健壮的分布式组织网络结构设计的建模与仿真分析  

Modeling and simulation on robust communication network structure of distributed organizations

作　　者：刘海斌[1] 胡斌[1] LIU Hai-bin HU Bin(School of management, Huazhong University of Science and Technology, Wuhan 430074, Chin)

机构地区：[1]华中科技大学管理学院,湖北武汉430074

出　　处：《管理工程学报》2017年第3期155-161,共7页Journal of Industrial Engineering and Engineering Management

基　　金：国家自然科学基金资助面上项目(71271093);国家自然科学基金重点资助项目(71131004)

摘　　要：借助Watts的卡夫曼网络结构模型构建了分布式组织的人际交互网络结构模型,然后基于NK模型和组织学习理论给出了学习绩效的运算。通过一系列的仿真实验检查了不同网络结构下的学习绩效。我们发现半隔离的网络结构是一种最优的模式,它能给组织带来较高的学习绩效,同时我们验证了该网络结构在组织规模、学习速度、动荡环境和成员关系变化情况下具有很强的健壮性。我们发现:(1)在动荡的环境下,一定程度的离职有益于组织的学习并能带来比稳定的成员关系还高的绩效,而且半隔离的网络结构比其他结构在减弱成员离职带来的负作用要强;(2)当个体的学习速度在区间[0.5,0.7]时,学习绩效是最高的,小于0.7时,半分隔的结构都是最优的,但是在高于0.7时,组织的学习绩效独立于沟通结构;(3)动态变化环境下,半分隔的组织结构能让组织保持高的长期学习收益;(4)分布式组织的学习绩效随着规模的增大而变小,任何一种规模下,半隔离的网络结构的绩效也都是最高的。With the development of communication and information technology, an increasing number of organizations are adopting the geographically dispersed structure. For distributed organizations, if each location can coordinately work together, dispersed work has no difference from the traditional organizational model. Because the communication among locations is pivotal to carry out the coordination, we focus on the design of communication structures of distributed organizations. Each location not only has its own internal interaction network, but also has the coordinating network with other locations. Thus, these networks form the model of overall communicating network structure of distributed organizations. From the network structure perspective, distributed organizations have the unique pattern. We use the simulation method to design the optimal personal structure that can give the high learning performance. Firstly, we review the existing relevant literatures that are related to this research topic, identify the research gap, and discuss our research topic. Secondly, we model the interpersonal communication network structures of distributed organizations based on Watts＇ caveman model. We further calculated organizational learning performance based on NK model and organizational learning theory. We design a series of simulation experiments to examine performance under different network structures and other factors. Finally, we find that the semi-isolated structure is an optimal model and can give the highest performance to an organization. Meanwhile, we verified the robustness of this structure in different sizes, learning rates, environmental turbulence, and turnover benefit through a series of simulations. We find that：（1） Under environmental turbulence and a certain rate of turnover benefit, an organization can have higher performance than the stable membership. Moreover, regardless of the degree of turnover benefit the semi-isolated network can help weaken the negative effect and ensure that an organizati

关 键 词：分布式组织 卡夫曼网络 NK模型 组织学习 沟通 

分 类 号：C936[经济管理—管理学]