基于两点累积信息的原／倒变量展开的对偶优化方法

THE DUAL OPTIMIZATION METHOD BY ORIGINAL/RECIPROCAL VARIABLES’EXPANSION BASED ON CUMULATIVE INFORMATION AI TWO POINTS

提出和论证了改善保守近似的展开“准则”及其定理；提出了基于两点累积信息的原／倒变量近似、论证了其凸凹性不变的优点；提出了应用它的对偶优化方法及其在桁架优化的实现；提出和证明了有关收敛性定理。

In this paper,we propose and prove following results:1.an expansion“criteria”of improving conservative approkimation and corresponding theorms;2.an approkimateEORCT(expansion of original/reciprocal variables based on cumulative informations attwo points)and its advantage of identical convex or concave character with primal function;3.a dual optimization method by using EORCT approximation and numerical collvegencyof its application to truss structures;4.relating convergent theorms of the optimizationmethol