Research On Model And Application Of Grey Incidence Decision Making Based On General Grey Number

Grey incidence decision model is an important branch of grey decision theory system.It is also an important component of grey system theory.It is widely used in social,economic,management,engineering and other fields.Study on grey relational decision theory is relatively mature,the research results are abundant in recent years,but from the perspective of subject development and other decision theory development,it still has space to expand in depth research.In particular,with the development of big data technology,the amount of information available is very large,while providing information sources are also very much,and the data structure becomes more and more complex,which makes the decision information for complexity and uncertainty.The general grey number is used to characterize the complexity and uncertainty of decision information.the grey incidence decision model is systematically studied,Based on the general grey number information.According to its modeling object,the grey incidence decision model is expanded from real number to interval grey number,and then to general grey number.At the same time,it is extended from one-dimensional general grey number to panel data type and then to multidimensional data sequence.It broadens the research depth of grey incidence decision model and expands the application scope of this method.The main research results are summarized as follows:(1).The general grey number ranking method and the distance measure problem.According to the nature of philosophy of general grey number,in accordance with the complexity of decision information and accurate expression of decision information requirements,the concept of the kernel expectation and the kernel variance of the general grey number are put forward,and then the distance measure of general grey number and the method of sorting expectation of kernel and variance of kernel for the general grey numbers are proposed based on the idea of general grey number and the essential features of Euclidean distance.(2).Study on grey relational decision model based on area.In view of the deficiency of the classical grey relational model,on the one hand,the correlation between different sequences is measured from the angle of the polygon area between the adjacent points of the two sequence.the polygon area is used as the correlation coefficient,because it can reflect the mutual influence between the indexes comprehensively,and can accurately reflect the distance and geometric similarity between the two sequence curves.On the other hand,The combination weight optimization model is constructed based on the principle of moment estimation.And based on the "function driven" and "differential drive",and combined with the grey relational model,the index weight determination model is constructed.The grey relational proximity decision model is constructed based on the above two aspects.(3).Study on grey incidence decision model of interval grey number based on information decomposition.On the premise of information without losing,Interval grey number is decomposed into real type sequence of "white sequence " and "grey sequence" based on the information decomposition method.In addition,Considering the consistency of transformation,a consistency coefficient model is constructed.Thegrey correlation model,projection decision model and the consistency coefficient of the model are anastomosing and integrated,The Grey Incidence Decision Model of interval grey number and the consistent projection decision model of interval grey number are constructed based on information decomposition.(4).Study on grey incidence decision model based on general grey number.Because of the complexity of system evolution,its uncertainty is widespread.It is difficult to describe the characteristics of system development and evolution accurately with a real number or an interval grey number.In order to describe the characteristics of the system accurately,the concept of general grey number is proposed.According to the idea of kernelness and degree of greyness,the degree of incidence,the absolute and relative degree of incidence,the similitude and closeness of grey incidence and the corresponding decision model are constructed based on the general grey number.(5).Study on grey dynamic incidence decision model based on general grey number.in view of the problem of optimal decision making in the dynamic development of the scheme,the whole decision is viewed as a multi stage(three stage),namely,static,semi dynamic and dynamic.Firstly,the ideal effect vectors are given in three states;Secondly,the absolute degree of grey incidence and relative degree of grey incidence are constructed in each stage.At last,the three stage's relative degree of grey incidence,he absolute degree of grey incidence and synthetic degree of grey incidence model are constructed.Because of the dynamic process,the precision of decision mainly depends on the improvement of prediction accuracy.Therefore,the grey prediction model is improved from the background value of grey forecasting model and the grey forecasting model itself angle.It mainly involves the optimization model of GM(1,1)model background value,and improves the grey envelope prediction model and the three time variant parameter discrete grey prediction model.(6).Study on grey incidence decision model of general grey number based on panel data.Relational decision model of general grey number will is extended to panel data.On the one hand,the similarity and nearness between two grey lines are measured form the grey fold line's slope and the area between the grey fold lines the angle.On the other hand,according to the distance measure of general grey number,the distance calculation formula between the corresponding points of two line is constructed.The Deng's relational degree model is extended to grey panel data,and the grey panel data correlation model and its decision algorithm are constructed based on time.(7).Study on grey incidence decision model of general grey number in project selection of scientific and technological enterprises.In the light of the characteristics of the selection and approval of scientific and technological projects,an index system for comprehensive evaluation of project selection is constructed.By using the Generalized Grey Incidence Decision Model,the correlation degree between the positive and negative ideal is calculated,and the relative association degree model is obtained.The three scientific and technological projects are evaluated and sorted.