Research On Multi-Information Intensity Characteristic Condition Evaluation Of Variable Speed Hydraulic Pump

In actual work,the hydraulic pump is o ften affected by factors such as working resistance and dynamic load,causing fluctuations in the rotational speed,resulting in amplitude modulation phenomena,resulting in a significant non-stationarity of the signal,together with the incompleteness of the single signal characteristic information,so that the evaluation accuracy of the hydraulic pump evaluation model is low.Therefore,it is very necessary to study the method of multi-information status assessment for hydraulic pumps under variable speed conditions.Aiming at the problems of the rotational speed fluctuation and amplitude modulation of hydraulic pump,this paper uses multi-scale modal filtering method to filter the multi-source signal of hydraulic pump.First,the vibration signal,outlet flow signal,and pressure signal of the pump housing under variable operating conditions are collected through vibration sensors,flow sensors,and pressure sensors.Second,using time-domain order analysis methods based on EEMD theory to convert the three time-domain non-stationary signals into angular stationary signals respectively,solving the problem of rotational speed fluctuations.Then,three kinds of angular domain stationary signals were separately filtered by multi-scale morphology,eliminating the phenomenon of amplitude modulation and obtaining three kinds of filtered signals with rich fault feature information.The vibration intensity has a good effect on the evaluation of the degree of failure.At present,there are formulas for calculating the vibration intensity in the time domain and frequency domain,but there is no formula for calculating the vibration intensity in the angular domain.In this paper,starting from the definition of vibrational intensity,we refer to the process of deriving the vibration intensity calculation formulas in the time and frequency domain,and derive the calculation formula of the vibration intensity in the angular domain.It is used to extract three kinds of filter signal intensity characteristic factors with rich fault feature information in the angular domain,which increases the completeness of the status information of hydraulic pump status evaluation.In this paper,starting from the fault mechanism,the development mechanism of the wear process of the inner edge of the shoe is described.The flow leakage and volumetric efficiency of the slide of the plunger pump are theoretically analyzed,and the leakage flow of the shoe is correlated with the wear of the inner edge of the shoe.The law of change divides the working state of the pump and the wear stage of the sliding shoe.The intensity characteristics of pump casing vibration signal,outlet flow signal and pressure signal at each stage of variable-speed plunger pump shoe wear were extracted,and the multiple information intensity evaluation vector sample set that can be used to evaluate the wear stage of the shoe was determined.Using BP neural network and D-S evidence theory to establish a multi-information fusion assessment model,and achieve the fusion of multiple information intensity features for the wear stage of plunger pumps in variable speed conditions.The test sample was used to verify the accuracy of the evaluation model,and a high-accuracy evaluation result was obtained.The validity of the multiple information intensity feature state assessment method for the variable-speed hydraulic pump was proved.