考虑传送带中断时间的混流装配线排程优化研究

发布时间：2018-05-06 02:17

本文选题：多目标优化 + 差分进化算法　；参考：《苏州大学》2016年硕士论文

【摘要】：作为车间调度问题的一个分支,混流装配线的排序问题对混流装配线能否合理高效地发挥作用有着至关重要的影响。合理的排序方案可降低生产成本,降低在制品及产成品库存甚至达到零库存,实现准时化生产,可以有效提高企业的市场竞争力。混流装配线的排序问题已被公认是NP难问题,几十年来一直受到研究者们的广泛关注。本文研究了混流装配线的单目标和多目标优化问题。针对以往研究只考虑了传送带中断时间对混流装配线的影响而忽视了工人空闲时间造成的成本浪费,提出给传送带中断时间和工人的空闲时间分配不同的权重系数,将其作为整体目标进行优化。同时考虑了产品切换时间对传送带中断时间的影响,据此建立了相应的混流装配线单目标优化模型,分别采用差分进化算法和遗传算法进行求解,并基于仿真实验对两种算法的优劣进行了分析和比较。基于上述单目标优化模型,分别以传送带中断时间和工人空闲时间最小化、产品切换时间最小化和提前/延后完工时间最小化为目标建立了混流装配线多目标优化模型。针对差分进化算法和双种群差分进化算法在求解多目标问题时容易丢失部分非支配解的缺陷,提出了一种改进双种群差分进化算法。利用一个辅助种群保存选择过程中被舍弃的非支配解,并将辅助种群中的个体与父代种群和主种群的个体随机比较,选择较优的个体进入下一代,通过这种方式降低非支配解的舍弃率。为了尽可能保存迭代过程中产生的非支配解,提出在算法迭代一定的次数时,选择父代种群、主种群和辅助种群中的非支配个体直接进入下一代遗传。最后通过与NSGAII、差分进化算法以及双种群差分进化算法进行比较,验证了所提出的改进双种群差分进化算法在求解多目标问题时具有的良好性能。
[Abstract]:As a branch of job shop scheduling problem, the scheduling problem of mixed flow assembly line plays an important role in whether the mixed flow assembly line can play a role reasonably and efficiently. Reasonable sequencing scheme can reduce the production cost, reduce the inventory of WIP and finished products to zero inventory, and realize just-in-time production, which can effectively improve the market competitiveness of enterprises. Mixed-flow assembly line scheduling problem has been recognized as NP-hard problem, has been widely concerned by researchers in recent decades. In this paper, the problem of single objective and multi-objective optimization for mixed flow assembly line is studied. In view of the fact that the previous studies only considered the influence of the belt interruption time on the mixed flow assembly line and neglected the cost waste caused by the workers' idle time, the paper proposed that different weight coefficients should be assigned to the conveyor belt interruption time and the workers' idle time. Optimize it as an overall goal. At the same time, considering the influence of the product switching time on the belt interruption time, the corresponding mixed flow assembly line single objective optimization model is established, and the differential evolution algorithm and genetic algorithm are used to solve the problem, respectively. The advantages and disadvantages of the two algorithms are analyzed and compared based on the simulation experiments. Based on the above single-objective optimization model, a multi-objective optimization model for mixed-flow assembly line is established with the objectives of minimization of belt interruption time and workers' idle time, minimization of product switching time and minimization of advance / delay completion time. In order to solve the problem of multi-objective problems, a modified two-species differential evolutionary algorithm is proposed to solve multi-objective problems, which is easy to lose part of the non-dominated solution of the differential evolution algorithm and the two-species differential evolution algorithm. Using an auxiliary population to preserve the abandoned non-dominant solution in the selection process, and comparing the individual in the auxiliary population with the parent population and the individual of the main population at random, the better individuals are selected to enter the next generation. In this way, the abandonment rate of the nondominant solution is reduced. In order to preserve the non-dominated solution generated in the iterative process as much as possible, it is proposed that the parent population, the non-dominant individuals of the main population and the auxiliary population be selected to enter the next generation of heredity directly when the algorithm iterates a certain number of times. Finally, compared with NSGA II, differential evolution algorithm and double population differential evolution algorithm, the proposed improved double population differential evolution algorithm has good performance in solving multi-objective problems.
【学位授予单位】：苏州大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TB497

【相似文献】