船舶建造中不规则分段空间调度算法研究

发布时间：2018-05-10 07:25

本文选题：船舶建造 + 空间资源调度　；参考：《哈尔滨工业大学》2017年硕士论文

【摘要】：在当前造船模式下,空间资源被视为制约船舶生产效率的瓶颈资源之一,其调度问题牵扯时间、空间多个维度,复杂性较高,是目前船舶企业普遍关注的重点。因此本文针对船舶不规则分段建造空间资源的调度问题展开研究,并在此基础上设计了船舶建造项目管理原型系统。首先考虑到船舶建造过程中成组技术应用的重要性,对分段的聚类问题展开了研究。将分段聚类问题转化为不规则多边形的聚类问题进行求解,以分段平面几何图形的轮廓特征为相似判断依据建立了几何相似模型,并给出了相似度计算方法;在此基础上研究基于模拟退火的聚类算法以提高聚类结果的质量;同时研究基于中心特征的增量聚类方法,提高分段聚类实际应用中的效率。其次在分段聚类的基础上,研究给定分段集合特定时间下的静态空间调度问题。针对船舶分段空间布局的特点给出了离散化的几何信息处理方式,考虑相似分段集中布局以及剩余不可用空间最小化建立了空间静态调度的数学模型;在遗传算法的基础上结合启发式策略提出了静态调度的求解算法,通过仿真实验验证了算法在在求解速度和质量方面的优越性。随后在静态调度基础上考虑时间因素研究分段动态空间调度问题。考虑分段加工的时序约束,以场地负载均衡和工期最短为目标建立了动态空间调度数学模型;针对此类问题解空间过大的特点,通过采取基于调度序列的编码方式和基于顶点定位的空间定位方法来缩小解空间,结合遗传算法和模拟退火算法设计了混合求解算法,算例表明了算法具备较好的搜索效率和求解质量。最后以船舶建造过程中系统的需求分析为依托,设计了船舶建造项目管理原型系统,针对系统体系结构、系统功能以及数据库信息模型进行了详细设计。综上所述,本文主要研究不规则分段聚类及其静态、动态空间调度问题的求解方法,对已有算法作出了改进使之更加适用于实际问题的求解,本文研究具有理论及实用价值,能够为船舶实际生产提供一定的理论及技术软件支持。
[Abstract]:In the current shipbuilding mode, space resources are regarded as one of the bottleneck resources which restrict the efficiency of ship production. The scheduling problem involves time, space dimension and complexity, which is the focus of the shipbuilding enterprises. In this paper, the scheduling problem of irregular segmented building space resources is studied, and a prototype system of ship construction project management is designed. Firstly, considering the importance of the application of group technology in ship building, the clustering problem of segmentation is studied. The piecewise clustering problem is transformed into the irregular polygon clustering problem. The geometric similarity model is established based on the contour feature of the segmented plane geometry and the similarity calculation method is given. On this basis, the clustering algorithm based on simulated annealing is studied to improve the quality of clustering results, and the incremental clustering method based on central features is studied to improve the efficiency in practical application of segmented clustering. Secondly, on the basis of piecewise clustering, the static space scheduling problem for a given piecewise set is studied. According to the characteristics of ship segment spatial layout, a discrete geometric information processing method is presented. The mathematical model of spatial static scheduling is established considering the similar piecewise centralized layout and the minimization of the remaining unusable space. Based on genetic algorithm (GA) and heuristic strategy, a static scheduling algorithm is proposed, and the superiority of the algorithm in solving speed and quality is verified by simulation experiments. Then based on the static scheduling, considering the time factor, the piecewise dynamic space scheduling problem is studied. Considering the timing constraints of piecewise machining, a mathematical model of dynamic spatial scheduling is established with the objective of site load balance and shortest time limit, and the solution space of this kind of problem is too large. By adopting the coding method based on scheduling sequence and the spatial location method based on vertex location to reduce the solution space, a hybrid solution algorithm is designed by combining genetic algorithm and simulated annealing algorithm. The example shows that the algorithm has good search efficiency and solution quality. Finally, based on the requirement analysis of the system in the process of ship construction, the prototype system of ship construction project management is designed, and the system architecture, system function and database information model are designed in detail. To sum up, this paper mainly studies irregular piecewise clustering and its solving methods of static and dynamic space scheduling problems, and improves the existing algorithms to make them more suitable for solving practical problems. The research in this paper has theoretical and practical value. Can provide certain theory and technical software support for ship actual production.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U671

【参考文献】