虚拟制造单元重构中资源选择问题研究

为了在保证交货期的前提下，尽可能有效地对制造资源进行优化组合，降低制造费用，提出了根据产品开发的需要进行单元重构。在产品多工艺路径的制造方式下，建立了以满足工件交货期为前提，制造费用最小化，利润最大化的资源选择与分配问题的数学模型，并给出了与传统资源选择方法完全不同的一种蚁群优化与遗传变异操作相结合的算法。该资源选择首先产生若干个多工艺路径中可能的资源配置，再以此为初始解空间，采用基于动态信息索更新策略的蚁群优化算法，进行寻优搜索，并应用遗传算法中的交叉和变异策略，拓展可行解空间，直至得到满意解。通过实例表明，蚁群遗传算法收敛速度快，并能得到最优解。     

敏捷制造单元重构算法

敏捷制造单元动态重构的实质是制造资源的优化选择及制造单元的构建。首先运用规划理论解决了针对待加工任务的可利用制造资源优化选择问题。其次，运用模糊数学理论，提出了基于粒度适中原则的模糊聚类方法，完成制造资源及待加工工件的聚合，实现了敏捷制造单元的构建。实例表明，设计的两阶段重构算法可有效地完成敏捷制造单元的重构过程，实现资源面向任务的动态优化。     

敏捷制造单元动态重构算法的研究

提出了一种基于整数规划和排序聚类算法的敏捷制造单元动态重构算法,该算法为两步,首先根据加工任务、工艺规程和允许的制造资源整合,通过整数规划算法,确定由哪些制造资源以何种工艺规程完成加工任务,然后,根据制造资源承担加工任务的工艺相似性,通过排序聚类算法,将制造资源聚合成不同的敏捷制造单元。最后,对该算的性能和特点进行了分析,并用实例加以验证。     

改进蚁群算法下激光切割加工工艺优化研究

针对传统算法下激光切割加工工艺速度慢的问题,提出改进蚁群算法下激光切割加工工艺优化设计,根据激光切割加工工艺原理,选择激光切割加工工艺参数,在此基础上对穿孔点进行确定,并引用蚁群算法,确定激光切割加工路径,选择出最短路径,以此实现对激光切割加工工艺的优化。为保证此次设计的优化方法具有一定的实际应用意义,与改进前的加工工艺进行了对比,结果表明,该优化方法能减少激光器在每个加工轮廓之间移动所需要的时间,并且通过蚁群算法能够更快得到最优加工路径。     

蚁群算法及灰色理论在制造资源配置中的应用

为了优化网络化制造环境下的制造资源配置问题,提出了一种将灰色关联理论和蚁群算法相结合的资源优化选择求解方式.在该求解方式中,首先根据工艺规划将零件加丁任务分解为按照时间先后排列的工序集;然后在每个工序节点上利用灰色关联理论解决多目标决策问题中的优势,通过多层次灰色关联系数的确定,筛选出一定数量满足要求的候选制造资源,从而缩小问题域的范围;最后利用蚁群算法从运输成本角度出发,寻找零件加工最优的制造资源选择路径,以实现网络化制造中制造资源的优化配置.     

基于蚁群算法塑模孔群加工路径优化

孔群加工是塑料模具制造中的重要工艺。针对塑模型芯孔群加工问题,提出了一种基于蚁群算法的改进优化方法,通过正交试验优选初始化参数,利用改进后的蚁群算法对塑料模具型芯孔群加工进行路径优化,以解决大规模不规则孔群加工时,空行程过多,加工效率低等问题。与标准遗传算法和蚁群算法相比,孔群加工路径平均值分别降低了6.0%和2.4%。与传统分批加工路径相比缩短了50.1%。实际路径优化效率较好,提高了加工效率。     

基于二元增强学习架构的可重构制造系统调度

构建基于重构和调度二元增强学习架构的调度系统,借助重构增强学习系统的行为来实现资源配置、重构制造单元的功能,借助调度增强学习系统的行为来为各制造单元安排加工任务,实现优化各产品加工路径和加工顺序的功能.重构和调度增强学习系统通过状态转移、行为选择和报酬获取进行联系.提出结合函数泛化器的自适应步长增强学习算法的学习机制和学习步长调整机制,通过实验分析了函数泛化器的性能,验证了该算法解决一类重人型的可重构制造车间调度问题的有效性.     

双资源约束的航天结构件车间生产调度方法

针对航天结构件生产调度过程中工序加工时间随操作人员能力级别而变化,关重件和关键工序特定设备和人员要求约束等问题,构建综合考虑关键设备和人员双资源约束的柔性车间作业调度数学模型,解决航天结构件生产设备资源维、人员资源维、工序信息维的描述。为实现双资源约束下的柔性调度,提出嵌套式蚁群-遗传混合算法,针对结构件加工工序设备和人员双资源选择问题,建立资源选择问题与蚁群并行搜索的映射关系,设计蚁群算法实现资源选择;为提高算法搜索性能,利用遗传算法求解当前蚂蚁资源选择下的调度问题,在避免人员资源使用冲突前提下实现设备加工工序的优化排序,使得关键设备利用率最大,总完工时间最短,并通过综合考虑蚁群和遗传算法阶段的求解目标进行蚁群算法信息素的更新,大大提高算法获得较优解的性能。最后,通过航天结构件车间实际案例进行算法测试,测试结果表明,嵌套式蚁群-遗传混合算法能有效地求解双资源约束的航天结构件车间生产调度问题,可降低资源总负荷,提高关键设备利用率,避免人员资源冲突,具有良好的综合调度性能。     

基于费效模型的单元重构及优化

制造系统的合理单元构建、单元间布局可以有效的提高制造系统的生产效率、缩短生产准备时间以及缩短交付期。本文研究多制造单元构成的制造系统在多工艺、多加工路径条件总费用最小化问题。首先采用自适应遗传算法的算子设计,对制作单元数及工件数进行有效编码;其次运用精英保留策略对以编码的多个制造单元进行优化重组,减少工件的跨单元次数,使其加工和运输成本达到最优;最后应用模拟算例,验证模型和算法的有效性。     

基于混合算法的反射面天线面板装配序列规划

为了改善反射面天线面板的装配变形,从装配工艺入手,基于蚁群算法和遗传算法,结合传统的天线反射面装配方法,提出一种天线装配序列规划的混合算法。该算法利用蚁群算法快速得到初始种群,随后使用遗传算法对初始种群进行优化,根据所得优化解生成蚁群算法中路径上的信息素,通过加速蚁群算法最优解信息的积累来更快地得到最优解;同时,建立反射面装配的有限元仿真模型,利用该模型及时对得到的最优解(即装配序列)进行面向装配过程的面板装配变形动态仿真,将仿真结果返回算法中,进一步校正算法并得到最优解。以某工程抛物面天线的面板装配为例,验证了所提混合算法的正确性。     

基于GA的敏捷制造系统资源选择研究

针对资源选择提出了一种以最小的竞标花费和交通运费为优化目标的算法,该算法分两步完成.首先,通过对时间因素的排除,将0-1整数规划问题转化成2次指派问题;其次,在引入虚拟资源和虚拟任务之后,提出了一种基于遗传算法的优化算法.通过引入虚拟资源或虚拟任务,使得该算法种群的编码、初始化,以及交叉算子的设计等都变得非常简单,从而大大提高了运算速度.最后的实例分析也表明了该算法的有效性.     

A new manufacturing resource allocation method for supply chain optimization using extended genetic algorithm

In distributed manufacturing environments, the real competitive edge of an enterprise is directly related to the optimization level of its supply chain deployment in general, and, in particular, to how it allocates diverse manufacturing resources optimally. This is faced with increasing challenges caused by the conflicting objectives in manufacturing integration over distributed manufacturing resources. This paper presents a new manufacturing resource allocation method using extended genetic algorithm (GA) to support the multi-objective decision-making optimization for supply chain deployment. A new multi-objective decision-making mathematical model is proposed to evaluate, select, and sequence the candidate manufacturing resources allocated to sub-tasks composing the supply chain, by dealing with the trade-offs among multiple objectives including similarity, time, cost, quality, and service. An extended GA approach with problem-specific two-dimensional representation scheme, selection operator, crossover operator, and mutation operator is proposed to solve the mathematical model optimally by designing a chromosome containing two kinds of information, i.e., resource selection and resource sequencing. A case study is carried out to demonstrate the effectiveness and efficiency of the proposed approach.     

基于蚁群遗传算法优化的质量矩拦截弹非线性预测控制研究

以所建立的质量矩拦截弹数学模型为基础,通过对模型合理的简化,得到一个耦合的非线性动力学系统,利用非线性预测控制设计了拦截弹飞行控制系统.针对非线性预测控制算法的参数一般采用试凑的方法进行优化设计,提出一种基于蚁群遗传算法的非线性预测控制参数优化策略,同时应用遗传算法对蚁群算法的蚂蚁路径进行优化,并给出了这一算法的具体实现步骤,最后将该优化方法应用于拦截弹飞行控制系统.     

Concurrent tolerance design for manufacture and assembly with a game theoretic approach

This paper proposes a game theoretic approach for concurrent tolerance design considering the practical manufacturing process and assembly process, to achieve lower manufacturing cost and good product assemblability. The optimization objectives and constraints for concurrent tolerance design for manufacture and assembly are discussed, and the key technologies for concurrent tolerance design with game theory are investigated, including the strategy set division, payoff calculation, and Nash equilibrium evolution. An optimization algorithm is proposed to achieve concurrent tolerance design with a game theoretic approach, and the genetic algorithm is used to conclude the Nash equilibrium of the game for manufacture and assembly. Finally, a case study is given to verify the proposed approach.     

基于遗传算法的生产周期和生产费用双目标双向车间调度问题的研究

优化车间生产调度的目的不仅仅是为了缩短生产周期,更重要的是为了提高生产效率,降低生产成本。采用改进的遗传算法结合神经网络,以生产周期和生产费用为双目标,实施调度优化。编排车间作业采用两种不同的调度方法,即满足关键工件交货期的倒排法和综合考虑生产周期和生产成本的顺排法———双向双目标调度的方法。调度的策略是在确保客户交货期要求的前提下,充分利用现有资源,缩短产品生产周期、降低企业生产成本,提高企业的市场竞争力。     

多项目调度问题研究

现有多项目调度研究一般都基于不切实际的假定——资源在多个项目间进行传递时不需要花费额外的时间和成本。在对传递时间进行分析的基础上,建立了考虑资源传递时间的多项目调度问题的数学模型,并针对模型设计了结合并行调度的混合遗传算法,在保证项目任务紧前关系与资源约束的条件下,能够合理优化多项目的资源分配,从而有效地缩短多项目的总工期,通过实例验证了考虑传递时间的必要性,同时实例也证明了算法的有效性。     

Research on collaborative conceptual design based on distributed knowledge resource

Conceptual design is an early stage of the whole product development process, which involves the generation of solutions to satisfy the design requirements. It is a crucial issue on how to find proper principle solutions from enormous, isomeric, and even distributed knowledge resources repositories and then ally with these correlative distributed knowledge resources providers to collaborative conceptual design. Design is a task-solving process using corresponding resource, so it does not always have an agency in product design. This paper proposes a model of collaborative conceptual design based on distributed knowledge resources, which is centered on design tasks and based on dynamic knowledge alliance without agency. After a model of knowledge resource unit and a model of Web services based on knowledge resource are proposed, a virtual quantitative and qualitative knowledge resource representation based on knowledge grid is then put forward. The mixed knowledge resource repository is also proposed to support collaborative conceptual design. The strategies of knowledge register, publish, storage, reorganization, orientation, and problem solving based on mixed knowledge resource repository are then discussed in detail. The data exchange mode of principle solutions is then described. The selection strategy of dynamic knowledge alliance members is established. The framework of distributed knowledge resource is also put forward to support collaborative conceptual design. The process of collaborative conceptual design based on distributed knowledge resource is then proposed. The conceptual design of new devices for decreasing resistance of amphibious vehicles is given as an example, which demonstrates that the methodology is obviously helpful to produce valuable principle solutions with distributed knowledge for product innovation.     

网格任务调度算法的研究

提出了一种基于遗传算法和Min-Min的网格调度算法,该算法主要分为资源搜索和任务分配两部分。首先,利用遗传算法从网格的可用资源中快速地搜索出最优的或是近最优的资源;然后从任务队列中取一组任务,利用Min-Min算法把任务分配给已搜索到的资源上去执行。仿真实验结果证明:对于大规模资源的调度该算法是有效的,能够减少了任务执行时间,改善网格任务调度的效率。     

Parallel Test Tasks Scheduling and Resources Configuration Based on GA-ACA

A Genetic Algorithm-Ant Colony Algorithm(GA-ACA),which can be used to optimize multi-Unit Under Test(UUT)parallel test tasks sequences and resources configuration quickly and accurately,is proposed in the paper.With the establishment of the mathematic model of multi-UUT parallel test tasks and resources,the condition of multi-UUT resources mergence is analyzed to obtain minimum resource requirement under minimum test time.The definition of cost efficiency is put forward,followed by the design of gene coding and path selection project,which can satisfy multi-UUT parallel test tasks scheduling.At the threshold of the algorithm,GA is adopted to provide initial pheromone for ACA,and then dual-convergence pheromone feedback mode is applied in ACA to avoid local optimization and parameters dependence.The practical application proves that the algorithm has a remarkable effect on solving the problems of multi-UUT parallel test tasks scheduling and resources configuration.