Research On Degradation State Evaluation Method Of Axial Piston Pump Under Variable Speed

As an important part of hydraulic system,axial piston pump has the advantages of high integration,diversified control and wide range of speed and load regulation.When it is used in high pressure,large flow and other occasions,its safe operation often affects the reliability of the whole equipment.Due to the harsh environment and complex working conditions,the speed of piston pump is often non-stationary,and the change of speed will make it difficult to extract features and analyze reliability.Traditional feature extraction methods based on constant speed operation,such as statistical index methods about time-domain,frequency-domain,time-frequency-domain,are not so suitable due to the modulation and frequency modulation effect of variable speed operation.Therefore,it is of practical value to complete the degradation state assessment of piston pump under variable speed.Firstly,aiming at the shortcomings of traditional time-frequency analysis methods,such as energy divergence and ridge ambiguity,this paper introduces a method based on spline frequency modulated wavelet transform.This method uses spline function instead of linear chirp kernel to reduce the severity of ill conditioned normal equations.It can be used for time-frequency analysis of large-scale signals with continuous and highly concentrated instantaneous frequencies.Through the comparison of simulation experiments,this method can achieve good results under high noise conditions and achieve accurate estimation of signal time-frequency analysis.Secondly,aiming at the problem that the traditional time-frequency analysis method can not accurately specify the number of signal modes,fixed bandwidth parameters,and initialization of mode instantaneous frequency,the ACMP method is introduced.The method is verified by the variable speed data obtained from the simulation test,and compared with VMD and VNCMD,the superiority of this method is proved.Thirdly,a method combining SCT,ACMP and order analysis is proposed to complete the angle domain signal reconstruction and reconstructed signal analysis.The simulation results show that the method can effectively avoid the influence of speed transformation on vibration signal.In addition,through the construction of piston pump degradation test platform and data processing,the results obtained by this method are verified to be in line with the actual situation and have good representativeness.Finally,the peak index,pulse index,kurtosis index,order spectrum energy and order amplitude in the angle domain and order domain are selected as the degradation characteristics to establish the classification model.XGBoost classification model is introduced and compared with ANN,SVM and gbdt algorithm.It is found that this method can reduce the calculation time and ensure high recognition accuracy.In this paper,the experimental study on the wear degradation of the valve plate is carried out,and the degradation state evaluation of the axial piston pump under variable speed is completed,and good results are achieved.This paper lays a theoretical foundation for the signal processing of variable speed piston pump,and has practical application value in degradation assessment.