基于遗传模拟退火算法的布局优化研究

为提高矩形件排样算法的利用率与时间效率，提出将遗传算法和模拟退火算法融合优化的矩形排样算法。采用带符号的十进制编码,依据矩形件长宽比和面积而生成基因序列用于建立初始种群，以随机产生若干排样顺序与排样尺寸不一的个体，并以利用率为适应度函数，修改后的最低水平线搜索算法作为排样策略，保证较优个体得以保留，减少闲置区域的产生。采用10 组随机产生的矩形数据将本算法与现有文献提出的GA 算法进行对比实验，实验结果显示：该算法有效地提升了排样结果的利用率与时间效率。

On Layout Optimization Based on Genetic Simulated Annealing Algorithm

Based on the integration of the genetic algorithm (GA) and the simulated annealingalgorithm, an improved lowest horizontal line (ILHL) algorithm is presented in order to improveutilization and stability of the rectangular packing algorithm. In this algorithm, a signed decimalencoding is utilized to generate the gene sequence in accordance with the length-width ratio and thearea of the rectangle, which is employed to establish the initial population. The improved lowesthorizontal line algorithm adopts the best individuals from a number of random sequences withdifferent nesting orders and layout sizes, uses utilization rate as the fitness function and reduces theidle area. In this paper, a contrast experiment is operated to compare ten groups of rectangular datarandomly generated by ILHL with those generated by GA proposed in the current literature. Theexperiment results show that our algorithm (ILHL) can effectively improve the utilization rate andtime efficiency of the packing results.