知识转移对群体决策偏好收敛速度的影响  

Effect of Knowledge Propagation on Preference Convergence in Group Decision-Making

作　　者：杨雷[1] 姜明月[1] 

机构地区：[1]华南理工大学工商管理学院,广东广州510640

出　　处：《工业工程》2010年第3期6-12,共7页Industrial Engineering Journal

基　　金：广东省自然科学基金资助项目(8151064101000007);广东省教育厅人文社会科学研究创新团队项目(B16N908002f)

摘　　要：群体决策往往是一个动态过程,个体间的知识转移会使决策者的偏好改变,因此对决策过程进行动态考量更切合实际。提出了基于惯性粒子群算法的偏好收敛模型,探讨知识转移对群体决策动态偏好收敛过程的影响。将知识分为内部自有知识和外部转移知识,分别讨论了这两类知识及不同知识转移量对偏好收敛速度产生的影响,并对比分析了确定与随机因素条件下收敛次数的差异。结果表明在其他条件相同的情况下,如果群体内知识转移速度越大,那么群体最终达成观点一致性的程度越高。随机性模型比确定性模型经过更多次的仿真步长才能达到群体观点一致收敛。计算实例表明本文提出的偏好收敛动态模型可应用于员工绩效考核,能够为动态群体决策偏好收敛分析和评价提供依据。Group decision-making is a dynamic process and knowledge propagation in the group members can change the preference of the decision maker.Based on inertia particle swarm optimization（PSO） algorithm,a dynamic preference convergence model is proposed to explore the effect of knowledge propagation on the dynamic preference convergence process.Knowledge is divided into internal and external knowledge.With the model,the effect of different types of knowledge and their propagation capability on preference convergence is analyzed by simulation.Simulation results show that the higher the speed of knowledge propagation is,the more consistent the decision made is in the group.It also shows that,under a stochastic environment,it needs more simulation steps to achieve consistent convergence.A case study of employee performance assessment is presented to show the application of the proposed method.

关 键 词：动态群体决策 知识转移 惯性粒子群优化算法 偏好收敛 

分 类 号：C934[经济管理—管理学]