摘要
以变角度纤维的起始角、终止角和层合板的层铺顺序为设计变量,弯曲刚度为设计目标,优化变角度纤维层合板。在优化过程中,对应层铺顺序的设计变量通过实数编码和置换策略转换成整数变量,实现离散空间和连续空间之间的转换。由于离散变量和连续变量的优化进程差异,本文提出一种改进的自适应差分进化算法,对不同类型的优化变量采取不同的自适应变异算子。相关算例计算结果表明:无论是对层合板的局部优化还是全局优化,改进的自适应差分进化算法都比差分进化算法和自适应差分进化算法计算精度高、收敛速率快,同时可以有效避免“早熟”现象。对于变角度纤维层合板弯曲刚度优化问题,本文提出的改进的自适应差分进化算法是一种相对高效的可行方法。
The starting angle of the variable angle fiber,the ending angle and the lamination order of the laminate are the design variables,and the bending stiffness is the design goal to optimize the variable angle fiber laminate.In the optimization process,the design variables corresponding to the layering order are converted into integer variables by real coding and replacement strategies to realize the conversion between discrete space and continuous space.Due to the difference of optimization process between discrete variables and continuous variables,Improved Adaptive Differential Evolution Algorithm is proposed in this paper,and different adaptive mutation operators are adopted for different types of optimization variables.The calculation results of related examples show that,regardless of the local optimization and global optimization of the laminate,Improved Adaptive Differential Evolution Algorithm is more accurate than Differential Evolution Algorithm and Adaptive Differential Evolution Algorithm,and can effectively avoid"Premature"phenomenon.For the optimization of the bending stiffness of variable angle fiber laminates,Improved Adaptive Differential Evolution Algorithm proposed in this paper is a relatively efficient and feasible method.
作者
吴双华
陈童
尹冠生
WU Shuang-hua;CHEN Tong;YIN Guan-sheng(School of Science,Chang′an University,Xi′an 710064,China)
出处
《玻璃钢／复合材料》
CAS
北大核心
2019年第11期12-17,共6页
Fiber Reinforced Plastics/Composites
基金
国家自然科学基金(11402035)
长安大学学生创新开放实验室项目(2018CXSY06)
关键词
弯曲刚度
变角度纤维层合板
编码
变异算子
自适应差分进化算法
bending stiffness
variable angle fiber laminate
coding
mutation operator
adaptive differential evolution algorithm
