基于最大熵的离散变量框架结构两相优化设计

根据两相优化设计的思路和离散变量的特点,应用最大熵准则,通过引入附加约束条件,交替进行乘子相和结构相优化,将离散变量框架优化设计问题等价地化为连续变量框架优化设计问题求解.在考虑应力约束、位移约束和尺寸约束等多个约束及多个工况的条件下,成功地对离散变量框架进行了优化设计. 算例结果表明:这种方法的迭代公式简单 ,收敛速度快;此法可以有效地应用于工程实际,尤其是多约束、多工况的大型工程结构优化问题.

Two-Stage Frame Structure Optimum Design with Discrete Variables Based on Maximum Entropy

Because the present algorithm is difficult even unable to solve the problem of structural optimization design of discrete variables, based on the thinking of two-stage optimization, discrete variable and maximum entropy, through introducing additional constraint condition, this problem is equivalently converted to the frame optimization design problem of continuous variables, which can be resolved by carrying out multiple-stage and structure-stage optimization process alternatively. Multiple loading cases and multiple constraints including stress constraints, displacement constraints and dimension constraints are all considered in this paper. Under this condition, the discrete variables frame structures are successfully designed and optimized. Examples show that this algorithm is concise and the iteration is converged rapidly, which can be applied to practical engineering efficiently, especially to the structure optimization of multiple loading cases and multiple constraints.