橡胶悬置元件结构参数优化设计方法

由于橡胶悬置元件的结构比较复杂,截面不规则,无法用传统的优化方法对其结构优化.利用遗传算法和神经网络相结合的策略对橡胶悬置元件的几何结构参数进行优化,即用神经网络学习算法建立橡胶悬置元件几何结构参数与其三个方向刚度的非线性全局映射关系,获得遗传算法求解结构优化问题所需的目标函数,用遗传算法进行优胜劣汰的寻优搜索运算,求出最优解.优化结果表明,橡胶悬置元件结构参数优化设计方法是可行的.

OPTIMAL DESIGN METHOD FOR STRUCTURAL PARAMETERS OF A RUBBER MOUNT

With irregular sections and complex shape,the structure of a rubber mount cannot be optimized using conventional optimal methods.A new method is used to optimize structural parameters of the rubber mount,it adopts the strategy of combining Genetic Algorithms(GA) and Neural Network(NN).Using the learning algorithm of NN,a nonlinear global mapping function can be established for the relation between its geometric parameters and its stiffnesses in x,y and z directions.The objective function of GA can be obtained synchronously.Optimal solution can be found using search operation of GA.The results show that the optimal design method proposed for structural parameters of the rubber mount is feasible.