Thestructure,functionandrecognitionmethodofanaxisorbitauto-recognizingsystemarepresentedinthispaper.Inordertomakethebestuseofinformationofformatanddynamiccharacteristicsofmarinesteamturbineaxisorbit,thestructureandfunctionsorneuralnetworkareappliedtothissystem,whichcanbeusedtoauto-recognizeaxisorbitofthesystemturbinerotorusingBPneuralnetwork.