复杂研发项目流程和组织的建模与优化研究

发布时间：2018-01-24 22:07

本文关键词： 复杂研发项目项目管理设计结构矩阵(DISM) 多领域矩阵(MDM)优化聚类　出处：《北京科技大学》2016年博士论文　论文类型：学位论文

【摘要】：当今,在技术和市场的双重驱动下,新产品的功能越来越强大,部件越来越复杂,产品的复杂性可以更好地满足消费者的需求并给企业带来商业契机,然而,产品的复杂性导致研发流程更加复杂,流程的复杂性导致组织结构和团队间的沟通也变得更为复杂。因此,对于复杂研发项目而言,困难不仅仅来自于设计任务本身,也来自于管理的复杂性,如何高效地管理研发项目部件、流程和组织之间的复杂关系,对于研发项目的成功起着至关重要的作用。本文基于设计结构矩阵(DSM)和质量功能展开(QFD)等方法,对复杂研发项目中的流程和组织进行建模与优化,主要研究内容如下：(1)为分析复杂研发项目的功能、部件和流程之间的耦合依赖关系,提出了将质量功能展开法和设计结构矩阵相结合,构建了基于QFD和DSM的研发项目多级瀑布模型,以分析关键设计参数的变更对产品功能、部件及流程的影响。首先,建立了产品功能和关键设计参数之间关系的质量屋,分析关键设计参数变更对产品功能的影响程度,确定功能之间的依赖关系,由此构建功能DSM模型,实现从QFD到功能DSM的量化转换,将产品研发与前端市场需求相结合,有利于提高新产品市场认可度。其次,基于多领域矩阵(MDM)进行跨领域元素之间依赖关系建模研究,构建“功能-部件-流程”跨领域元素之间相互映射关系准则和依赖度测量模型,得到反映流程间耦合关系的流程DSM。(2)进一步,对采用上述瀑布模型得到的流程DSM进行排序优化,综合考虑流程优化过程中可能对优化结果产生影响的多种因素,不仅考虑了反馈点数和反馈距离,也考虑迭代过程中返工概率、返工影响强度对项目的影响,及活动的成本和持续时间对排序结果的影响,以最小化“总协调成本”为目标函数,运用遗传算法并采用Matlab软件编程,对流程DSM进行排序优化,并采用DSM仿真方法评估项目优化后的效果,结果表明,本文所采用的优化方法可以有效降低项目成本,缩短工期,提高项目绩效。(3)设计活动间的重叠是研发项目的基本特征,重叠可以缩短工期,但也会导致返工风险,并增加与重叠并行性相关的成本。为探索重叠活动间的多次沟通对返工风险的影响,本文构建了基于多次沟通-风险的重叠模型,根据重叠活动间上下游活动的具体特性,建立了上游活动的信息变更率和下游活动的返工影响敏感度函数,从而探索多次沟通对返工风险的影响,在此基础上构建了活动返工风险和返工量计算模型及与重叠进行性相关的成本模型,并以成本节约值最大化为目标函数,求解得得到重叠活动间最佳的沟通次数和沟通间隔。进一步,建立描述项目所有活动间祸合关系的流程DSM矩阵,采用遗传算法对多次沟通-风险的重叠模式下的流程DSM进行排序优化并对优化结果进行仿真分析。(4)为分析产品功能、部件、技术措施及执行团队之间错综复杂的依赖关系,提出了扩展的域映射矩阵(EDMM)的概念,即在传统功能和部件DMM的基础上,考虑技术措施对功能和部件的影响,构建基于技术措施的功能-部件EDMM,并根据技术措施和功能、部件之间的关系,推导技术措施之间的间接依赖关系。进一步,改进了组织DSM聚类方法,通过对组织架构的聚类优化,使得类内的信息沟通最大化、沟通成本最小化,而不同类之间沟通最小化,从而形成利于沟通的组织模式,降低组织的复杂性。
[Abstract]:Today, in the dual drive technology and market, more and more powerful components of new products, more and more complex, the complexity of products can better meet the needs of consumers and bring business opportunities, however, because of the complexity of product development process more complicated process because of the complexity of organizational structure and team communication to become more complex. Therefore, the complex development project, not only from the difficult design task itself, but also from the complexity of management, how to effectively manage the project components, the complex relationship between process and organization, for project success plays a vital role. In this paper, based on design structure matrix (DSM) and Quality Function Deployment (QFD) method, modeling and optimization of complex development project in the process and organization, the main research contents are as follows: (1) for the analysis of complex R & D project function, coupling between components and process dependencies, the quality function deployment method and design structure matrix combination, constructs the QFD and DSM projects of multi-level waterfall model based on the analysis of key design parameters change to the function of products, components and processes. First, establish the house of quality the relationship between product function and key design parameters, analysis of key design parameters change on the degree of influence of product function, determine the dependencies between functions, thus constructing the function model of DSM, realize the conversion from QFD to quantify the function of DSM, combining the front end product development and market demand, improve new product market acceptance. Secondly, based on the multi domain matrix (MDM) modeling of the dependence between the cross domain elements, constructing "the criterion of mutual mapping between functional components and process of cross domain relation of elements And the dependence of the measurement model, the coupling relationship between the reflected process flow of DSM. (2) further, sorting optimization of the waterfall model obtained by adopting the process of DSM, a variety of factors which may influence the optimization process considering the results of the optimization process, not only consider the number of feedback and feedback distance, also consider the rework probability iterative process in effect, rework impact on the project, and the effect of activity cost and duration of the ranking results, to minimize the total cost of coordination "as the objective function, using genetic algorithm and using Matlab software programming, sort the optimization process of DSM, and uses DSM simulation method to evaluate the effect of optimization project. The results show that and the optimization method adopted in this paper can effectively reduce the project cost, shorten the construction period, improve project performance. (3) the design activity overlapping is a basic feature of research and development projects Overlap, can shorten the construction period, but also lead to rework risk and increase the cost, and overlapping related. In order to explore the communication between multiple overlapping activities of rework risk, this paper constructs the overlap model based on risk - repeated communication, according to the specific characteristics of overlapping activities between upstream and downstream activities, establish rework effect of sensitivity function of upstream activity information change rate and downstream activities, so as to explore the multiple effects of communication rework risk, on the basis of the cost model and the model with overlapping of related activities and the amount of rework rework risk calculation, and to save the cost of value maximization as the objective function, solving get overlap among the activities of the best communication frequency and communication interval. Further, the process of DSM matrix to describe the coupling relationship between all project activities, using genetic algorithm to repeatedly communicate risk overlap Under the mode of process DSM ranking optimization and simulation analysis of the optimization results. (4) for the analysis of product features, components, perplexing dependencies between technical measures and implementation team, proposed domain mapping matrix expansion (EDMM) concept, which is based on the traditional functions and components of DMM, considering the influence of Technology measures on the function and components of the building components technical measures function based on EDMM, and according to the technical measures and functions, the relationship between parts, between the derivation of technical measures of indirect dependency relations. Further, improved DSM clustering method, the clustering optimization of organizational structure, make class communication of information maximization communication, cost minimization, and minimize the communication between different classes, so as to form the organization mode to facilitate communication, reduce the complexity of the organization.

【学位授予单位】：北京科技大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：F273.2

【相似文献】