Research On Consensus Reaching Method By Considering Personalized Individual Semantics And Consistency Control

Group decision making(GDM)widely exists in the economic,political,cultural and other fields of human society,and has become one of the daily activities of human beings.Due to the uncertainty of the decision making environment and the ambiguity of the human cognition,decision makers are more and more inclined to use linguistic terms to express their opinions in GDM,which is linguistic GDM problems.In linguistic GDM problems,different decision makers may have different understandings of the same linguistic term,i.e.,personalized individualized semantics(PISs).Modeling decision makers’ PISs in linguistic GDM can obtain more reasonable and realistic decision making results.In PIS-based linguistic GDM problems,group consensus and individual consistency are two important research issues.However,most of the existing consensus reaching models only consider how to adjust decision makers’ preference relations to promote group consensus and ignore the individual consistency,which results in that individual consistency may be destroyed by using these consensus reaching models and further leads to unreasonable decision making results.To do so,this thesis aims to propose some consistency control and consensus reaching methods in PIS-based linguistic GDM with linguistic preference relations(LPRs).The main work of this thesis is summarized as follows:(1)For the problems existing in the optimization model of determining decision makers’ PISs,a consistency improvement model by considering PISs is proposed.To be specific,aiming at minimizing the deviation between decision maker’s initial and adjusted LPR,a consistency improvement model based on minimum adjustment and PISs is established.By solving this model,an LPR with acceptable additive consistency can be derived.(2)For linguistic GDM problems with LPRs,a consensus reaching model by considering PISs and consistency control is proposed.First,in light of the PIS-based consistency improvement model proposed in this thesis,an LPR with acceptable additive consistency is obtained.Afterwards,an optimization model of determining decision makers’ PISs is utilized to transform LPRs into fuzzy preference relations and then consensus levels are measured.On this basis,to meet different needs of different decision making problems,an automated feedback adjustment method and an interactive feedback adjustment method by considering PISs and consistency control are designed to ensure that LPRs’ individual consistency levels will not be destroyed while improving the group consensus level.(3)The implementation process of the proposed models and algorithms are illustrated in detail by taking electric vehicle charging station selection problems as an example.Furthermore,through simulation experiments,the proposed algorithms are compared with two algorithms which do not consider PISs and consistency control,respectively.Experimental results show that the proposed two algorithms can improve the group consensus level while avoiding destroying LPRs’ individual consistency levels.The proposed models and algorithms enrich the research of PIS-based linguistic GDM problems,which can be utilized to deal with some practical decision making problems,such as enterprise investment plan selection,electric vehicle charging station selection and supplier selection.