Research On State Identification And Invalid Prediction About Axial Plunger Pump Based On The Performance Degradation Data

Axial piston pump has many advantages,such as compact structure,high power density,high output pressure and long working life.It has been widely used in various fields,such as aerospace,special vehicles,engineering machinery,ship machinery and agricultural machinery,and so on,which are related to national security and people's life.As the "heart" of the whole hydraulic system,the performance of axial piston pump not only affects the reliability of the hydraulic system,but also has a decisive influence on the safe operation of the whole system.Therefore,it is of great significance to the identification and failure prediction research of the performance degradation state of the piston pump.Based on the existing analysis of testing programs for friction pair and combining with the test purpose,the experimental scheme of the performance degradation of the plunger pump is established,and the vibration signals and flow signals are collected in 6 different states of the plug pump disks based on the LabVIEW program experiment collection system,which provide the original data for the subsequent state recognition and failure prediction.Aiming at characteristic that the vibration signals of ram pump usually have strong nonlinearity and low signal to noise ratio,this paper combines the two methods of variational modal decomposition and multi-scale permutation entropy.Firstly,the vibration signal of the piston pump is decomposed to get the modal component containing the degenerate feature information,and then the entropy is arranged by the nonlinear processing method of the time series,and the modal fractions of the 6 different states are calculated.The entropy of the multi-scale arrangement of the quantity is used to determine the corresponding entropy characteristic vector in a state.Finally,the Gauss process classification training set is set up by the feature vector to realize the identification of the piston pump degenerate state.By analyzing the wear process and wear mechanism of the distributor,a mathematical model for the relationship between the leakage flow of the plunger pump and the wear volume of the distributor plate is established.Using Gauss process regression to process experimental data,we compare the degradation trajectories and the least squares fitting trajectories,and determine their superiority.Finally,the fitting curve is applied to predict the failure of piston pump due to the wear of the distributor plate.