Abstract:

The artificial bee colony(ABC) algorithm,a relatively recent bio-inspired approach mimicking the behavior of real bee colony,was applied to deal with the optimization problems of automobile structure.The metamodel of objective and constrains were gotten through combination of design of experiment and sequential response surface method.Then,the optimum design was obtained by the modified artificial bee colony algorithm.It can reduce the computing cost by the metamodeling techniques.Finally,a typical example was selected to proof this method.The comparison results between the simulated and experimental values show that this method has enough precision and satisfies the engineering practical demands.

Key words: modified artificial bee colony algorithm, structural optimization, design of experiment, sequential response surface method

摘要：

将蜂群算法应用于汽车结构件的优化问题。先由试验设计和序列响应面法构建目标函数及约束条件的代理模型,再应用改进的蜂群算法求解最优设计。在优化过程中调用的是代理模型,显著减少了有限元计算次数,提高了优化效率。最后,选取典型实例对该算法进行验证,比较预期值与实际值的结果表明,该算法具备了足够的求解精度,能够满足工程实际要求。

关键词: 改进蜂群算法, 结构优化, 试验设计, 序列响应面法