多工位装配序列粒子群优化算法

针对传统单工位装配序列求解上的不足,将粒子群算法应用于多工位多目标装配序列优化的求解,提出一种面向复杂多工位产品的装配序列优化方法。采用优先序列图(Assembly precedence graph,APG)来描述零件间的优先约束关系,构建优先关系矩阵、装配干涉矩阵、工位能力表和装配信息表,描述装配部件干涉及工位之间的关系;给出粒子群算法编码体系和装配关系算法模型表达方法;综合考虑装配操作成本、装配工具更换成本和装配夹装变更成本和运输成本的影响,提出有工程意义的适应度函数的表达式;根据APG生成随机的可行初始装配序列,并利用粒子群算法(Particle swarm algorithm,PSO)对装配序列和装配工位进行优化。以飞机起落架装配序列规划实例验证多工位粒子群装配序列优化算法有效性。     

基于遗传模拟退火融合算法的船舶分段装配序列优化

针对复杂船舶分段装配序列规划问题,提出基于遗传模拟退火算法的分段装配序列规划求解方法,综合考虑分段装配中的工艺约束和几何约束,建立以分段装配所需时间和消耗成本为优化目标的问题模型,并为模型求解设计了遗传模拟退火融合算法,将模拟退火算法的局部搜索能力与遗传算法的快速全局搜索能力相结合,达到快速收敛到全局最优解的目的.通过实例验证了该算法的有效性.     

面向装配系统可靠性分析的车身夹具系统优化设计

为描述多工位车身装配系统的渐进衰退特性,建立了面向车身产品尺寸质量和夹具系统要素的装配系统可靠性模型。结合夹具设计稳定性参数约束,提出了面向白车身装配系统可靠性的车身多工位焊装过程的夹具布局优化方法;在给定夹具布局下构建了定位销制造成本函数,提出多工位夹具定位销磨损率的优化模型并对其进行优化分配。将上述方法应用于侧围装配案例,分析发现优化后装配系统可靠性衰退过程得到较大改善,并有效提升了夹具系统的使用寿命。     

敏捷化开发环境下产品装配建模研究

为了实现以自顶向下为主、自底向上为辅的双向产品设计方法,本文提出了一种面向产品族的混合装配模型。该模型由装配骨架模型、层次模型、约束关系模型、零件属性模型及参数化设计模型组成。通过引入高层次的装配骨架模型反映设计者的设计意图并自动向下游传播装配约束,实现自顶向下面向过程的设计方法;通过引入低层次的约束关系模型记录零件之间的各种约束关系,检验、改进和完成最终装配模型的重建,实现自底向上的面向结果的设计方法;通过引入参数化设计模型完成装配体的变形设计,实现以往知识的重用。最后用实例说明该混合装配建模方法的有效性。     

基于有向割集分解的装配序列生成方法

基于有向图理论,提出了装配序列规划新方法。把有向图理论引入到装配模型表达,提出了有向装配连接图模型,并给出了有向装配连接图的性质。通过对有向装配连接图的割集分解计算,实现了装配序列生成。此方法比无向割集分解方法计算量小、产生无效序列少。为进一步剔除无效序列研究了单调非线性装配意义下基于有向割集分解的装配序列生成算法,并讨论了约束的加入和表达方法。最后给出了一个装配实例,证明了算法的有效性。     

基于约束蚁群算法的装配序列优化

针对装配序列优化问题,定义装配约束条件下的蚂蚁转移概率函数和启发式信息函数,提出了面向装配序列优化的约束蚁群算法,以获得相对较优的装配序列。改进了算法的参数设置,利用信息素残留系数的动态变化以及优化转移概率的公式,控制算法的迭代速度,避免了算法过快收敛导致的局部最优解。运用CATIA二次开发方式获取了干涉矩阵等装配信息模型,并以一个典型实例分析了约束蚁群算法的特性,验证了算法的可行性和可靠性,实例证明了约束蚁群算法的有效性。     

基于改进粒子群算法的多工位装配序列规划

针对现有单工位装配序列规划结果难以满足工位需求的问题,提出一种基于改进粒子群算法的多工位装配序列规划方法。建立多工位装配模型来描述零部件的几何信息及其与工位的关系;用装配序列可行性、装配方向一致性、装配聚合性及工位间平衡性4个评价指标构建适应度函数。为解决一般粒子群算法易陷入局部最优解的问题,对惯性权重进行了改进,提出粒子相似度和相似度阈值的概念,并通过相似度阈值控制粒子的变异,提高了算法的全局搜索能力。以某型发动机为装配实例,验证了改进粒子群算法应用于多工位装配序列规划的可行性;同时将该算法和遗传算法、一般粒子群算法进行比较,证明了该算法的优越性。     

基于网络截集的装配生产线瓶颈识别模型

针对飞机装配生产线设计阶段现有的瓶颈识别方法存在的不足,提出了基于网络截集的装配生产线瓶颈识别模型。以最大流最小截理论为基础对装配生产线瓶颈进行了定义;通过分析工位顺序和工位间的关联关系,建立了包含工位和装配任务关系的有向装配网络模型,将装配生产线上的工位分为装配起点、装配终点和装配中间点三部分进行描述;构造装配有向网络的邻接矩阵,建立了装配生产线瓶颈识别数学模型。以一个7个工位装配生产线瓶颈识别为例,讨论了所提方法的应用过程,并与同类方法相比,说明了方法的有效性。     

融合连接矩阵与模糊集合的装配序列智能规划

针对产品装配环节中过于依靠传统人工规划和二维图纸装配、智能化不足等问题,提出了一种基于连接矩阵与模糊集合的装配序列智能规划方法。提取产品CAD模型中的装配信息和约束信息,生成产品本体模型,建立已知装配序列产品结构信息规则库和典型结构库。典型结构或者完全匹配的产品直接输出装配序列;相似结构通过相似匹配重用典型结构装配规划特征进行装配序列推理,非典型非相似结构应用基于模糊集的装配知识推理得到其装配序列。基础装配序列通过基于连接关系的装配序列优化方法进行二次规划,得到完整装配序列,实现产品装配序列智能规划。通过圆柱齿轮减速器装配序列智能规划应用实例验证了方法的可行性。     

考虑装配关系复杂性的多目标装配线平衡优化方法

针对装配关系复杂性影响工人工作效率,进而产生瓶颈工位的问题,提出一种考虑装配关系复杂性的改进型多目标装配线平衡优化方法。建立了一种合理测度作业元素装配关系复杂性的方法,继而定义装配关系复杂性平滑系数这一指标。以生产节拍、平滑系数和装配关系复杂性平滑系数为优化目标,工序优先关系和工作站数为约束,建立了多目标装配线平衡优化方法模型。采用遗传算法作为优化算法,并对遗传算法的交叉环节利用模糊聚类算法进行改进,判断选作交叉的两个个体的近亲概率,在一定概率上防止了近亲交叉,增加了种群繁衍的多样性。通过算例验证了所提平衡优化方法的有效性,为制定考虑装配关系复杂性的平衡方案提供了一种可行方法。     

基于遗传算法的飞机装配序列规划优化方法

在基于割集的装配顺序规划方法中,随着零件的增加,装配序列数呈指数增加,由此导致算法的组合爆炸问题。为此,提出了基于．遗传算法的装配序列规划优化方法,研究了零件装配信息的基因组表示方法,将参与装配的零件的基因组排列成序,实现装配顺序的染色体表示。采用模糊集理论建立了适应度函数,并对装配序列进行评价和优化。以某型飞机机翼产品为例,验证了技术的有效性。     

Integrated assembly and machining planning for electronic products using a genetic algorithm approach

To produce an electronic product, both assembly operations and machining operations are required in the process plan. In most cases, the assembly operations and machining operations need to be combined in a continued order with an integrated sequence. This is different from the traditional process planning approaches in which machining operations and assembly operations are separated as two independent tasks with no interactions. For an electronic product, the two types of operations and the associated costs may affect each other in an interactive way. Therefore, the sequence planning of assembly operations and machining operations must be analyzed with an integrated model. In this research, a graph-based model is presented to represent the assembly and machining operations in an integrated model. The related operation cost functions are developed to evaluate the costs for the integrated assembly and machining sequences. The integrated sequence planning problem is solved using a genetic algorithm approach with an objective of lowest operation costs. As a result, the assembly operations and machining operations can be planned in an integrated sequence suitable for producing electronic products. The result shows that the developed method using the genetic algorithm approach is efficient for solving the integrated sequence planning problem. Example products are demonstrated and discussed.     

Imperialist competitive algorithm for assembly sequence planning

Automated generation of all feasible assembly sequences for a given product is highly desirable in manufacturing industry. Many research studies in the past decades described efforts to find more efficient algorithms for assembly sequence planning. Imperialist competitive algorithm for assembly sequence planning is presented in this paper. Population individuals called countries are in two types: colonies and imperialists that all together form some empires. Each assembly sequence is encoded into the country. The proposed algorithm is tested and compared with genetic algorithm and particle swarm optimization. Results show that imperialist competitive algorithm can improve the quality in solution searching and upgrade the opportunity to find optimal or near-optimal solution for assembly sequence planning.     

基于改进蚁群算法的装配序列规划研究

装配序列规划能解决装配过程中所出现的速度和质量问题,降低生产成本。通过考虑装配影响因素,对装配体零件之间关系进行数学建模,提出一种解决该问题的改进蚁群算法。算法提出用加强装配关系矩阵来表示零件之间存在的强约束关系;用加强随机状态转移规则,使蚂蚁能更快地寻找到最优或者近优解;完善信息素更新规则,使算法能更好的利用正反馈机制。通过实例证明,该改进型算法能实际有效地解决装配序列规划问题。     

An effective integration approach toward assembly sequence planning and evaluation

An integration strategy for assembly sequence planning and sequence scheme evaluation is proposed. This strategy can be used to plan a reasonable assembly sequence, to optimize a sequence scheme, and to predict whether a collision will occur between the assembly tool and assembled components by considering factors like target components and assembly resources. A hybrid method is presented for assembly sequence modeling that combines human-computer interactive operations to manually build a hierarchical assembly sequence main model and a hybrid graph method to automatically generate sub-assembly sequence schemes of the main model. An optimization algorithm based on time-cost is introduced to handle a best candidate components selection. This relieves the problem of limited capability found when handling large size assembly models with traditional methods. The essential issues involved in system implementation are discussed as well; these include a representation method for the assembly consequence model, an optimization model of assembly sequence planning, and an object-oriented system architecture model employed with multi-agent technology for visually evaluating the assembling process. This system, KM computer-aided assembly process planning, KMCAAPP, has been developed on the basis of our previous work, KMCAD3; KMCAAPP uses the presented approach. KMCAAPP can be integrated with CAD model from KMCAD3D. A case study shows that the presented approach can use large CAD assembly models and delivers a feasible and effective way to integrate the assembly sequence planning process with scheme evaluation by visually evaluating the assembling process. This allows the identification of design errors in a timely manner and mitigates economic loss.     

An intelligent interactive approach for assembly process planning based on hierarchical classification of parts

Assembly process planning is a difficult problem in the product design and development process. In order to reduce the difficulty of assembly process planning (APP), this paper presents an approach to finding a practical plan for assembly process through the intelligent interactive method based on hierarchical classification of parts. First, two hierarchical classification principles towards interactive disassembly are presented based on the key assembly constraint (KAC). Then, the initial assembly relationship of the product is readjusted to a new hierarchical structure which is more suitable for assembly planning according to a novel hierarchical classification algorithm. On the basis of the new hierarchy of product, an intelligent disassembly-based approach is presented, which not only takes full advantage of the ant colony algorithm for searching and constructing an initial assembly sequence with three available assembly principles, but also makes full use of the experience and knowledge of the human being for achieving a better assembly process. The intelligent disassembly approach can achieve an optimized assembly sequence, paths and scheme necessary information regarding the required assembly fixtures and tools. Finally, a case study shows that the proposed approach is feasible and practical.     

基于模拟退火算法的装配序列生成与优化

分析了一个优良装配序列应满足的装配约束条件,从几何可行和工艺优良两方面对装配序列进行评价;介绍了影响装配序列工艺性能的四大因素及其量化方法。在此基础上,建立了基于模拟退火算法的装配序列优化规划模型,并对算法的性能和速度进行了分析、处理。最后通过齿轮泵的应用实例说明了该优化算法的有效性。     

基于遗传算法的工步优化排序方法

针对数控加工中心上零件加工工步的排序问题,以辅助加工时间最短为优化目标,使用遗传算法对零件在一次装夹情况下的加工工步进行优化排序。提出了使用特征关系图和特征高度描述待加工特征之间加工的优先顺序、采用工步优先关系矩阵校验工步序列合理性的方法。论述了初始群体的生成、遗传算子以及工步优化排序的过程和算法。实际应用表明,该方法可有效提高工艺规划系统中工步的优化排序能力。     

Unification of the a priori inconsistencies checking among assembly constraints in assembly sequence planning

Sequence planning generation is an important problem in assembly line design. A good assembly sequence can help to reduce the cost and time of the manufacturing process. This paper focuses on assembly sequence planning (ASP) known as a hard combinatorial optimization problem. Although the ASP problem has been tackled via even more sophisticated optimization techniques, these techniques are often inefficient for proposing feasible assembly sequences that satisfy the assembly planners’ preferences. This paper presents an approach that makes easier to check the validity of operations in assembly process. It is based on a model of the assembly planners’ preferences by means of strategic constraints. It helps to check a priori the consistency of the assembly constraints (strategic and operative constraints) given by the assembly system designers before and while running an assembly plan generation algorithm. This approach reduces the solution space significantly. A case study is presented to demonstrate the relevance of the proposed approach.     

Research on a kind of assembly sequence planning based on immune algorithm and particle swarm optimization algorithm

In order to solve the assembly sequence planning of a certain type of product (e.g., product is composed of different types of parts and each type has several identical parts), first of all, the rule of nomenclature for this type of product is designed. Secondly, geometric feasibility and coherence are designed as constraint conditions, and these two constrains and processability are combined with each other as the objective function. Finally, the proposed novel method under the name of immune particle swarm optimization algorithm is tested. The results show that the proposed immune particle swarm algorithm is an efficient approach to solve the assembly sequence planning problem.