Technology And Application Research On Prognostics And Health Management Of An Aero-engine Anti-surge Control System

It has always been a major issue for aviation-developed countries of safeguarding flight safety from a technical perspective.Focusing on state monitoring and fault diagnosis techniques for the safety of engine flight,it has continued to be an important development since the 1950s.Today,equipped with prognostics and health management system has become an important symbol and a prominent feature of advanced fighter engines.As a modern ground attacking machine that can meet the depth of attacking combat radius and all-weather precision strike capability,an aircraft has become the backbone of aviation ground attack aircraft,and has very important strategic and tactical significance in national air defense and enemy deterrence.This type of aircraft is equipped with two aero-engines and belongs to gas turbine engines.The compressor has a common unstable working condition called surge.If the consequences of surging are extremely serious,the damage to the parts will be minor,and the safety will be jeopardized.In order to avoid surge during compressor operation,an anti-surge control system is installed on the engine.Compared with many control systems such as engine fuel system,oil system,and starting system,anti-surge control system has a more complex control relationship.The failure rate is higher.For this reason,this paper takes the Air Force Weapons Scientific Research Project“Resesrch on the prognostics and health management technology of aero-engine”as the research background.Through in-depth analysis of the working principle of the anti-surge control system,the three key parameters reflecting the working state of the system are determined to obtain the reflection prevention.The key parameters of the operating state of the anti-surge control system and the failure time reflecting the system reliability were designed and developed.Three key parameter measurement circuits were developed and system status monitoring methods were researched and developed.The key parameters reflecting the working state of the anti-surge control system were fully utilized and the system reliability was reflected.The reliability of aero-engine anti-surge control system was analyzed by the failure time,and the working status and remaining life of the system were predicted.Based on the above research results,an aero-engine ground integrated health management system was designed and developed.The main research content is:?1?On the basis of an in-depth analysis of the working principle of the anti-surge control system,three key parameters that reflect the working status of the system are determined.This paper starts with the basic principle of surge in axial compressors used in aero-engines,the conditions that cause surging,and the measures to prevent surge,and analyzes in depth the structure,working principle,and control laws of an anti-surge control system for aero-engines.According to the control law,the engine inlet temperature?T₁?,high-pressure compressor rotor speed?N_H?and adjustable guide vane angle?IGV?are determined to reflect the three key parameters of anti-surge control system.?2?In order to obtain the three key parameters reflecting the working status of the anti-surge control system and the failure time reflecting the system reliability,a key parameter measurement circuit was designed and developed,and a system condition monitoring method was researched and developed.In order to obtain the three key parameters that reflect the working status of the anti-surge control system,a measurement circuit for the engine inlet temperature,high-pressure compressor rotor speed,and adjustable guide vane angle was designed and developed.Three key state parameter measurements were analyzed through experiments.Accuracy,for the problem that measurement results are disturbed by noise,the denoising method based on wavelet analysis is studied.The signal-to-noise ratio and smoothness are used as evaluation indicators.The method of 6-layer Daubechies wavelet soft threshold denoising is determined through simulation and comparison.In order to obtain the failure time that reflects the reliability of the system,an anti-surge control system condition monitoring method was researched and developed,and the system control law curve reflected by the key state parameters of each anti-surge control system was used as a judgment basis.If the normal working boundary conditions were exceeded,record the flight time is the failure time of the anti-surge control system.?3?Based on the failure time that reflects the reliability of the system,the reliability of an anti-surge control system for aero-engine is analyzed.To fully understand the reliability of anti-surge control system for aero-engine,this paper proposes a reliability model estimation method based on Weibull distribution,support vector regression and artificial bee colony algorithm.The parameters of the Weibull distribution model are estimated by using the fault time that reflects the reliability of the system.By comparison,it is found that the maximum likelihood estimation of the Weibull distribution model under large sample conditions has large parameter bias and the effect of the least squares estimation is not ideal.This paper proposes that The Weibull distribution model parameter estimation method based on support vector regression is used to evaluate the root mean square error.The effectiveness of this algorithm is proved by simulation calculation.In order to further improve the estimation effect,the optimization based on artificial bee colony algorithm is proposed.Support vector machine penalty parameters and insensitive loss function methods still use the root mean square error as the evaluation index.The effect of the optimization algorithm is proved by simulation calculations.Finally,using a Weibull distribution model estimation result,an aero-engine anti-surge is analyzed of control system reliability.?4?Based on the key parameters reflecting the working status of the anti-surge control system,the state prediction of anti-surge control system status of aero-engineIn order to accurately predict the current working state of an aero-engine anti-surge control system,this paper presents a state prediction method based on self-organizing feature map neural network,hidden Markov model and simulated annealing algorithm.Aiming at the problem of hidden Markov model training in each state,the data classification method based on self-organizing feature mapping neural network was studied.The key parameters reflecting the working state of anti-surge control system were classified,and discrete observations in each state were generated.Sequence,for the observation sequence can not intuitively reflect the state of the system,study the construction of the prediction model based on hidden Markov model method to track the system performance degradation process;for the Baum-Welch algorithm training hidden Markov model parameters easily fall into local minimum for the value problem,an improved training algorithm combining the local optimization and fast convergence rate Baum-Welch method with the simulated degradation algorithm with the slow convergence rate of the global optimization ability is proposed,which improves the accuracy of the model estimation through simulation.The calculation validates the effectiveness of the algorithm.Based on the above state prediction model,a state prediction method of an anti-surge control system for aero-engine is studied and formulated.The prediction method is applied through specific examples.?5?Predicting the remaining life of anti-surge control system of aero-engine based on the failure time reflecting the reliability of the anti-surge control system and the key parameters reflecting the working state of the system.In order to grasp the remaining life of an aero-engine anti-surge control system,this paper proposes a residual life prediction method based on a combination of the proportional hazards model,hidden Markov model and particle swarm optimization algorithm.The relationship between the system failure rate and its working time and its concomitant variables is established by using the proportional hazards model.The state transition process of each concomitant variable is described by hidden Markov model;based on the particle swarm optimization algorithm,the reliability of the system is reflected.The failure time and key parameters reflecting the working state of the system estimate the parameters of the proportional failure rate model,and proposes an improved algorithm based on the LDW algorithm and the NLDW algorithm to solve the problem that the particle swarm optimization algorithm is easily trapped in the local solution.The simulation calculation shows that the comparison effectiveness of the algorithm;based on the above-mentioned ratio failure rate model,a method for predicting the remaining life of an aero-engine anti-surge control system was studied and formulated,and the remaining life of the system was calculated through specific examples.In order to translate the above research results into practical applications,an aero-engine ground integrated health management system was designed and developed based on prognostics and health management techniques.The program adopts a combination of Visual Basic and Matlab programming methods,including data management,state monitoring,fault prediction,evaluation reporting and other functions.Through the collection of system failure time and performance status information,accurately predict its working status and remaining life,and gives a status evaluation report.