【作者】 张红宇；

【导师】 高阳；

【作者基本信息】 中南大学 ， 管理科学与工程， 2010， 博士

【摘要】 再制造(remanufacturing)是废旧品的处理方式之一,是废弃物循环利用的重要途径,它对于发展循环经济,创建环境友好型和资源节约型社会,具有重要意义。当前,再制造在我国正步入实际运作阶段,相关生产管理问题日益凸显。现有研究证实,再制造生产系统中的不确定性以及由此引起的可变性,使得传统计划和控制工具难以直接应用。因此,研究不确定性因素影响下的再制造生产计划与调度问题,对实现再制造生产活动的有效管理,推进再制造生产实践,是十分必要的。为此,本研究从通过信息共享以降低再制造中的不确定性影响出发,对再制造信息共享及生产计划与调度进行研究,主要工作和创新如下：(1)针对当前再制造生产的特点和需求,提出了一种面向再制造信息共享的对等网(peer-to-peer,简称P2P)模型ReMIS。为适应再制造网络的高度动态性,模型采用基于自然簇和主题簇构成的两层混合结构,上层由主题对等体组成结构化的主题网络,下层由普通对等体组成P2P网络。ReMIS采用基于主题簇的资源分类和检索方法,该方法在处理海量再制造资源信息的检索请求时,具有较好的查询效率,并能减少再制造生产中的词汇语义冲突。同时,应用发布／订阅技术,可使再制造企业获取实时信息成为可能。查询性能分析表明,ReMIS可满足再制造生产系统的需求。(2)提出了再制造信息共享激励模型。借鉴管理信息系统阶段划分理论,并根据事物发展的规律以及再制造业的发展特点,定义了ReMIS的三个发展阶段。在阶段Ⅰ中,以一个拆卸商和一个加工商为例,应用委托代理理论,构建了激励拆卸商信息共享的激励模型,给出了求解方法。通过对模型进行分析与讨论,说明了激励机制有助于提高拆卸商共享信息的积极性,减少因生产计划变更而增加的额外成本。在阶段Ⅱ中,以回收商为对象,运用多维博弈理论,构建了多个回收商同时进行企业信誉、共享信息和产品市场价格博弈的模型。算例分析表明,企业间的多维博弈可促进企业进行信息共享,而企业在信息共享上的投入,可促进ReMIS的建设,并激励企业获得更大的利润。这为再制造中激励机制的设计提供了理论指导,也为制定推动再制造业发展的措施提供了参考。在阶段Ⅲ中,针对可能出现的道德风险,提出了一种基于信誉机制的规避策略。(3)提出了考虑废旧品可购买量约束的再制造生产计划模型及求解方法。针对现阶段我国再制造企业面临废旧品供应不稳定而且数量严重不足的问题,应用信息熵预测市场供应量,提出了一种考虑废旧品可购买量约束的再制造环境下的单产品生产计划模型,给出了求解方法,并通过共享信息修正和调整后续生产计划,使之动态适应再制造生产的变化性,以优化生产计划。算例分析表明,该方法可用于求解当前废旧零部件最大可购买量有限的再制造生产计划问题,能够及时修正和调整生产计划,以响应外部环境变化。(4)针对工件加工路径的可变性特点,构建了再制造生产中的多目标生产调度模型,提出了一种基于可变长工序编码的改进遗传算法进行求解,设计了异常染色体的识别和重构方法,以及相应的遗传算子。由于采用可变长工序编码,该算法可实现随机工序数目和随机工序顺序情况下再制造生产调度问题的优化求解,仿真实验验证了该算法的可行性和有效性,表明该方法可用于求解再制造生产中Job-Shop调度问题。更多还原

【Abstract】 As one way of waste disposal and important means of cyclic utilization of waste, the remanufacturing is of great significance to develop recycling economy, as well as to create environment-friendly and resource-saving society. At present, the remanufacturing is entering into practice stage in China, so that the problems involved in production management are becoming more and more prominent. The informed research has confirmed that the traditional planning and control tools can not be directly applied to remanufacturing, because of the uncertainty in remanufacturing system and the variability causing by additional uncertainty. Therefore, it is extremely necessary to study the remanufacturing production planning and scheduling under the influence of uncertain factors, which could be useful to manage the production activities effectively and to promote remanufacturing practice powerfully. Consequently, considering the adoption of information sharing to reduce the impact of uncertainty in the remanufacturing, production planning and scheduling in remanufacturing systems are studied in this paper, and the main work and innovation are as follows.(1) A model for remanufacturing information sharing based on P2P network is proposed, which is named ReMIS and aimed at the current characteristics and real demands of remanufacturing production. For the requirements of the highly dynamic remanufacturing network, the model employs two-layer hybrid structure based on physics clusters and topic clusters, in which upper layer is topic network composed of topic peers, lower layer is P2P network composed of common peers. ReMIS adoptes cluster-based resource sort-and-retrieval mechanisms, which can insure high query performance and reduce vocabulary semantic conflicts in remanufacturing production, when dealing with mass retrieval requests about remanufacturing resource information. Additionally, the application of publish/subscribe mechanism makes it possible for remanufacturing enterprise to obtain real-time information. The analysis of query performance indicates that ReMIS network was feasible and highly-efficient, and it can meet the demands of remanufacturing production systems.(2) An incentive model of remanufacturing information sharing is put forward. Employing the stage division theory of management information system, three stages of ReMIS are defined, according to the objective laws of development of things, as well as the development characteristics of remanufacturing industry. At the first stage, an example involved with a disassembly contractor and a processor is proposed. By using principal-agent theory, the incentive model is built to encourage disassembly contractor to share information, and the solution method is given. The analysis and discussion of this model illustrate that incentive mechanism is helpful to improve the enthusiasm of disassembly contractor to share information, and to cut down the added extra cost resulting from the changing of production plan. At the second stage, taking recycler as the object of study, using multi-dimensional game theory, a model is designed in which multi-recyclers carry on game based on enterprise reputation, sharing information and product market price simultaneously. The example analysis shows that multi-dimensional game among the enterprises can promote enterprises to share information, and the inputs of information sharing can speed up the construction of ReMIS, and encourage enterprises to expect greater profits. These can provide theoretical guidance for the design of incentive mechanism in remanufacturing, and provide a reference for establishing measures to accelerate the development of our country’s remanufacturing industry. At the third stage, in view of some potential moral risk, an avoidance strategy is presented on the basis of credit standing mechanism.(3) A production planning model in remanufacturing environment which takes waste purchase capacity constraint into account, and its solution method are proposed. Aimed at the problem that the waste supply from Chinese remanufacturing enterprise is unstable and severely insufficient at the current stage, utilizing information entropy to predict market supply quantity, a single-product production planning model and its solution method is proposed under the remanufacturing environment which takes waste purchase capacity constraint into account. Moreover, through amending and adjusting follow-up production plan based on the shared information, the model is able to dynamically adapt to the changes of remanufacturing production, so the production plan can be optimized. The analysis by example shows that this method is suitable for solving current remanufacturing production planning problem of that purchase capacity of waste parts are limited, and it can timely amend and adjust production plan in response to the changes of the external environment.(4) A remanufacturing production scheduling model and its solution method are put forward. In view of variability of job process path, a multi-objective remanufacturing production scheduling model is proposed, and an improved genetic algorithm (GA) based on variable-length operation encoding method is proposed to solve this model. Then the GA operators and methods to identify chromosomes validity and reconstruct them are designed. Owing to the using of variable-length operation encoding method, the algorithm can achieve the optimal solution of remanufacturing production scheduling problem under the environment of random number of operation and order of operation. In the end, The simulation results verify the feasibility and validity of the algorithm, and shows that this method can be applied to solve Job-Shop scheduling problem for remanufacturing production.更多还原