具有混合变量的桁架结构多目标形状模糊优化设计

分别以桁架杆件截面面积和可动结点坐标为离散和连续设计变量,以结构总重量最小、结点位移最小和结构固有频率最大为目标函数,进行桁架结构形状多目标优化设计。采用的改进遗传算法用二进制编码方式对两类不同变量进行编码,用凝聚函数将问题的多项约束转化成单个约束,用凝聚选择策略和竞争保留机制保证有潜力的解被优先选择,最终得到问题的理想解和有效解。最后从工程实际出发,根据模糊相似优先比方法,从多组有效解中寻找出最优解。

Multi-Objective Fuzzy Shape Optimization Design with Mixed Variables of the Truss

With the section area of truss rod component and the movable node coordinates as the discrete and continuous design variables,an optimal design is made of the multi-objective shape truss structure with minimum total weight,minimum node displacement and maximum fixed frequency as the objective.The improved genetic algorithm is adopted to code two groups of different variables with the binary coding means.And then,the aggregating function is used to transform the multi-constrains into the single constraint,and the aggregating selection strategy and competitive mechanism are used to assure the potential solutions selected preferentially.Finally,as of engineering practical conditions,the optimal solution is determined from the effective solutions with the theory of fuzzy similarity priority ratio method.