Research On Robustness Of Interval-valued Some Triple I Fuzzy Reasoning

Due to its ability to model linguistics and system uncertainties,interval-valued fuzzy reasoning has become very popular in various fields.This dissertation is a robustness study on interval-valued fuzzy inference.The main research contents of this thesis are outlined as follows:First,the research on interval-valued(1,2,1)-[?,?] type triple I algorithms.Interval-valued triple I algorithms based on fuzzy modus ponens and fuzzy modus tollens,were extended to the interval-valued(1,2,1)-[?,?] type triple I algorithms.The corresponding interval-valued(1,2,1)-[?,?] type triple I solutions are given.The robustness of interval-valued(1,2,1)-[?,?] type triple I algorithms are proved.The sensitivity of(1,2,1)-[?,?] type triple I solutions based on interval-valued Lukasiewicz implication and R0 implication are given.As main corollaries,the sensitivity of(1,2,1)-? type universal triple I methods based on classical sets are given.Second,the research of fuzzy reasoning based on interval-valued Schwe izer-Sklar t-norms.By the combination of interval-valued Schweizer-Sklar t-norms and triple I principles for fuzzy reasoning,give the interval-valued triple I solutions based on interval-valued Schweizer-Sklar t-norms.The robustness of the triple I algorithms based on Schweizer-Sklar intervalvalued t-norms(m ?(0,?))are proved.It is shown that the quality of interval-valued fuzzy reasoning algorithms depends on the selection of int erval-valued fuzzy connectives.Third,the robustness of interval-valued triple I algorithms based on normalized Minkowski distance is investigated.Based on the normalized Minkowski distance,the perturbation of interval-valued G?del implication and the corresponding t-norm,?ukasiewize implication and the corresponding t-norm,Goguen implication and the corresponding t-norm are discussed,respectively.The robustness of triple I algorithms based on interval-valued G?del,?ukasiewize and Goguen implication are proved.