An aircraft environmental control system heat exchanger fault image recognition method and system

Abstract

The application discloses an aircraft environmental control system heat exchanger fault image recognition method and system, acquires a heat exchanger image through a double RGB camera, carries out first pretreatment on the image, and constructs a three-dimensional model of the heat exchanger image; acquires a heat exchanger image through an infrared camera, carries out second pretreatment on the image, and generates a rainbow temperature map; fits the rainbow temperature map with the three-dimensional model, generates a rainbow temperature embedded three-dimensional model, acquires the temperature of a plurality of feature points in a region to be analyzed in the three-dimensional model, carries out first fault judgment to obtain a first fault, finds repair information according to fault information, and generates the rainbow temperature embedded three-dimensional model in the manner of fitting the rainbow temperature map through the construction of the three-dimensional model, analyzes the temperature abnormal range in all directions, and locates the fault type in combination with contour information.

Description

Technical fields:

[0001] This invention belongs to the field of computer data processing, and in particular relates to a method and system for image recognition of heat exchanger faults in an aircraft environmental control system. Background technology:

[0002] The aircraft environmental control system (ECS) is the primary system ensuring the normal operation of aircraft electronic equipment. Its main purpose is to ensure the safety of pilots and passengers while maintaining a comfortable environment under changing flight conditions, and to provide a suitable working environment for the increasingly complex electronic and electrical equipment on board. It has functions such as anti-icing / de-icing and defogging. The ECS includes the cabin pressure control system, cabin temperature control system, and cabin ventilation control system. To ensure safe flight in extreme weather conditions and to meet ever-increasing passenger comfort requirements, aircraft ECSs have become increasingly complex. With increased operating time, the probability of various malfunctions increases, and various signs of failure emerge, such as valve wear and degradation, turbine blade breakage, and heat exchanger blockage. Failure to address these malfunctions promptly can not only cause discomfort to passengers but also reduce the efficiency of onboard equipment, increase system energy consumption, and shorten equipment lifespan.

[0003] While research on overall aircraft failure prediction has been conducted for many years, research on failures in the aircraft environmental control field is still in its infancy. In particular, fault diagnosis for heat exchangers in the air-cooled equipment of aircraft environmental control systems has not yet been developed, leading to increased maintenance workload. Improving the accuracy of fault diagnosis for heat exchangers in the air-cooled equipment of aircraft environmental control systems has become an urgent technical problem to be solved. Summary of the Invention

[0004] To address the issues of slow efficiency and poor accuracy in fault diagnosis of heat exchangers in air-cooled equipment of aircraft environmental control systems, this method proposes an image recognition method for faults in heat exchangers of air-cooled equipment in aircraft environmental control systems. The method involves acquiring heat exchanger images using dual RGB cameras, performing a first preprocessing step to construct a 3D model of the heat exchanger image, acquiring heat exchanger images using an infrared camera, performing a second preprocessing step to generate a rainbow temperature map, fitting the rainbow temperature map to the 3D model, and obtaining the temperature of multiple feature points in the area to be analyzed within the 3D model. The temperature of these feature points and their average value are then compared with a first threshold temperature to obtain the difference results. A first fault is determined based on the difference results. If the difference exceeds a first boundary value, a temperature anomaly is identified. If the difference exceeds a second boundary value, the temperature exceeds a limit, triggering an alarm signal. A second fault is determined based on the contour information of a local image of the heat exchanger within a sliding window. Finally, the final fault information is obtained based on the first and second faults. By constructing a three-dimensional model and fitting a rainbow temperature map, a rainbow temperature embedded three-dimensional model was generated, which can comprehensively analyze the temperature anomaly range and locate the fault type by combining contour information. The multi-dimensional progressive judgment method improves the judgment range of temperature anomalies, and the combination of dynamic threshold judgment method improves the accuracy of temperature anomaly judgment. By enhancing the acquired images, the model building accuracy is improved, effectively solving the fault diagnosis of heat exchangers at the remote end.

[0005] The technical solution adopted by the present invention to solve the above technical problems is as follows:

[0006] A method for image recognition of faults in heat exchangers of air-cooled equipment in an aircraft environmental control system includes the following steps:

[0007] S1. After acquiring the heat exchanger image through dual RGB cameras, the image is preprocessed in the first stage to construct a three-dimensional model of the heat exchanger image; after acquiring the heat exchanger image through an infrared camera and preprocessing the image in the second stage, a rainbow temperature map is generated.

[0008] S2. Fit the rainbow temperature map to the three-dimensional model to generate a rainbow temperature embedded three-dimensional model;

[0009] S3. Obtain the temperature of multiple feature points in the region to be analyzed in the three-dimensional model; the feature points are random sampling points in the red area of ​​the rainbow temperature three-dimensional model, or points marked by the user in the region to be analyzed.

[0010] S4. Perform a difference calculation between the temperature of the multiple feature points and the average temperature and the first threshold temperature to obtain the difference result;

[0011] S5. Based on the difference result, a first fault is determined and a first fault is obtained; if the difference exceeds the first boundary value, the temperature is determined to be abnormal; if the difference exceeds the second boundary value, the temperature is determined to be out of limit and an alarm signal is triggered.

[0012] S6. Based on the local image contour information of the heat exchanger image in the sliding window, a second fault is determined, and the second fault is obtained.

[0013] S7. Obtain the final fault information based on the first and second faults.

[0014] Furthermore, the step of acquiring heat exchanger images using dual RGB cameras specifically includes the following steps:

[0015] S1A. Acquire the heat exchanger image captured by the first RGB camera in the dual RGB camera setup;

[0016] S1B. Using a 30*30 pixel matrix as the unit, start from the upper left corner of the heat exchanger image acquired in step S1A and traverse and verify the image until the lower right corner of the image is reached.

[0017] For example, if the image size is 90*90, the traversal order according to the 30*30 pixel matrix is ​​as follows:

[0018] The coordinates of the top left corner of the image are defined as (0,0), and the coordinates of the bottom right corner are defined as (90,90).

[0019] The rectangle positions in the first traversed image are: (0,0) as the top left corner and (30,30) as the bottom right corner;

[0020] The rectangle positions in the second traversal of the image are: (30,0) as the top left corner and (60,30) as the bottom right corner;

[0021] The rectangle positions in the third traversal of the image are: (60,0) as the top left corner and (90,30) as the bottom right corner;

[0022] The rectangle positions in the fourth traversal of the image are: (0,30) as the top left corner and (30,60) as the bottom right corner;

[0023] This process continues until the last rectangle is reached: (60, 60) is the top left corner and (90, 90) is the bottom right corner.

[0024] The method for traversal verification is as follows:

[0025] When the 30*30 pixel matrix traverses to the rectangle J in the image with (i,j) as the top-left corner coordinate and (i+30,j+30) as the bottom-right corner coordinate, calculate the average value J of the R component of all pixels in the matrix.R The average value of the G component J G The average value of component B, J B ;

[0026] Get the average value P of the R component of all pixels within the previous traversed rectangle P of rectangle J. R The average value P of the G component G The average value P of component B B ;

[0027] Get the average value Q of the R component of all pixels within the next rectangle Q that is traversed from rectangle J. R The average value of the G component Q G The average value Q of component B B ;

[0028] when Then, the image at the coordinate position of rectangle J in the heat exchanger image acquired in step S1A is replaced by the image at the same coordinate position as rectangle J in the heat exchanger image acquired by the second RGB camera of the dual RGB cameras.

[0029] S1C. Repeat step S1B until all positions of the heat exchanger image acquired in step S1A have been processed, and use the processed heat exchanger image as the heat exchanger image acquired by the dual RGB cameras.

[0030] Furthermore, the first threshold temperature is determined by the aircraft's flight status.

[0031] Furthermore, the first image preprocessing includes:

[0032] S11. After grayscale processing, the original heat exchanger image is low-pass filtered.

[0033] S12. Subtract the grayscale heat exchanger image from the blurred heat exchanger image after low-pass filtering to generate a heat exchanger image containing high-frequency components.

[0034] S13. Enhancement is achieved by magnifying the high-frequency heat exchanger image and superimposing a grayscale heat exchanger image.

[0035] Furthermore, the construction of the three-dimensional model of the heat exchanger image includes:

[0036] S14. Perform stereo calibration on the two RGB cameras.

[0037] S15. Correct the distortion and stereoscopic correction of the enhanced left and right heat exchanger images.

[0038] S16. Perform stereo matching on the corrected left and right heat exchanger images to calculate the disparity map.

[0039] S17. The three-dimensional point cloud information of the measured object is calculated by combining the disparity map obtained by stereo matching with the reprojection matrix after stereo correction, and the three-dimensional structural information of the object is restored.

[0040] Furthermore, by acquiring images of the heat exchanger using an infrared camera and performing a second preprocessing of the images, a rainbow temperature map is generated, including:

[0041] S21. Perform mean filtering on the heat exchanger image. Since the heat exchanger image is mainly composed of Gaussian noise, mean filtering is used.

[0042] The mean filtering model is as follows: Where f(x,y) is the heat exchanger image, which is an M*N matrix, and h(x,y) is the low-pass impulse response, which is an L*L matrix.

[0043] S22. Perform pseudo-color encoding on the mean-filtered heat exchanger image to generate rainbow-colored and iron-red heat exchanger images.

[0044] Furthermore, based on the difference results, a first fault is determined, and the first fault is obtained, including:

[0045] S51. Determine if the difference is greater than the first boundary value. If so, proceed to S52.

[0046] S52. Determine if the rate of change of the difference is greater than 0. If so, determine that the temperature fault is in the area to be analyzed.

[0047] The first threshold temperature is determined by the aircraft's flight status, including ambient temperature, aircraft altitude, flight speed, and speed of sound.

[0048] An image recognition system for faults in heat exchangers of air-cooled equipment in an aircraft environmental control system is characterized by comprising: an image acquisition module, a model building module, a temperature analysis module, and a fault analysis module connected sequentially via wired or wireless networks.

[0049] The image acquisition module is used to acquire images of the heat exchanger through dual RGB cameras, perform a first preprocessing on the images, and construct a three-dimensional model of the heat exchanger image; and to acquire images of the heat exchanger through an infrared camera, perform a second preprocessing on the images, and generate a rainbow temperature map.

[0050] The model building module is used to fit the rainbow temperature map with the three-dimensional model to generate a rainbow temperature embedded three-dimensional model.

[0051] The temperature analysis module is used to acquire the temperature of multiple feature points in the area to be analyzed in the 3D model, and to perform difference calculations between the temperature of the multiple feature points and the average temperature and a first threshold temperature to obtain the difference result; to perform a first fault judgment based on the difference result to obtain a first fault; if the difference exceeds a first boundary value, it is determined that the temperature is abnormal; if the difference exceeds a second boundary value, it is determined that the temperature exceeds the limit and an alarm signal is triggered; and to perform a second fault judgment based on the local image contour information of the heat exchanger image in the sliding window to obtain a second fault.

[0052] The fault analysis module is used to obtain the final fault information based on the first and second faults, and to find repair measures based on the fault information.

[0053] A computer-readable storage medium storing a computer program, characterized in that a processor executes the computer program to implement a method for image recognition of faults in the heat exchanger of an air-cooled equipment in an aircraft environmental control system.

[0054] A terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, is characterized in that the processor executes the computer program to implement a method for image recognition of faults in the heat exchanger of an air-cooled equipment in an aircraft environmental control system.

[0055] The beneficial effects of this invention are as follows:

[0056] 1) By enhancing the acquired images, the accuracy of model building is improved, effectively solving the problem of fault diagnosis of heat exchangers at the remote end.

[0057] 2) By constructing a three-dimensional model and fitting the rainbow temperature map, a rainbow temperature embedded three-dimensional model is generated, which can comprehensively analyze the temperature anomaly range and locate the fault type by combining the contour information.

[0058] 3) The multi-dimensional progressive judgment method has improved the judgment range of temperature anomalies, and the combination of dynamic threshold judgment method has improved the accuracy of temperature anomalies.

[0059] 4) When acquiring images using dual RGB cameras, the image acquired by the first RGB camera is verified using a template matrix. If any abnormalities are found, the image acquired by the second RGB camera is used to fill or correct them, thereby improving the accuracy of image acquisition and providing strong data support for subsequent fault diagnosis.

[0060] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above description and other objects, features and advantages of the present invention more obvious and understandable, preferred embodiments are provided and described in detail below. Attached Figure Description

[0061] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings.

[0062] Figure 1 A structural diagram of a method and system for image recognition of heat exchanger faults in an aircraft environmental control system. Detailed Implementation

[0063] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0064] In the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," "fixing," etc., should be interpreted broadly. For example, they can refer to a connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0065] Example 1

[0066] A method for image recognition of faults in heat exchangers of air-cooled equipment in an aircraft environmental control system, characterized in that the method includes the following steps:

[0067] S1. After acquiring the heat exchanger image through dual RGB cameras, the image is preprocessed in the first stage to construct a three-dimensional model of the heat exchanger image; after acquiring the heat exchanger image through an infrared camera and preprocessing the image in the second stage, a rainbow temperature map is generated.

[0068] S2. Fit the rainbow temperature map to the three-dimensional model to generate a rainbow temperature embedded three-dimensional model;

[0069] S3. Obtain the temperature of multiple feature points in the region to be analyzed in the 3D model;

[0070] S4. Perform a difference calculation between the temperature of the multiple feature points and the average temperature and the first threshold temperature to obtain the difference result; the first threshold temperature is determined by the aircraft flight status.

[0071] S5. Based on the difference result, a first fault is determined and a first fault is obtained; if the difference exceeds the first boundary value, the temperature is determined to be abnormal; if the difference exceeds the second boundary value, the temperature is determined to be out of limit and an alarm signal is triggered.

[0072] S6. Based on the local image contour information of the heat exchanger image in the sliding window, a second fault is determined, and the second fault is obtained.

[0073] S7. Obtain the final fault information based on the first and second faults.

[0074] The first image preprocessing includes:

[0075] S11. After grayscale processing, the original heat exchanger image is low-pass filtered.

[0076] S12. Subtract the grayscale heat exchanger image from the blurred heat exchanger image after low-pass filtering to generate a heat exchanger image containing high-frequency components.

[0077] S13. Enhancement is achieved by magnifying the high-frequency heat exchanger image and superimposing a grayscale heat exchanger image.

[0078] The construction of a 3D model of the heat exchanger image includes:

[0079] S14. Perform stereo calibration on the two RGB cameras.

[0080] S15. Correct the distortion and stereoscopic correction of the enhanced left and right heat exchanger images.

[0081] S16. Perform stereo matching on the corrected left and right heat exchanger images to calculate the disparity map.

[0082] S17. The three-dimensional point cloud information of the measured object is calculated by combining the disparity map obtained by stereo matching with the reprojection matrix after stereo correction, and the three-dimensional structural information of the object is restored.

[0083] The process of acquiring images of the heat exchanger using an infrared camera, and generating a rainbow temperature map after a second preprocessing of the images includes:

[0084] S21. Perform mean filtering on the heat exchanger image. Since the heat exchanger image is mainly composed of Gaussian noise, mean filtering is used.

[0085] The mean filtering model is as follows: Where f(x,y) is the heat exchanger image, which is an M*N matrix, and h(x,y) is the low-pass impulse response, which is an L*L matrix; in this example, L is 3.

[0086] S22. Perform pseudo-color encoding on the mean-filtered heat exchanger image to generate rainbow-colored and iron-red heat exchanger images.

[0087] The pseudo-color coding model is as follows:

[0088]

[0089] Where R(x,y), G(x,y), and B(x,y) are the values ​​of the red, green, and blue components of the pseudo-color, respectively, f(x,y) is the grayscale value of the original image, and T... R T G T B These represent the mapping relationships between red, green, and blue values ​​and grayscale values, respectively.

[0090] Furthermore, fitting the rainbow temperature map to the 3D model involves matching the coordinates of the highest temperature and multiple sampling points to the 3D model. Starting from the coordinates of the high-temperature region, the rainbow temperature is generated by diffusion based on the material's thermal conductivity parameters to embed the 3D model.

[0091] Furthermore, based on the difference results, a first fault is determined, and the first fault is obtained, including:

[0092] S51. Determine if the difference is greater than the first boundary value. If so, proceed to S52.

[0093] S52. Determine if the rate of change of the difference is greater than 0. If so, determine that the temperature fault is in the area to be analyzed.

[0094] Furthermore, the first threshold temperature is determined by the aircraft's flight status, including determining the first threshold temperature based on ambient temperature, aircraft flight altitude, flight speed, and speed of sound.

[0095] The fault information includes the type of fault, such as dirt or corrosion on the surface of the heat exchanger, pipe blockage, leakage, or severe damage to the seals; the coordinate range of the fault location, the temperature at the fault location, and the fault level.

[0096] Based on the fault information, query the repair measures on the cloud server.

[0097] Example 2

[0098] An image recognition system for faults in heat exchangers of air-cooled equipment in an aircraft environmental control system is characterized by comprising: an image acquisition module, a model building module, a temperature analysis module, and a fault analysis module connected sequentially via wired or wireless networks.

[0099] The image acquisition module is used to acquire images of the heat exchanger through dual RGB cameras, perform a first preprocessing on the images, and construct a three-dimensional model of the heat exchanger image; and to acquire images of the heat exchanger through an infrared camera, perform a second preprocessing on the images, and generate a rainbow temperature map.

[0100] The model building module is used to fit the rainbow temperature map with the three-dimensional model to generate a rainbow temperature embedded three-dimensional model.

[0101] The temperature analysis module is used to acquire the temperature of multiple feature points in the area to be analyzed in the 3D model, and to perform difference calculations between the temperature of the multiple feature points and the average temperature and a first threshold temperature to obtain the difference result; to perform a first fault judgment based on the difference result to obtain a first fault; if the difference exceeds a first boundary value, it is determined that the temperature is abnormal; if the difference exceeds a second boundary value, it is determined that the temperature exceeds the limit and an alarm signal is triggered; and to perform a second fault judgment based on the local image contour information of the heat exchanger image in the sliding window to obtain a second fault.

[0102] The fault analysis module is used to obtain the final fault information based on the first and second faults, and to find repair measures based on the fault information.

[0103] A computer-readable storage medium storing a computer program, characterized in that a processor executes the computer program to implement a method for image recognition of faults in the heat exchanger of an air-cooled equipment in an aircraft environmental control system.

[0104] A terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, is characterized in that the processor executes the computer program to implement a method for image recognition of faults in the heat exchanger of an air-cooled equipment in an aircraft environmental control system.

[0105] The beneficial effects of this invention are as follows:

[0106] 1) By enhancing the acquired images, the accuracy of model building is improved, effectively solving the problem of fault diagnosis of heat exchangers at the remote end.

[0107] 2) By constructing a three-dimensional model and fitting the rainbow temperature map, a rainbow temperature embedded three-dimensional model is generated, which can comprehensively analyze the temperature anomaly range and locate the fault type by combining the contour information.

[0108] 3) The multi-dimensional progressive judgment method has improved the judgment range of temperature anomalies, and the combination of dynamic threshold judgment method has improved the accuracy of temperature anomalies.

[0109] 4) When acquiring images using dual RGB cameras, the image acquired by the first RGB camera is verified using a template matrix. If any abnormalities are found, the image acquired by the second RGB camera is used to fill or correct them, thereby improving the accuracy of image acquisition and providing strong data support for subsequent fault diagnosis.

[0110] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for image recognition of heat exchanger faults in an aircraft environmental control system, characterized in that, The method includes the following steps: S1. After acquiring the heat exchanger image through dual RGB cameras, the image is preprocessed in the first stage to construct a three-dimensional model of the heat exchanger image; after acquiring the heat exchanger image through an infrared camera and preprocessing the image in the second stage, a rainbow temperature map is generated. The process of acquiring images of the heat exchanger using an infrared camera, and generating a rainbow temperature map after a second preprocessing of the images includes: Mean filtering is applied to the heat exchanger image because the heat exchanger image is mainly composed of Gaussian noise. The mean filtering model is as follows: ,in This is an image of a heat exchanger, and it is an M*N matrix. , is an L*L matrix; in this example, L is 3; The mean-filtered heat exchanger image is pseudo-color encoded to generate rainbow-colored and iron-red heat exchanger images. The pseudo-color coding model is as follows: ； Wherein, are the values ​​of the red, green, and blue components of the pseudo-color, respectively. T represents the grayscale values ​​of the original image. R T G T B These represent the mapping relationships between red, green, and blue values ​​and grayscale values, respectively. Fitting the rainbow temperature map to the 3D model involves matching the coordinates of the highest temperature and multiple sampling points to the 3D model. Starting from the coordinates of the high-temperature region, the rainbow temperature is generated by diffusion based on the material's thermal conductivity parameters to embed the 3D model. S2. Fit the rainbow temperature map to the three-dimensional model to generate a rainbow temperature embedded three-dimensional model; S3. Obtain the temperature of multiple feature points in the region to be analyzed in the 3D model; S4. Perform a difference calculation between the temperature of the multiple feature points and the average temperature and the first threshold temperature to obtain the difference result; S5. Based on the difference result, a first fault is determined and a first fault is obtained; if the difference exceeds the first boundary value, the temperature is determined to be abnormal; if the difference exceeds the second boundary value, the temperature is determined to be out of limit and an alarm signal is triggered. S6. Based on the local image contour information of the heat exchanger image in the sliding window, a second fault is determined, and the second fault is obtained. S7. Obtain the final fault information based on the first and second faults.

2. The method for image recognition of heat exchanger faults in an aircraft environmental control system according to claim 1, characterized in that: The first threshold temperature is determined by the aircraft's flight status.

3. The method for image recognition of heat exchanger faults in an aircraft environmental control system according to claim 1, characterized in that: The first image preprocessing includes: S11. After grayscale processing, the original heat exchanger image is low-pass filtered. S12. Subtract the grayscale heat exchanger image from the blurred heat exchanger image after low-pass filtering to generate a heat exchanger image containing high-frequency components. S13. Enhancement is achieved by magnifying the high-frequency heat exchanger image and superimposing a grayscale heat exchanger image.

4. The method for image recognition of heat exchanger faults in an aircraft environmental control system according to claim 1, characterized in that: The construction of the three-dimensional model of the heat exchanger image includes: S14. Perform stereo calibration on the two RGB cameras. S15. Correct the distortion and stereoscopic correction of the enhanced left and right heat exchanger images. S16. Perform stereo matching on the corrected left and right heat exchanger images to calculate the disparity map. S17. The three-dimensional point cloud information of the measured object is calculated by combining the disparity map obtained by stereo matching with the reprojection matrix after stereo correction, and the three-dimensional structural information of the object is restored.

5. The method for image recognition of heat exchanger faults in an aircraft environmental control system according to claim 1, characterized in that: Based on the difference results, the first fault is determined, including: S51. Determine if the difference is greater than the first boundary value. If so, proceed to S52. S52. Determine if the rate of change of the difference is greater than 0. If so, determine that the temperature fault is in the area to be analyzed. The first threshold temperature is determined by the aircraft's flight status, including ambient temperature, aircraft altitude, flight speed, and speed of sound.

6. A fault image recognition system for a heat exchanger in an aircraft environmental control system, characterized in that, The system includes: an image acquisition module, a model building module, a temperature analysis module, and a fault analysis module, which are connected sequentially via wired or wireless networks. The image acquisition module is used to acquire images of the heat exchanger through dual RGB cameras, perform a first preprocessing on the images, and construct a three-dimensional model of the heat exchanger image; and to acquire images of the heat exchanger through an infrared camera, perform a second preprocessing on the images, and generate a rainbow temperature map. The process of acquiring images of the heat exchanger using an infrared camera, and generating a rainbow temperature map after a second preprocessing of the images includes: Mean filtering is applied to the heat exchanger image because the heat exchanger image is mainly composed of Gaussian noise. The mean filtering model is as follows: ,in This is an image of a heat exchanger, and it is an M*N matrix. , is an L*L matrix; in this example, L is 3; The mean-filtered heat exchanger image is pseudo-color encoded to generate rainbow-colored and iron-red heat exchanger images. The pseudo-color coding model is as follows: ； in, These are the numerical values ​​of the red, green, and blue components of the pseudo-color, respectively. T represents the grayscale values ​​of the original image. R T G T B These represent the mapping relationships between red, green, and blue values ​​and grayscale values, respectively. Fitting the rainbow temperature map to the 3D model involves matching the coordinates of the highest temperature and multiple sampling points to the 3D model. Starting from the coordinates of the high-temperature region, the rainbow temperature is generated by diffusion based on the material's thermal conductivity parameters to embed the 3D model. The model building module is used to fit the rainbow temperature map with the three-dimensional model to generate a rainbow temperature embedded three-dimensional model. The temperature analysis module is used to acquire the temperature of multiple feature points in the area to be analyzed in the 3D model, and to perform difference calculations between the temperature of the multiple feature points and the average temperature and a first threshold temperature to obtain the difference result; to perform a first fault judgment based on the difference result to obtain a first fault; if the difference exceeds a first boundary value, it is determined that the temperature is abnormal; if the difference exceeds a second boundary value, it is determined that the temperature exceeds the limit and an alarm signal is triggered; and to perform a second fault judgment based on the local image contour information of the heat exchanger image in the sliding window to obtain a second fault. The fault analysis module is used to obtain the final fault information based on the first and second faults, and to find repair measures based on the fault information.

7. A computer-readable storage medium storing a computer program, characterized in that, The processor executes the computer program to implement the aircraft environmental control system heat exchanger fault image recognition method as described in any one of claims 1-5.

8. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the aircraft environmental control system heat exchanger fault image recognition method as claimed in any one of claims 1-5.

Images