Aircraft plume feature extraction and correlation analysis method

Abstract

The embodiment of the present disclosure relates to a method for extracting and analyzing the correlation of aircraft plume features. The method comprises: obtaining image data and pressure data of the plume; extracting pressure curve features from the pressure data to obtain pressure curve feature points; preprocessing the plume image data to obtain a plume region; extracting Mach ring features from the plume region to obtain optimal Mach ring features; extracting shape features from the plume region to obtain optimal plume shape features; and performing multi-dimensional fusion analysis on the optimal Mach ring features, the optimal plume shape features and the pressure curve feature points to obtain an analysis result. The embodiment of the present disclosure can quickly and efficiently extract Mach ring features and plume shape features in any plume image, and quickly migrate to the correlation analysis of plume features of different types of aircraft. The multi-dimensional information can be aligned and fused for analysis, and the multi-dimensional information alignment is up to microseconds, and the analysis accuracy is high.

Description

Technical Field

[0001] This disclosure relates to the field of image processing technology, and in particular to a method for extracting and analyzing the features of an aircraft exhaust plume. Background Technology

[0002] In today's world, technological innovation has become a major battleground for international strategic competition, with unprecedented intensity surrounding technological dominance. Aircraft are devices that fly within or outside the atmosphere (space), encompassing three main categories: aircraft, spacecraft, and rockets and missiles. A nation's aircraft technology directly reflects its strategic goals, comprehensive national strength, integrated scientific and industrial capabilities, and its ability to mobilize systems and integrate resources. With the rapid development of science and technology, particularly aerospace technology, new opportunities for competition and cooperation have emerged in the global race for dominance in aerospace, including manned spaceflight, deep space exploration, satellite navigation, space science, and commercial launches. China's ambition to become a leading power in aircraft technology is timely.

[0003] Aircraft engines are the most advanced products in the equipment manufacturing field, representing a nation's technological prowess and comprehensive national strength. They have long been considered core technologies influencing national air transport, defense security, and maintaining strategic advantages. As the "heart" of an aircraft, the engine is a highly complex aerodynamic, thermodynamic, and rotating machine. A direct evaluation of engine performance and reliability is crucial in the design and development of aircraft engines. Extracting and analyzing the characteristics of the engine exhaust plume provides a direct reflection of the engine's operation and is a key technical means in aircraft performance and reliability testing. There are two main methods for analyzing aircraft exhaust plume characteristics: one is to analyze the exhaust plume based on physical mechanisms, verify its reliability through numerical models, and simulate the combustion process of the exhaust plume temperature field. The other method involves designing different sensors or temperature and radiance detection equipment to detect relevant characteristics in the aircraft engine exhaust plume region, and then having researchers with specialized knowledge independently analyze the correlation between the data. While these methods can obtain the characteristics of the aircraft exhaust plume, they are limited by factors such as engine model, the detection capabilities and accuracy of the testing instruments, and therefore cannot be universally applied to the analysis of exhaust plume characteristics in most aircraft. Furthermore, barriers exist between different feature data, making it difficult to integrate information from different dimensions for exhaust flame feature analysis. Summary of the Invention

[0004] To avoid the shortcomings of existing technologies, this application provides a method for extracting and analyzing the characteristics of aircraft exhaust plumes, which solves the problems of poor universality and difficulty in multi-dimensional information fusion analysis in existing technologies.

[0005] According to an embodiment of this disclosure, a method for extracting and analyzing the characteristics of an aircraft exhaust plume is provided, the method comprising:

[0006] Acquire exhaust flame data; wherein, the exhaust flame data includes image data and pressure data;

[0007] Pressure curve features are extracted from the pressure data to obtain pressure curve feature points;

[0008] The exhaust flame image data is preprocessed to obtain the exhaust flame region in the exhaust flame image data;

[0009] Mach ring features are extracted from the exhaust region to obtain the optimal Mach ring features;

[0010] Shape features are extracted from the exhaust region to obtain the optimal exhaust shape features;

[0011] The optimal Mach ring feature, the optimal tail flame shape feature, and the pressure curve feature points are subjected to multi-dimensional fusion analysis to obtain the analysis results.

[0012] The step of acquiring the exhaust flame data includes:

[0013] The exhaust flame data is acquired using a pressure sensor;

[0014] The entire exhaust plume area is captured using a high-speed camera and a thermal imager to obtain exhaust plume image data.

[0015] The step of extracting pressure curve features from the pressure data to obtain pressure curve feature points includes:

[0016] A two-end approach smoothing algorithm is used to perform preliminary smoothing on the pressure curve of the pressure data.

[0017] Obtain the pressure curve feature points in the pressure curve; wherein, the pressure curve feature points include at least the ignition pressure peak, the maximum pressure, and the pressure oscillation segment.

[0018] The step of preprocessing the exhaust plume image data to obtain the exhaust plume region in the exhaust plume image data includes:

[0019] Adaptive filtering is applied to the exhaust flame image data to remove interference noise.

[0020] The exhaust region in the denoised exhaust image data is extracted with fine precision.

[0021] The step of extracting Mach ring features from the exhaust region to obtain the optimal Mach ring features includes:

[0022] Based on the physical mechanism of Mach rings, threshold screening is performed on the intensity values ​​of thermal imagers, and the optimal Mach ring feature of the exhaust flame region is selected according to the formula.

[0023] The step of extracting Mach ring features from the exhaust region to obtain the optimal Mach ring features further includes:

[0024] Extract the area, maximum temperature, and average temperature of the region inside and between Mach rings.

[0025] The step of extracting shape features from the exhaust region to obtain the optimal exhaust shape features includes:

[0026] The foreground and background replacement operation is performed on the exhaust flame image data to continuously expand the exhaust flame region into a background region. After each replacement, the score of the exhaust flame region is calculated according to the formula to obtain the optimal exhaust flame shape feature.

[0027] The step of extracting shape features from the exhaust region to obtain the optimal exhaust shape features further includes:

[0028] Extract the area, maximum temperature, and average temperature of the entire exhaust region.

[0029] The step of performing multi-dimensional fusion analysis on the optimal Mach ring feature, the optimal tail flame shape feature, and the pressure curve feature points to obtain the analysis results includes:

[0030] The optimal Mach ring feature, the optimal tail flame shape feature, and the pressure curve feature points are subjected to multi-dimensional fusion analysis, and the analysis results are visualized.

[0031] The method also includes:

[0032] Based on the characteristic points of the pressure curve, determine the key moments of the tail flame combustion and extract the UNIX timestamps corresponding to the key moments of attention.

[0033] Based on the UNIX timestamp and the data capture offset time, extract the tail flame image data corresponding to that time.

[0034] The technical solutions provided by the embodiments of this disclosure may include the following beneficial effects:

[0035] In the embodiments of this disclosure, the above-described method for extracting and analyzing the features of aircraft exhaust plumes can, on the one hand, quickly and efficiently extract Mach ring features and exhaust plume shape features from any exhaust plume image, and rapidly transfer them to the correlation analysis of exhaust plume features of different aircraft models. On the other hand, it can align and fuse multi-dimensional information, with multi-dimensional information alignment reaching the microsecond level and high analysis accuracy. Attached Figure Description

[0036] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure. It is obvious that the drawings described below are merely some embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0037] Figure 1 This diagram illustrates the steps of a method for extracting and analyzing the characteristics of an aircraft exhaust plume in an exemplary embodiment of this disclosure.

[0038] Figure 2 This diagram illustrates the acquisition of exhaust plume data of an aircraft in an exemplary embodiment of this disclosure.

[0039] Figure 3 A flowchart illustrating the method for extracting and analyzing the characteristics of an aircraft exhaust plume in an exemplary embodiment of this disclosure is shown. Detailed Implementation

[0040] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that this disclosure will be more comprehensive and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0041] Furthermore, the accompanying drawings are merely illustrative diagrams of embodiments of this disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities.

[0042] This example implementation first provides a method for extracting and analyzing the characteristics and correlation of an aircraft's exhaust plume. (Reference) Figure 1 As shown, the method for extracting the characteristics of the aircraft's exhaust plume and analyzing its correlation may include steps S101 to S106.

[0043] Step S101: Obtain the tail flame data; wherein, the tail flame data includes image data and pressure data;

[0044] Step S102: Extract pressure curve features from the pressure data to obtain pressure curve feature points;

[0045] Step S103: Preprocess the exhaust flame image data to obtain the exhaust flame region in the exhaust flame image data;

[0046] Step S104: Extract Mach ring features from the exhaust region to obtain the optimal Mach ring features;

[0047] Step S105: Extract shape features from the exhaust region to obtain the optimal exhaust shape features;

[0048] Step S106: Perform multi-dimensional fusion analysis on the optimal Mach ring feature, the optimal tail flame shape feature, and the pressure curve feature points to obtain the analysis results.

[0049] The aforementioned methods for extracting and analyzing the characteristics of aircraft exhaust plumes offer several advantages. First, they enable the rapid and efficient extraction of Mach ring features and plume shape characteristics from any exhaust plume image, which can then be quickly transferred to the correlation analysis of exhaust plume features from different aircraft models. Second, they allow for the alignment and fusion analysis of multi-dimensional information, achieving microsecond-level alignment and high analysis accuracy.

[0050] Below, we will refer to Figures 1 to 3 The steps of the above-described method for extracting and analyzing the characteristics of aircraft exhaust plumes in this example embodiment will be described in more detail.

[0051] In one embodiment, the step of acquiring the exhaust flame data includes: acquiring the exhaust flame data using a pressure sensor; and capturing the complete exhaust flame area using a high-speed camera and a thermal imager to acquire the exhaust flame image data.

[0052] Specifically, such as Figure 2 As shown, a pressure sensor is installed at the exhaust nozzle to acquire pressure data of the exhaust; a high-speed camera and a thermal imager are simultaneously installed on the same side to capture the complete exhaust area and acquire exhaust image data.

[0053] In one embodiment, such as Figure 3 As shown, the step of extracting pressure curve features from the pressure data to obtain pressure curve feature points includes: using an end-to-end approach smoothing algorithm to perform preliminary smoothing on the pressure curve of the pressure data; obtaining the pressure curve feature points in the pressure curve; wherein, the pressure curve feature points include at least the ignition pressure peak, the maximum pressure, and the pressure oscillation segment.

[0054] Specifically, a two-end approach smoothing algorithm is used to initially smooth the pressure curve. Then, the key moments in the curve, such as the ignition pressure peak, the maximum pressure, and the pressure oscillation segment, are determined.

[0055] A two-end approximation smoothing algorithm is used to initially smooth the pressure curve. The formula for the two-end approximation smoothing algorithm is as follows:

[0056] data[i]=max(data[i], data[i+1])i∈[0,MaxIndex)

[0057] data[i]=max(data[i], data[i-1])i∈[MaxIndex,len(data))

[0058] Where MaxIndex is the index corresponding to the maximum value point of the pressure curve, data is the tail flame pressure data, and len represents the length of the data obtained.

[0059] Then, based on the difference in slope between data points, the regions of the ignition pressure peak, maximum pressure, and pressure oscillation segment in the curve are determined. The metric function is:

[0060]

[0061] Among them, Y i Xi represents the pressure value of the tail flame at the i-th node. 为 At the time corresponding to the i-th node, K before K represents the forward slope. after This represents the backward slope.

[0062] In one embodiment, such as Figure 3 As shown, the step of preprocessing the exhaust flame image data to obtain the exhaust flame region in the exhaust flame image data includes: applying adaptive filtering to the exhaust flame image data to remove interference noise in the exhaust flame image data; and finely extracting the exhaust flame region in the denoised exhaust flame image data.

[0063] Specifically, firstly, adaptive median filtering is applied to the acquired exhaust plume image data to remove interference noise. Then, the exhaust plume region in the image is refined, and the foreground region where the exhaust plume is located is detected and preserved.

[0064] In one embodiment, such as Figure 3 As shown, the step of extracting Mach ring features from the exhaust region to obtain the optimal Mach ring features includes: based on the physical mechanism of Mach rings, thresholding the intensity values ​​of the thermal imager, and selecting the optimal Mach ring features for scoring the exhaust region according to a formula.

[0065] Specifically, adaptive threshold Mach ring feature extraction is used for the exhaust flame region in the exhaust flame image data.

[0066] First, based on the physical mechanism of Mach rings, a threshold filtering method is applied to the data intensity values ​​of the exhaust plume image to calculate the maximum temperature intensity (Max) in the exhaust plume region. Then, an adaptive threshold (thre) is selected, and data locations in the exhaust plume image region with intensity values ​​greater than Max * thre are identified. A connected component determination algorithm is used to merge adjacent data locations. Finally, the following formula is applied:

[0067]

[0068] Calculate the Mach ring feature score at the current threshold. In the formula, n represents the number of Mach rings detected, InnerDis(x) represents the inner diameter of the currently detected Mach rings, and InterDis(x) represents the distance between Mach rings. This represents the average distance between Mach rings. Repeat the above steps, selecting the threshold with the highest Mach ring feature score as the final output. Simultaneously, extract the area, highest temperature, and average temperature information of the regions inside and between Mach rings.

[0069] In one embodiment, such as Figure 3 As shown, the step of extracting shape features from the exhaust flame region to obtain the optimal exhaust flame shape features includes: performing a foreground-background replacement operation on the exhaust flame image data, continuously expanding the exhaust flame region into a background region, calculating the score of the exhaust flame region according to a formula after each replacement, and obtaining the optimal exhaust flame shape features.

[0070] Specifically, an adaptive shape feature extraction method is adopted for the exhaust flame region in the exhaust flame image data. A foreground-background replacement operation is performed on the exhaust flame image, continuously expanding the exhaust flame region into a background region. After each replacement, the score of the exhaust flame region is calculated according to a formula, and the exhaust flame shape feature with the optimal score is obtained.

[0071] First, calculate the maximum and minimum temperature intensity values ​​(Max and Min) of the exhaust flame region, and adaptively select a threshold (thre). Then, set the intensity values ​​of data with intensity values ​​greater than Max*thre to 1, and the intensity values ​​of regions with intensity values ​​less than Max*thre to 0, obtaining binary data of the exhaust flame shape. Next, perform foreground contour number detection on the binary data. If the detected contour number is greater than 1, reselect the threshold; otherwise, proceed according to the following formula:

[0072]

[0073] Calculate the score for the exhaust region and find the exhaust shape feature that achieves the optimal score. In the formula, M and N represent the height and width of the image, respectively, P(i,j) represents the pixel value in the i-th row and j-th column, and μ represents the mean of the image. Repeat the above steps and select the threshold with the highest Mach ring feature score as the final output.

[0074] Simultaneously, the area, maximum temperature, and average temperature of the entire exhaust flame are extracted.

[0075] In one embodiment, such as Figure 3 As shown, the step of performing multi-dimensional fusion analysis on the optimal Mach ring feature, the optimal tail flame shape feature, and the pressure curve feature points to obtain the analysis results includes: performing multi-dimensional fusion analysis on the optimal Mach ring feature, the optimal tail flame shape feature, and the pressure curve feature points, and visualizing the analysis results.

[0076] Specifically, firstly, multi-dimensional data are aligned along the time dimension. Based on the characteristic points of the pressure curve, the key moments of focus for the exhaust flame combustion are determined, and the corresponding UNIX timestamps are extracted. Based on the UNIX timestamps and the data acquisition offset, exhaust flame images corresponding to these moments are extracted from high-speed data and thermal imager data. Finally, the Mach ring characteristics, shape characteristics, and temperature information of different feature domains within the exhaust flame at that moment are extracted from the exhaust flame image data. The multi-dimensional information is then fused, analyzed, and the results visualized.

[0077] The effectiveness of this application can be further illustrated by the following experiments.

[0078] 1. Experimental conditions

[0079] This application is based on the central processing unit. The experiment was conducted using Python on an i7-10700F 2.9GHz CPU, 16GB of RAM, and a Windows 10 operating system.

[0080] The data used in the experiment came from the Internet.

[0081] 2. Experiment Content

[0082] The accuracy of Mach ring feature extraction was tested as follows. Table 1 shows the test results for the axial position of the front end of the Mach ring, where the predicted value is the result extracted using the method in this application, and the true value is the Mach ring feature position annotated by actual researchers. The deviation between the two was calculated experimentally. As can be seen from the table, the feature deviation value extracted by this application is less than 1.3%, and the error value is almost negligible.

[0083] Table 1

[0084]

[0085]

[0086] Table 2 presents the experimental results for the prediction accuracy of the number of Mach rings, the accuracy of the extracted tail flame area features compared to the actual results, and the accuracy of the selection of feature points on the pressure curve in tail flame feature extraction. As can be seen from the table, the method of this application can effectively extract relevant features of the aircraft tail flame and perform correlation analysis.

[0087] Table 2

[0088] 98％ 95％ 96％

[0089] The aforementioned method for extracting and analyzing aircraft plume features demonstrates greater versatility. This application decouples the strong coupling between aircraft type and physical mechanism, enabling the extraction and correlation analysis of plume features for all observable aircraft with plumes. Furthermore, this application comprehensively utilizes plume information from multiple dimensions for fusion analysis, resulting in more accurate extraction of plume features and more comprehensive extraction of keyframe plume information.

[0090] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. In addition, those skilled in the art can combine and integrate the different embodiments or examples described in this specification.

[0091] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the appended claims.

Claims

1. A method for extracting and analyzing the characteristics and correlation of aircraft exhaust plumes, characterized in that, The method includes: Step S101: Acquire tail flame data; wherein, tail flame data includes image data and pressure data; Step S102: Extract pressure curve features from the pressure data to obtain pressure curve feature points; Step S103: Preprocess the exhaust flame image data to obtain the exhaust flame region in the exhaust flame image data; Step S104: Extract Mach ring features from the exhaust region to obtain the optimal Mach ring features; Based on the physical mechanism of Mach rings, threshold filtering is performed on the intensity values ​​of thermal imagers. The maximum temperature intensity (Max) in the exhaust flame region is calculated, and an adaptive threshold (thre) is selected. Intensity values ​​greater than Max in the exhaust flame image region are statistically analyzed. The data locations of the three nodes are determined; a connected component determination algorithm is used to merge adjacent data locations. The Mach ring feature at the current threshold is obtained according to formula (1). ; （1） Where n represents the number of Mach rings currently detected, InnerDis(x) represents the inner diameter of the currently detected Mach ring, and InterDis(x) represents the distance between Mach rings. This represents the average distance between Mach rings; Repeat the above steps and select the threshold with the highest Mach ring feature score as the optimal Mach ring feature; Simultaneously, the area, maximum temperature, and average temperature of the regions inside and between Mach rings are extracted; Step S105: Extract shape features from the exhaust region to obtain the optimal exhaust shape features; Perform foreground-background replacement operation on the tail flame image data to continuously expand the tail flame area into the background area. After each replacement, calculate the score of the tail flame area according to formula (2) to obtain the tail flame shape feature with the best score. （2） in, M and N P( represents the height and width of the image, respectively) i,j ) indicates the first i line, number j Column pixel values, μ The mean of the image is represented; repeat the above steps and select the threshold with the highest Mach ring feature score as the optimal tail flame shape feature; Simultaneously, the area, maximum temperature, and average temperature of the entire exhaust region are extracted; Step S106: Perform multi-dimensional fusion analysis on the optimal Mach ring characteristics, optimal exhaust flame shape characteristics, and pressure curve characteristic points to obtain analysis results; specifically including: Align the optimal Mach ring features, optimal tail flame shape features, and pressure curve feature points in the time dimension. Determine the key moment of tail flame combustion based on the pressure curve feature points and extract the corresponding UNIX timestamp. Extract the exhaust flame image data corresponding to the time specified in the UNIX timestamp and the data capture offset. Extract the Mach ring features, tail shape features, and temperature information of different feature domains within the tail flame from the tail flame image data at that moment; The Mach ring characteristics, tail shape characteristics, and temperature information of different feature domains within the region at that moment were fused, analyzed, and visualized.

2. The method for extracting and analyzing the characteristics of aircraft exhaust plumes according to claim 1, characterized in that, The steps for obtaining exhaust flame data include: Data on the exhaust plume is obtained using a pressure sensor; The entire exhaust plume area was captured using a high-speed camera and a thermal imager to obtain exhaust plume image data.

3. The method for extracting and analyzing the characteristics of aircraft exhaust plumes according to claim 1, characterized in that, The step of extracting pressure curve features from pressure data to obtain pressure curve feature points includes: A two-end approach smoothing algorithm is used to perform preliminary smoothing on the pressure curve of the pressure data. Obtain the characteristic points of the pressure curve in the pressure curve; among which, the characteristic points of the pressure curve include at least the ignition pressure peak, the maximum pressure, and the pressure oscillation segment.

4. The method for extracting and analyzing the characteristics of aircraft exhaust plumes according to claim 1, characterized in that, The step of preprocessing the exhaust plume image data to obtain the exhaust plume region in the exhaust plume image data includes: Adaptive filtering is applied to the exhaust flame image data to remove interference noise from the exhaust flame image data; The exhaust region in the denoised exhaust image data is extracted with fine precision.

Images