离散变量结构优化设计序列定界组合算法研究

Application of a sequential delimitative and combinatorial algorithm to discrete optimum design of structures

首先给出了离散变量结构优化设计的数学模型，然后将总组合数按照设计变量的取值分类，把原规划问题分解成若干个维数更低、许用离散集更少的子问题；在算法设计过程中，将离散变量定界组合算法的思想应用于该问题，给出了定界的新方法． 若干算例表明，该方法可以大大减少组合搜索的次数，提高解题效率．

The idea of the sequential delimitative combinatorial algorithm is used to the optimum design of structure with discrete variables. First, the optimal problem is converted to several sequential sub\|problems with lower dimensions and similar structures by recurrent method. In the procedure of computation ,new combinations are generated by using a multi\|level generating method and the sequential delimitative combinatorial algorithm is used to search optimum solution. In the procedure of delimitation,both the constraint functions and united constraint function are used to delimited so that the computational efficiency is very higher than only using united function to delemited.Several examples show that the algorithm has higher efficiency.