【摘要】 讨论了一类非凸二维结构函数 f (x)在数字空间 Z2 中的逼近分解问题 ,利用 BP网络来确定分解基元的选择与组合 ,使结构函数的分解保持拓扑一致性 ,最后给出这类非凸二维结构函数的一个逼近分解方法更多还原

【Abstract】 In this paper, we discuss an approach decomposition for a class of non convex two demension structural function f(x) in digital space Z\+2 Using BP neural network determines the selection and combination of decomposition basis elements, which makes the decomposition of structural function remain topological Finally, an approach decomposition method for the non convex two demension structural function is given更多还原