【作者】 朱华炳；

【导师】 刘光复；

【作者基本信息】 合肥工业大学 ， 机械制造及自动化， 2005， 博士

【摘要】 本文结合合肥车桥有限责任公司的H1后桥车间实际规划项目以及安徽省自然基金课题,对制造业生产物流的静态规划和动态调度技术进行了系统地分析和研究,重点研究机械制造业生产物流的规划方法。在分析制造业的不同生产物流系统的基础上,将其本质属性归纳为节点守恒属性、复制属性、约简属性、关系属性和信息映射属性。将物流系统分为决策层、管理层和控制层三个层次及分析指标体系、系统表达体系、评价指标体系等三个体系进行分析研究。 在传统的面向所有行业的玛格数定义基础上,针对机械制造业的物料特点,提出适合机械制造业的OPTIZ玛格系数概念,用于制造系统的当量物流量的计算。 采用基本Petri网对制造物流系统各个子系统进行建模,提出了面向对象的着色赋时Petri网(OCTPN)的物流系统建模方法,该方法既保留了贴近人们思维习惯的图形符号描述法,又改变了因系统复杂所带来的模型复杂状况。同时,对制造加工及物流系统进行集成建模,建立了包含制造加工系统和物流系统的层次化模型,并就汽车后桥车间物流系统规划的实例基于OCTPN方法进行建模和分析。采用E-Factory、Witness软件进行计算机辅助建模和仿真,对H1后桥车间生产物流系统进行设计与优化。 采用遗传算法对车间调度问题分别进行单目标优化和多目标优化,较好地解决了遗传算法在解传统Job Shop问题中因近亲繁殖所导致的早熟现象。研究了AGV调度系统的调度规则及路径规划,采用遗传算法进行路径优化得到较优的行走路线。更多还原

【Abstract】 Combining the project from the H1 rear bridge shop in Hefei Automobile Bridge Co.Ltd and the nature science foundation of Anhui province, this paper analyzes and researches the static planning technologies and the dynamic schedule technologies of production logistics in manufacturing industry, putting emphasis on the research of planning methods of production logistics in mechanical manufacturing industry. Based on the analysis of different production logistics systems of manufacturing industry, the essential attributes of production logistics system are concluded as the node conservation attribute, copy attribute, reduction attribute, relation attribute and information mapping attribute. To analyze and study the logistics system, the system is divide into decision-making hierarchy、 management hierarchy and control hierarchy, and also the analysis index system, the expression system and the assessment index system.Based on the traditional definition of magnified, the definition of OPTIZ magnified which is suitable for mechanical manufacturing industry is put forward and used to calculate the equivalent logistics amount of manufacturing system.Adopting the basic Petri net, the sub-systems models of the manufacturing logistics system are established. The logistics system modeling method with O-O colored timed Petri net (OCTPN) is put forward. This model method not only has the figure and symbol description mode which is coincident with human being’s thinking mode, but also avoids the model complexity caused by the system complexity. At the same time, the manufacturing and logistics system is modeled, the hierarchy model including the manufacturing process model and the logistics system model is established. For the example of logistics planning in the automobile rear bridge shop, the system is modeled and analyzed with OCTPN method. E-Factory and Witness are used to computer-aided model and simulate, to design and optimize the production logistics system in the H1 automobile rear bridge shop.GA is adopted to carry out the single-objective optimization and the multi-objective optimization of the shop schedule. The premature convergence problem of GA caused by inbreeding in solving traditional Job Shop problem is well solved. The schedule rules and route planning of AGV schedule system are studied, and the more optimized routing is obtained with GA.更多还原