Abstract: AninformationfusionmethodwasproposedbasedonDSevidencetheoryandBayestheoryforlifetimepredictionofimportantaerospaceequipment,momentumwheel.Firstly,multisourcelifeinformationwerecollectedandminedtoobtainthepriordistributionofthemomentumwheel′slifetime,inordertobuildDSevidencecollections.Secondly,DScombinationrulewasusedtoobtainreasonableweightallocationforpriordistributions.Afterthat,fusionposteriordistributionwasfiguredoutandthevaluesoflifeparameterswerealsoestimated.Finally,accordingtoparameters′estimation,thelifetimepredictionformomentumwheelwasderived.Simulationresultshowsthatthepredictionusingtheproposedmethodisclosertothereallifetimemeasurements.

Key words: lifetimeprediction, multisourceinformationfusion, DSevidencetheory, Bayesmethod, Momentumwheel