Research On The Technology Of Gas Turbine Bearing Fault Diagnosis And Fault Prediction

With the development of gas turbine technology and the industrial production needs, gas turbine in some industrial production status of the rising. Gas turbine failure not only maintenance need to spend, and delay industrial production will be more big losses. Therefore, it is necessary to gas turbine fault diagnosis and prediction, especially the fault of the bearing parts. In this paper, according to the characteristics of the gas turbine bearing on the fault diagnosis and fault prediction research, and based on LabVIEW established the system platform. The following elements are completed:(1)Based on the vibration method diagnose the bearing fault. According to the rotation speed of the bearing, failure frequency characteristics in different occasions is deduced. For bearing diagnosis acceleration data, using the fast Fourier transform method and wavelet change method extraction bearing fault characteristic frequency. In the different bearing fault condition, contrast the characteristics extracted the effectiveness of the method of frequency.(2)Based on intelligent neural network method diagnose bearing fault. According to different bearing fault cases, vibration acceleration data with different characteristic of frequency domain and time domain features, wavelet packet method extracting bearing frequency domain range of value. Calculated in time domain range is representative of the value. The time domain and frequency domain characteristic value are characteristic vector of neural network.The probability neural network was used to identify the bearing fault.(3)Based on sequence forecast and neural network method predict the bearing fault. According to the theory of phase space reconstruction construct bearing parameters. The BP neural network was used to predict the related parameters of gas turbine bearing though single parameter and point. The result of predict was the neural network’s input. The error percentage of the model output was judging the state of bearing.(4)Based on the Lab VIEW establishes the gas turbine bearing fault diagnosis and fault forecasting system platform. The Lab VIEW software of NI was selected as the development platform for program design. The initial program system design, user login procedure, information management, operation the interface procedures, fault diagnosis module program, fault prediction module procedure were designed. These modules in the main VI diagram of the according to certain logic relationship together, they form a complete diagnosis and predict the development platform.