基于均匀变形和混合智能算法的框架结构抗震优化设计

Aseismic optimization design of a frame structure based on uniform deformation and a hybrid intelligent algorithm

结构在地震作用下发生均匀损伤或均匀变形可以避免出现薄弱层和集中损伤,且结构整体抗震性能可得到全面提升。以此为目标对结构进行抗震优化设计具有重要的理论和工程意义,但相关的理论和规律性研究尚不够系统深入。针对目前结构优化算法计算效率较低的不足,提出基于差分进化算法和粒子群算法的混合智能算法,并给出结构优化计算流程。针对弯剪型框架结构,以弹性动力时程分析中结构各层的层间相对位移相等为优化目标,基于混合智能算法对结构楼层刚度进行优化设计,获得了楼层刚度的最优分布规律。在此基础上研究了地震动随机性和梁柱刚度比等因素对优化结果和刚度分布规律的影响。基于曲线拟合和等效理论分析,建立楼层最优刚度比和等效最优截面尺寸比的多项式函数。通过与现有研究成果的对比,验证所建立的经验函数具有较高的精度,可为面向均匀变形的结构优化设计提供准确有效的参考依据并可提高计算效率。

If the structural story drifts are equal when a frame is subjected to earthquake,the occurrence of weak story and concentrated damage can be avoided,and the overall seismic performance can be improved in an all-round way.It has great theoretical and engineering significance to optimize the seismic design according to above object,but the relevant theoretical and regular research is not intensive.In view of the shortcomings of the low efficiency of present structural optimization algorithms,a hybrid intelligent algorithm based on differential evolution algorithm and particle swarm optimization algorithm was proposed,and the structural optimization calculation process was given.For a bend-shear frame structure,the optimization objective is to ensure equal interlayer displacement,the story stiffness was optimal designed based on the hybrid intelligent algorithm,and the optimal distribution of floor stiffness was obtained.On this basis,the influence of the randomness of ground motion and the stiffness ratio of beam and column to the optimization results and the distribution of stiffness was studied.The polynomial function of the optimal stiffness ratio of the floor and the ratio of the equivalent optimum cross section size was established.By comparing with the existing research results,it is verified that the established empirical function has high accuracy,and an accurate and effective reference basis for the optimization design for uniform deformation is provided and the calculation efficiency is improved.