【作者】 李斌；

【导师】 钟毅芳；

【作者基本信息】 华中科技大学 ， 机械设计及理论， 2007， 博士

【摘要】 为适应全球化的市场竞争环境和日益增长的用户个性化需求,越来越多的企业采用大规模定制的生产模式进行产品设计和制造,大规模定制被认为是21世纪的主要制造模式。本文结合实际课题对面向大规模定制的产品配置设计和制造策略进行了研究,具体研究了研究了产品配置建模和求解、产品配置优化、基于构件复用的产品配置设计和配置管理以及装配线优化调度等面向大规模定制的关键使能技术。建立配置模型是配置求解的基础,本文首先对基于产品结构的配置模型进行了研究,分析了产品结构模型的组成,对配置规则进行分类,指出产品配置的一般过程和配置方案优化的概念,并对配置优化和装配线优化调度的算法基础——多目标遗传算法进行了研究。分析了约束满足问题和产品配置的关系,研究了约束满足问题和关系数据库理论关系,提出可采用关系代数理论进行配置求解;研究了一般约束满足问题解决配置问题存在的缺点,提出一种动态CSP方法建立产品配置模型,增加用户约束和层次约束,减小搜索空间,提高求解效率。分析常用配置求解方法存在的问题,指出传统的方法只能得到可行的配置方案,难以进行配置优化,并提出一种产品优化配置方法,建立了相关数学模型,提出采用一种多目标遗传算法进行求解,取得了较好的效果。根据生产准备的特点和软件复用的原理,提出一种基于构件复用的产品设计原理框架;研究了构件复用模式下的产品配置管理,分析了产品数据的分层结构,提出了一种将项目关系树结构转换为关系的方法,定义了结构状态和非结构状态的状态规则;研究了配置版本管理模型和版本分类管理方法,提出一种版本管理方法解决面向产品结构的版本管理和产品历史数据的追踪。在制造方面,研究了装配线的基本原理、特征和线平衡等基本理论,说明装配线在大规模定制生产模式下的重要性,并指出装配线调度存在的问题。结合实际企业情况,提出一种多目标优化调度方法,设计了多目标优化的目标函数,采用相关产品信息编码作为在制品唯一表示和装配调度指令,同时提出一种多目标遗传算法进行求解。在工程应用方面,从软件实现的角度分析了基于产品结构配置建模、配置优化、装配线优化调度以及构件复用等关键模块,并将这些研究成果应用于东风汽车和昌河汽车等工程项目中,取得了较好的应用效果。更多还原

【Abstract】 The complex product markets of the twenty-first century will demand the ability to quickly and globally deliver a high variety of customized products. Mass customization (MC) is the key strategy to meet this challenge. MC does not only represent product development, but also represent product manufacturing. Based on several projects, the whole dissertation focuses on MC, chiefly including some key enabling technologies to support the implementation of MC effectively.Configuration-Oriented product model (COPK) is used to describe the composition and components’relationship in the product; it is the basis of product configuration. The characteristics of COPK are analyzed, several chief concepts are defined, and the process of product configuration based on COPK is provided. The multi-objective genetic algorithm is analyzed.Traditional CSP′s limitation of knowledge representation in product configuration is analyzed. A Type of Product Configuration Method Based on Dynamic Constraint Satisfaction Problem (PDCSP) is purposed. It has more capability for representing knowledge and less searching than traditional CSP.In order to solve the problem of multi-objective optimization of production configuration, a multi-objective genetic algorithm is developed, and then the coding and decoding representation of the solution as well as the calculation of the fitness function are designed. The performance comparison of the proposed genetic algorithm with three other genetic algorithms is made. It is shown that the proposed genetic algorithm outperforms the others.Following the idea of software reuse and taking a product of assembly type into consideration, this dissertation put forward the general principles of new product development based on component and architecture reuse, and describes their key techniques such as component/architecture relevance and assembly constraint check. In addition, a MC oriented product version management method is proposed.In order to solve the scheduling problems of automobile assembly line in the environment of mass customization, a multi-objective optimization scheduling method is described and its mathematical formulas are provided; a multi-objective genetic algorithm is designed for finding near-Pareto or Pareto optimal solutions of the problems. And then the coding and decoding representation of the solutions as well as the calculations of fit ness functions are designed. A new genetic evaluation and selection mechanism is proposed.Finally, this dissertation introduces the implantation of product configuration and assembly line scheduling in some projects, which testified the advancement and practicability of research achievements in this dissertation.更多还原