Research On Consensus Decision-making Models With Heterogeneous Linguistic Information And Their Application

People advocate giving full play to the collective wisdom in order to improve the accuracy of decision-making.As a result,group decision-making(GDM)could be preferred by the administrator instead of individual decision-making.Traditionally,only a selection process on alternatives,where the preferences of each expert over the alternatives are gathered effectively and the best alternative or subset of alternatives is chosen,has been applied to the solving of GDM problems.However,failure to sufficiently take into account the agreement level among experts,the resolution scheme obtained in this way may not be accepted by all experts within the group.For this reason,consensus reaching process(CRP)was introduced as an additional phase in the resolution of GDM problem.Achievement of opinion consensus is the precondition of performing selection process and the guarantee of obtaining the accepted best alternative or subset of alternatives for the group.Morever,since experts from different fields prefer to characterize the degree of their preferences in a very natural way,the judgments of issues presented by experts are often in the heterogeneous form of linguistic information(for example,multi-granularity linguistic preference information,uncertain linguistic preference information,unbalanced fuzzy linguistic preference information,etc.)Some effective linguistic consensus models based on different CRPs have been proposed to deal with GDM problems in heterogeneous linguistic contexts since the consensus model with the linguistic preference information was first put forward in 1990 by Herrera et al.Now,the research on consensus decision-making with heterogeneous linguistic information has catched the attention from researchers at home and abroad,and the presented models and methods have been widely used in optional alternative selecting,medical diagnosis,emergency decision-making,and other fields.Nevertheless,so far,there exist deficiencies(either in theory or in application)that need to be improved for consensus models and decision-making methods under heterogeneous linguistic information.According to the research ideas from the shallower to the deeper,this thesis on the information aggregation and consensus decision-making under heterogeneous linguistic information environment is mainly divided into six parts as following.(1)A consensus measure problem for heterogeneous group(that is,experts use different linguistic habits and have different degrees of importance)is studied in the process of consensus decision making.From the perspective of a consensus coordinator,the optimistic level of reaching a consensus in the group about preferences over paired comparisons of the objectives can be characterized by a consensus satisfaction degree index of heterogeneous group.An OWA operator based on consensus satisfaction degree,whose weights are established by a novel determining weight model,is presented to aggregate the preference similarities of experts.In the last,a consensus degree calculation method for heterogeneous group is developed.(2)Group consensus decision making problems with multi-granularity linguistic preference information are studied.From two aspects of individual and group,implied importance degree information of experts under preference information is fully exploited,and determined by an optimization model which is constructed based on the individual consistent degree and the similarity degree between individuals and group.A linguistic consensus model is presented,where non-consensus preferences identification and correction process is guided by experts' importance degree.Finally,a GDM method which is a guarantee of reaching a certain degree of group consensus before aggregating expert opinions is developed.(3)Group consensus decision making with uncertain additive linguistic preference information is investigated.The uncertain additive linguistic preferences are transformed into the uncertain 2-tuple preferences.The individual consistency measure and the individual consensus deviation measure are defined,respectively.An optimization model to determine initial weights on experts is established with the help of the above two definitions.Then,the aggregated fuzzy evaluating matrix and the aggregated group preference are constructed by using possibility degree between the two uncertain 2-tuple preferences.The measure of expert cumulative consensus contribution to group and the group consensus measure are developed.By penalizing weights of the least cooperating experts,group consensus is achieved,without forcing the experts to modify their opinions.Finally,a group consensus decision-making approach is developed,and a ranking order of alternatives and the best alternative are obtained.(4)Adaptive consensus decision-making approaches are investigated to deal with GDM problems with two types of linguistic preference contexts.For linguistic group consensus decision-making problems with heterogeneous group(different linguistic importance of experts),two measures of group consensus are defined to reflect respectively the level of agreement among all the experts and the difference degree between the experts' individual opinions and the collective one.Based on the consensus measure indicators,an optimization model of linguistic scales granulation is constructed.An improved PSO algorithm is established to obtain the optimal cutoff points for the linguistic scales.A GDM method is developed,and a ranking order of alternatives and the best alternative are obtained.Then,an adaptive consensus model based on linguistic information granulation is presented for group consensus decision making problems with unbalanced fuzzy linguistic preference information.A granular representation of unbalanced linguistic terms is concerned with the interval formation of a family of information granules over the unit interval.In the case of the cutoffs of information granules unknown,individual consistent degree and group consensus degree of granulation are defined.An optimization model to determine the optimal cutoffs of information granules is established with the help of the above two definitions.Group consensus is achieved by constantly adjusting individual preferences through the optimization of the cutoffs.Finally,a GDM method which is a guarantee of reaching a certain degree of group consensus before aggregating expert opinions is proposed.(5)Linguistic multi-attribute GDM problems with multidimensional preference information in the form of pairwise alternatives and incomplete weight information are studied.A fuzzy regret/rejoice function and a computational formula for the perceived utility of alternative decisions are defined.A perceived utility value-based group consistency index(which reflects the total consistency)and a group inconsistency index(which represents the total inconsistency)for pairwise rankings of alternatives based on regret theory and an a priori multidimensional preference order given by the decision-makers are proposed.Then,under the circumstances of an unknown fuzzy ideal solution,a mathematical programming model to determine the optimal attribute weights and a defuzzified fuzzy ideal solution,with the idea of the Linear Programming Technique for Multidimensional Analysis of Preference(LINMAP),is set up.Finally,a new GDM method considering regret aversion of the decision-makers is developed.The ranking order of alternatives is obtained based on the decision-makers' optimal comprehensive perceived utility values.(6)Applied research on emergency management for emergencies occurring in the production system of steel enterprise.After reviewing research status on emergency GDM methods and design methods of alternative emergency plans,this thesis constructs the design idea of alternative contingency plans for emergencies occurring in the production system of steel enterprise.Based on the linguistic preference information of decision-making experts on the comparison of each two alternatives,the proposed method(granulation adaptive consensus decision-making method)is used to rank several contingency decision-making plans and is verified the rationality and feasibility by methods comparison.