Intermittent　faults　have　the　characteristics　of　short　duration,　repeated　occurrence　and　unobvious　fault　features.　If　the　intermittent　fault　in　aircraft　fuel　system　can　not　be　diagnosed　accurately　and　promptly,　it　will　cause　serious　accidents.　A　method　of　intermittent　fault　diagnosis　based　on　support　vector　machine　(SVM)　is　presented　.　A　simulation　platform　based　on　Flowmaster　is　built　to　simulate　the　intermittent　fault　in　the　aircraft　fuel　system.　The　fault　signal　features　in　the　fault　system　are　extracted　according　to　the　time　series.　These　fault　features　are　trained　as　the　training　data　of　SVM.The　training　results　show　that　this　method　can　accurately　diagnose　and　locate　intermittent　fault　in　aircraft　fuel　system.　©　2020　The　Institution　of　Engineering　and　Technology.