Abstract: To research the group decision-making process of collaborative building and to provide evidence for the design and implementation of collaborative systems,a subgroup decision-making process model with a structure of nearest neighbor coupled network was proposed by using Markov chain and Stochastic Petri Net.The transition probability matrix and its recursive formula were given to achieve the similar structural expansion.Based on the bounded rationality theory,a probability function was applied to describe the changes of subjects'decision-making capacity,and the questionnaire was designed to obtain the function parameters.By taking subgroup decision-making process with three subjects as an example,the state transition probability matrix of the subgroup and the discussion times required to form a consensus were calculated.The effectiveness of the proposed model was verified with Monte Carlo method.

Key words: knowledge management, collaborative building, subgroup decision-making process, Markov chain, stochastic Petri net

摘要： 为研究协同中的群体决策过程,进而为协同系统的设计和实施提供依据,以马尔科夫链和随机Petri网为工具,提出一种结构为最近邻耦合网络的子群决策随机过程模型,并给出了转移概率矩阵及其递推公式,以实现类似结构的扩展。根据有限理性理论,该模型采用一个概率函数描述主体的决策能力,以体现主体决策时的随机特性。以某系统中的三人子群决策过程为例,计算了子群状态转换概率及形成共识所需的讨论次数,并利用蒙特卡洛方法验证了该模型的有效性。

关键词: 知识管理, 协同创作, 子群决策过程, 马尔科夫链, 随机Petri网