Aero-engine Health Monitoring Technology Based On Electrostatic Induction

With the success of the first flight of the C919 large aircraft,this marked a new era for the rapid development of China's aviation industry.The aero-engine as the "heart" of the aircraft is the most important of all system components.Based on this,on-line monitoring and diagnosis of the aeroengine core system(gas lines,oil lines,etc.)are particularly important.However,at present,there are not very good solutions in our country and even in the world.The electrostatic monitoring technology is a technology that has the ability of on-line monitoring diagnosis and prediction.In recent years,the technology has been applied to aeroengine research more and more.The principle is that when the engine air circuit or oil system failure will produce a large number of abnormally charged particles,compared to the normal production of such particles with large particle size and high concentration characteristics,carry the charge is much larger than the normal range,Electrostatic sensors are used to monitor the charge level of abnormal particles and to evaluate the performance of engine components.The paper,aero-engine online monitoring is difficult to achieve,and used electrostatic induction is proposed to monitor aeroengine components.The main research contents are as follows: 1.Introduced the commonly used monitoring methods and electrostatic monitoring technology theoretical foundation,analyzed the aeroengine gas and oil path of particle failure sources and charging mechanism;2.Completed the sensor structure model design,And the mathematical model is established.The spatial distribution of the sensitivity of the sensing probe is analyzed by finite element analysis.The main factors affecting the sensitivity of the probe are studied,and the optimal design of the probe model is completed;3.High sensitivity preamplifier circuit,and further signal processing,including rectification and low-pass filter design,on the basis of a second-stage amplifier circuit design,and finally in order to facilitate data storage and read the signal was carried out A / D conversion;4.In order to realize the monitoring of aeroengine components and simulate the real test environment,set up a simulation test stand,measured a large number of experimental data,from the charged particle size and concentration of the two major factors are analyzed,the first is through Fitting analysis of particle size and monitoring signals gives approximate quantitative relationships;then focus on particle concentration and monitoring signals were analyzed,the six characteristic signals of monitoring signals were extracted,and the order of reduction was obtained by the method of factor analysis.The order of the reduced order was used as input.The BP neural network method based on genetic algorithm optimization was established mathematical model,through the optimal training obtained the relationship between the signal and concentration.