Research On Rotating Machine Order Analysis Method Based On Vold-Kalman Tracking Filter

Order analysis is one of the most important techniques of rotating machinery involving vibration signal processing and fault diagnosis. Currently, uniform angular intervals re-sampled plus FFT order tracking method is applied. This way can only get order spectrum and order tracking spectrum, but can not get order extraction in time domain. In addition, rotating machinery is composed of many rotating parts. Orders of different shafts cross each other in frequency. However, conventional order tracking method can only refer to one reference frequency, extracted order waveform be interrupted by crossing order that cause distortion. Order tracking technology based on Vold-Kalman tracking filter can well solve these two problems. This paper based on Vold-Kalman tracking filter researches on rotating speed computation and extraction order components. Finally order analysis system has been developed. That is very significant for rotating machinery characteristic analysis.Rotating speed computation is not only one of key technologies of order analysis, but also necessary condition for implemented Vold-Kalman order tracking. According to whether has tachometer, rotating speed computation is researched by two ways. When the speed pulse signal is provided, a calculation of the speed curve is achieved using first order digital differential method and the lease square of five points three times smoothing approach. It greatly reduces amount of calculation and removes speed error or defect.For the situation of provided vibration signal but without tachometer pulses signal, a novel speed curve calculation method based on energy centrobaric IFE method is proposed. Compared with based on STFT spectral map IFE, processed signal meets the criterion of full-period sampling so as to frequency resolution changes by self adapting. Introducing energy centrobaric method, it can correct the error, improve the accuracy of analysis and reduce interference between close spectral lines.Order components Extraction in timedomain is implemented using Vold-Kalman tracking filtering method. This method can separate order wave in time domain without phase bias. It allows beat free extraction of close and crossing orders in systems with multiple shafts using Preconditioned Conjugate Gradient (PCG) method. A novel order tracking method for rotating machinery based on tacholess and Vold-Kalman order tracking filtering technique is proposed. Compared with conventional methods of Vold-Kalman ordering tracking method, the proposed method is a complement for existing methods, which has high precision in algorithms with no aid of tachometer and other special hardware.Finally, an Vold-Kalman order analysis systems for rotating machinery is developed, with VC++ as the development platform. The main functions of this instrument include: rotational speed, order re-sampled, FFT, order spectrum, STFT, Vold-Kalman tracking filtering and so on. This system has friendly interface and is operated easy. Experiment result shows that this system can effectively extract and analyze characteristic information of non-stationary vibration signal in rotating machinery.