Research On Aggregated Models Of Uncertain Information Based On Normal Cloud Model

With the scientific decision-making, democratic management and the actual complexity of theproblem, more and more decision-making prefer to use the way of brainstorming, and groupdecision-making has been a hot research. In the decision-making process, due to the limitedrationality of human thinking, the uncertainty of the external environment and the complexity ofdecision-making problems, finding the appropriate method to integrate different preferenceinformation and choosing the appropriate way to express decision makers’ information are the focusof research in the decision-making field.Randomness and fuzziness is the most critical feature in all kinds of uncertainty feature, andpeople paid more attention on this. Academician Li Deyi proposed the concept of cloud model basedon the probability theory and fuzzy mathematics, by a specific structure construction algorithm, cloudmodel achieved uncertain conversion between qualitative concept and quantitative expression, and ithas been an important way to express uncertain information. We propose using of normal cloud modelto express decision-makers’ preference information, and a method based on normal cloud model isgiven to solve the uncertain multiple attribute decision problems.In this paper, according to the shortcomings of the existing normal cloud model similaritycalculation methods, we proposed a new comprehensive similarity calculation method, which takinginto account normal cloud model graphical features as well as digital features. The characteristic ofcomprehensive similarity has been studied, and by comparing with the existing similarity comparisonmethods proved the effectiveness and practicability of this new calculation method.For group decision making problem, normal cloud model is proposed to express the languageinformation of decision makers, and by using the new similarity calculation method of cloud model tomine potential link between policy makers from program decision vector and attribute decision vectordual perspective. By constructing normal cloud model similarity matrix and its equivalence matrix toachieve decision-makers clustering, and propose the use of a new method of normal cloud model toachieve similar policymakers gathered information integration.A new cloud distance calculation method is proposed, which based on random cloud dropletsgenerated. According to the classification of decision makers, inter-class decision makers’ weightsestimation model is established based on cloud distance calculation. A new normal cloud modelintegration method which considering decision makers’ weight is proposed, and multi-class decisionmakers’ integrated cloud decision matrix is generated, a method based on progress is proposed to sortthe alternatives.Finally, the situation that weight of the decision makers are not given is considered. normal cloudmodel deviation of entropy is defined, which is used to measure the difference between two normalcloud models. In addition, a grey target decision-making method based on cloud model deviation ofentropy is established, and the comprehensive evaluated value of each alternative is calculated by theaggregated model of this paper.