Abstract: Inordertodetecttheabnormalsubsequenceinthepseudoperiodictimeseriesofspacecrafttelemetry，asubsequencesegmentationalgorithmonthewiderangeofmaximumcyclewindowwasproposedbasedonthetwoconventionalsegmentationmethods.Then，ananomalysubsequencedetectionmethodbasedonanoptimizedsequencemodelwasbuiltbyusingthedoublequantilerangescriteriontosetthethresholdofthedistancedetection.Thesatelliteexperimentalresultsshowthatthemethodcaneffectivelydetecttheabnormalsubsequenceofsatellitetelemetrydata.

Key words: anomalydetection, satellitetelemetry, pseudoperiodic, timeseries, abnormalsubsequence, doublequantilerangescriterion