Conflict evidence synthesis method based on confidence divergence

Abstract

The invention discloses a conflict evidence synthesis method based on confidence divergence. The method comprises the following steps: introducing the confidence divergence to represent the difference and conflict degree between evidence sources, so as to determine the credibility of each evidence source, employing information entropy to represent the uncertainty of each evidence source, and calculating the weight of each evidence source. According to the method, the evidence credibility and the evidence weight are combined, and the evidence source is preprocessed and corrected, so that the influence of highly conflicted and unimportant evidences in the synthesis process can be effectively reduced, the corrected evidence source is synthesized by applying a Dempster combination rule, an anti-visual unreasonable result is prevented from being generated, the rationality of an actual decision is ensured, and the reliability of the actual decision is improved. The conflict evidence synthesis result can be used in the fields of target recognition, fault diagnosis, multi-criterion decision and the like, and has important theoretical significance and application value; the method has the advantages of better convergence, robustness and effectiveness when conflict evidence is processed and synthesized.

Description

technical field [0001] The invention relates to a method for synthesizing conflict evidence, in particular to a method for synthesizing conflict evidence based on confidence divergence. Background technique [0002] Due to the complexity of real decision-making problems, the decision-making data is often uncertain, inaccurate, and incomplete. rationality. Evidence theory, as an uncertain reasoning theory, continuously shrinks hypotheses through evidence accumulation, and has significant advantages in multi-source data fusion. The evidence theory was proposed by Dempster and promoted and perfected by Shafer. It is also called Dempster–Shafer (D-S) evidence theory. Its core is the Dempster combination rule, which satisfies weaker conditions than Bayesian probability theory and breaks through the need for a priori Knowledge constraints, able to directly deal with uncertain, imprecise and incomplete multi-source data. Therefore, evidence theory has attracted the attention of ...

AI Technical Summary

Problems solved by technology

Although some scholars have used methods such as evidence distance and reliability entropy to modify evidence sources, these methods are still insufficient in the convergence and accuracy of conflict evidence synthesis.

Images