《智能制造系统车间生产计划与调度》由会员分享，可在线阅读，更多相关《智能制造系统车间生产计划与调度（131页珍藏版）》请在金锄头文库上搜索。
1、南京航空航天大学博士学位论文智能制造系统车间生产计划与调度姓名：袁坤申请学位级别：博士专业：机械电子工程指导教师：朱剑英20070501南京航空航天大学博士学位论文 i摘 要 车间生产计划与调度，是体现企业竞争力的一个重要方面；其研究具有重要的现实意义和一定的理论价值。研究工作主要内容及创新点如下： 为满足实际生产计划与调度优化目标的多样性需求，提出了一种综合优化目标模型，实现了多目标、模糊目标、特殊性要求子目标等综合目标的评价；试验结果表明该模型合理、可行，并在后续的研究内容中得到成功应用。 为了加快迭代速度、避免早熟，提出了个体优化群体多样化的遗传算法；在该算法中，提出了一种新的矩阵编码及
2、其对应的遗传算子操作方法，并给出了一种可保证群体多样性的 GA 参数控制方法；试算结果显示，该算法性能优良。 为了提高全局寻优的能力及效率，整合了遗传算法的选择、交叉和变异算子，提出了解的覆盖面较大的集成算子遗传算法。 该算法采用一种新的矢量编码及对应的算子操作方法；三个多目标 FJSP 试算结果，均优于国际最新文献和近期 IEEE 成员文献的结果。 为了获得求解计划与调度集成优化所需的快速算法，提出了一种个体粒子与群体粒子搜索经验的表示、学习和整合方法，并给出了算法的具体实现；多个标准算例试验结果及其分析显示，该算法计算速度较快。 在上述理论的基础上，为某汽车变速箱企业装配车间，建立了车间生
3、产计划与调度集成优化模型； 提出了一种分层搜索优化策略和两层内嵌结构的集成算子遗传算法；给出了该车间生产计划与调度集成优化的具体结果；结果分析表明，所给模型及其求解方法正确可行，其优化结果远好于生产实际中的结果。 建立了该企业轴类热后加工工段生产计划与调度集成优化模型；提出了一种直接搜索的优化策略，并给出了求解方法；计算结果与实际情况的比较表明，所给模型及其求解方法是合理有效的。 本课题的研究，受到国家自然科学基金重大项目“支持产品创新的先进制造技术中若干基础性研究” （项目编号：59990470）资助。 关键词关键词：作业车间调度, 计划与调度, 多目标, 集成优化, 遗传算法, 粒子群优化
4、 智能制造系统车间生产计划与调度 iiABSTRACT The shop production planning and scheduling optimization is very important because of its theoretical and practical significance. The subject is studied in this thesis and several innovations are presented. A new objective estimate model is presented for shop production p
5、lanning and scheduling.It can deal with the optimization with multi-objective, uncertain objective, and the sub-objective demand of each job. An example of scheduling is given, which proves this method is feasible and reasonable. In order to accelerate the convergence speed and to avoid the earlines
6、s, an individual optimization and colony diversity genetic algorithm(IOCDGA) is presented. In IOCDGA, a new matrix representation and the basic manipulation for the genetic operator are presented, and the colony average distance and the colony entropy are difined, and the crossover probability and t
7、he mutation probability are controlled by the colony diversity. Computational results show that the algorithm is applicable and the number of iterations is less. For the sake of improving the search ability and efficiency, an integrated operator genetic algorithm(IOGA) is presented, where a new vect
8、or representation is presented, and genetic operators are adjusted and united. Computational results are better than the international latest research results. The discrete particle swarm optimization is studied for the need of high efficiency and speediness algorithm in SPPS integration optimizatio
9、n, in which the representation and learn and footing method are presented for the search experince of the individual and colony particle. The implemention of the algorithm is given. Computational results show that the algorithm is better in speed. A general integrated optimization scheme of shop pro
10、duction planning and scheduling(SPPS) is present for an automobile gearbox enterprise. The integrated optimization model of SPPS for the assembly workshop, and the representation of solution of optimization and the analysis of solution space are present. A policy of hierarchical search and the imple
11、mention of embed IOGA with two-hierarchy are given.The enterprise production data are computed, and the results is fairer far than the production practice. The integrated optimization model of SPPS for the section of Machining of shaft after heat treatment is presented. A policy of direct search and
12、 the implemention of IOGA are 南京航空航天大学博士学位论文 iiigiven. Examples prove these methods are available and efficient. The study is supported by key project of National Natural Science Foundation. Key words: job shop scheduling, production planning and scheduling, multi-objective, integration optimization
13、, genetic algorithm, particle swarm optimization. 南京航空航天大学博士学位论文 vii图清单 图 2.1 模糊生产周期隶属度函数14 图 2.2 模糊交付期隶属度函数 15 图 2.3 W%=(1,0,0)时的甘特图18 图 2.4 W%=(0,1,0)时的甘特图19 图 2.5 W%=(0,0,1)时的甘特图19 图 2.6 W%=(0.4,0.3,0.3)时的甘特图 20 图 3.1 GOMR 编码示例23 图 3.2 基于矩阵编码的 IPOX 步骤一示例24 图 3.3 基于矩阵编码的 IPOX 步骤二示例24 图 3.4 基于矩阵编码的 IPOX 步骤三示例25 图 3.5 基于矩阵编码的 IPOX 步骤四示例
