A method for predicting abnormal flameout of scramjet engine

Abstract

The present invention belongs to the technical field of scramjet engine control, and is specifically a method for predicting abnormal flameout of a scramjet engine, which mainly consists of principal component analysis of real-time combustion chamber pressure, extraction of maximum principal component of pressure oscillation, standard deviation calculation, flameout risk degree calculation and flameout risk feedback. According to the measured combustion chamber pressure curve, real-time analysis is performed to obtain the flameout risk degree value of the scramjet engine and feed it back to the flight control system for processing. The method realizes the flameout risk prediction of the scramjet engine and ensures that the scramjet engine works outside the flameout boundary and operates stably and reliably.

Description

Technical Field

[0001] This invention belongs to the field of scramjet engine control technology, and more specifically, relates to a method for predicting abnormal shutdown of scramjet engines based on principal component analysis and a shutdown risk assessment function. Background Art

[0002] As a power plant for hypersonic vehicles, the scramjet engine possesses high specific impulse at Mach 6 and above, offering significant advantages over other power cycle systems. Therefore, the scramjet engine has become one of the core technologies for hypersonic vehicles, and research on its mechanisms and applications has become a hot topic in recent years.

[0003] Due to factors such as the intake environment, flight conditions, and equivalence ratio, scramjet engines are at risk of stalling during cruise. If the engine stalls during cruise, the aircraft will lose power directly, causing unpredictable consequences.

[0004] After decades of theoretical research, the basic principles of scramjet engines have been understood both domestically and internationally. However, in recent years, scramjet engines have been in the verification flight phase, and their flameout limitation and protection systems are not yet perfect. There is currently no publicly available literature on abnormal flameout prediction criteria and early warning systems for scramjet engines. Abnormal flameout prediction for scramjet engines refers to assessing the degree of flameout risk when the engine is operating at the flameout boundary and promptly feeding this assessment back to the flight control system to prevent the engine from entering an abnormal flameout state. Scramjet engines differ from traditional aero engines. Due to their unique operating environment and conditions, monitoring and predicting abnormal flameout states presents significant challenges. These challenges are mainly reflected in the following aspects: scramjet engines operate in a supersonic combustion state; the engine flameout process exhibits nonlinear, unsteady, and abrupt characteristics; and abnormal flameout occurs within milliseconds. Obvious characteristics are not easily observed before abnormal flameout in scramjet engines. Therefore, data mining methods are needed to extract pre-flameout features, and a dedicated evaluation function needs to be designed to quantitatively assess the degree of flameout risk, providing abnormal flameout prediction and feedback. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a predictive method for abnormal flameout in scramjet engines, and meets the specific usage requirements of scramjet engines.

[0006] The specific technical solution of the present invention is as follows:

[0007] A method for predicting abnormal flameout in a scramjet engine includes the following steps:

[0008] (1) Preprocessing of the measured pressure curve of the engine combustion chamber:

[0009] (1.1) Downsampling is performed on the raw data of the engine combustion chamber, retaining 50 sampling points per second;

[0010] (1.2) The trend term was eliminated from the processed combustion chamber pressure data;

[0011] (2) Principal component analysis of combustion chamber pressure:

[0012] (2.1) The combustion chamber pressure data of the scramjet engine after step (1) is decomposed using principal component analysis.

[0013] (2.2) Calculate the variance contribution rate of the main components after decomposition and extract the components with the largest pressure pulsation amplitude;

[0014] (3) Engine risk assessment:

[0015] (3.1) Calculate the average pressure in the combustion chamber over a fixed time period;

[0016] (3.2) Calculate the standard deviation of pressure during a fixed time period in the combustion chamber;

[0017] (3.3) Calculate the engine stall risk assessment function;

[0018] (4) Feedback mechanism for risk level values:

[0019] (4.1) Feedback the flameout risk level of the scramjet engine to the flight control system;

[0020] (4.2) The flight control system adjusts the aircraft attitude and refuels the engine according to the degree of engine shutdown risk, so that the engine can get out of the shutdown boundary as soon as possible.

[0021] As a preferred method, in the principal component analysis step (2.1) of the combustion chamber pressure, principal component analysis in unsupervised learning is used to extract the principal components of the measured combustion chamber pressure of the scramjet engine.

[0022] In step (2.2), the principal component with the largest combustion chamber pressure pulsation amplitude is selected. y max This component reflects the combustion oscillation characteristics of the combustion chamber, and large-scale combustion pressure pulsations are more likely to cause the risk of flameout.

[0023] As a preferred method, in step (3.3), the risk of engine stalling is assessed.

[0024] For the largest principal component y max Perform fixed step size NThe standard deviation is calculated using the following formula: ;

[0025] In the formula, p i The largest principal component y max The data values ​​in the middle, μ For fixed step size N The average value of the combustion chamber pressure, i.e.

[0026]

[0027] The absolute value of the difference between the combustion chamber pressure before and during combustion is used as the calculation benchmark for the degree of engine stall, and dimensionless calculations are performed. The mapping relationship characterizing the degree of engine stall risk is defined as follows: ;

[0028] In the formula, The average pressure in the combustion chamber before engine combustion; The average pressure in the combustion chamber when the engine is operating normally. This is a function for assessing the risk of engine stalling.

[0029] As a preferred approach, scramjet engines operate in a supersonic combustion state, and the flameout process is completed within milliseconds. The characteristics preceding abnormal flameout in scramjet engines are not easily observed directly. To predict abnormal flameout in scramjet engines, it is necessary to mine statistical patterns from the engine's combustion chamber pressure data. This invention employs principal component analysis (PCA) in unsupervised learning to extract principal components from the measured combustion chamber pressure of a scramjet engine.

[0030] In step (2.2), it is assumed that the measured combustion chamber pressure x yes m A 3D random variable, where Σ is the combustion chamber pressure. x The covariance matrix Σ has eigenvalues ​​of Σ as follows: The eigenvectors corresponding to the eigenvalues ​​are respectively The combustion chamber pressure x No. k The main components of combustion are:

[0031]

[0032] β mk It is the m-th eigenvector of the k-th principal component of combustion. x m yes x The m-th column vector;

[0033] x No. kThe variance of the principal components of combustion is:

[0034] ;

[0035] definition x The k Combustion main components y k The variance contribution rate y k The ratio of the variance to the sum of all variances is calculated using the following formula:

[0036] .

[0037] The beneficial effects of the present invention compared with the prior art are:

[0038] This invention employs principal component analysis (PCA) to extract the main components of engine combustion chamber pressure. Components with larger pressure pulsations represent the engine's combustion oscillation characteristics. This PCA method filters out data that has little impact on abnormal engine shutdown, further improving the accuracy of engine shutdown prediction. A shutdown risk function is used to quantify the degree of engine shutdown risk, providing specific quantitative indicators. Feeding the predicted engine shutdown risk value back to the flight control system can effectively prevent abnormal shutdown during the operation of the scramjet engine. Attached Figure Description

[0039] Figure 1 This is the pressure change curve of the combustion chamber after pretreatment.

[0040] Figure 2 These are the third-order principal components obtained through principal component analysis.

[0041] Figure 3 This is a curve for evaluating the degree of engine shutdown. Detailed Implementation

[0042] The following describes the embodiments of the present invention through specific examples. Those skilled in the art will readily understand the other advantages and benefits of the present invention from the disclosure herein. The present invention may also be implemented or applied through various other specific embodiments, and the details in this specification may be modified or altered based on different viewpoints and applications without departing from the spirit of the present invention.

[0043] An embodiment provides a method for predicting abnormal shutdown of a scramjet engine, comprising the following steps:

[0044] (1) Preprocessing of the measured pressure curve of the engine combustion chamber:

[0045] (1.1) Downsampling is performed on the raw data of the engine combustion chamber, retaining 50 sampling points per second;

[0046] (1.2) Remove trend terms from the processed combustion chamber pressure data;

[0047] (2) Principal component analysis of combustion chamber pressure:

[0048] (2.1) The combustion chamber pressure data of the scramjet engine after step (1) is decomposed using principal component analysis.

[0049] (2.2) Calculate the variance contribution rate of the main components after decomposition and extract the components with the largest pressure pulsation amplitude;

[0050] (3) Engine risk assessment:

[0051] (3.1) Calculate the average pressure of the combustion chamber over a fixed time period. ;

[0052] (3.2) Calculate the standard deviation of the pressure in the combustion chamber for a fixed time length. Where, p i The largest principal component y max The data values ​​in the middle, μ For fixed step size N The average value of the combustion chamber pressure.

[0053] (3.3) Calculate the engine stall risk assessment function;

[0054] (4) Feedback mechanism for risk level values:

[0055] (4.1) Feedback the flameout risk level of the scramjet engine to the flight control system;

[0056] (4.2) The flight control system adjusts the aircraft attitude and refuels the engine according to the degree of engine shutdown risk, so that the engine can get out of the shutdown boundary as soon as possible.

[0057] As a preferred method, in the principal component analysis step (2.1) of the combustion chamber pressure, principal component analysis in unsupervised learning is used to extract the principal components of the measured combustion chamber pressure of the scramjet engine.

[0058] In step (2.2), the principal component with the largest combustion chamber pressure pulsation amplitude is selected. y max This component reflects the combustion oscillation characteristics of the combustion chamber, and large-scale combustion pressure pulsations are more likely to cause the risk of flameout.

[0059] As a preferred method, in step (3.3), the risk of engine stalling is assessed.

[0060] For the largest principal component y max Perform fixed step size N The standard deviation is calculated using the following formula: ;

[0061] Where, p i The largest principal component y max The data values ​​in the middle, μ For fixed step size N The average value of the combustion chamber pressure, i.e.

[0062]

[0063] The absolute value of the difference between the combustion chamber pressure before and during combustion is used as the calculation benchmark for the degree of engine stall, and dimensionless calculations are performed. The mapping relationship characterizing the degree of engine stall risk is defined as follows: ;

[0064] Where, The average pressure in the combustion chamber before engine combustion; The average pressure in the combustion chamber when the engine is operating normally. This is a function for assessing the risk of engine stalling.

[0065] As a preferred approach, scramjet engines operate in a supersonic combustion state, and the flameout process is completed within milliseconds. The characteristics preceding abnormal flameout in scramjet engines are not easily observed directly. To predict abnormal flameout in scramjet engines, it is necessary to mine statistical patterns from the engine's combustion chamber pressure data. This invention employs principal component analysis (PCA) in unsupervised learning to extract principal components from the measured combustion chamber pressure of a scramjet engine.

[0066] In step (2.2), it is assumed that the measured combustion chamber pressure x yes m A 3D random variable, where Σ is the combustion chamber pressure. x The covariance matrix Σ has eigenvalues ​​of Σ as follows: The eigenvectors corresponding to the eigenvalues ​​are respectively The combustion chamber pressure x The k The main components of combustion are:

[0067] ;

[0068] β mk It is the m-th eigenvector of the k-th principal component of combustion. x m yes xThe m-th column vector;

[0069] x The k The variance of the principal components of combustion is:

[0070]

[0071] definition x The k Combustion main components y k The variance contribution rate y k The ratio of the variance to the sum of all variances is calculated using the following formula:

[0072]

[0073] like Figure 1 As shown. First, the pressure curve of the scramjet engine combustion chamber needs to be acquired. After downsampling and trend term elimination preprocessing, the engine combustion chamber pressure curve is... Figure 1 The curve shown is the combustion chamber pressure curve under normal engine operation. After 333 seconds, the engine experienced an abnormal shutdown.

[0074] Then, principal component analysis was used to decompose the combustion chamber pressure data of the scramjet engine to obtain the main components of the scramjet engine combustion chamber pressure. Figure 2 The diagram shows the three main components of combustion chamber pressure. y 1 , y 2 and y 3 The variance contribution rates of the three components were 71.23%, 19.89%, and 8.87%, respectively, among which the principal components... y 3 It has the largest pressure pulsation amplitude, which can better reflect the combustion oscillation characteristics of scramjet engines and the main components y 3 Defined as y max When the pressure in the combustion chamber of a scramjet engine fluctuates significantly, its combustion stability is poor, and the risk of flameout is greater.

[0075] For principal components y 3 Standard deviation with fixed time step S The calculation, and the standard deviation S Incorporating into the engine stall risk assessment function f d In the process, the abnormal shutdown assessment curve of the scramjet engine was obtained, as shown in the figure. Figure 3 As shown. Figure 3 It can be seen that after 80 seconds, the amplitude of the engine stall risk assessment function increases significantly and two peaks appear, indicating that the engine has a significant risk of stalling.

[0076] The risk assessment of the scramjet engine's shutdown level is fed back to the flight control system, which then performs actions such as adjusting the aircraft's attitude and refueling the engine based on the degree of shutdown risk, so that the engine can get out of the shutdown boundary as soon as possible.

[0077] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A method for predicting abnormal flameout of a scramjet engine, characterized in that The steps include: (1) Preprocessing of measured curve of engine combustion chamber pressure: (1.1) Downsample the raw data of the engine combustion chamber and retain 50 sampling points per second; (1.2) Eliminate the trend item of the processed combustion chamber pressure data; (2) Principal component analysis of combustion chamber pressure: (2.1) Decomposing the combustion chamber pressure data of the scramjet engine after being processed in step (1) using the principal component analysis method; (2.2) Calculate the variance contribution rate of the main components after decomposition and extract the component with the largest pressure pulsation amplitude; (3) Engine risk assessment: (3.1) Calculate the average pressure of the combustion chamber for a fixed time length; (3.2) Calculate the standard deviation of the pressure in the combustion chamber for a fixed time length; (3.3) Calculate the engine stall risk assessment function; In step (3.3), the engine flameout risk level is assessed. For the largest principal component y max Fixed step size N The standard deviation calculation formula is ; In the formula, p i The largest principal component y max The data values ​​in μ For fixed step length N The average value of the combustion chamber pressure, ； The absolute value of the difference between the combustion chamber pressure before and during combustion is used as the calculation basis for the degree of engine flameout, and a dimensionless calculation is performed. The mapping relationship that characterizes the degree of engine flameout risk is defined as ; In the formula, is the average pressure in the combustion chamber before engine combustion; It is the average pressure of the combustion chamber when the engine is working normally. It is the engine flameout risk assessment function; (4) Feedback link of risk level value: (4.1) Feedback the flameout risk level of the scramjet engine to the flight control system; (4.2) The flight control system will adjust the aircraft attitude and refuel the engine according to the degree of flameout risk, so that the engine can get out of the flameout boundary as soon as possible.

2. The method for predicting abnormal flameout of a scramjet engine according to claim 1 is characterized in that: In the principal component analysis step (2.1) of the combustion chamber pressure, the principal component analysis method in unsupervised learning is used to extract the principal components of the measured combustion chamber pressure of the scramjet engine. In step (2.2), select the principal component with the largest combustion chamber pressure pulsation amplitude y max .

3. The method for predicting abnormal flameout of a scramjet engine according to claim 1, characterized in that: In step (2.2), it is assumed that the measured combustion chamber pressure x yes m dimensional random variable, Σ is the combustion chamber pressure x The covariance matrix of Σ has eigenvalues ​​of The eigenvalues ​​and corresponding eigenvectors are , then the combustion chamber pressure x No. k The main components of combustion are: ； β mk is the mth eigenvector of the kth combustion principal component, x m yes x The mth column vector of ; x No. k The variance of the combustion principal component is: ； definition x No. k Combustion main component y k The variance contribution rate is y k The ratio of the variance of to the sum of all variances is calculated as follows: 。

Images