正问题的求解效率和计算稳定性在结构优化设计中是至关重要的。在正问题求解中采用了一种实时算法——一缩减基法。从理论上介绍了缩减基法,并对缩减基法计算结果的误差进行估计;然后以蜂窝梁结构为例,得到修正设计参数后结构的缩减基法结果及该计算结果的误差限。与有限元结果相比较,缩减基法不仅计算速度快,而且计算精度也很高,采用微遗传算法进行优化时,每一次迭代时调用缩减基法计算结果,大大缩短了设计周期,提高了工作效率。
Efficiency and stability of forward problem are important in structural optimum design. A real-time algorithm, called reduced-basis method (RBM) ,was applied to solve the forward problem herein. Firstly, RBM was introduced theoretically, and the calculation error estimation can be also provided. Taking a castellated beam for example,the results computed from RBM and error bound were obtained. Compared with results from the finite element method, RBM can obtain accurate results lastly. After vaying the parameters, computation of forward problem was needed to carry on in the optimization process. Reduced-basis method are combined with micro genetic algorithm, which saves time and improves the efficiency.