The Research Of Gear Box Fault Diagnosis Based On Embedded System

Gearbox is key component widely used in machinery and equipment.Real-time monitoring and fault diagnosis of gearbox have significant ineconomic and security implications for industrial production. In this paper, amonitoring and diagnosis system is developed based on DSP in order to realizediagnosis of gearbox, and empirical mode decomposition algorithm embeddeddiagnostics system, and applied to characteristic quantity extraction of commongearbox failure, to achieve the purpose of fault diagnosis.Embedded fault diagnosis system is a kind of integrate the signalacquisition, data processing and feature extraction based on embedded hardwareplatform to perform at an organic whole，it can realize automatic fault diagnosisfunctions and a variety of methods of communication. In this paper, DSPTMS320F28335of TI Company as the core processor, and integrate theacquisition module and communication module circuit, make the system canrealize analog acquisition, digital quantity input/output and speed measurement function, also achieving a serial communication functions.The basic principle and algorithm of EMD method is mainly introducedin this paper, according to the implementation process of the EMD algorithm,the overall algorithm flow is designed, and discuss the extreme value point andthe IMF component judgment standard of EMD algorithm. This paper also usesthe model design of the MATLAB/Simulink, the MATLAB code of EMDalgorithm generate embedded code and is applied to fault diagnosis system.Theexperiment firstly collect gear box experiment data and get each IMFcomponent, and by using the EMD energy extracting feature extraction of faultinformation characteristic value, the BP neural network were used to detect theuse of this method in the gear box fault diagnosis effect. The experimentalresults show that this system can extract the fault information effectively, so asto further achieve the goal of fault diagnosis.