Experimental Study On Wear Accelerated Life Of External Gear Pump

As a typical hydraulic pump,the gear pump plays a very important role in the fields of engineering machinery,walking machinery,ships and aerospace.To ensure that the gear pump works under normal conditions,it is particularly important to predict its life.However,the traditional full-cycle life test method has the disadvantages of long test time and large capital consumption.Therefore,this article studies the accelerated life test method of the gear pump.Research on wear mechanism and accelerated test methods.This paper introduces the characteristics and working principle of the external gear pump,analyzes its wear degradation mechanism,and obtains the main degradation mechanism affecting its life.At the same time,the accelerated life test method of gear pump is introduced,and the accelerated life test bench is designed and built.Research on signal noise reduction and reconstruction.Based on the simulation signal,the variational modal decomposition method(VMD)is used for noise reduction and reconstruction analysis,and compared with the ensemble empirical mode decomposition method(EEMD)and the improved ensemble empirical mode decomposition method(MEEMD)The analysis proves the applicability of VMD method for noise reduction in this paper.Extraction of degradation fusion indicators.Based on the gear pump data obtained in this experiment,the VMD-Hilbert is used to first reduce the noise,and then the degradation performance index extraction based on the time domain,frequency domain,time frequency domain characteristic parameters is extracted,and the factor analysis method is used to Multi-feature parameters are fused and reduced to obtain the degradation fusion index of the gear pump.In order to increase its robustness,the Holt dual parameter linear exponential smoothing method is used to smooth the degradation fusion index to achieve accurate prediction of the degradation fusion index.Life prediction.The use of Bayesian regularized radial basis function neural network(Trainbr-RBFNN)to predict the life of the gear pump,the network model needs to be trained before the prediction,the training data is the degradation fusionindex and output flow of the gear pump;Model,the test data is substituted into the model,and the prediction result of the input feature vector is obtained through complex operations within the model.Using the time conversion formula to convert the life time of the gear pump under accelerated stress to normal stress,the actual life time and predicted life time under normal stress are obtained,which shows the accuracy of Trainbr-RBFNN in predicting the remaining life.