A method for analytical redundancy design of variable cycle engines

Abstract

The present invention relates to the technical field of aviation engine safety design, and specifically discloses a variable cycle engine resolution margin design method, which includes: determining the performance health parameters of the engine gas path components and the nonlinear mathematical model of the variable cycle engine; obtaining the state parameters of the variable cycle engine adaptive dynamic model based on the calculation of the extended Kalman filter; calculating the resolution margin of the variable cycle engine according to the state parameters of the variable cycle engine adaptive dynamic model; determining the engine fault type according to the resolution margin of the variable cycle engine and the sensor fault diagnosis threshold; determining the corresponding engine dynamic model reconstruction fault sensor according to the engine fault type, and replacing the fault sensor signal in the engine sensor measurement according to the signal of the fault sensor reconstructed by the determined engine dynamic model. The variable cycle engine resolution margin design method provided by the present invention can effectively improve the reliability of sensor fault diagnosis.

Description

Technical Field

[0001] The invention relates to the technical field of aviation engine safety design, and in particular to a variable cycle engine analytical redundancy design method. Background Art

[0002] With the development of modern flight technology, traditional turbojet and turbofan engines are difficult to meet the needs of aircraft under different flight missions due to their relatively fixed working modes. Variable cycle engines have become the preferred aviation power unit of the new generation due to their unique advantages. Variable cycle engine (VCE) is a gas turbine engine that changes its thermodynamic cycle by changing the geometry, size and position of certain engine components. The control of variable cycle engines is still mainly based on the Full Authority Digital Electronic Control (FADEC) system. The normal operation of engine sensors is an important prerequisite for the safe and stable operation of the engine digital control system. In particular, the control parameters of variable cycle engines have increased by more than 40% compared to traditional engines. Engine sensors are very prone to failure when working in high temperature and high pressure environments. At present, the main ways to improve the fault tolerance of engine control systems are hardware redundancy and analytical redundancy.

[0003] The main problems caused by hardware redundancy are complex structure, large volume, high cost, and increased engine weight, which reduces the aircraft payload. The method of introducing analytical redundancy can partially replace hardware redundancy. After the engine sensor fails, the analytical signal is used to replace the fault signal, which improves the fault tolerance of the control system while reducing the cost and weight of hardware redundancy. Therefore, it is of great significance to establish a system for diagnosing, isolating and reconstructing engine sensor faults to improve the reliability of variable cycle engine control systems and reduce costs and weight. Since analytical redundancy depends on the analytical mathematical model of the engine, the accuracy of the model is crucial for the diagnosis and reconstruction of engine sensor faults. Performance degradation will inevitably occur during the life cycle of the engine, resulting in a mismatch between the model and the real engine. In general analytical redundancy research, the problem of insufficient accuracy of sensor fault diagnosis and reconstruction signals caused by engine performance degradation is not considered, resulting in low reliability of sensor fault diagnosis.

[0004] Therefore, how to improve the reliability of sensor fault diagnosis becomes a technical problem that needs to be solved urgently by those skilled in the art. Summary of the invention

[0005] The present invention provides a variable cycle engine analytical redundancy design method to solve the problem of low reliability of fault diagnosis of sensors existing in the related art.

[0006] As one aspect of the present invention, a method for designing analytical redundancy of a variable cycle engine is provided, which comprises:

[0007] Determine the performance health parameters of engine gas path components and the nonlinear mathematical model of variable cycle engines;

[0008] Performing calculation based on extended Kalman filtering according to the engine gas path component performance health parameters and the nonlinear mathematical model to obtain state parameters of the variable cycle engine adaptive dynamic model;

[0009] Calculating the analytical margin of the variable cycle engine according to the state parameters of the variable cycle engine adaptive dynamic model;

[0010] Determining an engine fault type according to the analytical margin of the variable cycle engine and a sensor fault diagnosis threshold;

[0011] The corresponding engine dynamic model reconstruction fault sensor is determined according to the engine fault type, and the signal of the fault sensor reconstructed according to the determined engine dynamic model replaces the fault sensor signal in the engine sensor measurement.

[0012] Furthermore, the performance health parameters of the engine gas path components and the nonlinear mathematical model of the variable cycle engine are determined, including:

[0013] The nonlinear mathematical model of the double-external variable cycle engine is established according to the engine gas path component analysis method, wherein the expression of the nonlinear mathematical model is: ,

[0014] in, k Represents the time parameter, x k Indicates k The system state quantity at time, u k Indicates k The system input at time y k Indicates k The sensor measurement value at the time, f( ) represents the state transfer equation of the variable cycle engine model, g( ) represents the measurement equation of the variable cycle engine model, Indicates k The state noise at the moment, Indicates k The measurement noise at each moment;

[0015] The degree of change in component performance after a sensor failure occurs is quantified according to the engine gas path component performance health parameter, wherein the engine gas path component performance health parameter h Selected as the efficiency variation coefficient of the rotating parts SEi and flow rate variation coefficient SW i , ,

[0016] in, Indicates the actual efficiency of the component, Indicates the actual flow of the component, represents the ideal value of component efficiency, represents the ideal value of the component flow, i Indicates the number of a part.

[0017] Furthermore, the expression of the nonlinear mathematical model k System state quantity at time x k =[N L , N H ] T , where N L Indicates the low pressure rotor speed, N H Indicates the high pressure rotor speed;

[0018] No. k System input at time u k =[ W f , A 8], where W f Indicates the fuel supply to the main combustion chamber. A 8 represents the throat area;

[0019] No. k Sensor measurement value at time y k =[N L , N H , T 21 , P 21 , T 15 , P 15 , T3, P3, T5, P5], where N L Indicates the low pressure rotor speed, N H Indicates the high pressure rotor speed, T 21 Indicates the fan outlet section temperature, P 21 Indicates the fan internal outlet cross-sectional pressure, T 15 Indicates the temperature of the front mixing chamber, P 15 represents the front mixing chamber pressure, T3 represents the compressor outlet temperature, P3 represents the compressor outlet pressure, T5 represents the low-pressure turbine outlet temperature, and P5 represents the low-pressure turbine outlet pressure.

[0020] Furthermore, according to the engine gas path component performance health parameters and the nonlinear mathematical model, a calculation based on extended Kalman filtering is performed to obtain state parameters of the variable cycle engine adaptive dynamic model, including:

[0021] The engine gas path component performance health parameters h Extending to the system state quantity of the nonlinear mathematical model, we get x =[ N L , N H , h T ] T ;

[0022] According to the calculation process of the extended Kalman filter, the time update calculation and the measurement update calculation are respectively performed to obtain the estimated value of the performance health parameter of the engine gas path component;

[0023] The estimated values ​​of the performance health parameters of the engine gas path components are input into the variable cycle engine model, and the characteristics of the corresponding rotating components are corrected so that the variable cycle engine model can track the output of the real engine and obtain the state parameters of the variable cycle engine adaptive model.

[0024] Further, according to the calculation process of the extended Kalman filter, the time update calculation and the measurement update calculation are respectively performed to obtain the estimated value of the performance health parameter of the engine gas path component, including:

[0025] The computation of the time update involves computing the prior state estimate and the prior state covariance matrix:

[0026] ,

[0027] ;

[0028] in, P k Indicates k System state quantity at time x k The covariance of Q represents the variance of the state noise, represents the Jacobian matrix;

[0029] The measurement update calculation includes calculating the Kalman filter gain matrix and updating the prior state and covariance matrix according to the measurement signal to obtain the posterior estimate and covariance matrix:

[0030] ,

[0031] ,

[0032] ,

[0033] in, P k Indicates k System state quantity at time x k The covariance of represents the Kalman filter gain, represents the Jacobian matrix, represents the variance matrix of the measurement noise, represents the identity matrix;

[0034] in, .

[0035] Further, the analytical margin of the variable cycle engine is calculated according to the state parameters of the variable cycle engine adaptive dynamic model, including:

[0036] Create an N-dimensional buffer and store the system state within N steps , ;

[0037] For the system state quantity within N steps Calculate the average value and get k The average value of the system state quantity at time ,

[0038] ,

[0039] Wherein, N is a natural number greater than 1;

[0040] The state parameters of the variable cycle engine adaptive dynamic model are updated according to the estimation result of the system state quantity to obtain the analytical margin of the variable cycle engine.

[0041] Further, the state parameters of the variable cycle engine adaptive dynamic model are updated according to the estimation result of the system state quantity to obtain the analytical margin of the variable cycle engine, including:

[0042] When the sensor is not faulty or the pressure sensor is faulty, k System state quantity at time x k Update the state parameters of the variable cycle engine adaptive dynamic model to obtain the analytical redundancy of the variable cycle engine :

[0043] ;

[0044] When the speed sensor and temperature sensor fail, k The average value of the system state quantity at time The state parameters of the variable cycle engine adaptive dynamic model are updated to obtain the analytical redundancy of the variable cycle engine:

[0045] ,

[0046] in, Indicates k Estimated value of the adaptive dynamic model of a variable cycle engine at time instant.

[0047] Further, determining the engine fault type according to the resolution margin of the variable cycle engine and the sensor fault diagnosis threshold includes:

[0048] The covariance of the error between the sensor value of the variable cycle engine and the estimated value of the variable cycle engine adaptive dynamic model is calculated according to the analytical margin of the variable cycle engine:

[0049] The first j The sensor in k The measured value at the moment is replaced by Middle j The sensor in k The estimated value of the adaptive dynamic model of the variable cycle engine at time is obtained :

[0050] ,

[0051] For each sensor, the k The sensor measurement value at the moment k The covariance of the error between the estimated values ​​of the adaptive dynamic model of the variable cycle engine at time :

[0052] ,

[0053] in, represents a constant-valued diagonal matrix;

[0054] According to the covariance and sensor fault diagnosis thresholds to determine the type of engine fault.

[0055] Further, according to the covariance and sensor fault diagnosis thresholds to determine the type of engine fault, including:

[0056] The covariance comparing with a sensor fault diagnosis threshold;

[0057] If the covariance is not greater than the sensor fault diagnosis threshold, it is determined that no fault has occurred in the variable cycle engine;

[0058] If the covariance of the nth row is are all less than the sensor fault diagnosis threshold, and the covariance of the other n-1 rows If both are greater than the sensor fault diagnosis threshold, it is determined that the nth sensor is faulty;

[0059] The calculation formula of the sensor fault diagnosis threshold is:

[0060] ,

[0061] in, represents the sensor fault diagnosis threshold, k Indicates k time, s Indicates the sliding buffer length, represents the magnification factor, Represents the high-pressure rotor speed of the variable-cycle engine adaptive dynamic model.

[0062] Further, determining the corresponding engine dynamic model reconstruction fault sensor according to the engine fault type, and replacing the fault sensor signal in the engine sensor measurement according to the signal of the determined engine dynamic model reconstruction fault sensor, including:

[0063] When the speed sensor and / or temperature sensor fails, k The average value of the system state quantity at time Run the engine dynamics model to reconstruct the faulty sensor:

[0064] ;

[0065] When the pressure sensor fails, k System state quantity at time x k Run the engine dynamics model to reconstruct the faulty sensor:

[0066] ;

[0067] Reconstruct the fault sensor signal by reconstructing the engine dynamic model Replace the nth fault sensor signal in the engine sensor measurement with the nth signal in: .

[0068] The variable cycle engine analytical margin design method provided by the present invention establishes an adaptive dynamic model of the variable cycle engine by extending the Kalman filter, and the analytical margin of the variable cycle engine is adaptively calculated based on the sensor fault diagnosis threshold, providing a feasible solution for effectively improving the reliability of sensor fault diagnosis and the accuracy of analytical margin when the variable cycle engine suffers from performance degradation. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the present invention but do not constitute a limitation of the present invention. In the accompanying drawings:

[0070] Figure 1 The present invention provides a flow chart of the variable cycle engine analytical redundancy design method.

[0071] Figure 2a The ground single culvert mode P provided by the present invention 15 Schematic diagram of the adaptive model output for sensor failure simulation.

[0072] Figure 2b The ground single culvert mode P provided by the present invention 15 Schematic diagram of signal reconstruction comparison for sensor failure simulation.

[0073] Figure 3a The ground single culvert dynamic deceleration process T provided by the present invention 15 Schematic diagram of the adaptive model output for sensor failure simulation.

[0074] Figure 3b The ground single culvert dynamic deceleration process T provided by the present invention 15 Schematic diagram of signal reconstruction comparison for sensor failure simulation.

[0075] Figure 4a The ground double culvert mode T provided by the present invention 21 Schematic diagram of the adaptive model output for sensor failure simulation.

[0076] Figure 4b The ground double culvert mode T provided by the present invention 21 Schematic diagram of signal reconstruction comparison for sensor failure simulation.

[0077] Figure 5a The ground double culvert dynamic acceleration process N provided by the present invention H Schematic diagram of the adaptive model output for sensor failure simulation.

[0078] Figure 5b The ground double culvert dynamic acceleration process N provided by the present invention H Schematic diagram of signal reconstruction comparison for sensor failure simulation. DETAILED DESCRIPTION

[0079] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0080] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present invention.

[0081] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so as to describe the embodiments of the present invention described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0082] The safety and reliability of aircraft engine measurement components is an important prerequisite for the stable operation of the engine control system. The control variables of variable cycle engines increase by 20% to 40% compared with traditional engines. Therefore, it is of great significance to carry out research on analytical redundancy technology of variable cycle engine control system and design sensor fault diagnosis, isolation and signal reconstruction system of variable cycle engine control system to improve the reliability and fault tolerance of variable cycle engine control system. The analytical redundancy technology depends on the accuracy of the engine mathematical model. The performance degradation of the engine will inevitably occur during the service life, which will lead to mismatch between the model and the engine.

[0083] Therefore, in view of the variable cycle engine's requirements for control system reliability and fault tolerance, a variable cycle engine analytical redundancy design method is provided in this embodiment. Figure 1 is a flow chart of a method for designing analytical redundancy of a variable cycle engine according to an embodiment of the present invention. Figure 1 As shown, including:

[0084] S110, determining performance health parameters of engine gas path components and a nonlinear mathematical model of a variable cycle engine;

[0085] In the embodiment of the present invention, it specifically includes:

[0086] The nonlinear mathematical model of the double-external variable cycle engine is established according to the engine gas path component analysis method, wherein the expression of the nonlinear mathematical model is: ,in, k Represents the time parameter, x k Indicates k The system state quantity at time, u k Indicates k The system input at time y k Indicates k The sensor measurement value at the time, f( ) represents the state transfer equation of the variable cycle engine model, g( ) represents the measurement equation of the variable cycle engine model, Indicates k The state noise at the moment, Indicates k The measurement noise at each moment;

[0087] The degree of change in component performance after a sensor failure occurs is quantified according to the engine gas path component performance health parameter, wherein the engine gas path component performance health parameter h Selected as the efficiency variation coefficient of the rotating parts SE i and flow rate variation coefficient SW i , ,

[0088] in, Indicates the actual efficiency of the component, Indicates the actual flow of the component, represents the ideal value of component efficiency, represents the ideal value of the component flow, i Indicates the number of a part.

[0089] More specifically, the expression of the nonlinear mathematical model k System state quantity at time x k =[N L , N H ] T , where N L Indicates the low pressure rotor speed, N H Indicates the high pressure rotor speed;

[0090] No. k System input at time u k =[ Wf , A 8], where W f Indicates the fuel supply to the main combustion chamber. A 8 represents the throat area;

[0091] No. k Sensor measurement value at time y k =[N L , N H , T 21 , P 21 , T 15 , P 15 , T3, P3, T5, P5], where N L Indicates the low pressure rotor speed, N H Indicates the high pressure rotor speed, T 21 Indicates the fan outlet section temperature, P 21 Indicates the fan internal outlet cross-sectional pressure, T 15 Indicates the temperature of the front mixing chamber, P 15 represents the front mixing chamber pressure, T3 represents the compressor outlet temperature, P3 represents the compressor outlet pressure, T5 represents the low-pressure turbine outlet temperature, and P5 represents the low-pressure turbine outlet pressure.

[0092] It should be noted that the engine gas circuit component performance health parameters can reflect the performance health status of the engine gas circuit components, such as sudden gas circuit failure or gradual performance degradation, etc. The engine gas circuit component performance health parameters are used to quantify the degree of sudden change and gradual degradation of the performance of each component after a failure occurs.

[0093] S120, performing calculation based on extended Kalman filtering according to the engine gas path component performance health parameters and the nonlinear mathematical model to obtain state parameters of the variable cycle engine adaptive dynamic model;

[0094] In the embodiment of the present invention, the following may be specifically included:

[0095] The engine gas path component performance health parameters h Extending to the system state quantity of the nonlinear mathematical model, we get x =[ N L , N H , h T ] T ;

[0096] According to the calculation process of the extended Kalman filter, the time update calculation and the measurement update calculation are respectively performed to obtain the estimated value of the performance health parameter of the engine gas path component;

[0097] The estimated values ​​of the performance health parameters of the engine gas path components are input into the variable cycle engine model, and the characteristics of the corresponding rotating components are corrected so that the variable cycle engine model can track the output of the real engine and obtain the state parameters of the variable cycle engine adaptive model.

[0098] In an embodiment of the present invention, the engine gas path components may specifically include mechanical components such as a fan, a compressor, and a turbine.

[0099] For example, in the embodiment of the present invention, i = 5 as an example, the component health parameters are limited by the location and number of actual engineering measurement parameters, and the health parameter vector is selected. h =[SE1 SW1 SE2 SW2 SE3 SW3 SE4 SW4] T , expand the health parameters to the state quantity, x =[ N L , N H , h T ] T . The superscript T represents transpose.

[0100] Further, according to the calculation process of the extended Kalman filter, the time update calculation and the measurement update calculation are respectively performed to obtain the estimated value of the performance health parameter of the engine gas path component, including:

[0101] The computation of the time update involves computing the prior state estimate and the prior state covariance matrix:

[0102] ,

[0103] ;

[0104] in, P k Indicates k System state quantity at time x k The covariance of Q represents the variance of the state noise, represents the Jacobian matrix;

[0105] The measurement update calculation includes calculating the Kalman filter gain matrix and updating the prior state and covariance matrix according to the measurement signal to obtain the posterior estimate and covariance matrix:

[0106] ,

[0107] ,

[0108] ,

[0109] in, P k Indicates k System state quantity at time x k The covariance of represents the Kalman filter gain, represents the Jacobian matrix, represents the variance matrix of the measurement noise, represents the identity matrix;

[0110] in, .

[0111] S130, calculating the analytical margin of the variable cycle engine according to the state parameters of the variable cycle engine adaptive dynamic model;

[0112] In the embodiment of the present invention, an N-dimensional buffer is established to store the system state quantity within N steps. , ;

[0113] For the system state quantity within N steps Calculate the average value and get k The average value of the system state quantity at time ,

[0114] ,

[0115] Wherein, N is a natural number greater than 1;

[0116] The state parameters of the variable cycle engine adaptive dynamic model are updated according to the estimation result of the system state quantity to obtain the analytical margin of the variable cycle engine.

[0117] In the embodiment of the present invention, taking N=10 as an example, a 10-dimensional buffer is established to store the state quantity x within 10 steps, and the average is calculated.

[0118] ,

[0119] .

[0120] Further, the state parameters of the variable cycle engine adaptive dynamic model are updated according to the estimation result of the system state quantity to obtain the analytical margin of the variable cycle engine, including:

[0121] When the sensor is not faulty or the pressure sensor is faulty, k System state quantity at time x kUpdate the state parameters of the variable cycle engine adaptive dynamic model to obtain the analytical redundancy of the variable cycle engine :

[0122] ;

[0123] When the speed sensor and temperature sensor fail, k The average value of the system state quantity at time Update the state parameters of the variable cycle engine adaptive dynamic model to obtain the analytical redundancy of the variable cycle engine :

[0124] ,

[0125] in, Indicates k Estimated value of the adaptive dynamic model of a variable cycle engine at time instant.

[0126] It should be understood that, in the embodiment of the present invention, the analytical margin of the variable cycle engine is the estimated value of the adaptive dynamic model of the variable cycle engine.

[0127] S140, determining the engine fault type according to the resolution margin of the variable cycle engine and the sensor fault diagnosis threshold, including:

[0128] The covariance of the error between the sensor value of the variable cycle engine and the estimated value of the variable cycle engine adaptive dynamic model is calculated according to the analytical margin of the variable cycle engine:

[0129] The first j The sensor in k The measured value at the moment is replaced by Middle j The sensor in k The estimated value of the adaptive dynamic model of the variable cycle engine at time is obtained :

[0130] ,

[0131] For each sensor, the k The sensor measurement value at the moment k The covariance of the error between the estimated values ​​of the adaptive dynamic model of the variable cycle engine at time :

[0132] ,

[0133] in, represents a constant diagonal matrix, generally the standard deviation of the measurement noise;

[0134] According to the covariance and sensor fault diagnosis thresholds to determine the type of engine fault.

[0135] In the embodiment of the present invention, according to the covariance and sensor fault diagnosis thresholds to determine the type of engine fault, including:

[0136] The covariance comparing with a sensor fault diagnosis threshold;

[0137] If the covariance is not greater than the sensor fault diagnosis threshold, it is determined that no fault has occurred in the variable cycle engine;

[0138] If the covariance of the nth row is are all less than the sensor fault diagnosis threshold, and the covariance of the other n-1 rows If both are greater than the sensor fault diagnosis threshold, it is determined that the nth sensor is faulty.

[0139] Specifically, the calculation formula of the sensor fault diagnosis threshold is:

[0140] ,

[0141] in, represents the sensor fault diagnosis threshold, k Indicates k time, s Indicates the sliding buffer length, represents the magnification factor, Represents the high-pressure rotor speed of the variable-cycle engine adaptive dynamic model.

[0142] S150, determining a corresponding engine dynamic model reconstruction fault sensor according to the engine fault type, and replacing the fault sensor signal in the engine sensor measurement with the signal of the fault sensor reconstructed according to the determined engine dynamic model.

[0143] In the embodiment of the present invention, when the speed sensor and / or the temperature sensor fails, k The average value of the system state quantity at time Run the engine dynamics model to reconstruct the faulty sensor:

[0144] ;

[0145] When the pressure sensor fails, k System state quantity at time x k Run the engine dynamics model to reconstruct the faulty sensor:

[0146] ;

[0147] Reconstruct the fault sensor signal by reconstructing the engine dynamic model Replace the nth fault sensor signal in the engine sensor measurement with the nth signal in:

[0148] .

[0149] In summary, the variable cycle engine analytical margin design method provided in the embodiment of the present invention establishes an adaptive dynamic model of the variable cycle engine by extending the Kalman filter, and the analytical margin of the variable cycle engine is adaptively calculated based on the sensor fault diagnosis threshold, which provides a feasible solution for effectively improving the reliability of sensor fault diagnosis and the accuracy of analytical margin when the variable cycle engine suffers from performance degradation.

[0150] Under the above processing steps, the present invention carries out the desktop simulation of single and double external culverts of the variable cycle engine numerical control system analytical redundancy, which is in line with expectations. In the single and double external culvert modes of the variable cycle engine, the sensor bias failure is simulated after the engine gas path performance is degraded, and the sensor failure is diagnosed and the signal is reconstructed to verify the effectiveness of sensor fault diagnosis and the accuracy of reconstructed signals after the engine performance is degraded. In the simulation, the amplification factor γ=2.3 and the sliding buffer length s=2 are taken.

[0151] At the ground point, the variable cycle engine mode selection valve is closed, and the front and rear adjustable duct ejector areas are reduced to put the engine in single duct operating mode.

[0152] Sensor P when compressor flow rate degrades by 2% at 1.25s 15 At 2.5s, a -3% bias fault occurs. Some outputs of the adaptive model, sensor fault indication parameters with and without adaptive model, and signal reconstruction simulation results are shown in Figure 2. Figure 2a and Figure 2b As shown. Figure 2a As shown in the figure, it can be seen that the adaptive model tracks the output of the real engine well after the engine gas path performance degradation occurs. Figure 2b It can be seen that the reconstructed signal without the adaptive model deviates significantly from the true signal, and due to misdiagnosis, the reconstructed signal returns to the default value ( Figure 2b As pointed out in , its average error relative to the true value is 0.4%, while the average error of the reconstructed signal using the adaptive model is 0.1%. Compared with the reconstructed signal without model adaptation, the latter more accurately reflects the true signal value.

[0153] At 1.25s, the fan efficiency degrades by 2%. From 2.5s to 10s, the engine undergoes dynamic deceleration. At 3.75s, the fan efficiency degrades by 2%. 15 The sensor has a -3% bias fault, and the dynamic simulation results are as follows Figure 3a and Figure 3b As shown. Figure 3a It can be seen from the dynamic deceleration process of a single culvert on the ground that the adaptive model is able to track the output of the real engine during both performance degradation and dynamic processes. Figure 3b This reflects that the reconstructed signal using the adaptive model matches the real signal more accurately than the reconstructed signal without the adaptive model, which illustrates the effectiveness and accuracy of the redundancy strategy in the dynamic process under the single culvert mode.

[0154] At the ground point, the variable cycle engine opens the mode selection valve, enlarges the front and rear adjustable duct ejector areas, increases the duct flow rate, increases the bypass ratio, and puts the engine in dual duct working mode.

[0155] Sensor T in case of 1% degradation of high pressure turbine efficiency at 1.25s 21 A -3% bias fault occurs at 2.5s. The gas path performance estimation results, partial output of the adaptive model, sensor fault indication parameters, and signal reconstruction simulation results are shown in Figure 2. Figure 4a and Figure 4b As shown. Figure 4a It can be seen that the variable cycle engine is in a lower state in the dual outer bypass mode than in the single outer bypass mode. After the high-pressure turbine efficiency degrades by 1%, the adaptive model accurately tracks the output of the real engine. Figure 4b It can be seen that although the reconstructed signal of the faulty sensor without the adaptive model coincides with the real signal, it can be seen that T 21 The sensor does not change significantly after the high-pressure turbine efficiency degrades. The reconstructed signal using the adaptive model not only coincides with the real signal but also has the same jitter trend as the real signal.

[0156] At 1.25s, the compressor efficiency degrades by 2%. During 2.5s to 10s, the engine undergoes dynamic acceleration. At 3.75s, the N H The sensor has a -3% bias fault, and the dynamic simulation results are as follows Figure 5a and Figure 5b shown. Figure 5a The simulation results of the ground double culvert dynamic acceleration process once again demonstrate the superiority of the analytical redundancy strategy using adaptive model design. Figure 5b It can be seen that there is a steady-state error between the reconstructed signal without the adaptive model and the true value, and the reconstruction accuracy is significantly lower than that of the reconstructed signal with the adaptive model. The simulation results verify the effectiveness of the analytical redundancy method in the dual-containment mode.

[0157] In summary, the variable cycle engine analytical redundancy design method provided by the present invention establishes a variable cycle engine adaptive model by adopting an extended Kalman filter, designs a variable cycle engine analytical redundancy logic strategy on this basis, and introduces threshold adaptive dynamic calculation. Simulation analysis shows that this method can effectively improve the reliability of sensor fault diagnosis and the accuracy of analytical redundancy when the variable cycle engine suffers from performance degradation.

[0158] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present invention, but the present invention is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A variable cycle engine analytical redundancy design method, characterized in that: include: Determine the performance health parameters of engine gas path components and the nonlinear mathematical model of variable cycle engines; Performing calculation based on extended Kalman filtering according to the engine gas path component performance health parameters and the nonlinear mathematical model to obtain state parameters of the variable cycle engine adaptive dynamic model; Calculating the analytical margin of the variable cycle engine according to the state parameters of the variable cycle engine adaptive dynamic model; Determining an engine fault type according to the analytical margin of the variable cycle engine and a sensor fault diagnosis threshold; Determine the corresponding engine dynamic model reconstruction fault sensor according to the engine fault type, and replace the fault sensor signal in the engine sensor measurement with the signal of the fault sensor reconstructed according to the determined engine dynamic model; Determine the performance health parameters of the engine gas path components and the nonlinear mathematical model of the variable cycle engine, including: The nonlinear mathematical model of the double-external variable cycle engine is established according to the engine gas path component analysis method, wherein the expression of the nonlinear mathematical model is: ,in, k Represents the time parameter, x k Indicates k The system state quantity at time, u k Indicates k The system input at time y k Indicates k The sensor measurement value at the time, f( ) represents the state transfer equation of the variable cycle engine model, g( ) represents the measurement equation of the variable cycle engine model, Indicates k The state noise at the moment, Indicates k The measurement noise at each moment; The degree of change in component performance after a sensor failure occurs is quantified according to the engine gas path component performance health parameter, wherein the engine gas path component performance health parameter h Selected as the efficiency variation coefficient of the rotating parts SE i and flow rate variation coefficient SW i , ,in, Indicates the actual efficiency of the component, Indicates the actual flow of the component, represents the ideal value of component efficiency, represents the ideal value of the component flow, i Indicates the number of the component; Calculating the analytical margin of the variable cycle engine according to the state parameters of the variable cycle engine adaptive dynamic model includes: Create an N-dimensional buffer and store the system state within N steps , ; For the system state quantity within N steps Calculate the average value and get k The average value of the system state quantity at time , , Wherein, N is a natural number greater than 1; updating the state parameters of the variable cycle engine adaptive dynamic model according to the estimation result of the system state quantity to obtain the analytical redundancy of the variable cycle engine; The state parameters of the variable cycle engine adaptive dynamic model are updated according to the estimation result of the system state quantity to obtain the analytical margin of the variable cycle engine, including: when the sensor is not faulty or the pressure sensor is faulty, according to the first k System state quantity at time x k Update the state parameters of the variable cycle engine adaptive dynamic model to obtain the analytical redundancy of the variable cycle engine : ; When the speed sensor and temperature sensor fail, k The average value of the system state quantity at time Update the state parameters of the variable cycle engine adaptive dynamic model to obtain the analytical redundancy of the variable cycle engine : ,in, Indicates k The estimated value of the adaptive dynamic model of the variable cycle engine at the moment; Determining the engine fault type according to the resolution margin of the variable cycle engine and the sensor fault diagnosis threshold includes: The covariance of the error between the sensor value of the variable cycle engine and the estimated value of the variable cycle engine adaptive dynamic model is calculated according to the analytical margin of the variable cycle engine: The first j The sensor in k The measured value at the moment is replaced by Middle j The sensor in k The estimated value of the adaptive dynamic model of the variable cycle engine at time is obtained : , for each sensor, calculate the k The sensor measurement value at the moment k The covariance of the error between the estimated values ​​of the adaptive dynamic model of the variable cycle engine at time : ,in, represents a constant-valued diagonal matrix; According to the covariance and sensor fault diagnosis thresholds to determine the type of engine fault; According to the covariance and sensor fault diagnosis thresholds to determine the type of engine fault, including: The covariance comparing with a sensor fault diagnosis threshold; If the covariance is not greater than the sensor fault diagnosis threshold, it is determined that no fault has occurred in the variable cycle engine; If the covariance of the nth row are all less than the sensor fault diagnosis threshold, and the covariance of the other n-1 rows If both are greater than the sensor fault diagnosis threshold, it is determined that the nth sensor is faulty; The calculation formula of the sensor fault diagnosis threshold is: , in, represents the sensor fault diagnosis threshold, k Indicates k time, s Indicates the sliding buffer length, represents the magnification factor, Represents the high-pressure rotor speed of the variable cycle engine adaptive dynamic model; The method comprises: determining a corresponding engine dynamic model to reconstruct a fault sensor according to the engine fault type, and replacing a fault sensor signal in the engine sensor measurement according to a signal of the fault sensor reconstructed by the determined engine dynamic model, including: when a speed sensor and / or a temperature sensor fails, k The average value of the system state quantity at time Run the engine dynamics model to reconstruct the faulty sensor: ; When the pressure sensor fails, k System state quantity at time x k Run the engine dynamics model to reconstruct the faulty sensor: ; Reconstruct the fault sensor signal by reconstructing the engine dynamic model Replace the nth fault sensor signal in the engine sensor measurement with the nth signal in: .

2. The method according to claim 1, characterized in that The expression of the nonlinear mathematical model is k System state quantity at time x k =[N L , N H ] T , where N L Indicates the low pressure rotor speed, N H Indicates the high pressure rotor speed; No. k System input at time u k =[ W f , A 8], where W f Indicates the fuel supply to the main combustion chamber. A 8 represents the throat area; No. k Sensor measurement value at time y k =[N L , N H , T 21 , P 21 , T 15 , P 15 , T3, P3, T5, P5], where N L Indicates the low pressure rotor speed, N H Indicates the high pressure rotor speed, T 21 Indicates the fan outlet section temperature, P 21 Indicates the fan internal outlet cross-sectional pressure, T 15 Indicates the temperature of the front mixing chamber, P 15 represents the front mixing chamber pressure, T3 represents the compressor outlet temperature, P3 represents the compressor outlet pressure, T5 represents the low-pressure turbine outlet temperature, and P5 represents the low-pressure turbine outlet pressure.

3. The method according to claim 1, characterized in that: According to the engine gas path component performance health parameters and the nonlinear mathematical model, calculation based on extended Kalman filtering is performed to obtain state parameters of the variable cycle engine adaptive dynamic model, including: The engine gas path component performance health parameters h Extending to the system state quantity of the nonlinear mathematical model, we get x =[ N L , N H , h T ] T ; According to the calculation process of the extended Kalman filter, the time update calculation and the measurement update calculation are respectively performed to obtain the estimated value of the performance health parameter of the engine gas path component; The estimated values ​​of the performance health parameters of the engine gas path components are input into the variable cycle engine model, and the characteristics of the corresponding rotating components are corrected so that the variable cycle engine model can track the output of the real engine and obtain the state parameters of the variable cycle engine adaptive model.

4. The method according to claim 3, characterized in that According to the calculation process of the extended Kalman filter, the time update calculation and the measurement update calculation are respectively performed to obtain the estimated value of the performance health parameter of the engine gas path component, including: The computation of the time update involves computing the prior state estimate and the prior state covariance matrix: ， ； in, P k Indicates k System state quantity at time x k The covariance of Q represents the variance of the state noise, represents the Jacobian matrix; The measurement update calculation includes calculating the Kalman filter gain matrix and updating the prior state and covariance matrix according to the measurement signal to obtain the posterior estimate and covariance matrix: ， ， ， in, P k Indicates k System state quantity at time x k The covariance of represents the Kalman filter gain, represents the Jacobian matrix, represents the variance matrix of the measurement noise, represents the identity matrix; in, .

Images