双种群混合遗传算法求解航空复合材料柔性调度问题

Two-population hybrid genetic algorithm for flexible job-shop scheduling problem in aerospace composites

考虑航空复合材料柔性车间调度中的运输约束,以最小化完工时间为目标,建立调度模型,提出一种改进的双种群混合遗传算法进行求解。根据问题特点,基于工序排序、机器选择和运输约束3个子问题,设计三层实数编码以及对应解码方案。采用混合初始化提高种群质量,进化过程中采用交叉算子执行全局搜索,为双种群设计基于机器负载平衡和变邻域的局部搜索,提高全局和局部搜索能力。与对比算法相比10个测试算例中BPRD指标取得9个最优,APRD指标全部取得最优,t检验显著性有明显差异,验证算法的优越性。将算法应用于航空复合材料车间中,实现实际生产的调度,验证算法的可行性。

Considering the transportation constraints in the scheduling of flexible job-shop for aerospace composite materials,a scheduling model was established with the goal of minimizing the completion time to solve this model,an improved two-population hybrid genetic algorithm was proposed.Based on the characteristics of the problem,a three-layer real number encoding and the corresponding decoding scheme were designed for the three sub-problems of operation sequencing,machine selection,and transportation constraints.A mixed initialization was adopted to improve the population quality,and a crossover operator was employed for global search during the evolutionary process.A local search strategy based on machine load balancing and variable neighborhood was designed for the two-population,enhancing both global and local search capabilities.Compared with the contrast algorithm,the BPRD index for 9 out of the 10 test examples is obtained optimally,the APRD index is obtained optimally for all examples,and the t-test shows a significant difference,verifying the superiority of the proposed algorithm.The algorithm was applied to an actual aerospace composite materials manufacturing system,realizing the scheduling of practical production activities and verifying its feasibility.