基于人工蜂群算法和有限元强度计算的集装箱船剖面结构优化

Optimization of the section structure of container ship based on artificial bee colony algorithm and finite element strength calculation

  • 摘要:
      目的  现有基于有限元强度计算的结构优化研究大多采用改写单元节点信息文件来实现参数化建模的方法,为解决在船体剖面结构优化过程中难以考虑型材数量变化的问题,提出一种基于参数化几何建模分析和人工蜂群(ABC)算法的船舯剖面结构优化方法。
      方法  首先,在Matlab平台编写蜂群算法,并基于ABAQUS内核语言Python建立能够在其CAE模块中生成几何模型的脚本文件;其次,建立能够提交有限元计算和读取结果的Python脚本文件,通过将算法每次生成的解改写到脚本对应位置完成几何模型的更新,后台调用ABAQUS并依次运行脚本文件;最后,将计算结果返回到Matlab平台中进行校核,完成参数化几何建模与有限元分析。
      结果  以4600 TEU集装箱船在总纵弯矩作用下的舱段剖面结构优化为例验证了该方法的可行性,得到集装箱船舱段结构减重达18.7%。
      结论  经对比分析,在设定条件下基于有限元的优化方法比基于规范的优化方法更加充分。

     

    Abstract:
      Objectives  The method of rewriting the element node information file to realize parametric modeling is commonly employed in existing structural optimization based on finite element (FE) strength calculation, but it remains difficult to consider variations of the profile number in hull section structure optimization. To this end, a container ship structure optimization method based on FE strength calculation is proposed using an artificial bee colony (ABC) algorithm and parametric FE modeling method.
      Methods  First, the bee colony algorithm is written on the Matlab platform, and a script file which can generate the geometric model in its CAE module is established based on the ABAQUS kernel in the Python language. A Python script file which can submit the FE calculation and read the results is established. The geometric model is updated by rewriting the solution generated by the algorithm to the corresponding position in the script, then ABAQUS is called in the background and the script files are run in turn. Finally, the calculation results are returned to Matlab for verification, and the parametric geometric modeling and FE analysis are completed.
      Results  The feasibility of this method is verified by taking the section structure optimization of a 4600 TEU container ship as an example. It is found that the weight reduction of the container ship cabin structure reaches 18.7%.
      Conclusions  When the results of the FE strength optimization and code optimization are compared and analyzed, under the set conditions, the FE optimization method is more sufficient than that based on code.

     

/

返回文章
返回