【作者】 余彪；

【导师】 赵韩；

【作者基本信息】 合肥工业大学 ， 工业工程， 2013， 博士

【摘要】 信息技术、制造技术的快速发展以及经济全球化的不断深入给企业带来了巨大的影响。一方面,在这些新技术的支持下,企业的运行效率得到了极大的提高；但是另一方面,由于这些新技术的推动,客户的需求变得越来越个性化并且全球化竞争使得这种趋势日益明显。在这种情况下,有效即时地处理客户的个性化需求成为企业赢得竞争的关键。大规模定制生产模式正是在这种背景下应运而生。大规模定制的目标是以大规模生产的成本、质量和效率生产个性化的产品,以满足客户的个性化需求。自从这种新的制造模式提出以来,就得到了全世界无数学术界以及工业界研究者们的青睐。大量有效的方法和技术被提出并被应用到大规模定制中并且获得了巨大的成果。尽管如此,大规模定制中还有大量的问题有待研究。本研究在分析当前国内外大规模定制研究现状的基础上,结合产品生命周期管理以及并行工程的相关理念和技术,对传统的客户化产品设计技术-产品配置设计从理论和实现技术上进行了扩展,提出了产品生命周期配置的概念。产品生命周期配置的目标是对所考虑的产品生命周期过程同步进行配置与优化。论文首先研究了产品生命周期配置的架构,在分析其分布式特征的基础上,提出了基于Multi-Agent技术的产品生命周期配置框架。在此框架中,各种参与者通过Agent协同,共同完成对客户订单的配置。针对产品生命周期配置涉及到各种来自不同部门、具有不同知识和教育背景的参与者,他们之间缺少直接交流的专业语言,研究中提出了基于语义WEB本体论的产品生命周期信息模型,从而支持不同的参与者以及软件Agent之间无障碍地沟通交流。除此之外,为了方便产品生命周期配置问题求解,研究中还给出了另外一种基于AND-OR树或者AND-OR图的产品生命周期建模方法。本研究总共研究了四种不同的产品生命周期配置过程,它们分别是：传统的产品配置；客户化产品设计、制造过程规划以及制造资源分配并行配置；客户化产品设计、制造过程规划以及维护过程规划并行配置：客户化产品设计与供应链设计并行配置。针对这四种不同的产品生命周期配置方法,本研究提出了四种不同的求解算法,它们分别是：基于创新自适应量子遗传算法的产品配置、基于多种群协同遗传规划的客户化产品设计、制造过程规划以及制造资源分配并行配置、基于多种群协同遗传规划的客户化产品设计、制造过程规划以及维护过程规划并行配置以及基于并行协同量子遗传算法的客户化产品设计与供应链设计并行配置。研究中以具体的案例对这些算法进行了验证,并且给出了它们与其它算法的对比实验。实验结果表明,研究中提出的算法均具有非常优秀的寻优能力和计算效率。为了验证所提出的产品生命周期配置理论与技术,研究中还实现了一个基于Multi-Agent的产品生命周期配置原型系统。文章最后还以一个家具公司的客户化窗户配置作为应用案例,对基于创新自适应量子遗传算法的产品配置理论以及所实现的原型系统进行了验证和应用。应用结果表明,所提出的理论和方法能够帮助企业有效地提高产品配置效率、客户满意度以及企业业务执行效率,从而提高企业的竞争能力。更多还原

【Abstract】 The development of information technology and manufacturing technology and the deepening of economic globalization have dramatically impacted the competition of enterprises. On the one hand, the efficiency of enterprise has been improved by the supports of the new technologies; on the other hand, the customers’ requirements become more and more personalized due to these new technologies and the global competition accelerates this trend. In this condition, the enterprise has to manage the customers’ individual requirements effectively and timely to win the competition against other competitors.Mass customization emerges under such situations. Its objective is to provide individual products for individual customers’ requirements by the cost, quality and efficiency of mass production. Considerable researches have been received from both academic and industrials in the past decades. A lot of methods and technologies have been proposed to achieve the objective of mass customization. Though these progresses, there are still a lot of problems need to be studied further in this field. In this work, based on the literatures review, a new kind method-product lifecycle configuration for mass customization is proposed to extend the traditional customized product design method-product configuration. By combining the theories of product lifecycle management, concurrent engineering and product configuration, the proposed product lifecycle configuration aims to achieve synchronous configuration and optimization of product lifecycle processes.Firstly, the framework of product lifecycle configuration is analyzed, and due to the distributed nature of product lifecycle configuration, a Multi-Agent based product lifecycle configuration framework is proposed in the work. All participants of product lifecycle configuration can work cooperatively to configure customized products for individual customers’ requirements under this framework.Because the participants of product lifecycle configuration come from different departments and they have different knowledge about the product and different educational backgrounds, they cannot cooperative to work effectively. In this work, a semantic web ontology based product lifecycle information model is introduced to help the participants for their communications, and also the software Agents can communicate by the ontology. And for problems solving in product lifecycle configuration, an AND-OR tree or AND-OR graph based model is also proposed for the product lifecycle information modeling.For the product lifecycle configuration, four kinds of configuration processes are considered in this work, and they are traditional product configuration, concurrent configuration of customized product design, manufacturing process planning and manufacturing resource allocation, concurrent configuration of customized product design, manufacturing process planning and maintenance process planning, and concurrent configuration of customized product design and supply chain design.To solve the four kinds of product lifecycle configurations effectively, four kinds of algorithms are proposed, and they are novel adaptive quantum genetic algorithm based product configuration, multi-population based cooperative genetic programming based concurrent configuration of customized product design, manufacturing process planning and manufacturing resource allocation, multi-population based cooperative genetic programming based concurrent configuration of customized product design, manufacturing process planning and maintenance process planning, and parallel cooperative quantum genetic algorithm based concurrent configuration of customized product design and supply chain design. Case studies are also conducted to demonstrate the effectiveness of proposed algorithms. And to illustrate the advantages of the proposed algorithms, four groups of comparative experiments are created for each algorithm. The experimental results show that all the proposed algorithms have very good search abilities for optimality and high efficiency on computing.To demonstrate the effectiveness of the proposed methodologies for product lifecycle configuration, a Multi-Agent based prototype system for product lifecycle configuration is implemented by Java language in this work. And at last, an industrial case study about a customized window in a furniture company is used as an application of the prototype system and also is used to validate the effectiveness of proposed product configuration methods. The results of the application show that the proposed theories and methods can improve the efficiency of the configuration design, the customer satisfaction and the business process of the enterprises effectively, and further to promote the abilities of competitions.更多还原