This paper proposes a memetic algorithm (MA) with novel semi-constructive crossover and mutation operators (MASC) to minimize makespan in permutation flowshop scheduling problem (PFSP). MASC combines the strengths of genetic algorithm (GA), simulated annealing (SA), and Nawaz–Enscore–Ham (NEH) algorithm. The aim is to enhance GA in identifying promising areas in the search space, whose local optima will be subsequently located by SA. This is achieved by means of novel crossover and mutation operators that construct chromosomes by using two different types of genes: static and dynamic genes. MASC is tested on the well-known Taillard’s benchmark instances. The proposed operators are compared with traditional operators. The results show that the proposed operators produce considerable improvements. These improvements reach up to 20.79% in the average relative error of best solution and 11.86% in the average relative error of average solution. MASC is compared with fourteen well-known and state-of-the-art algorithms. These algorithms include MA, whale optimization, ant colony optimization, particle swarm optimization, artificial bee colony, monkey search, and iterated greedy. The results show that MASC outperforms all the compared algorithms except three iterated greedy algorithms. Moreover, the improvement in the average relative error of best solution achieved on the best-so-far MA is 37.92%. Therefore, MASC can be considered as one of the best-so-far methods for PFSP.

本文提出了一种具有新颖的半结构化交叉和变异算子（MASC）的模因算法（MAMET），以最大程度地减少置换流水车间调度问题（PFSP）的生成时间。MASC结合了遗传算法（GA），模拟退火（SA）和Nawaz–Enscore–Ham（NEH）算法的优势。目的是增强GA在搜索空间中确定有前途的区域的能力，SA会随后确定其局部最优值。这是通过使用两种不同类型的基因（静态和动态基因）构建染色体的新型交叉和变异算子来实现的。MASC已在著名的Taillard基准实例上进行了测试。将拟议的运营商与传统运营商进行比较。结果表明，提出的算子产生了很大的改进。这些改进最多可达到20个。最佳解决方案的平均相对误差为79％，平均解决方案的平均相对误差为11.86％。将MASC与14种众所周知的最新算法进行了比较。这些算法包括MA，鲸鱼优化，蚁群优化，粒子群优化，人工蜂群，猴子搜索和迭代贪婪。结果表明，除了三种迭代贪婪算法外，MASC的性能均优于所有比较算法。此外，在迄今为止最好的MA上，最佳解决方案的平均相对误差提高了37.92％。因此，MASC被认为是迄今为止最先进的PFSP方法之一。鲸鱼优化，蚁群优化，粒子群优化，人工蜂群，猴子搜索和迭代贪婪。结果表明，除了三种迭代贪婪算法外，MASC的性能均优于所有比较算法。此外，在迄今为止最好的MA上，最佳解决方案的平均相对误差提高了37.92％。因此，MASC被认为是迄今为止最先进的PFSP方法之一。鲸鱼优化，蚁群优化，粒子群优化，人工蜂群，猴子搜索和迭代贪婪。结果表明，除了三种迭代贪婪算法外，MASC的性能均优于所有比较算法。此外，在迄今为止最好的MA上，最佳解决方案的平均相对误差提高了37.92％。因此，MASC被认为是迄今为止最先进的PFSP方法之一。