船体平面分段建造装配序列规划与装配线平衡方法研究

发布时间：2018-03-12 15:38

本文选题：装配序列规划（ASP）　切入点：装配线平衡（ALB）　出处：《上海交通大学》2014年博士论文　论文类型：学位论文

【摘要】：船舶制造中的装配工作一般耗时最久、需要的关键资源最多。其中，分段装配工时大概占船体装配总工时的45%左右，分段装配成本在船体装配总成本中所占的比例更是高达50%以上。因此，研究船体分段装配过程中的最优化问题，对缩短船体建造周期、减少装配成本以及提高装配质量具有重大意义。本文以船体平面分段装配为对象，研究了分段装配序列规划与装配线平衡方法，为解决复杂产品的装配最优化问题提供决策依据。研究工作主要包括以下内容：（1）基于实例推理（CBR）与基于几何约束推理相结合的平面分段装配序列规划方法。船体分段结构复杂，所包含的零部件众多，给可行装配序列推理造成很大的困难。另外，在结构组成上，组成分段的零部件之间没有严格的几何约束关系，满足几何约束的可行序列数量较大，不利于序列的评价及优选。本文利用基于实例推理的方法，参照与装配问题最相似实例的装配顺序，生成装配问题的初始优先约束关系，然后结合基于约束的推理，生成装配体完整的装配序列，并利用遗传算法进行装配序列选优。论文用具体加工算例验证了该方法的有效性。（2）批量型装配线平衡问题建模及求解方法研究。在分段部件装配过程中，各部件不仅要满足加工时间、平台占用面积等的约束，还需符合一定的完工优先关系。以装配线劳动生产率最大化及工作站负荷均衡为优化目标，在合理假设的基础上，，建立批量型装配线平衡问题的数学模型。本文利用文化基因算法对问题进行求解，寻找最优的分段部件装配计划。通过染色体的解码过程实现任务的智能分批，设计了序列自动调整算子，保证生成的最优解符合优先关系约束。另外，在算法框架中加入局部搜索算子，加快了算法的求解过程。最后通过分段部件装配的实际算例验证算法的有效性。该问题的解决，对于具有优先关系约束的任务割分或排序问题的求解具有参考价值。（3）混合型装配线平衡问题建模及求解方法研究。船体平面分段装配流水线平衡的目的是寻找最优的加工任务排序，实现总任务的生产周期最短、任务拖期数量最少以及工作站负荷均衡等目标。为了获取该多目标优化问题的Pareto最优解，本文在合理假设的基础上，构建了混合型装配线平衡问题的数学模型。提出了一种改进的多目标遗传算法，设计了个体适应值的分配策略及Pareto解集更新机制，使得算法能够快速地获取满足优化目标的多组非劣解。最后通过船体平面分段装配实例验证了算法的有效性。混合装配线平衡问题的解决，可以为多目标组合优化问题非劣解的求取提供一定的参照。（4）产品装配序列规划及装配线平衡系统集成方法研究。研究了序列规划及装配线平衡的关系，构建了船体平面分段装配最优化应用系统，并以国内某造船厂平面分段装配的实际加工数据对系统进行了验证。综上，本文给出了一整套船体平面分段装配过程最优化的解决方法，对企业缩短分段装配工时、节约生产成本提供了决策支持。
[Abstract]:The general assembly work of Shipbuilding's longest and most critical resource needs. Among them, sub assembly time probably accounted for about 45% of the total hull assembly, sub assembly cost in hull assembly in the total cost of the proportion is up to more than 50%. Therefore, optimization research on hull assembly process. To shorten the construction cycle of the hull, reduce assembly cost and is of great significance to improve the quality of assembly.
In this paper, segmented assembly sequence planning and assembly line balancing method are studied based on segmented assembly of hull plane, which provides decision basis for solving assembly optimization problems of complex products.
(1) based on case-based reasoning (CBR) plane combined with geometric constraints based on piecewise assembly sequence planning method. Hull structure is complex, contains many components, causing great difficulties for feasible assembly sequence reasoning. In addition, in structure, no strict geometric constraint relationship between the composition of some parts meet, the geometric constraint of feasible sequence number is large, is not conducive to the sequence evaluation and optimization. This paper uses the method of case based reasoning, assembly sequence and assembly problems referring to the similar case, the initial generation of assembly precedence relationship problems, and then combined with constraint based reasoning, generating assembly assembly sequence integrity, and assembly sequence optimization by genetic algorithm. The specific processing examples verify the effectiveness of the method.
(2) batch type assembly line balancing problem modeling and solving methods in the study. The segmentation of each component in the assembly process, not only to meet the processing time, the platform occupied area and other constraints, also need to meet certain priority to the assembly line. The completion of labor productivity maximization and work station load balancing as the optimization goal, based on reasonable assumptions, establish mathematical model of batch type assembly line balancing problem. This paper uses the memetic algorithm to solve the problem, finding the optimal segment assembly plan. The task of intelligent batch through the decoding process of chromosomes, designed to automatically adjust the sequence of operators, ensure that the generated optimal solution with precedence constraints. In addition, join a local search operator in the algorithm frame, to speed up the algorithm. Finally, through the actual effectiveness of the segmentation assembly example verification algorithm. The problem The solution is of reference value for solving the problem of task cutting or sorting with priority relation constraints.
(3) research on Modeling and solving method of mixed model assembly line balancing problem. Hull plane block assembly line balance is objective to find the optimal sorting processing tasks, achieve total task production cycle short, the minimum number of tardy tasks and workstation load balancing goals. To obtain the Pareto optimal solutions of multi-objective optimization problem in this paper, on the basis of rational hypothesis, constructs a mathematical model for mixed model assembly line balancing problem. This paper proposed an improved multi-objective genetic algorithm, the design of individual adaptation strategies and Pareto value of solution set update mechanism, the algorithm can quickly obtain a satisfying non inferior solution. Finally, the optimization goal the plane through the hull block assembly examples verify the effectiveness of the algorithm. To solve the mixed model assembly line balancing problem, for multi-objective optimization non dominated solutions for providing certain Reference.
(4) research on system integration method of product assembly sequence planning and assembly line balance. To study the relationship between sequence planning and assembly line balancing, planar hull block assembly optimization application system is constructed, the verification of the system data and the actual processing by a domestic shipyard plane block assembly.
In conclusion, this paper gives a complete solution to the optimization of hull plane sectional assembly process, which provides decision support for shortening segmental assembly time and saving production cost.

【学位授予单位】：上海交通大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：U671

【参考文献】