基于遗传算法的复合材料层合壳荷载识别

LOAD IDENTIFICATION FOR A COMPOSITE LAMINATED SHELL USING GENETIC ALGORITHM

  • 摘要: 该文提出了将遗传算法(Genetic Algorithm)与有限元分析相结合的方法,用来解决反问题中壳体结构静荷载的作用位置及大小的识别。首先,利用有限元法计算出结构若干观测点的电荷,建立以该电荷为变量的目标函数,然后以荷载位置和大小作为参数进行遗传算法的二进制串编码,利用该优化方法迭代反算出施加在结构上的荷载位置和大小。另外,因计算电荷值和实验值之间存在一定误差,以一节点为研究对象,往计算电荷值中加入10%的白噪声干扰,计算结果表明有干扰和无干扰情况下的荷载的识别结果相对误差很小。

     

    Abstract: This article proposes a method which combines the Genetic Algorithm with FEM, using to identify concentrated load in the shell structure. First, the electric charges of certain observation points are calculated by FEM and the objective function is established by taking the electric charge as variable. Then,the binary strings are established by taking the load position and force as parameters. Finally, the load are calculated by using this optimized method. Moreover, because there is certain error between the analog computation electric charges and the actual value, a node is studied by putting 10% noise toward the analog computation electric charge, and the result shows that the relative error is very small under situations with or without disturbance.

     

/

返回文章
返回