Research On A Series Of Typical Faults Feature Extraction Methods For Axial Piston Pump Bearings

The axial piston pump is one of the most core parts in the hydraulic fluid power system.Axial piston pumps are always in harsh working conditions such as heavy loads and high running speed,and unexpected fault will inescapably occur in key machine elements(such as bearings,pistons).When the axial piston pump is working,every rotation period is accompanied by the reciprocating motion of pistons,which will induce violent periodic impulses.Most of the fault information is submerged,and it brings challenges to fault diagnosis.In this paper,the axial piston pump is taken as the research object,and the noise suppression and fault information separation in the vibration signals are deeply studied.The main research contents of this paper are arranged as follows:1.Aiming at the problem that the natural periodic impulses in vibration signal of axial piston pump are too violent to separate fault impulses,a fault frequency bands location method based on improved fast spectral correlation(Fast-SC)algorithm is proposed.A new indicator named kurtosis enhanced spectral entropy(KESE)is exhibited to locate the fault frequency bands from the whole spectral frequency band,and the squared enhanced envelope spectrum(SEES)is employed to further extract feature frequencies.The simulation and experimental results show that the method can quickly and accurately separate the bearing fault information from the vibration signal and identify the fault type.2.Aiming at the randomness of artificial determination and lack of prior knowledge in common deconvolution methods parameters,a fault feature extraction method based on adaptive multipoint optimal minimum entropy deconvolution adjusted(MOMEDA)and Teager energy operator(TEO).Multipoint kurtosis is utilized to identify deconvolution periods of periodic impulse components effectively and the improved advance-retreat algorithm is used to optimize the filter length of MOMEDA.This method can overcome the uncertainty of parameter setting and improves the deconvolution accuracy.Compared with the traditional deconvolution method,this method has more advantages in the practical application ofaxial pistonpump vibration signals.3.Aiming at the components selection and noise residue of signal decomposition method,a novel method based on the fuzzy entropy assisted improved singular spectrum decomposition(SSD)denoising is proposed to extract fault features.Fuzzy entropy of each component is calculated to screen the components with fault information and remove the natural periodic impulses components.The soft threshold denoising algorithm is applied to the remaining components.This method can be used to detect the complex vibration signal of the hydraulic pump.