representation of L P-optimized mathematical model is established. In order to solve the problem of the validity of the component matrices, the component matrices are based on the location-based genetic operations. Taking into account the independence of the number of flammable toxins, stochastic genetic operations are performed on the combustible poison matrices. Classical hereditary manipulations of the spin matrix are performed. Analog traveling salesman problem, completed L P-optimized genetic algorithm coding and decoding. And can form a complete core fuel management program with the corresponding pile physics calculation program.