Research On Vibration Mechanism And Pattern Recognition Diagnosis Method Of Axial Piston Pump

As the main power source in the hydraulic system,the axial piston pump will directly affect the normal operation of the entire hydraulic equipment system.Therefore,this paper takes axial piston pump as the research object,and studies the fault mechanism analysis and pattern recognition fault diagnosis method.The content is as follows:Starting from the working principle of the axial piston pump,the in-depth analysis of the cause of the fault vibration is carried out,and the force analysis of the important parts and stress concentration positions inside the pump body is carried out to find the transmission of the internal fault vibration signal.The path and the final bearing force point of the vibration,and on this basis,the fault vibration signal of the axial piston pump is collected and processed.Using the newly constructed wavelet threshold function for the collected fault vibration signal to perform noise reduction processing to obtain the noise signal after noise reduction.The decomposition principle of the local mean decomposition algorithm(LMD)is analyzed and studied.The related methods are used to improve the endpoint effect and the excessive decomposition amount.The original PF component feature information is guaranteed while the endpoint effect problem is alleviated and the LMD is improved.The ability to decompose the algorithm.Combined with the denoising advantage of wavelet new threshold function and the decomposition characteristics of local mean algorithm,an energy feature extraction method based on NW-LMD decomposition algorithm is proposed.The experimental results are verified by comparison with LMD decomposition algorithm to prove the fault characteristics of the algorithm.The superiority of energy extraction.Introducing BP neural network algorithm in the diagnosis system.For the BP neural network algorithm,the local minimum value and the convergence speed are too slow.The particle swarm optimization algorithm is used to optimize the BP neural network weight and threshold and find the global The optimal solution is used to improve the running speed and convergence precision of BP neural network algorithm,and a new fault diagnosis method based on PSO-BP neural network algorithm is formed.This method is used to simulate the fault characteristic energy information of the axial piston pump.Finally,the experimental data prove that the PSO-BP neural network algorithm is compared with the traditional BP neural network algorithm in the axial piston pump fault diagnosis.Aspects have faster,more accurate fault identification capabilities.Through the testing of some actual sample data,the proposed fault diagnosis method based on NW-LMD algorithm and PSO-BP neural network algorithm has practical significance,which is important for fault diagnosis of axial piston pump in complex working conditions.Realistic meaning.