Research On Lubricating Oil Abrasive Particles Monitoring Method Based On Coaxial Capacitive Sensor Network

Engine health management is an important means to prevent system or component failure,ensure engine safety and reduce maintenance costs by monitoring the performance status of the engine and predicting the trend of engine performance.As an important element of engine health management,on-line lubricant wear particle monitoring technology can obtain the internal wear condition of engine by monitoring wear particles in lubricating oil,which can provide timely warning of sudden failure and is of great significance for ensuring the safety of aircraft.The main contents of the research in this paper are as follows:1)This article summarizes the current research status of off-line detection and online monitoring of lubricant oil wear particles,compares and analyzes the advantages and disadvantages of typical lubricant wear particles monitoring sensors,and proposes a new technical idea by using coaxial capacitive sensing network for online monitoring of oil wear particles based on the coaxial capacitive sensors previously proposed by the group.2)This paper presents a theoretical analysis of sensors with different types of plates in the same sub-probe space within a coaxial capacitive sensing network,which includes non-parallel plate electrodes and parallel curved electrodes.The relationship between the capacitance signal of wear particles and the sensing network structure is also further clarified,which lays the theoretical foundation for the simulation and experiment.3)In this paper,a coaxial capacitive sensing network model was established in COMSOL software,and the parameters of the coaxial capacitive sensing network,including the structural parameters of the sensing network and the applied voltage of the excitation electrode plate,were determined by simulation.After determining the sensing network parameters,the effects of different sizes and numbers of abrasive particles on different capacitance values of the sensing network were simulated,and the results showed that the simulated capacitance values of different types of electrode plates increased as the size and number of abrasive particles increased.4)In this paper,a coaxial capacitive multi-channel sensing network is prepared,and an experimental platform is designed for on-line monitoring of lubricant particles,and then an algorithm based on EMD and DWT is proposed to extract the monitored signals,and the relationship between the size and number of wear particles and the output value of the sensing network is verified.5)The capacitance data of different types of electrodes obtained from the simulation are used for SVM training to generate the wear particle classification model,including the qualitative classification model of abrasive particles under actual working conditions and the qualitative classification model of abrasive particles used in experiments,and the experimental data are imported into the model for verification,which finally shows that the coaxial capacitive sensing network can achieve quantitative identification and qualitative classification of wear particles of lubricating oil.