基于分布估计算法求解混合流水车间调度问题

发布时间：2018-04-04 04:07

本文选题：混合流水车间调度　切入点：分布估计算法　出处：《大连交通大学》2014年硕士论文

【摘要】：近些年来,随着社会的逐步发展,科学技术的重要性在社会的生产发展中日益凸显。通过利用科学技术的进步与发展来提高生产效率、降低生产成本、提高企业竞争力越来越受到各领域的重视。所谓车间调度,就是要分配现有资源来满足企业的正常、有序、快速生产,也就是说要分配好工件、设备的加工顺序及与时间的关系,以达到设备的最大化利用和提高生产效率的目的。因此,深入地研究车间调度问题有着重要的科研及生产价值。目前社会处在高速发展阶段,越来越多的车间调度出现在社会生产中,也随之出现了很多智能优化算法。分布估计算法(Estimation of Distribution Algorithm, EDA)是近年来新兴的一种进化算法,它将建立概率模型和采样引入到进化的过程中,而不是遗传算法里的变异和交叉操作。分布估计算法是通过概率模型进行的全局搜索优化。这样能够避免一些传统优化算法的缺陷出现。分布估计算法的优化效率,优化性能,优化效果等方面表现的比其他很多算法更优秀,更符合实际需求。本文分析了置换的流水车间调度问题和更为复杂的混合流水车间调度问题特点、分类等情况,研究了目前出现的一些主要其他算法对于这类问题的求解并分析这些算法求解这类问题的优点和不足之处。本文将基于分布估计算法这一基础算法求解流水车间调度问题。根据分析出的其他算法对于求解这类问题的优点和不足之处,为了到达发扬这些优势并避免那些不足处的母体,本文对分布估计算法进行了适当的改进,改进的分布估计算法有着更高的优化效率,更强的优化性能,更好更优秀的优化效果。最后本文通过车间调度中经典的Rec类问题数据和具体车间调度实例对改进的分布估计算法进行了测试和性能的验证试验,并将测试结果很好很直观的与其他算法进行了比较,结果数据都显示出了改进的分布估计算法的优良的性能。然后通过模拟调度系统的实现,更进一步验证了改进的分布估计算法的有效性和优越性。
[Abstract]:In recent years, with the gradual development of society, the importance of science and technology has become increasingly prominent in the development of social production.By using the progress and development of science and technology to improve production efficiency, reduce production costs and improve the competitiveness of enterprises, more and more attention has been paid to various fields.The so-called workshop scheduling is to allocate existing resources to meet the normal, orderly and rapid production of enterprises, that is to say, to allocate the processing order of jobs and equipment and their relationship with time.In order to maximize the use of equipment and improve production efficiency.Therefore, the in-depth study of job shop scheduling has important scientific research and production value.At present, the society is in the high speed development stage, more and more job shop scheduling appears in the social production, also appeared many intelligent optimization algorithms.Estimation of Distribution algorithm (EDAA) is a new evolutionary algorithm in recent years. It introduces probabilistic model and sampling into evolutionary process, rather than mutation and crossover operation in genetic algorithm.Distribution estimation algorithm is a global search optimization based on probabilistic model.In this way, the defects of some traditional optimization algorithms can be avoided.The optimization efficiency, optimization performance and optimization effect of the distributed estimation algorithm are better than many other algorithms and meet the actual needs.In this paper, we analyze the characteristics and classification of income job shop scheduling problem and the more complex hybrid income job shop scheduling problem.The advantages and disadvantages of some other algorithms for solving this kind of problems are studied and the advantages and disadvantages of these algorithms are analyzed.In this paper, the basic algorithm based on the distribution estimation algorithm is used to solve the income job shop scheduling problem.According to the advantages and disadvantages of other algorithms for solving this kind of problems, in order to develop these advantages and avoid the disadvantages of the matrix, this paper makes a proper improvement on the distribution estimation algorithm.The improved distribution estimation algorithm has higher optimization efficiency, better optimization performance and better optimization effect.Finally, this paper tests and verifies the performance of the improved distributed estimation algorithm by using the classic Rec class problem data and specific job shop scheduling examples, and compares the test results with other algorithms directly.The results show that the improved distribution estimation algorithm has good performance.Then, the effectiveness and superiority of the improved distribution estimation algorithm are further verified by the implementation of the simulation scheduling system.
【学位授予单位】：大连交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TB497

【参考文献】