Abstract: Motor is an indispensable tool in machining.The various faults often take place in the motor operation.In order to ensure the safe running of the motor,it is particularly important for on-line monitoring of the motor running state.Three-phase asynchronous motor was regarded as the research object.The important parameters of the motor running (vibration,sound,speed and temperature) were obtained by the sensor acquisition.The motor running features were extracted by the time-frequency analysis and energy analysis,and those features constituted feature vectors.The state recognition model was built according to BP neural network training method.The motor running state was evaluated according to BP neural network pattern recognition method.The visualization monitoring system was constructed by using LabVIEW software.The motor running parameters and the states were real-time displayed in the visual interface,and the on-line intelligent monitoring was completed.

Key words: motor, state, neural network, monitoring