A hybrid uncertainty analysis method based on random set theory

摘要

In view of the fact that hybrid uncertainty presentation and propagation with dissonance or imprecision information in risk assessment, a variable steps size discrete random set method is proposed. Various types of incomplete and inaccurate information were transformed to random set formation by the proposed framework. A hybrid uncertainty propagation model is built based on random extension principle, the lower and upper CDF of outputs are calculated. In order to decrease dissonance between uncertainty parameters, D-S evidence combination principle is used to merge the collision information. Variable step size parameter discretization strategy was proposed based on ADM and ODM for reducing the truncated tail relative errors. In conclusion, take the response of a mass-spring-damper system acted on by a harmonic forcing an example, the effectiveness and feasibility of the proposed method is validated.