摘要
主要针对柔性作业车间调度问题进行求解,利用改进粒子群算法作为求解方法,以最小化最大完工时间(C_(max))作为该问题的求解目标.在算例的选取上,选用作业车间调度问题的8*8经典算例和柔性作业车间调度问题的Brandimarte算例对提出的算法进行验证.改进粒子群算法由遗传算法和粒子群算法构成,遗传算法具有较好的全局搜索能力,但搜索过程中收敛的精度不高,粒子群算法由于其寻优特性,在搜索过程中速度较快,但容易陷入局部最优,综合考虑两者的优缺点,将遗传算子引入粒子群算法中,采用交叉搜索的方式,调整惯性权重以及变异的方式使粒子进化,当粒子群进化到一定程度后,对部分粒子进行变异处理从而避免算法陷入局部最优解,同时可以提高粒子群算法的收敛精度.依据柔性作业车间调度问题的特点,在经过多次变换种群规模以及迭代次数后,求解出最适合柔性作业车间调度问题的最优解.
We mainly solved the flexible job shop scheduling problem, using the improved particle swarm optimization as the solution method, and minimizing the maximum completion time(C_(max)) as the solution goal of the problem. In the selection of the calculation example, the job shop scheduling problem 8*8 Classic example and Brandimarte example of flexible job shop scheduling problem were selected to verify the proposed algorithm. Improved particle swarm algorithm was composed of genetic algorithm and particle swarm algorithm. Genetic algorithm had better global search ability, but it converged in the search process. The accuracy of the particle swarm algorithm was not high. Due to its optimization characteristics, the particle swarm algorithm was faster in the search process, but it was easy to fall into the local optimum. Considering the advantages and disadvantages of the two, the genetic operator was introduced into the particle swarm algorithm, and cross search was adopted. The method of adjusting the inertia weight and the method of mutation made the particles evolve. When the particle swarm evolved to a certain level, some particles were mutated to avoid the algorithm falling into the local optimal solution, and at the same time, the convergence accuracy of the particle swarm algorithm could be improved. According to flexibility, the characteristics of the job shop scheduling problem, after many changes in the population size and the number of iterations, the optimal solution that was the most suitable for the flexible job shop scheduling problem was solved.
作者
吴晓雯
郑巧仙
WU Xiaowen;ZHENG Qiaoxian(College of Computer and Information Engineering Hubei University,Wuhan 430062,China)
出处
《湖北大学学报（自然科学版）》
CAS
2022年第5期501-507,共7页
Journal of Hubei University：Natural Science
基金
国家自然科学基金(61803149)资助。
关键词
遗传算法
粒子群算法
柔性作业车间调度
genetic algorithm
particle swarm optimization
flexible job shop scheduling problem
