基于多目标遗传算法的板料拉深成形工艺参数优化设计

Process Parameters Optimization of Sheet Metal Forming in Drawing Process Based on Multi-objective Genetic Algorithm

以多种工艺参数(压边力、摩擦因数等)作为优化变量,多种成形缺陷(起皱、破裂等)作为优化目标,结合多目标遗传算法和数值模拟,建立了板料拉深成形工艺参数的优化设计模型。为了减少数值模拟的次数,利用人工神经网络建立了各种工艺参数和模拟结果之间的映射关系,大大提高了优化的效率。以汽车消声器为例,对其拉深成形工艺参数进行了优化,通过对优化结果进行数值模拟可以看出,该优化参数完全避免了各种缺陷的产生,这说明该优化算法具有较好的优化结果。

An optimization model was established with MOGA and numerical simulation, in which several process parameters (e.g. blank-holding force, friction coefficient) were optimization variables, and several forming problems(e.g. crinkling, cracking) were optimization objections. An ANN was built to connect process parameters and simulation results which improved the optimization efficiency. At last a model of the exhaust muffler was provided. The result shows that the forming problems can be avoided via numerical ...