Modeling Method Of Association Clustering Based On Grey Information

The continuous development of society brings endless problems,and many theories and methods of uncertain system emerge in endlessly.Grey system theory is widely used in all aspects,but the existing literature shows that grey relational clustering model has some limitations in dealing with some practical problems.Aiming at the problems of different data types,incomplete information and low accuracy of the model,this paper studies the association clustering modeling method under grey information.Firstly,the grey information association model based on Hamming distance is constructed.Secondly,the DT type association model based on multi-source heterogeneous information is constructed.Finally,the grey cloud association clustering model is constructed.It is also used to study agricultural drought in Henan Province.The main achievements are as follows:(1)Aiming at the problem that the observation value is three parameter interval grey number and the grey correlation sequence does not consider time,the Hamming distance formula based on decision preference coefficient of three parameter interval grey number is given,which transforms the three parameter interval grey number into real number.According to Deng’s correlation degree,the dynamic correlation model of factor characteristic panel data type is three parameter interval grey number,and the correlation model based on Hamming distance under grey information is constructed.The model is applied to the decision-making evaluation of the factors affecting soil water content,and the key factors affecting soil water content are obtained,which shows the effectiveness of the model.(2)In view of the fact that the type of observation data is not unique and the data has obvious time dimension and object dimension,this paper considers the grey association method based on the fusion of panel data and grey number.Based on the perspective of proximity,the total displacement of development level and the total displacement of development rate are constructed.The fractional function is transformed into an exponential function to improve the resolution of the existing model.On this basis,a DT type association model with grey multi-source heterogeneity is proposed based on panel data.In view of the problem that the data information of regional agricultural drought assessment is multi-source and heterogeneous,and the static data can’t fully describe the real characteristics of drought influencing factors,this model is applied to the drought assessment of five cities in Henan Province,and compared with other models,which shows the effectiveness of DT correlation model.(3)Aiming at the problem that the whiteness weight function in the existing grey clustering evaluation model can’t reflect the grey,fuzzy and random characteristics at the same time,which makes the decision result inconsistent with the objective reality.A multi-attribute multi-stage decision-making model is proposed to solve the dynamic multi-attribute decision-making problem with unknown index weight and time weight and interval grey number.The index weight is determined by the grey correlation degree between each index and the ideal index,and the comprehensive weight is obtained by combining the subjective weight.Using the principle of maximum deviation and maximum entropy to solve the time weight,the traditional whitening weight function is transformed into grey cloud whitening weight function,which fully reflects the coexistence of multiple decision conclusions,and the grey cloud association clustering model is constructed.Taking 18 cities in Henan Province as the research object,the drought degree under 15 indicators was analyzed.The analysis results accord with the actual situation of drought disaster in Henan Province,and verify the rationality of the model and method.