基于遗传算法的多目标柔性资源调度研究

发布时间：2018-01-30 00:55

本文关键词： 柔性车间调度问题多目标优化改进遗传算法指数分布　出处：《天津大学》2014年硕士论文　论文类型：学位论文

【摘要】：在全球发展深化的趋势下,我国制造业正在面临着巨大的冲击,所以提高生产效率以及资源利用率,促进转型升级已经成为制造业发展的必然趋势。其中柔性车间调度问题(FJSP)是该领域的经典问题,而多目标柔性车间调度问题在调度时要衡量各个目标,增加了调度的难度,是原有FJSP的深化。此类研究已经成为了制造业的重点问题。在现有传统调度问题的研究中,一般认为工件加工时间是定值,但是结合生产实际情况来看,加工时间往往是波动的,而不可能是单一的定值。所以本文在详细文献综述的基础之上,拟提出一种改进的遗传算法来解决此问题,以优化现有算法。本文的研究将基于加工时间服从指数分布假设的多目标柔性车间调度问题为对象,旨在平衡机台负荷和减小订单延误时间。在研究过程中将OOC编码与RC编码技术结合用于产生初始化群体,在保证种群多样性的同时提高算法的效率。同时考虑到遗传算法本身的局限性,采用线性变换,适应度变换,自适应交叉、变异,使遗传算法前期保持种群的多样性,避免早熟,后期提高搜索的效率,更快找到最优解。经过算法验证,本文提出的改进遗传算法能很好地实现柔性车间调度问题的多目标优化,并在算法性能以及算法收敛速度方面均优于自适应遗传算法与标准遗传算法,有效的优化了遗传算法,对制造业实践的生产活动具有一定的指导意义。
[Abstract]:Under the trend of global development, China's manufacturing industry is facing a huge impact, so improve production efficiency and resource utilization. Promotion of transformation and upgrading has become an inevitable trend in the development of manufacturing industry, in which flexible shop scheduling problem (FJSP) is a classic problem in this field. The multi-objective flexible job shop scheduling problem has to measure each target in scheduling, which increases the difficulty of scheduling. It is the deepening of the original FJSP. This kind of research has become the key problem in the manufacturing industry. In the existing research of traditional scheduling problem, it is generally considered that the processing time of the workpiece is a fixed value, but according to the actual situation of production. Processing time is often fluctuating, but can not be a single fixed value. Therefore, based on the detailed literature review, this paper proposes an improved genetic algorithm to solve this problem. In order to optimize the existing algorithms, the research of this paper will be based on the hypothesis of processing time from the exponential distribution of multi-objective flexible job shop scheduling problem as an object. In order to balance the load of the machine and reduce the delay time of the order, the OOC coding and RC coding technology are combined to generate initialization group in the research process. Considering the limitations of genetic algorithm, linear transformation, fitness transformation, adaptive crossover and mutation are adopted. The genetic algorithm can keep the diversity of population, avoid precocity, improve the efficiency of search and find the optimal solution more quickly. The improved genetic algorithm proposed in this paper can achieve multi-objective optimization of flexible job shop scheduling problem, and is superior to adaptive genetic algorithm and standard genetic algorithm in the performance and convergence speed of the algorithm. The genetic algorithm is optimized effectively, which has a certain guiding significance to the production activities of manufacturing practice.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TB497;TP18

【参考文献】