Research On The Utility Consensus Decision-Making Based On Multi-Attribute Interaction

Multi-attribute consensus decision making is a hot topic in the field of group decision making.It is usually a process in which multiple decision makers reach consensus through multiple rounds of consultation and opinion updating on multiple decision attributes.However,due to the complexity of practical problems,the relationships among multiple attributes are not independent of each other as hypothesized in the existing research,but there are known or unknown correlations that restrict and influence each other.The existence of association relationship between attributes affects the process of opinion adjustment in group decision making,and then has a non-negligible impact on the consensus result.In addition,in the process of consensus reaching,the consensus utility of decision makers is also an important factor affecting the final consensus result,which are related to the smooth implementation of the consensus program.Therefore,this paper will consider the simple pairwise association relationship and complex multi-level association network among multiple attributes in consensus decision making,and discuss the utility consensus decision making problem considering attribute association relationship.The main research work of this paper is as follows:(1)A maximum utility consensus model considering pairwise association of attributes under budget constraints is constructed.This paper studies the consensus decision making problem of how to maximize utility with limited budget under multi-attribute association.Firstly,the minimum cost consensus model based on attribute association is constructed by introducing Choquet integral for the association relationship between multiple attributes.Secondly,considering the decision maker’s utility and consensus cost budget constraints,an attribute associated maximum utility consensus model with budget constraints is constructed.In order to ensure that the model has the optimal solution,an interaction relationship optimization model is constructed to determine the value range of the interaction weight between attributes.Finally,the effectiveness of the consensus model is verified in the context of the contract between the government and the farmers.(2)The maximum utility consensus model considering complex attribute association network under budget constraint is constructed.In the existing complex multi-attribute group decision-making environment,there are often hierarchical and networked relationship among decision attributes,which will affect the adjustment of decision makers’ opinions in the process of consensus formation.Firstly,the DEMATEL-ISM method was introduced to build the maximum similarity optimization model,generate the directed association network between attributes under the group opinion,and determine the hierarchical association network of attributes.Secondly,the weight of each level and attribute weight were calculated based on attribute centrality and cause degree in DEMATEL,and the individual utility function of multi-attribute association was constructed accordingly.Secondly,the utility updating method of decision makers in the process of consensus is first constructed,and based on this,the maximum utility consensus model considering multi-level attribute association is constructed.Finally,the effectiveness of the model is proved by numerical example and comparative analysis.The result of research shows that the pairwise interaction between single-layer attributes has an impact on consensus cost and consensus budget.Among them,the redundant relationship will increase consensus cost,the complementary relationship will reduce consensus cost,and both of them will improve consensus utility.In addition,the stronger the interaction relationship is,the greater the corresponding impact on consensus,and in contrast,the impact of redundant relationship on consensus is greater than that of complementary relationship on consensus.When the scale of attributes is expanded,the unknown complex relationships among attributes can be classified by the improved DEMATEL-ISM method.The multi-level interaction between attributes has the same influence on the opinion adjustment of attributes at the same level.Moreover,due to the constraints of multi-level association,the consensus opinion at this time is often farther away from the upper and lower limits of the opinion interval than the independent attribute,resulting in a smaller total utility.