Gas path performance degradation evaluation method based on exhaust temperature of aero-engine turbine

Abstract

The invention discloses a gas path performance degradation evaluation method based on the exhaust temperature of an aero-engine turbine, and relates to the technical field of aero-engine maintenance. According to the method, based on a time sequence neural network method, a mapping relation between inlet environment parameters and control parameters of an engine and engine exhaust temperature or collected engine section parameters is established by utilizing reference data of engine installation operation, and subsequent operation data of the engine is substituted into a previously trained baseline model; and the difference value between the output result of the baseline model and the real operation data of the engine reflects the performance attenuation condition of the engine gas path. Compared with the prior art, the evaluation method can evaluate the performance degradation condition of the engine by using very limited data acquired in the operation process of the installed engine, has the advantages of simplicity and reliability, and provides technical support for performance degradation evaluation of the gas path of the engine.

Description

Technical Field

[0001] This invention relates to the field of aircraft engine maintenance technology, and more specifically, to a method for assessing the performance degradation of the gas path based on the turbine exhaust temperature of an aircraft engine. Background Technology

[0002] As engine usage time increases, dust and other contaminants adhere to the surfaces of airflow components, roughening the surfaces of bleed air engine blades and leading to a decline in engine airflow performance. Current methods for assessing the degree of this performance decline, such as calculating the inflection point temperature to reflect the performance degradation of aero-engines (CN112651624B), have limitations. This method fails to accurately identify engine performance degradation for aircraft with different control schemes and does not fully consider the impact of different air intakes on engine performance degradation after installation. Another method (CN111382522A) calculates the installed thrust of the engine based on the aircraft's takeoff run data. However, this method measures the total thrust of two engines and distributes thrust to each engine based on the low-pressure rotor speed, failing to accurately reflect the performance degradation of individual engines. Furthermore, the calculation of the total thrust of two engines uses aircraft design data for parameters such as wheel friction coefficients. As the aircraft's usage time increases, the actual aircraft condition will inevitably deviate from the design data, resulting in a discrepancy between the total thrust calculated from takeoff run distance and the actual thrust value. Currently, in the civil aviation engine field, the data collected by the engine is converted into a K value (major engine manufacturers provide their own conversion formulas), and then the converted K value is compared with the engine's baseline value to obtain the deviation, thereby assessing the engine's performance degradation. However, due to factors such as differences in control laws between some engines and general civil aviation engines, the above conversion formula is not applicable to all engines, and there are cases where this method is not suitable for assessing the degree of engine performance degradation. Summary of the Invention

[0003] The purpose of this invention is to provide a method for evaluating the performance degradation of the gas path based on the turbine exhaust temperature of an aero-engine. This method can use data collected by existing sensors in the aero-engine to evaluate the overall performance degradation of the aero-engine's gas path, and has the advantages of simple operation and accurate results.

[0004] The embodiment of the present invention is achieved as follows: A method for assessing the degradation of gas path performance based on turbine exhaust temperature of an aero-engine, comprising: S1. Based on the temporal neural network algorithm, build an engine neural network model and establish the mapping relationship between environmental parameters, control parameters and target performance parameters; S2. Collect environmental parameters, control parameters, and target performance parameters from the first few flights after the engine starts operating as baseline data, train the engine neural network model, and obtain a personalized digital model of the gas path performance baseline. S3. Input the environmental and control parameters of the engine's subsequent operation into the personalized gas path performance baseline digital model, and calculate the theoretical values ​​of the target performance parameters; S4. Assess the amount of performance degradation in the engine's air passages by comparing the theoretical value with the actual measured value.

[0005] The beneficial effects of the embodiments of the present invention are: This invention provides a method for assessing the degradation of airflow performance based on turbine exhaust temperature in aero-engines. It utilizes a time-series neural network approach to establish a mapping relationship between engine inlet environmental parameters, control parameters, and engine exhaust temperature or collected engine cross-sectional parameters using baseline data from engine operation. By substituting subsequent engine operating data into a pre-trained baseline model, the difference between the baseline model output and the actual engine operating data reflects the degradation of engine airflow performance. Compared to existing technologies, this assessment method can evaluate engine performance degradation using very limited data collected during engine operation, offering advantages such as simplicity and reliability, and providing technical support for assessing engine airflow performance degradation. Attached Figure Description

[0006] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0007] Figure 1 A flowchart of a gas path performance degradation assessment method based on turbine exhaust temperature of an aero-engine provided by the present invention; Figure 2 The gas path performance degradation assessment method based on turbine exhaust temperature of an aero-engine provided in Embodiment 1 of the present invention is based on the assessment results of the low-pressure turbine exhaust temperature T6 of the engine when the engine is in an intermediate state. Figure 3 The gas path performance degradation assessment method based on turbine exhaust temperature of an aero-engine provided in this embodiment of the invention is based on the assessment results of the low-pressure turbine exhaust temperature T6 of the engine when the engine is in its maximum state. Figure 4The gas path performance degradation assessment method based on turbine exhaust temperature of an aero-engine provided in Embodiment 2 of the present invention is based on the assessment results of the high-pressure compressor outlet pressure P31 and the low-pressure turbine outlet pressure P6 when the engine is in an intermediate state. Figure 5 The gas path performance degradation assessment method based on turbine exhaust temperature of an aero-engine provided in Embodiment 2 of the present invention is based on the assessment results of the high-pressure compressor outlet pressure P31 and the low-pressure turbine outlet pressure P6 when the engine is in its maximum state. Detailed Implementation

[0008] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of the present invention. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of the present invention.

[0009] The following provides a detailed description of the gas path performance degradation assessment method based on aero-engine turbine exhaust temperature provided by this invention. A complete flowchart can be found in the provided text. Figure 1 As shown.

[0010] This invention provides a method for evaluating the performance degradation of the gas path based on the turbine exhaust temperature of an aero-engine, comprising: S1. Based on the temporal neural network algorithm, build an engine neural network model and establish the mapping relationship between environmental parameters, control parameters and target performance parameters; S2. Collect environmental parameters, control parameters, and target performance parameters from the first few flights after the engine starts operating as baseline data, train the engine neural network model, and obtain a personalized digital model of the gas path performance baseline. S3. Input the environmental and control parameters of the engine's subsequent operation into the personalized gas path performance baseline digital model, and calculate the theoretical values ​​of the target performance parameters; S4. Assess the amount of performance degradation in the engine's air passages by comparing the theoretical value with the actual measured value.

[0011] Optionally, temporal neural network algorithms include Long Short-Term Memory (LSTM) networks, basic recurrent neural networks (RNNs), and CNN-LSTM hybrid networks.

[0012] Furthermore, the environmental parameters include flight altitude H, Mach number Ma, and engine inlet total temperature T1; the control parameters include fan physical speed N1, fan equivalent speed N1hs, compressor physical speed N2, compressor equivalent speed N2hs, fan guide vane angle a1, compressor guide vane angle a2, throttle lever angle PLA, and engine exhaust nozzle angle D8.

[0013] Optionally, the target performance parameter is at least one of the following: engine low-pressure turbine exhaust temperature T6, pressure parameter collected from the engine airflow passage section, and temperature parameter collected from the engine airflow passage section. The pressure and temperature parameters collected from the engine airflow passage section need to be determined based on the engine model; they are not applicable to some engines where no pressure or temperature sensors are installed in the airflow passage section.

[0014] Furthermore, in step S2, when collecting baseline data, the number of sorties collected shall not be less than three. Specifically, this depends on the training results. Preferably, the engine neural network model is trained until the absolute value of the steady-state error of the engine's typical state, as determined by the engine neural network model itself, does not exceed 1%, at which point the training is completed, and a personalized digital model of the gas path performance baseline is obtained.

[0015] Furthermore, when the target performance parameter is the low-pressure turbine exhaust temperature of the engine or the temperature parameter collected from the cross-section of the engine airflow passage, the engine neural network model is trained until the absolute value of the steady-state error of the engine neural network model in the typical state corresponding to the engine does not exceed 1%, and the difference between the actual value and the calculated value of the target performance parameter is less than 5℃, then the training is completed.

[0016] This personalized digital model of the airflow performance baseline represents the performance state of the engine after installation without any performance degradation, and serves as a benchmark for future comparisons of engine airflow performance degradation.

[0017] Furthermore, when the engine is installed on another aircraft or at another installation location on the same aircraft, considering the impact of differences in the aircraft's air intake on the engine's airflow performance, it is necessary to re-collect data from the first few flights when the engine is installed on the latest aircraft location and retrain to obtain a new personalized airflow performance baseline digital model.

[0018] Furthermore, the formula for calculating the performance degradation of the gas path is as follows:

[0019] In the formula, P represents the gas path performance degradation of the target performance parameter, and X...实际值 X represents the actual measured value of the target performance parameter. 理论值 The target performance parameters are theoretical values ​​calculated using a personalized gas path performance baseline digital model.

[0020] Since the data types (environmental parameters and control parameters of engine operation) input to the personalized airflow performance baseline digital model are consistent with the data types (environmental parameters and control parameters) of the actual engine operation, the output of this personalized airflow performance baseline digital model represents the target performance value corresponding to the condition that the engine has not experienced airflow performance degradation. If the target performance value is the engine exhaust temperature T6, then the difference between the T6 output by the personalized airflow performance baseline digital model and the T6 value collected by the engine sensors represents the degree of engine airflow performance degradation. If the target value is the engine cross-sectional parameter collected by the engine sensors (e.g., compressor outlet pressure P31), then the difference between the engine cross-sectional parameter output by the personalized airflow performance baseline digital model and the actual value collected by the sensors at the engine cross-section partially reflects the cause of engine airflow performance degradation.

[0021] Furthermore, given the inherent error between the output values ​​of the target performance parameters of the engine typical state model and the corresponding actual measured values ​​during the training of the personalized engine airflow performance baseline model, it is necessary to further correct the airflow performance degradation. The correction formula is as follows: P' = P - P0 In the formula, P' is the corrected actual gas path performance degradation, P is the original gas path performance degradation, and P0 is the inherent error measured during the training phase of the engine neural network model using benchmark data.

[0022] Furthermore, when target performance parameters are collected from multiple different engine components, the contribution of each component to the overall engine performance degradation can be determined by analyzing the differences in the amount of performance degradation in the airflow path. For example, when the target performance parameters are the high-pressure compressor outlet pressure P31 and the low-pressure turbine outlet pressure P6, comparing the amount of performance degradation in the airflow path of these two components can determine the contribution of the high-pressure compressor and the low-pressure compressor to the overall engine airflow path performance degradation. The component with a larger amount of performance degradation has a greater contribution.

[0023] The following specific examples will provide further explanation. Example

[0024] This embodiment provides a method for evaluating the performance degradation of the gas path based on the turbine exhaust temperature of an aero-engine, which includes the following steps: S1. Based on the LSTM temporal neural network algorithm, a neural network model of the engine is built, and the mapping relationship between environmental parameters, control parameters and target performance parameters is established; Specifically, the environmental parameters include the engine inlet total pressure P1 and the engine inlet total temperature T1.

[0025] The control parameters include the physical fan speed N1, the equivalent fan speed N1hs, the physical compressor speed N2, the equivalent compressor speed N2hs, the fan guide vane angle a1, the compressor guide vane angle a2, the throttle lever angle PLA, and the engine exhaust nozzle angle D8.

[0026] The target performance parameter is the engine low-pressure turbine exhaust temperature T6.

[0027] S2. Collect environmental parameters, control parameters, and target performance parameters from the first few flights after the engine starts operating as baseline data to train the engine neural network model until the absolute value of the steady-state error of the engine's typical state in the self-test of the engine neural network model does not exceed 1%, and the difference between the actual value and the calculated value of the target performance parameter is less than 5℃ (4℃ in the intermediate state and 3℃ in the maximum state). The training is then completed, and a personalized gas path performance baseline digital model is obtained.

[0028] S3. Input the environmental and control parameters of the subsequent engine running time (0.1~1.0) t into the personalized gas path performance baseline digital model to simulate the theoretical low-pressure turbine exhaust temperature T6 of the engine when the engine does not experience performance degradation under the same environmental and control parameters.

[0029] S4. Collect the actual value of the engine low-pressure turbine exhaust temperature under the same parameter conditions using a temperature sensor, and then use the formula...

[0030] Calculate the performance degradation of the gas path.

[0031] Calculate the amount of performance degradation in the air passage when the engine is in the intermediate state and at its maximum state.

[0032] S5. Correct the calculated gas path performance degradation by subtracting inherent errors. The inherent error for the intermediate state is 0.5% (corresponding to 4℃), and the inherent error for the maximum state is 0.4% (corresponding to 3℃). The results are as follows: Figure 2 and Figure 3 As shown in the figure, the calculated thrust Fr decay during long-term ground test of this engine model is selected as a comparison.

[0033] It can be seen that, regardless of whether the engine is in its intermediate or maximum state, using thrust Fr and exhaust temperature T6 as degradation parameters respectively, the degradation trends are basically consistent. The engine shows virtually no performance degradation during the (0.4~0.5)t operating time because engine performance is adjusted during the (0.3~0.4)t period, and the calculated results are consistent with reality. The slow performance degradation during the (0.9~1.0)t operating time is due to airflow channel cleaning; therefore, the engine performance degradation is slow during this stage, and the model calculation results are consistent with reality. Example

[0034] This embodiment provides a method for evaluating the performance degradation of the gas path based on the turbine exhaust temperature of an aero-engine, which includes: S1. Based on the LSTM temporal neural network algorithm, a neural network model of the engine is built, and the mapping relationship between environmental parameters, control parameters and target performance parameters is established; Specifically, the environmental parameters include the engine inlet total pressure P1 and the engine inlet total temperature T1.

[0035] The control parameters include the physical fan speed N1, the equivalent fan speed N1hs, the physical compressor speed N2, the equivalent compressor speed N2hs, the fan guide vane angle a1, the compressor guide vane angle a2, the throttle lever angle PLA, and the engine exhaust nozzle angle D8.

[0036] The target performance parameters are the high-pressure compressor outlet pressure P31 and the low-pressure turbine outlet pressure P6.

[0037] S2. Collect environmental parameters, control parameters, and target performance parameters from the first few flights after the engine starts operating as baseline data, train the engine neural network model until the absolute value of the steady-state error of the engine in the typical state of the engine does not exceed 1% in the self-test of the engine neural network model, and complete the training to obtain a personalized digital model of the gas path performance baseline.

[0038] S3. Input the environmental and control parameters of the engine's subsequent running time (0.1~1.0) t into the personalized gas path performance baseline digital model to simulate the theoretical high-pressure compressor outlet pressure P31 and low-pressure turbine outlet pressure P6 when the engine does not experience performance degradation under the same environmental and control parameters.

[0039] S4. Collect the actual values ​​of the high-pressure compressor outlet pressure P31 and the low-pressure turbine outlet pressure P6 under the same parameter conditions using pressure sensors, and then use the formula...

[0040] Calculate the performance degradation of the gas path.

[0041] Calculate the amount of performance degradation in the air passage when the engine is in the intermediate state and at its maximum state.

[0042] S5. Correct the calculated gas path performance degradation by subtracting inherent errors. The result is as follows: Figure 4 and Figure 5 As shown in the figure, the calculated thrust Fr decline during the long-term ground test of this engine model and the thrust at exhaust temperature T6 calculated in Example 1 are selected for comparison.

[0043] pass Figures 4-5 It can be seen that the overall performance of the engine's air circuit deteriorates with the increase of usage time. The performance degradation trend of model T6, which represents the overall performance, and model P31, which represents the performance of the components, is consistent with the performance degradation trend of the engine over time.

[0044] Furthermore, the degradation of engine model P31 was higher than that of model P6 throughout the process, indicating that the engine high-pressure compressor was the biggest contributor to the decline in the overall engine air circuit performance. The analysis results are consistent with the engine disassembly and inspection results after long-term testing.

[0045] In summary, this invention provides a method for assessing the degradation of airflow performance based on turbine exhaust temperature in aero-engines. It utilizes time-series neural network methods such as LSTM to establish a mapping relationship between engine inlet environmental parameters, control parameters, and engine exhaust temperature or collected engine cross-sectional parameters using benchmark data from engine operation. By substituting subsequent engine operating data into a pre-trained baseline model, the difference between the baseline model output and the actual engine operating data reflects the degradation of engine airflow performance. Compared to existing technologies, this assessment method can evaluate engine performance degradation using very limited data collected during engine operation, offering advantages of simplicity and reliability, and providing technical support for assessing engine airflow performance degradation.

[0046] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for evaluating the performance degradation of the gas path based on the turbine exhaust temperature of an aero-engine, characterized in that, include: S1. Based on the temporal neural network algorithm, build an engine neural network model and establish the mapping relationship between environmental parameters, control parameters and target performance parameters; S2. Collect the environmental parameters, control parameters, and target performance parameters from the first few flight sorties after the engine starts operating as baseline data, train the engine neural network model, and obtain a personalized digital model of the gas path performance baseline. S3. Input the environmental parameters and control parameters of the engine's subsequent operation into the personalized gas path performance baseline digital model, and calculate the theoretical value of the target performance parameter; S4. Assess the amount of performance degradation of the engine's air passage by the difference between the theoretical value and the actual measured value.

2. The identification method according to claim 1, characterized in that, The environmental parameters include flight altitude, Mach number, and engine inlet total temperature; the control parameters include fan physical speed, fan equivalent speed, compressor physical speed, compressor equivalent speed, fan guide vane angle, compressor guide vane angle, throttle lever angle, and engine exhaust nozzle angle.

3. The identification method according to claim 2, characterized in that, The target performance parameter is at least one of the following: engine low-pressure turbine exhaust temperature, pressure parameters collected from the engine airflow passage section, and temperature parameters collected from the engine airflow passage section.

4. The identification method according to claim 3, characterized in that, In step S2, when collecting the baseline data, the number of sorties collected shall not be less than 3.

5. The identification method according to claim 4, characterized in that, In step S2, the engine neural network model is trained until the absolute value of the steady-state error of the engine neural network model under typical conditions corresponding to the engine does not exceed 1%, and the training is completed to obtain the personalized gas path performance baseline digital model.

6. The identification method according to claim 5, characterized in that, In step S2, when the target performance parameter is the low-pressure turbine exhaust temperature of the engine or the temperature parameter collected from the cross section of the engine airflow passage, the engine neural network model is trained until the absolute value of the steady-state error of the engine neural network model in the typical state corresponding to the engine does not exceed 1%, and the difference between the actual value and the calculated value of the target performance parameter is less than 5℃, then the training is completed.

7. The identification method according to claim 6, characterized in that, When the engine is installed on another aircraft or at another installation location on the same aircraft, data from the previous few flights needs to be collected again to retrain the new personalized airflow performance baseline digital model.

8. The identification method according to claim 1, characterized in that, The formula for calculating the gas path performance degradation is as follows: ， In the formula, P represents the gas path performance degradation of the target performance parameter, and X... 实际值 X is the actual measured value of the target performance parameter. 理论值 The target performance parameter is the theoretical value calculated by the personalized gas path performance baseline digital model.

9. The identification method according to claim 8, characterized in that, The correction formula is as follows: P' = P - P0 In the formula, P' is the corrected actual gas path performance degradation, P is the original gas path performance degradation, and P0 is the inherent error measured during the training phase of the engine neural network model using the benchmark data.

10. The identification method according to claim 8, characterized in that, When the target performance parameters are collected from multiple different components of the engine, the degree of contribution to the engine performance degradation is determined by the difference in the amount of performance degradation of the air passage of each component.

Images