The Frequency Spectrum Refinement Analysis Of Steam Turbine Vibration Signal

Steam turbine is the rotating power machinery which converts energy of steam to mechanical power, mainly the prime mover used for generating electricity. In the long run, the turbine is prone to vibration fault, so it is necessary to detect the fault of the turbine in real time. In order to get the characteristic of narrow band, the frequency resolution is improved by using pectrum zooming technology. Embedded vibration signal analysis system based on spectrum thinning algorithm is designed in this paper.This paper analyzes the commonly used method of frequency spectrum thinning, including the CZT method, the complex modulation zoom-FFT method, the phase compensation method, The MATLAB simulation and performance comparison of the three algorithms are carried out, and their characteristics and application occasions are analyzed. Through the study of know, complex modulation zoom-FFT suitable for used in zoom multiples is little, the narrowband band, less conversion occasions; chirp Z transform is the biggest advantage is can be flexible to refine the adjustment band and magnification, the spectrum due to the fence effect caused by aliasing, chirp Z transform to achieve the purpose of very precise refinement; refinement method of phase compensation for the refinement of the full band spectrum. Comparing the three thinning algorithm, complex modulation ZOOM-FFT method in computational efficiency, accuracy and flexibility are ideal, so the selection of polyphony of ZOOM-FFT method as the system spectrum refinement of the algorithm.After comparing various vibration signal analysis system at home and abroad, the development scheme of the vibration signal spectrum thinning system based on embedded ARM9 as the core is determined. On the hardware side, the minimum system, the pretreatment circuit, the A/D conversion and the LCD display circuit are designed based on the Samsung Corp’s S3C2440. Vibration signal sensor using eddy current displacement sensor. Software design, the use of the embedded software system, embedded software system can be classified bottom layer, middle,top.underlying the preparation and transplantation of the buttons and LCD driver, cutting and transplantation of the Linux operating system in the middle layer, Completed the transplantation of the UCGUI3.90 graphics library.Prepared the application program and complex modulation ZOOM-FFT algorithm program in top layer. Finally, After a series of tests, the system can achieve the purpose of the frequency spectrum refinement of the vibration signal of the steam turbine, and the system also has the advantages of convenient carrying, low price, stable performance and practical function.