Rotating Machine Order Analysis System Based On Self-adaptive Computed Order Tacking

Order analysis is one of the most important techniques of rotating machinery involving vibration signal processing and fault diagnosis. Its characteristic is fully using the information of rotating speed. The vibration signal of rotating machinery generated during the acceleration or deceleration phase is the non-stationary signal, which the fundamental frequency changes with the time. Traditional spectral analysis methods based on FFT can not effectively extract and analyze the characteristic information of this phase. However the emergence of order analysis can resolve the problem well. The technique will be studied in this paper.Rotating speed computation is one of the key technologies of order analysis. Computing method of rotating speed based on one digital differential, single pulse subsection interpolation is proposed and implemented in this paper. This method firstly calculates the initial speed with one order digital differential, then through subsection interpolation, gets the rotating speed of every sampling moment, establishs the relationship of speed and vibration amplitude, and lays the foundation for the follow-up order analysis.Order tracking technology is one of the most important areas of order analysis. Based on the comparison and analysis the advantages and disadvantages of current order tracking methods, and combining the initial speed and circular motion equation, the adaptive calculation order tracking technology based on the circular motion equation is proposed and implemented in the paper. This technology according to the change of rotating speed, adaptive selects different method to calculate order tracking. When the change of speed is slow, we adopt the linear interpolation method to calculate order tracking; when the change of speed is large, we use constant angular acceleration equation to calculate order tracking; when the change of speed is dramatic, we use cubic spline interpolation method to calculate order tracking. The method improves the operating efficiency while guaranteeing the requirements of accuracy. Then order spectrum can be obtained after the FFT of order re-sampled vibration signal. In order to obtaining the time-domain waveform of interested order from order spectrum, order component extraction. based on Gabor time-frequency filtering technology is implemented.Time-frequency analysis is one of the commonly used methods for analysis the non-stationary vibration signals. But for rotating machinery, time-frequency analysis is difficult to reveal the relation between amplitude and rotating speed. With deep research on the principle of Short-Time Fourier Transformation (STFT), combining the rotating speed calculated by subsection interpolation, the algorithm of rotating speed moment adding window Fourier Transform is proposed and implemented in the paper, which provides the theoretical basis for computing rotating speed spectral map of rotating machinery.After through research on the existing order analysis techniques, with new ideas, new methods and new approaches proposed in this paper, the order analyzer system for rotating machinery is developed. The system that includes interface, results display and all kinds of arithmetic function is developed through VC6.0. Experimental test results show that the method which is researched and proposed in the paper is correct and effective, and the value of the developed system is obvious. It lays firm basement for promoting extensive use of order analysis in rotating machinery non-stationary vibration.