【作者】 陈凯；

【导师】 蒋祖华； 刘建峰；

【作者基本信息】 上海交通大学 ， 机械工程（工业工程）， 2016， 硕士

【摘要】 随着信息技术的高速发展,造船模式也在向智能化和精细化方向不断改进。船舶作为大宗商品,生产时需将其划分为若干个分段进行建造,分段从建造车间脱胎后,还需经过预舾装、涂装、总组等工艺才能到船坞进行搭载。由于工艺车间场地的限制,半成品分段经常需要运到分段堆场中暂存、检验和修补。分段在堆场上是用平板车进行运输的,由于分段体积庞大,在其行走的路径上必须保证足够的空间,否则对于阻挡分段的移动将消耗大量的成本。因此,在进行分段堆场调度时,应优化分段放置位置及分段任务执行顺序,从而降低移动阻挡分段的成本。针对分段在堆场中的调度问题,本文做了以下的研究:对于单个堆场的调度,为求得物流成本最小的堆场调度方案,构建了以最小化移动度为目标的堆场调度模型,并构造多链DNA遗传算法对任务分段的执行顺序及进场分段的放置位置进行优化。其中一个任务分段进、出场时,将堆场上堆位的邻接关系及分段占用情况转化为结点距离图,并用最短路径算法进行求解,从而得到阻挡分段最少的路径。针对路径上的阻挡分段,提出了五种移动策略,并在验证阶段对五种移动策略的调度结果进行比较。对于分段在多个堆场中的调度问题,考虑了分段的工艺流程,并根据堆位对不同工艺阶段分段的偏好设定优先级,构建了以移动度最小为目标的多堆场调度模型。在此基础上,将分段调度任务分解为多个阶段进行求解,在确定任务执行顺序、选择目标堆位、阻挡分段移动策略阶段分别提出两种策略,并在验证阶段将其与船厂现行策略的调度结果进行对比。最后,通过开发分段堆场调度信息系统,将构建的堆场调度模型及求解算法嵌入其中,并通过交互式界面实现对数据信息的管理及前台堆场调度模块的应用。更多还原

【Abstract】 With the development of information technology,the shipbuilding mode keeps changing toward the trend of intelligence and elaboration.As a kind of large-scale equipment,ships need to be divided into many blocks to be manufactured.After the block leaves the processing position in the manufacturing workshop,it requires pre-outfitting,painting and major assembling before it can be sent to the shipyard and equipped.Due to the limitation of workshop spaces,half-finished blocks often need to be moved to the block yard for temporary storage,inspect and repair.The blocks are transported using a flat transporter in the yards.Because of its large size,the flat transporter needs a lot of space on the route,otherwise the removal of many obstructive blocks will incur a lot of cost.In order to reduce the cost related with moving obstructive blocks,the scheduling of a block yard should optimize the placing positions of blocks as well as the execution order of block tasks.To solve the block yard scheduling problem,the theses has done the following research.For the scheduling of a single yard,construct a model minimizing the block move degree to obtain the schedule with the lowest logistic cost of block yard.The order of block tasks and the position of an inbound block are optimized by adopting a multi-chain DNA genetic algorithm.At the inbound and outbound stage of a block,by mapping the connective relation and occupation status of yard positions to a node distance graph,obtain the route with the least obstructive blocks using the shortest path algorithm.Propose five move strategies for the blocks obstructing the routes,and compare the results of the five strategies in the scheduling experiment.For the scheduling of multiple block yards,based on the work stages of ship blocks and the preference of yard positions to the blocks at different work stages,set the priority of block placing.Construct the scheduling model that takes the minimization of block move degree as the objective.Divide the scheduling task into multiple phases to solve,and propose two solutions strategies at each of the block task sequencing phase,target yard position selection phase,and obstructive block moving phase.The proposed strategies are evaluated with experiments and compared with the existing scheduling methods used by the factory.At last,through developing a block yard scheduling information system,apply the proposed model and algorithms to practice.The system contains interactive user interface to realize the management of background information as well as the application of foreground scheduling modules.更多还原