【作者】 聂书志；

【导师】 叶邦彦；

【作者基本信息】 华南理工大学 ， 机械制造及其自动化， 2010， 博士

【摘要】 在全球化和网络化的制造新环境下,网络化协同制造资源的优化配置是实现异地协同制造的关键技术,已经成为现代集成制造研究领域的热点问题之一。本文以网络化协同制造资源优化配置为背景,以制造资源为核心,深入研究制造资源对象的信息管理和建模问题,重点研究基于DNA遗传算法的制造资源的智能优化仿真问题,为实现制造资源的优化配置奠定了理论基础,并建立基于DNA遗传算法的网络化协同制造资源优化配置系统。主要研究内容如下:(1)针对协同制造资源优化配置的理论问题进行研究。由于目前的协同制造资源优化配置尚没有完善的和普遍适用的理论基础,从概念和过程等方面对制造资源进行系统的、多层次的基础理论研究,将协同制造的资源优化配置技术分为协同制造资源信息模型、协同制造资源优化配置模型、协同制造资源优化仿真算法三类,在此基础上,建立以制造资源为核心的制造资源优化配置概念模型和过程模型,开展制造资源的管理和基于UML的建模工作。(2)研究网络化协同制造资源的优化配置技术和评价指标。由于协同制造任务的复杂性,将制造资源优化配置技术从更为细微的资源能力信息层次来研究制造资源的选择,并建立更加科学化和精确化的定性指标评价方法,研究基于定量优化指标的面向候选制造资源集合的优化配置技术,将制造资源优化配置问题归结为传统制造模式下的多目标车间制造资源调度技术、虚拟企业或者动态联盟的伙伴选择技术以及网络化协同制造资源优化配置技术三个基本问题,并给出相应的求解方法。(3)研究基于DNA遗传算法的协同制造资源优化配置仿真算法。针对普通遗传算法存在编码简单且局部寻优能力差等特点,重点研究采用两类基于DNA计算的改进遗传算法来求解协同制造资源优化配置问题,RNA-SIGA算法采用四进制编码方式,使用分区选择方法,设计分区动态自适应的交叉、变异等操作;DNA-DIGA算法采用双单链编码方式,使用基于Pareto排序的个体密度选择操作,设计分类渐变和突变的交叉、变异等操作,通过推理证明两类算法的全局收敛性,接着采用典型测试函数对两类算法的性能进行仿真测试,并分别利用遗传算法和设计的两类算法对比求解三类优化配置问题,优化结果证明了算法的有效性。(4)研究网络化协同制造资源优化配置系统的设计与开发。提出协同制造资源优化配置系统的体系结构和业务流程,完成系统的功能设计与开发,实现了基于DNA遗传算法的协同制造资源优化配置技术的集成。分别以多目标车间协同资源优化、协同伙伴优选和网络化协同资源优化配置为例,给出制造资源优化配置实例的应用流程和运行结果,系统地验证本文所提出的理论、模型和方法的有效性。更多还原

【Abstract】 In the new globalization and network manufacturing environment, network collaborative manufacturing resources optimization deployment is the key technology to realize distributed collaborative manufacture; it has become one of the hot topics in the study field of modern integrated manufacturing. This study takes network collaborative manufacturing resources optimization deployment as a background, based on the manufacturing resource as the core, made a deep research about the information management and modeling problems of the manufacturing resources objects, mainly studied the Intelligent optimization simulation problems of the manufacturing resources which based on DNA genetic algorithm, laid a theoretical foundation for the optimize configuration of the manufacturing resources, and established network collaborative manufacturing resources optimization deployment system based on DNA genetic algorithm. The main research contents are as follows:(1) Study the theory basis of the collaborative manufacturing resources optimization deployment. Because there is no perfect and universally applicable theoretical basis for the current collaborative manufacturing resources optimization deployment, made systematic and multilevel theory basis research for the manufacturing resources from the aspects of concept and process; divided the collaborative manufacturing resources optimization deployment into three fields: collaborative manufacturing resources information model, collaborative manufacturing resources optimize configuration model and collaborative manufacturing resources optimize simulation algorithm, then established manufacturing resources optimization deployment conceptual model and process model with the manufacturing resources as the core; worked on the management of manufacturing resources and the model establishment based on UML.(2) Study on evaluation index and optimize configuration technology of network collaborative manufacturing resources. Due to the complexity of collaborative manufacturing, study the manufacturing resource selection from the more subtle resource capacity information levels, and establish more scientific and accurate qualitative index evaluation method, studied the optimize configuration technology--candidates for manufacturing resources collection which based on quantitative optimization index, attributed the manufacturing resources optimization allocation problem into three basic problems: the multi-objective workshops manufacturing resources scheduling technique in the traditional manufacturing model; the partner selection technology of the virtual enterprise or dynamic union; the network collaborative manufacturing resources optimization deployment; and presented the corresponding solving methods.(3) Study on simulation algorithm of collaborative manufacturing resources optimization deployment based on DNA genetic algorithm. For genetic algorithm has simple code and poor ability for local research, this paper mainly studies the method of using two kinds of improved genetic algorithm based on DNA computing to solve the problems of collaborative manufacturing resources optimization deployment. RNA-SIGA algorithm adopt four-digit-system encoding, use zoning selection method, design dynamically adaptive crossover and mutation operations; DNA-DIGA algorithm adopt sing-dual chain coding method, use the choose operation based on Pareto individual density sorting, design the crossover and mutation operations of classification gradient and mutations, proved the global convergence of these two algorithms by reasoning, using typical test function to make simulation test for these two kinds of algorithm, and by respectively using the genetic algorithm and the design algorithm to contrast solve three kinds of optimize configuration problems, the optimization results proved the effectiveness of the algorithm.(4) Study on design and development of network collaborative manufacturing resources optimization deployment system. Put forward the system structure and business process of network collaborative manufacturing resources optimization deployment system; complete the functions design and development of the system; realize the integration of the collaborative manufacturing resources optimization deployment which based on DNA genetic algorithm. Respectively set the multi-objective workshops collaborative resources optimization, collaborative partner selection and network collaborative manufacturing resources optimization deployment as the example, given the application process and operation results for the manufacturing resources optimization allocation, systemically verify the effectiveness of the theories, models and methods which proposed in this paper.更多还原