Failure Status Of Electro-hydraulic Servo Valve For Aero-engine Diagnosis

Electro-hydraulic servo systems combine the advantages of electrical and hydraulic systems,with the advantages of high control accuracy,fast system response,high system output power and the ability to flexibly process various signals.Therefore,electrohydraulic servo systems are widely used in the aerospace field,which requires high load power and fast response.The electro-hydraulic servo valve studied in this paper is a key control element of the aero-engineering system,which converts the small electrical signal input to the torque motor into a high-power hydraulic flow signal and pressure signal output through an internal multi-stage amplification mechanism to achieve precise control of the hydraulic actuator.Electro-hydraulic servo valves are characterized by high precision and high integration,and their failure situations are complex and difficult to identify effectively.This brings great hidden danger to the safe work of electro-hydraulic servo system,so it is especially important to study an efficient and reliable fault diagnosis algorithm for electrohydraulic servo valve.This paper focuses on the problem of fault diagnosis of the electro-hydraulic servo valve of the jet deflector plate,and explores an efficient,reliable and accurate fault diagnosis algorithm for the common types of faults of this servo valve,so as to provide an idea and an option to solve the problem of electro-hydraulic servo valve fault diagnosis.Firstly,the mathematical model of each component of the electro-hydraulic servo valve of the jet deflection version is established respectively by theoretical analysis to deeply understand and reveal its working principle,and thus obtain the state space equations of the whole valve;and theoretical analysis is conducted for the common failure forms of the electro-hydraulic servo valve to reveal the mechanism of its occurrence.Second,an artificial neural network algorithm is introduced in the investigation of fault diagnosis algorithm of electro-hydraulic servo valve,and a fault diagnosis algorithm combining convolutional neural network and gated recursive unit is proposed.Again,based on the mathematical model of the electro-hydraulic servo valve of the jet deflector plate,the simulation model of the electro-hydraulic servo valve is established by using the hydraulic simulation software,and the static characteristic curve of the electro-hydraulic servo valve is obtained through the simulation model,and some structural parameters inside the model are changed to simulate the common failure forms,and the changes of each curve before and after the failure are analyzed to provide reference and basis for the electro-hydraulic servo valve fault diagnosis algorithm and accelerated degradation experiment.Finally,in order to establish the correspondence between external fault forms and internal fault parts and to verify the effectiveness of the diagnosis algorithm proposed in this paper,an electro-hydraulic servo valve acceleration performance degradation experiment bench is built.The fault characteristics data of common faults of electrohydraulic servo valves are collected to clarify the characteristics of two fault forms of electro-hydraulic servo valve wear and blockage,and the data are used to train and verify the reliability and accuracy of the fault diagnosis algorithm proposed in this paper.