Research On Decision Model For Non-definite Multi-attribute Decision And Its Application

In the social activity, existing massive multi-attributes decision making problem. The multi-attributes decision making is important component in the modern policy decision science. It studies the following question:choosing the best plan from limited plans under many criterions. Until now, the people have already constructed many models for processing multi-objective decision making question, but the majority decision models are only suitable for the following policy-making background:When policy-markers exactly know various plans evaluation value under each evaluation attribute. However, in actual policy-making question, because the realistic question is complex or the information acquisition cost, the decision information is not precise, incomplete, fuzzy, in addition the policy-maker knowledge or information-handling capacity is limited, the policy-marker often only wants or can only supply the incomplete evaluation information. The above reasons causes sometimes to gain the precise decision information is very difficultly, even impossible. The people call this kind of multi-attributes decision making problem as the indefinite multi-attributes decision making question or the risk multi-attributes decision making question. Researching this kind of multi-attributes decision making problem stems from to deep the traditional multi-objective decision making question research, on the other hand is because decision model for this kind of decision-making question can satisfy actual policy-making needing, therefore the research on indefinite multi-attributes decision making question more and more receives the academic circle and the industrial circle interest.Because the decision information is not completeness, the indefinite multi-attributes decision making question comparing to the traditional multi-attributes decision making question is more complex. The prominent two difficult question is: (1)how to describe and process these indefinite decision information, to resolve this question relying on the theory researcher and the practice worker to develop the new processing non-determination information theory, to construct the new mathematical model; (2) policy-maker carries on the decision-making under the non-definite condition must undertake the corresponding risk, therefore the policy-maker risk predilection decided by the subjective and objective condition is the important policy-making factor. Therefore the following question is worth studying:to construct decision model in which this policy-making factor is one of the policy-making parameters, enable the decision model to meet policy-maker’s needs well. But in literature about the multi-objective decision making question research, there are rarely literatures to conduct the research to this question. This article mainly conducts the research from some different angles of view and thought to the above two questions. Summary and analyz domestic and foreign research about the non-definite multi- attributes decision making. From the theory development and practical application’s angle, done the further research to non-definite multi- attributes decision making question, constructed some new decision model, on the one hand to rich and complete multi-objective decision making theory, on the other hand to enable the decision model to satisfy actual policy-making needing.The main research content and innovation in this article as follows:(1) Introduce the multiple-valued intuition fuzzy set theory to the non-definite multi- attributes decision making question research. The multiple-valued intuitionistic fuzzy set is extending the traditional intuitionistic fuzzy set. Comparing the traditional ntuitionistic fuzzy set, the multiple-valued intuitionistic fuzzy set has better ability to descript the not definite, not precise and incomplete information. Now, the research on the multiple-valued intuition fuzzy set is at the start stage, this article conducts the research to fuse multiple-valued intuition fuzzy set’s many degrees of membership and the non-degree of membership, constructs some fusion method and carries on the analysis to these methods’good and bad points.(2) The similar measure of Vague value (set) is domain key technologies in the Vague set application. But in the majority of the existing research about the similar measure of Vague value (set), construct the method to measure similarity lacking from the macroscopic perspective. In this article, express the Vague value (set) in the three-dimensional space, so that explain the factors affecting the similar relationship between two Vague values (set) with the spatial geometry position relations. At last, constructed one kind method to measure the similarity between the Vague values (sets).(3) In the research on transforming vague set to fuzzy set, it is the key question to carry on the judgment to the vague-degree’s tendentiousness. Vague set’s vague-degree expresses one kind of indefinite degree, therefore the question transforming vague set to fuzzy set may regard as the venture decision under the definite condition. When the policy-maker carries on the venture decision, the policy-maker’s risk preferences decided by the subjective and objective condition is a very essential policy-making parameter. In this article, introduce the risk—income balancing weight into the transforming model, thus enable reflect the policy-maker’s risk preferences well in the transforming result. In addition, decompose the value of fuzzy set’s membership degree into the following two parts:the value of vague set’s real membership degree, the weighting sum of the risk value and the tendency value.(4) Considering policy-maker risk predilection, conduct the research to the non-definite multi-attributes decision making question, propose some decision molds according to the different policy-making background:The first kind of decision model:construct the sector value information fusion operator——Value Risk Weighting Similar Measure Operator (VRWSMO), and construct the sector multi-objective decision making model based on the VRWSMO operator.The second kind of decision model:construct the fuzzy linguistic value information fusion operator——Language Risk Weighting Similar Measure Operator (LRWSMO), and construct the indefinite language multi-objective decision making model based on the LRWSMO operator.The above two kind of decision model’s innovation as follow:Consider the policy-maker risk predilection in the decision-making process and construct three tuples (The lower limit similarity, risk degree, risk-income balance value) to portray the risk-degree existing in the decision-making process. The above two kind of decision model not only can consider the policy-maker risk predilection decided by the subjective and objective condition, moreover the policy-maker can adjust the corresponding policy-making parameter according to the policy-making environment change. The two decision models can adapt the fast changing social environment.(5) When the weight information of attributes is indefinite, how carries on processing to the indefinite weight information is worth studying. The traditional processing thought is that establish an optimized model based on certain criterion such as variance minimizing, take the solution of the optimized model as the attributes’weight value. The rationality and multiplicity of the criterion are the flaws of the traditional processing thought. Attempt to deal with the decision-making question, in which the weight values’ information only partly may know, from an opposite perspective in this article. Firstly suppose each plan is the synergy, extract the corresponding weight value’s supposition territory. Then judg match degree between each weight value’s supposition territory and the weight value’s objective territory, the plan owing the maximum match degree is the optimal plan.