SDN供需流的两阶段及其优化分析——基于木地板生产企业的视角

在传统企业向供需网企业转变过程中,作为SDN的一个供需流,其核心集中在对车间生产调度及物流配送的优化上.针对这两个问题,基于木地板生产企业的现状,本文引入了量子粒子群与模拟退火相结合的混合算法,以及一种求解物流系统的整合优化模型与求解的启发式算法分别对其进行研究,具有一定的实用价值.     

求解大规模生产批量问题的启发式算法

企业资源优化模型是多物料、多层、受多种能力约束、有启动时间和启动成本的生产批量问题,该问题是NP完全问题,求解十分困难。为此我们提出了一个新的启发式方法,通过交互求解线性规划松弛问题并应用改进的Silver-Meal方法处理批量来近似求解生产批量问题,并第一次将影子价格引入Silver-Meal方法的批量决策,数值实验表明新算法在不同规模问题上的有较好的表现。     

准时生产方式下混流装配线的调度问题

混流装配线的调度问题是 JIT生产方式中的一个重要问题 .本文建立了多级混流装配线的调度模型 ,其目标是使不同生产级上各零部件的消耗量尽可能保持均匀 ,并用遗传算法和模拟退火进行求解 .试验表明 :遗传算法和模拟退火算法的求解质量比“目标追随法”有所提高     

供应链中二级分销网络的优化设计模型

从供应链的集成和协调的角度出发 ,提出了在多工厂、多产品、多客户环境下 ,考虑需求分配的二级分销网络优化设计模型 .模型体现了在设计分销网络时 ,根据制造企业各分厂的生产能力和各客户区 (需求地 )对不同产品的需求情况 ,合理的将对产品的需求分配到各个分厂 ,制定相应的生产计划 ,以降低生产和分销环节的总费用 .文中还讨论了采用启发式算法同传统的分枝定界法相结合以提高问题的求解速度 ,并给出了一个数值例子 .     

带有多级制造环节的供应链能力分配方法及应用

针对生产过程具有多级制造环节的供应链能力分配问题,在考虑生产能力、市场容量、合同需求等约束的基础上,建立了混合整数非线性规划数学模型.基于约束满足技术进行问题求解,采用先确定后逆向回溯赋值的机制保证解的可行性.通过仿真实验和应用案例验证了模型与算法的可行性和有效性.     

基于约束的离散粒子群算法求解大规模指派问题

针对现有进化算法在求解传统指派问题时因取整而影响优化效果的问题,采用了一种基于AllDifferent约束的置换离散粒子群优化算法,该算法针对指派问题中各变量不能重复取值的特点,改进了算法的迭代方式,并引入了模拟退火的差解接受准则以提高优化效果,仿真算例表明改进后的算法在质量上和时间上更具有效性.     

两阶段法求解混装模式下的加热炉调度

针对目前加热炉调度模型少有考虑混装模式下加热炉调度优化的不足,建立了连铸-热轧混装-体化模式下的加热炉生产调度优化模型,并提出了基于贪婪算法和模拟退火算法的两阶段求解方法.生产数据测试表明该模型和算法能有效解决加热炉调度问题.     

蚂蚁算法在组合优化中的应用

蚂蚁算法是近年来新出现的一种随机型搜索寻优算法 ,自从在 TSP等著名问题中得到富有成效的应用之后 ,已引起越来越多的关注和重视 .本文进一步将这种新型的生物优化思想扩展到其他一些组合优化难题 ,包括目前尚缺乏有效求解手段的多目标组合优化问题 ,从实验上探索了蚂蚁算法的优化能力 ,获得了满意的效果     

公共服务平台下虚拟联盟成员选择机制及联盟企业间协同制造问题研究

本文主要研究了公共服务平台下虚拟联盟成员选择和联盟企业间协同生产的联合优化问题。该问题主要涉及到虚拟联盟战略层面和运营层面两个层面,旨在实现虚拟联盟成员选择、订单分配和生产排序等三个阶段的协同优化。在分析单一联盟企业的生产排序最优化性质的基础上,本文给出了涉及到顾客订单分配和联盟成员选择的结构化性质。考虑到联合优化问题的复杂性,引入并拓展了用于求解连续优化问题的树种算法,开发了嵌入二维离散编码策略和离散种子生成规则的离散树种算法,给出了结合前述结构化性质的适应度计算方法,并开展了一系列仿真对比试验,实验结果表明所提算法相较于离散萤火虫算法和集合蛙跳算法等离散智能优化算法而言,在求解质量和鲁棒性上具有一定的优越性。本文的主要创新点在于提出了横跨战略层面和运营层面的虚拟联盟协同模型,开发了新颖的离散树种算法,能够为虚拟制造联盟的组建和运行提供一定的指导意义。     

粒子群算法求解无能力约束生产批量计划问题

经典的粒子群优化算法是一个在连续的定义域内搜索数值函数极值的有效方法.目前,粒子群算法(particle swarm optimization,PS0)已经成为优化领域中的一个重要的优化工具,其应用在很多优化问题中都可以见到.虽然粒子群算法的应用范围已经十分广泛,但是关于应用其求解多级生产批量计划问题(multilevel lot-sizing problem,MLLs)的文章并不多见.文章提出结合遗传算法(genetic algorithm,GA)变异算子的混合粒子群优化算法(hybrid particle swarm optimizatjon,HPSO)求解无能力约束装配结构MLLS问题.通过实验验证了算法的可行性和有效性.     

基于遗传算法的虚拟企业协同资源优化问题研究

面向复杂零件的协同制造,以工艺流程为核心将协同制造任务进行分解,并有效利用"逻辑制造单元"和"逻辑加工路线"等概念描述复杂零件的协同制造任务,最终目标是形成基于复杂零件工艺流程的、可支撑异地协同生产的加工路线。对复杂零件协同制造的制造资源优化配置问题进行了数学分析和描述,建立了问题的目标函数与约束条件。本文以加工时间、运输费用和加工质量作为目标,约束条件包括顺序约束、释放期约束、时间约束、交货期约束、成本约束和质量约束,最终将资源优化配置问题归结为多目标优化问题,并利用遗传算法进行求解,得到了较为满意的结果。通过实例分析,将模型应用于某型号发动机叶片协同制造,说明采用本论文的模型可以有效解决复杂零件协同制造的资源优化配置问题。     

不完全信息下多阶段移动应用产品与服务合作的可更新契约

考虑不完全信息下由提供移动应用产品与服务的ERP厂商、APP厂商与客户企业组成的产品服务供应链,针对ERP厂商与APP厂商产品服务合作中存在的逆向选择与道德风险问题,构建两阶段的动态博弈模型,分析了APP厂商隐藏与披露能力信息对产品与服务质量、供应链系统绩效与ERP厂商利润的影响,提出了信息披露、收益共享的单阶段契约与存在契约变更的两阶段契约激励策略,比较了两种契约的价值。研究表明：通过在产品开发阶段观察APP厂商的契约菜单选择以及获得的产品质量信息,甄别APP厂商的能力信息,在服务阶段进行契约变更,有利于提高产品与服务质量,降低ERP厂商风险,使供应链系统利润与ERP厂商自身利润实现帕累托改善。最后,通过算例对研究结论进行了直观考察和说明。     

How can David orchestrate resources to enhance firm performance? A dynamic approach to coping with resource constraints

Given that the possession of abundant resources does not always guarantee superior firm performance, this study proposes a dynamic approach to enhancing firm value. Building on the resource orchestration thesis, we investigate how firms facing resource constraints enhance their capabilities over time by orchestrating resources and how this resource orchestration can play a role in improving firm performance. Using the dynamic perspective, this study identifies four resource orchestration modes, which are specified based on how a given firm dynamically allocates its resources to technology (enacted by R&D activities) and marketing (enacted by advertising activities). Accordingly, we claim that firms with resource constraints can enhance their performance by escalating or altering their resources over time. To test our hypotheses, we used a sample of 4078 small and medium-sized enterprises in manufacturing industries between 1984 and 2018. We found that focus escalation for technology positively affects firm performance, captured by Tobin's q, and that focus alternation toward either technology or market is positively related to firm performance. Finally, we discuss the theoretical and practical implications of our findings.     

企业家提升企业创新能力的路径研究

企业家是中小企业最为核心的要素之一,是创新的发起人和关键支持者,对创新能力的培育和提升起着决定性作用.现有对企业家和创新能力的研究均强调了企业家的重要作用,但对于企业家如何提升企业创新能力未能给出有力解释.本文基于创新过程,结合企业家的职能,提出企业家提升创新能力的四条路径:清晰科学的战略规划,引导创新方向;整合、配置资源,提供组织制度保障,建立创新基础;鼓励学习和知识交流,培育组织能力:构建创新型文化,提高员工创新意愿和创新积极性.本文并采用单案例研究方法,选取浙江省某中小企业为例,具体分析了企业家提升创新能力的四条路径在企业中的实践.     

Configurational conditions and Sustained Competitive Advantage: A fsQCA approach

Traditional manufacturing firms in emerging markets such as China are increasingly facing the challenge of digital transformation for Sustained Competitive Advantage (SCA). Underpinned by the resource orchestration theory, this study posits that organisational resources (relational and technological), transformation capability, and organisational characteristics are configurational conditions for SCA. Using fuzzy-set Qualitative Comparative Analysis (fsQCA), this study reveals multiple pathways – configurational resources and capabilities in sets of necessary and sufficient conditions – leading to SCA in a sample of 206 Chinese clothing manufacturing firms. Results suggest that relational resources are a necessary condition that fundamentally drives SCA; small firms can achieve SCA by combining relational resources and transformation capability, while portfolio technological resources are sufficient for large firms. For organisations with higher levels of global market engagement, both relational and portfolio technological resources are essential for achieving SCA. This study enhances understanding of the interdependence of the causal conditions (e.g., organisational resources and capability) in set relations with the outcome of SCA in different firms considering their size and level of global market engagement.     

Manufacturing Practices and Strategy Integration: Effects on Cost Efficiency,Flexibility, and Market‐Based Performance

Manufacturing plant managers have sought performance improvements through implementing best practices discussed in World Class Manufacturing literature. However, our collective understanding of linkages between practices and performance remains incomplete. This study seeks a more complete theory, advancing the idea that strategy integration and enhanced manufacturing capabilities such as cost efficiency and flexibility serve as intermediaries by which practices affect performance. Hypotheses related to this thesis are tested using data from 57 North American manufacturing plants that are past winners and finalists in Industry Week's“America's Best” competition ( Drickhamer, 2001 ). The results suggest that strategy integration plays a strong, central role in the creation of manufacturing cost efficiency and new product flexibility capabilities. Furthermore, strategy integration moderates the influences of product‐process development, supplier relationship management, workforce development, just‐in‐time flow, and process quality management practices on certain manufacturing capabilities. In turn, manufacturing cost efficiency and new product flexibility capabilities mediate the influence of strategy integration on market‐based performance. These findings have implications for practice and for future research.     

基于IT价值模型的企业信息化诊断研究

根据IT价值模型,本文从IT能力、流程绩效、动态能力和企业绩效四个层次构建了企业信息化诊断指标体系。在此基础上,可拓评价方法被用来判断各个指标状态和价值创造各个阶段的综合水平。结合评价结果和IT价值生成路径,企业可以对信息化实施的薄弱环节进行诊断。最后给出了一个具体的算例。     

Measuring Dynamic Capabilities: Practices and Performance in Semiconductor Manufacturing

Research on dynamic capabilities emphasizes the importance and role of organizational routines in explaining interfirm differences in performance. While performance differences are well documented, few empirical analyses explore the processes inside organizations that lead to dynamic capabilities or attempt to define and measure their performance effects. This paper examines one type of dynamic capability – the development and introduction of new process technologies in semiconductor manufacturing. This dynamic capability is an important source of competitiveness in the semiconductor industry, given the short product lifecycles, rapid price declines, and rapid technological advances that define the industry. Because much of the knowledge that underpins semiconductor manufacturing is idiosyncratic, firm-level R&D organization and information technology practices that facilitate problem solving and learning-based improvement provide important and enduring advantages. We derive models of the rate of improvement in manufacturing yield (i.e. the quality of production) and cycle time (i.e. the speed of production) following the development and introduction of new process technologies in manufacturing facilities, and test the empirical specifications of these models. The ways in which semiconductor manufacturers accumulate experience and articulate and codify knowledge within the manufacturing environment build new process development and introduction dynamic capabilities that improve performance.     

Production planning of mixed-model assembly lines: a heuristic mixed integer programming based approach

Process technology capabilities are becoming increasingly important as flexible manufacturing continues to be more prevalent, and as competition compels companies to provide expanded variety, at ever lower cost, so introducing plant and processes technological constraints. Model flexibility can also benefit from an appropriate production planning process, especially concerning mixed-model assembly lines, since it can facilitate master scheduling and line balancing activities, which are essential aspects of flexibility. Robust and practical planning approaches have to take into account two different aspects: the first consists in ensuring that the elaborated aggregate plan can be disaggregated into at least one detailed feasible plan for the realised demand, whereas the second in ensuring that this detailed plan is feasible at the operational level. This article faces the model flexibility challenge, reviewing and discussing the planning problem of a real world assembly manufacturing system, producing high volume and a variety of agricultural tractors and machines, analysing and resolving some important issues related to technological, organisational and managerial constraints. This article illustrates the implementation of an Advanced Planning System integrated with a mixed integer-programming model, which is solved by a new iterative heuristic approach capable of achieving interesting planning improvements for model-flexibility management.     

基于混合蚁群算法的生产系统设施规划问题研究

设施规划问题主要研究生产设备的布局规划,从而减小厂区内的物料搬运成本。一个有效的设施规划有利于生产过程中整体运作效率的提高。随着市场竞争的日趋激烈,市场环境处于不断的变化之中,制造企业需不断对设施布局进行重新规划来适应不断变化的市场环境对产品需求量的影响,并达到降低成本的目的。这一问题便需要用多阶段设施规划（MFLP）的方法来解决。本文提出了一种改进的混和蚁群算法（HACO）来解决带有财务预算约束的多阶段设施规划问题,并将此方法与其他一些典型的启发式算法进行了对比分析。结果表明,本文提出的HACO算法是求解带有财务预算约束的MFLP问题的一种有效的方法。