Open AccessA Multi-attribute Decision-making Optimization Algorithm Based on Conflict Evidence Fusion and Cloud Model
Author(s) -
Ni Zichun,
Xuan Wang,
Di Peng
Publication year - 2019
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/688/5/055014
Subject(s) - closeness , cloud computing , decision matrix , basis (linear algebra) , function (biology) , measure (data warehouse) , ideal (ethics) , vagueness , set (abstract data type) , reliability (semiconductor) , computer science , data mining , construct (python library) , ideal solution , mathematics , algorithm , artificial intelligence , operations research , mathematical analysis , philosophy , power (physics) , physics , geometry , epistemology , quantum mechanics , evolutionary biology , biology , programming language , fuzzy logic , operating system , thermodynamics
To eliminate fuzziness and uncertainty of linguistic comments in multi-attribute decision-making, this paper introduces D-S evidence theory into cloud model assessment. Golden section method is employed to convert experts’ linguistic comments into cloud decision-making matrix, and then criterion clouds of different levels in comment set are taken as the reference to determine the membership to each level of assessment, so as to construct the basic probability allocation function (mass function) for different experts in respect of different attributes in different schemes. Thereafter, conflict coefficient, Jousselme distance, and Pignistic probability distance are introduced on the basis of D-S evidence theory to define evidence conflict measure. By calculating evidence reliability and relative expert weight, mass functions for experts are modified and fused. In the end, mass functions of different schemes are fused on the basis of attribute weight, and compared with the mass functions of ideal cloud and negative ideal cloud. Therefore, optimal scheme will be determined by comparing average closeness.