Research On Cavitation State Analysis And Detection Method Of Axial Piston Pump

Axial piston pumps are widely used in various hydraulic systems.With the development of hydraulic pumps in the direction of high speed and high pressure,cavitation has become a key factor restricting the performance of axial piston pumps.The occurrence of cavitation will cause the mechanical performance of the hydraulic pump to decrease,causing the vibration of the hydraulic system to deteriorate,resulting in cavitation damage in the pump;severe cavitation will lead to the destruction of the mechanism and even the catastrophic consequences of machine damage and injury.Revealing the mechanism of cavitation and achieving effective detection of cavitation status has always been the focus and difficulty of research in academia and engineering.At present,although the research on cavitation theory has achieved some results,it is still under exploration.The research on the detection of the cavitation state of hydraulic pumps using oil as the transmission medium is still in the exploration stage,and it cannot provide a reliable basis for engineering practice.This dissertation takes axial piston pump as the research object,starting from two aspects,theoretical analysis and experimental research,and systematically studies the cavitation evolution process,cavitation external characteristics and cavitation state detection method of this type of pump.The work and results achieved are as follows:(1)Understand the axial piston pump structure and analyze the cavitation process.First of all,starting from the theory of cavitation combined with the fluid movement law in the axial piston pump,the whole process of the occurrence of the cavitation phenomenon of the axial piston pump is explained in detail from a microscopic perspective;secondly,the evolution of cavitation bubbles is established according to the cavitation dynamics The mathematical model of the process understands the cavitation process from a microscopic perspective.Finally,based on theoretical analysis,an axial piston pump cavitation test research system is designed and built.(2)Analysis and research on the external characteristics of the axial piston pump in the cavitation state.In this dissertation,according to the cavitation performance analysis,combined with the research ideas of signal analysis and camera method to analyze the cavitation state,the Lab VIEW data acquisition system is used to collect cavitation vibration signals,outlet flow signals and outlet pressure signals of different cavitation states.Detailed information on the state of the oil,such as the number and distribution of oil bubbles in the oil port,has carried out a more comprehensive measurement of the axial piston pump cavitation.By comparing and analyzing the detailed information of the oil state of the oil inlet,the critical judgment basis for the initial cavitation of the axial piston pump is proposed from the perspective of cavitation morphological characteristics,and the cavitation flow characteristics at the initial,development and severe stages of cavitation are obtained.Combined with the cavitation performance curve,the cavitation state of the axial piston pump is divided into normal state,cavitation development state and cavitation serious state according to the severity.Modern signal analysis method is used to compare the rule of the axial plunger pump’s external characteristics with the cavitation state.The analysis results can be used as an effective basis for cavitation state detection and provide a useful reference for the realization of cavitation state detection.(3)Research on the detection method of the cavitation state of the axial piston pump based on Extreme Learning Machine(ELM).In view of the fact that there are few researches on the detection of the cavitation state of the plunger pump,and there are problems such as weak cavitation signal,difficult feature extraction and low detection accuracy in the research process,an intelligent detection method for cavitation state is proposed.First,collect the external characteristics information under three cavitation conditions in each working condition.Secondly,various features of different signals are extracted and used as the input of the ELM intelligent detection model after dimensionality reduction,fusion and normalization to achieve high accuracy detection of the axial piston pump cavitation state.Finally,the ELM test results are compared with the output results of BP neural network and random forest(RF).The results show that the ELM model has more advantages in detecting the axial piston pump cavitation state.