| Through the study for present situation of the existing dynamic balancemeasuring system development and various methods of vibration signal processingadopted, it was found that the nonlinearity of signal measurement channel in thecurrent dynamic balance measurement and the existing digital filtering methodwhich could not quickly and accurately remove near frequency interference signalwere the main factors impacting on measurement system accuracy. In order toimprove the measurement precision of the dynamic balance machine, we musteliminate these influences.In view of these two problems, firstly this article adopted the support vectormachine algorithm for vibration signal measurement channel modeling andsimulating, made nonlinear error prediction and compensation and accomplishednonlinear correction. Secondly the author used harmonic wavelet transformalgorithm as the core to design the digital band-pass filter filtered out near frequencyinterference signal and improved the signal to noise ratio, Finally it was proposed toadopt discrete Fourier transform method after the whole cycle sampling to extractunbalance signal amplitude and phase based on Fourier transform principle,Simulation indicated that the proposed method had a higher precision ofmeasurement, faster operation speed, and overcomed the spectrum leakage of FFTmethod and low precision phase of cross-correlation method. The vibration signalprocessing and unbalanced signal extraction method proposed in this article aremore universal, and can be used for unbalanced signal measurement of the rotationof the rotor.Finally this paper summarized the propressiveness of the algorithm adoptedabove and its innovation and advantage in the application of dynamic balancemeasuring technology. These methods provided strong theoretical guidance for thedesigning of high precision dynamic balancing measurement system. In addition, theauthor put forward to research methods for high precision dynamic balancemeasurement of vibration signal. |
PDF Full Text Request