基于改进遗传算法的钣金折弯自动工序规划

Automatic procedure planning based on an improved genetic algorithm for sheet metal bending

为了提高钣金折弯工序规划的效率,需要保证在不干涉的前提下,为每个工步选择合适模具,使得模具更换、板料的掉头和翻转次数尽可能少。为此,采用一种改进的高效遗传算法对折弯工序进行优化设计。该方法将发生干涉的工序位置引入适值的计算,在初始种群中快速进化出更优个体,同时为了提高遗传算法的收敛速度,将每一代中的最优个体直接保存到下一代中。实验结果表明该算法相对传统方法更具有高效性,能够快速得到近似最优解的工序方案。

In order to ensure the efficiency of process planning, it needs to choose the appropriate tools for each step to make the number of tool, tool changes, part turn around and flip - overs minimal on the premise of non -interference. In this paper, an improved efficient genetic algorithm is developed to optimize the design of the bending procedure planning, which introduces the step position of interference to calculate the fit- ness and evolves the better individual in the initial population. Meanwhile, for the sake of improving the convergence rate, this algorithm saves the best solution of current generation to next generation. The re- sults show that the algorithm is highly efficient and can quickly obtain the approximate optimal sequences.