Computer Engineering and Applications ›› 2021, Vol. 57 ›› Issue (22): 182-189.DOI: 10.3778/j.issn.1002-8331.2007-0013

• Pattern Recognition and Artificial Intelligence • Previous Articles     Next Articles

Study on Application of Double Population Hybrid Genetic Algorithm in Cut Order Planning

DU Shouxin, WU Tao   

  1. School of Computer Science, Xi’an Polytechnic University, Xi’an 710048, China
  • Online:2021-11-15 Published:2021-11-16

双种群混合遗传算法的裁剪分床应用研究

杜守信,毋涛   

  1. 西安工程大学 计算机科学学院,西安 710048

Abstract:

In order to solve the problem of cut order planning in apparel production, a multi-objective mathematical model of cut order planning is established for optimization based on the influencing factors of production process and order requirements, and an improved double population particle swarm-genetic hybrid algorithm is used to solve the model. The hybrid algorithm divides the evolutionary population into a general population and an elite population, uses an improved genetic algorithm to search the evolutionary general population globally and screen the elite individuals, and combines the particle swarm optimization algorithm to evolve the elite population. Crossover and mutation ensure the diversity of the population, the optimization mechanism of particle swarm improves the speed of evolution, and the two groups constantly search for the optimal solution due to their cross influence during evolution. The experimental results show that the hybrid algorithm is stable in solving the multi-objective cut order planning problem. It has a faster optimization speed than the improved genetic algorithm. It reduces one cutting bed than the manual calculation method, and the cutting time is shortened by 5 min. The number of overcuts is reduced by 60%, which can adapt to different neets, and has certain application value in the actual production of cut order panning.

Key words: cut order planning, multi-objective, double population, genetic algorithm, particle swarm optimization, cross influence

摘要:

为解决服装生产中的裁剪分床计划问题,结合生产过程的影响因素和订单需求,建立了裁剪分床的多目标数学模型进行优化,使用一种改进的双种群粒子群-遗传混合算法对模型进行求解。混合算法将进化种群划分为普通种群和精英种群,利用改进的遗传算法来全局搜索进化普通群体并筛选精英个体,同时结合粒子群优化算法进化精英群体。交叉和变异保证种群的多样性,粒子群寻优机制提升进化速度,两种群在进化时交叉影响不断寻找最优方案。实验结果表明:混合算法在解决多目标的生产订单裁剪分床问题上表现稳定,相比改进的遗传算法有更快的寻优速度,比手工计算方法减少1个裁床,裁剪时间缩短5?min且超裁数量降低60%,可以适应不同目标需求,针对实际生产中的裁剪分床有一定的应用价值。

关键词: 裁剪分床计划, 多目标, 双种群, 遗传算法, 粒子群优化, 交叉影响