An engine fault diagnosis method, device, equipment and medium

Abstract

The application discloses an engine fault diagnosis method, device, equipment and medium, and relates to the technical field of engine analysis. The engine fault diagnosis method comprises the following steps: performing wave spectrum extraction on an engine exhaust plume hyperspectral image based on analysis wavelengths of different engine materials, and determining spectral values corresponding to the analysis wavelengths of different engine materials in the hyperspectral image; wherein the analysis wavelengths of different engine analysis materials correspond to different engine fault types; determining multi-dimensional spectral features of the hyperspectral image according to the spectral values corresponding to the analysis wavelengths of different engine materials; and performing fault diagnosis and identification on the multi-dimensional spectral features by using a pre-trained engine fault diagnosis model, and determining an engine fault diagnosis result. The method realizes the diagnosis of engine faults by analyzing engine exhaust plume spectra, and effectively identifies faults such as engine wear and ablation.

Description

Technical Field

[0001] This invention relates to the field of engine analysis technology, and in particular to an engine fault diagnosis method, device, equipment and medium. Background Technology

[0002] As the core of the transportation system, the engine's thrust chamber is one of the key components within the engine. Therefore, it is necessary to perform fault diagnosis on the engine thrust chamber during operation.

[0003] In the development of new engines, fault diagnosis includes: turbopump fault diagnosis, pipeline fracture fault diagnosis, reservoir fault diagnosis, thrust chamber fault diagnosis, generator fault diagnosis, etc. However, existing fault diagnosis data mainly comes from engine test vibration acceleration data, pulse pressure data, and speed data, but these data cannot provide direct and effective information for faults such as engine wear and ablation.

[0004] Therefore, how to diagnose engine wear and burning faults has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide an engine fault diagnosis method, device, equipment, and medium for diagnosing engine faults such as wear and burn-off.

[0006] To achieve the above objectives, the present invention provides the following technical solution: An engine fault diagnosis method includes: performing spectral extraction on a hyperspectral image of the engine exhaust flame based on the analysis wavelengths of different engine materials, and determining the spectral values ​​corresponding to the analysis wavelengths of different engine materials in the hyperspectral image; wherein the analysis wavelengths of different engine materials correspond to different engine fault types; determining the multidimensional spectral features of the hyperspectral image based on the spectral values ​​corresponding to the analysis wavelengths of different engine materials; and using a pre-trained engine fault diagnosis model to perform fault diagnosis and identification on the multidimensional spectral features to determine the engine fault diagnosis result.

[0007] In one optional embodiment of this application, the step of extracting the spectrum from the hyperspectral image of the engine exhaust based on the analysis wavelength of different engine materials, and determining the spectral values ​​corresponding to the analysis wavelengths of different engine materials in the hyperspectral image, includes: dividing the hyperspectral image of the engine exhaust into regions and determining the average spectrum of each region in the hyperspectral image of the engine exhaust; combining the average spectrum of each region in the hyperspectral image of the engine exhaust corresponding to different bands, and using interpolation to determine the spectral values ​​corresponding to the analysis wavelengths according to the analysis wavelengths of different engine materials.

[0008] In one optional embodiment of this application, the process of region division of the engine exhaust hyperspectral image includes: determining a reference hyperspectral image and performing contrast enhancement processing on the exhaust portion and background portion of the reference hyperspectral image; the engine exhaust radiation intensity in the corresponding band of the reference hyperspectral image is greater than the engine exhaust radiation intensity in other bands; obtaining a grayscale histogram of the contrast-enhanced reference hyperspectral image and performing maximum inter-class variance calculation on the grayscale histogram to determine a binarization threshold for the grayscale histogram; performing binarization processing on the grayscale histogram according to the binarization threshold to obtain a binary image of the grayscale histogram; performing morphological processing on the binary image to obtain an engine exhaust image of the reference hyperspectral image; determining the region division position of the engine exhaust hyperspectral image based on the engine exhaust image; and performing region division on the engine exhaust hyperspectral image according to the region division position.

[0009] In one optional embodiment of this application, the multidimensional image features of the hyperspectral image are determined by the following formula: ; ; ; ; ; in, , , , , These represent the spectral features of different dimensions of the hyperspectral image; j represents the analysis wavelength type of the hyperspectral image, j=p or h or q, where p represents the injection disk material, h represents the coating material, and q represents the inner wall material; n represents the total number of wavelengths of the injection disk material; m represents the total number of analysis wavelengths of the coating material; and r represents the total number of analysis wavelengths of the inner wall material. This represents the i-th analysis wavelength corresponding to the analysis wavelength type j in the hyperspectral image; represents the spectral value corresponding to the i-th analytical wavelength when the analytical wavelength type is j; c represents the spectral characteristic coefficient, c=1 when j=p, c=2 when j=h, and c=3 when j=q.

[0010] In one optional embodiment of this application, the engine fault diagnosis model is a support vector machine model, and different engine fault diagnosis models correspond to different engine fault types. The engine fault diagnosis model is trained in the following way: for an engine fault diagnosis model used to identify any fault type, based on the first sample spectral features corresponding to the fault type and the second sample spectral features corresponding to other fault types, determine the hyperplane coefficients that maximize the spatial distance between the first sample spectral features and the second sample spectral features; and construct the fault diagnosis model corresponding to the fault type based on the hyperplane coefficients.

[0011] In one optional embodiment of this application, different engine fault types correspond to different engine fault diagnosis models; the step of using pre-trained engine fault diagnosis models to perform fault diagnosis and identification on the multidimensional spectral features and determine the engine fault diagnosis result includes: sequentially inputting the multidimensional spectral features into each of the engine fault diagnosis models to obtain the engine fault type output by each engine fault diagnosis model; and determining the engine fault diagnosis result based on the engine fault diagnosis type output by each engine fault diagnosis model.

[0012] In one optional embodiment of this application, if the engine diagnostic model cannot identify the multidimensional spectral features, the method further includes: combining historical multidimensional spectral features of known engine fault types to perform neighbor classification identification on the multidimensional spectral features, and determining the fault diagnosis result corresponding to the multidimensional spectral features.

[0013] Compared with existing technologies, the engine fault diagnosis method provided by this invention extracts the spectrum from the hyperspectral image of the engine exhaust flame based on the analysis wavelengths of different engine materials, and determines the spectral values ​​corresponding to the analysis wavelengths of different engine materials in the hyperspectral image; wherein, the analysis wavelengths of different engine materials correspond to different engine fault types; based on the spectral values ​​corresponding to the analysis wavelengths of different engine materials, the multidimensional spectral features of the hyperspectral image are determined; using a pre-trained engine fault diagnosis model, fault diagnosis and identification are performed on the multidimensional spectral features to determine the engine fault diagnosis result. This method achieves engine fault diagnosis by analyzing the engine exhaust flame spectrum, and effectively identifies faults such as engine wear and ablation.

[0014] The present invention also provides an engine fault diagnosis device, comprising: The spectral value determination unit is used to extract the spectrum of the hyperspectral image of the engine exhaust flame based on the analysis wavelength of different engine materials, and determine the spectral values ​​corresponding to the analysis wavelengths of different engine materials in the hyperspectral image; wherein the analysis wavelengths of different engine analysis materials correspond to different engine fault types; the multidimensional feature determination unit is used to determine the multidimensional spectral features of the hyperspectral image based on the spectral values ​​corresponding to the analysis wavelengths of different engine materials; the model diagnosis unit is used to perform fault diagnosis and identification on the multidimensional spectral features using a pre-trained engine fault diagnosis model, and determine the engine fault diagnosis result.

[0015] Compared with the prior art, the beneficial effects of the engine fault diagnosis device provided by the present invention are the same as those of the engine fault diagnosis method described in the above technical solution, and will not be repeated here.

[0016] The present invention also provides an electronic device, comprising: a processor; a memory for storing executable instructions of the processor; the processor being configured to execute an engine fault diagnosis method by running the instructions in the memory.

[0017] Compared with the prior art, the beneficial effects of the electronic device provided by the present invention are the same as those of the engine fault diagnosis method described in the above technical solution, and will not be repeated here.

[0018] The present invention also provides a computer storage medium storing instructions, which, when executed, implement the above-described engine fault diagnosis method.

[0019] Compared with the prior art, the beneficial effects of the computer storage medium provided by the present invention are the same as those of the engine fault diagnosis method described in the above technical solution, and will not be repeated here. Attached Figure Description

[0020] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a flowchart of an engine fault diagnosis method provided in an embodiment of this application.

[0021] Figure 2 Hyperspectral image of engine exhaust provided in an embodiment of this application.

[0022] Figure 3 A schematic diagram of an agglomerative hierarchical clustering method provided in an embodiment of this application; Figure 4 This is a schematic diagram of the clustering of multidimensional sample spectral features in two-dimensional and three-dimensional space provided in the embodiments of this application.

[0023] Figure 5 This is a schematic diagram of k-nearest neighbor classification provided in an embodiment of this application.

[0024] Figure 6 This is a structural diagram of the engine fault diagnosis device provided in an embodiment of this application.

[0025] Figure 7 This is a schematic diagram of an electronic device structure provided in an embodiment of this application. Detailed Implementation

[0026] To facilitate a clear description of the technical solutions in the embodiments of the present invention, the terms "first" and "second" are used to distinguish identical or similar items with essentially the same function and effect. For example, the first threshold and the second threshold are merely used to distinguish different thresholds and do not limit their order. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and that the terms "first" and "second" are not necessarily different.

[0027] It should be noted that in this invention, the terms "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in this invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0028] In this invention, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can represent: a, b, c, a combination of a and b, a combination of a and c, a combination of b and c, or a, b, and c, where a, b, and c can be single or multiple.

[0029] This application provides an engine fault diagnosis method, apparatus, device, and medium, which will be described in detail in the following embodiments.

[0030] This application first provides an engine fault diagnosis method, please refer to... Figure 1 , Figure 1This is a flowchart of an engine fault diagnosis method provided in an embodiment of this application.

[0031] like Figure 1 As shown, the engine fault diagnosis method includes the following S101 to S103.

[0032] S101, based on the analysis wavelengths of different engine materials, perform spectral extraction on the hyperspectral image of the engine exhaust flame to determine the spectral values ​​corresponding to the analysis wavelengths of different engine materials in the hyperspectral image; wherein, the analysis wavelengths of different engine analysis materials correspond to different engine fault types.

[0033] During engine operation, fuel and oxidizer burn in the combustion chamber to produce high-temperature and high-pressure gas. After the gas converges through the nozzle, its speed gradually increases and reaches the speed of sound at the throat. Then, it expands and accelerates in the expansion section of the nozzle, reaching the speed of supersonic gas, and finally forms a plume through the nozzle exit.

[0034] Engine combustion chambers commonly use nickel-based superalloys or copper alloys as materials. Nickel-based alloys possess extremely strong high-temperature resistance, corrosion resistance, and oxidation resistance. Tungsten-copper alloys, on the other hand, are alloys based on tungsten and copper with other metals as auxiliary materials, possessing advantages such as high melting point, high boiling point, high strength, and high hardness. When nickel-based alloys and tungsten-copper alloys are worn away by high-speed combustion gases, the resulting abraded metallic substances will be mixed into the plume flame. These metallic substances emit characteristic spectra in the plume flame. Using the highly sensitive characteristic spectra of the metals as metal analysis spectra, even at low concentrations, these metal analysis spectra can be detected in the hyperspectral image of the engine exhaust flame.

[0035] Different materials exhibit different absorption, reflection, and transmission characteristics to electromagnetic waves of different wavelengths. In the embodiments of this application, the analysis wavelengths of the main metallic elements in nickel-based superalloys include 341.476 nm and 352.454 nm for nickel, 425.435 nm and 520.602 nm for chromium, 334.941 nm and 336.121 nm for titanium, and 371.994 nm and 385.991 nm for iron. In the analysis wavelengths of the main metallic elements in tungsten-copper alloys, the analysis wavelengths of tungsten include 400.875 nm and 407.436 nm for tungsten, 324.754 nm and 327.396 nm for copper, and 379.825 nm, 390.295 nm, and 386.410 nm for molybdenum.

[0036] Specifically, S101 includes: dividing the hyperspectral image of the engine exhaust flame into regions and determining the average spectrum of each region in the hyperspectral image of the engine exhaust flame; combining the average spectrum of each region in the hyperspectral image of the engine exhaust flame corresponding to different bands, and using interpolation to determine the spectral value corresponding to the analysis wavelength according to the analysis wavelength of different engine materials.

[0037] Please refer to Figure 2 , Figure 2 Hyperspectral image of engine exhaust provided in an embodiment of this application.

[0038] In practical applications, the hyperspectral image of the engine exhaust flame includes multiple spectral images in different bands. In order to avoid interference from data other than the engine exhaust flame in the hyperspectral image of the engine exhaust flame, it is first necessary to extract the exhaust flame part from the hyperspectral image of the engine exhaust flame.

[0039] Specifically, in this embodiment of the application, the region division of the hyperspectral image of the engine exhaust flame mainly refers to the region division of the hyperspectral image involving the engine exhaust flame, specifically including the following S1 to S6.

[0040] S1, determine a reference hyperspectral image, and perform contrast enhancement processing on the exhaust flame portion and background portion in the reference hyperspectral image; the engine exhaust flame radiation intensity in the corresponding band of the reference hyperspectral image is greater than the engine exhaust flame radiation intensity in other bands.

[0041] The reference hyperspectral image can be understood as the spectral image with a higher exhaust radiation intensity selected from hyperspectral images in various bands, and used as the reference image.

[0042] In one alternative implementation, a Gaussian kernel can be used to smooth the reference hyperspectral image in order to suppress high-frequency noise in the reference hyperspectral image.

[0043] S2, obtain the gray-level histogram of the reference hyperspectral image after contrast enhancement processing, and perform maximum inter-class variance calculation on the gray-level histogram to determine the binarization threshold of the gray-level histogram.

[0044] Gray-scale histograms are used to represent pixel brightness in hyperspectral images to facilitate the separation of the tail flame portion from other portions in a reference hyperspectral image.

[0045] In this embodiment, the calculation of maximizing the inter-class variance of the grayscale histogram is implemented based on the Otus algorithm: The specific Otus algorithm can be represented by the following formula (1): (1); in, Represents the variance between classes; and These represent the pixel percentages of the foreground and background in the grayscale histogram, respectively. and These represent the average pixel values ​​of the foreground and background, respectively; T represents the pixel brightness.

[0046] In practical applications, the pixel brightness of each pixel in the grayscale histogram is traversed to determine... The pixel brightness T corresponding to the maximum value is the binarization threshold.

[0047] S3. Based on the binarization threshold, the grayscale histogram is binarized to obtain a binary image of the grayscale histogram.

[0048] S3 above refers to retaining pixels in the grayscale histogram whose brightness is greater than or equal to the binarization threshold, and removing other pixels in the grayscale histogram whose brightness is less than the binarization threshold, so that only the engine exhaust flame is retained in the grayscale histogram.

[0049] S4, perform morphological processing on the binary image to obtain the engine exhaust flame image of the reference hyperspectral image.

[0050] S5. Based on the engine exhaust flame image, determine the region division position of the engine exhaust flame hyperspectral image.

[0051] S6, the hyperspectral image of the engine exhaust flame is divided into regions according to the region division position.

[0052] Morphological processing of a binary image refers to using an opening operation to first erode and then dilate the binary image, in order to remove small noise in the image while preserving the main shape of the engine exhaust flame.

[0053] Furthermore, considering the large size of the engine exhaust flame area, in order to facilitate the subsequent determination of the spectral values ​​of the analysis wavelength and improve the signal-to-noise ratio of the image, it is necessary to divide the exhaust flame area into several smaller regions, each with the same number of pixels.

[0054] After determining the region division location of the reference hyperspectral image, the division results of the reference hyperspectral image will be applied to hyperspectral images of other frequency bands outside the reference hyperspectral image.

[0055] Furthermore, after dividing the hyperspectral images of engine exhaust plumes in various frequency bands into regions, the spectral values ​​corresponding to the analytical wavelengths can be determined by interpolation based on the analytical wavelengths of different engine materials and the average spectra of each region in the hyperspectral images of engine exhaust plumes corresponding to different frequency bands.

[0056] Please refer to Figure 2 First, the average spectrum of all small regions in the hyperspectral image of the engine exhaust flame for each frequency band is calculated, and the average spectrum is normalized by min-max to map the average spectrum of each small region to between 0 and 1. Then, a region of interest is selected in the hyperspectral image of the engine exhaust flame, and the spectral value corresponding to the analysis wavelength is determined by interpolation according to the analysis wavelength of different engine materials.

[0057] For the embodiments of this application, the analysis wavelengths of the engine materials include the analysis wavelengths of the injection disc materials, the coating materials, and the inner wall materials. Furthermore, the spectral values ​​corresponding to the analysis wavelengths of the different engine materials can be expressed as follows: ; ; ; in, The values ​​represent the analysis wavelengths, p represents the injection disk material, n represents the total number of analysis wavelengths for the injection disk material, h represents the coating material, m represents the total number of analysis wavelengths for the coating material, q represents the inner wall material, and r represents the total number of analysis wavelengths for the inner wall material.

[0058] S102, determine the multidimensional spectral features of the hyperspectral image based on the spectral values ​​corresponding to the analysis wavelengths of different engine materials.

[0059] The multidimensional spectral features refer to the comprehensive analysis of the spectral values ​​corresponding to the analysis wavelengths of different engine materials, and the analysis of the different dimensions of spectral features exhibited by the engine materials in the spectral image.

[0060] Specifically, the multidimensional spectral features are determined by the following formulas (2) to (6): (2); (3); (4); (5); (6); in, , , , , These represent the spectral features of different dimensions of the hyperspectral image; j represents the analysis wavelength type of the hyperspectral image, j=p or h or q, where p represents the injection disk material, h represents the coating material, and q represents the inner wall material; n represents the total number of wavelengths of the injection disk material; m represents the total number of analysis wavelengths of the coating material; and r represents the total number of analysis wavelengths of the inner wall material. This represents the i-th analysis wavelength corresponding to the analysis wavelength type j in the hyperspectral image; represents the spectral value corresponding to the i-th analytical wavelength when the analytical wavelength type is j; c represents the spectral characteristic coefficient, c=1 when j=p, c=2 when j=h, and c=3 when j=q.

[0061] S103, using a pre-trained engine fault diagnosis model, perform fault diagnosis and identification on the multidimensional spectral features to determine the engine fault diagnosis result.

[0062] The engine fault diagnosis model is specifically a support vector machine (SVM) model. Its core is to find a hyperplane that can separate data points of different categories as clearly as possible. In the embodiments of this application, the data points are multidimensional spectral features obtained from the hyperspectral image of the engine exhaust flame.

[0063] For linear data points, in the process of constructing the engine fault diagnosis model, any hyperplane is described by the following formula (7): (7); Where w and b represent hyperplane coefficients, w is the normal vector of the hyperplane, and b is the bias term.

[0064] The optimization function is represented by the following formula (8) to ensure that all samples are correctly classified and the function margin is at least 1.

[0065] (8); in, The mathematical representation of the i-th multidimensional spectral feature. This represents the label category (i.e., engine fault type) corresponding to the i-th multidimensional spectral feature.

[0066] For nonlinear data points, a kernel function is introduced to map the nonlinear multidimensional spectral features originally in the original space to a higher-dimensional space. Then, the optimal hyperplane is solved in the higher-dimensional space to achieve the classification of multidimensional spectral features.

[0067] In practical applications, the kernel function can be a polynomial kernel function or a Gaussian kernel function. The polynomial kernel function can be represented by the following formula (9), and the Gaussian kernel function can be represented by the following formula (10): (9); (10); In the embodiments of this application, , Let represent the i-th and j-th multidimensional spectral features, respectively; d represents the order of the polynomial kernel function. Represents higher-dimensional space , Gaussian kernel; A parameter representing how the similarity between multidimensional spectral features varies with distance.

[0068] In one optional embodiment of this application, in order to determine the fault diagnosis result of the engine, a corresponding engine fault diagnosis model is set for different engine fault types.

[0069] For the multidimensional sample spectral features of engine fault category 'injector disc ablation', its label is set to 1, and the labels of other multidimensional sample spectral features are set to 0. Then, the hyperplane coefficients of the multidimensional sample spectral features of the two types of labels are determined, and then the engine fault diagnosis model for identifying injection disc ablation is determined based on the hyperplane coefficients.

[0070] For the multidimensional sample spectral features of engine fault category coating ablation, its label is set to 1, and the labels of other multidimensional sample spectral features are set to 0. Then, the hyperplane coefficients of the multidimensional sample spectral features of the two types of labels are determined, and then the engine fault diagnosis model for identifying coating ablation is determined based on the hyperplane coefficients.

[0071] For the multidimensional sample spectral features of engine fault category inner wall ablation, its label is set to 1, and the labels of other multidimensional sample spectral features are set to 0. Then, the hyperplane coefficients of the multidimensional sample spectral features of the two types of labels are determined, and then the engine fault diagnosis model for identifying inner wall ablation is determined based on the hyperplane coefficients.

[0072] For the multidimensional sample spectral features without engine faults, their label is set to 1, and the labels for other multidimensional sample spectral features are set to 0. Then, the hyperplane coefficients of the multidimensional sample spectral features with the two types of labels are determined, and then the engine fault diagnosis model for identifying the spectrum of fault-free engines is determined based on the hyperplane coefficients.

[0073] Before obtaining the multidimensional sample spectral features and training the SVM model, the engine fault category corresponding to each multidimensional sample spectral feature is first obtained by performing cluster analysis on the multidimensional sample spectral features.

[0074] Specifically, cluster analysis can employ agglomerative hierarchical clustering, a bottom-up hierarchical clustering method belonging to unsupervised learning clustering algorithms. It gradually merges similar samples or clusters to eventually form a tree-like clustering structure.

[0075] Please refer to Figure 3 , Figure 3 This is a schematic diagram of a cohesive hierarchical clustering method provided in an embodiment of this application.

[0076] Figure 3 The numbers shown on the horizontal axis can be understood as multiple discrete data points. At the beginning of clustering, each data point is first regarded as an independent cluster. In each clustering, the two closest clusters are found and merged into a new cluster until all data points reach the preset number of clusters.

[0077] The number of clusters preset for the spectral features of each multidimensional sample refers to the number of engine fault types. As mentioned above, the engine fault types include: injection disc ablation, coating ablation, inner wall ablation, and normal operation.

[0078] Please refer to Figure 4 , Figure 4 This diagram illustrates the clustering of multidimensional sample spectral features in two-dimensional and three-dimensional spaces, as provided in the embodiments of this application. In practical applications, the multidimensional sample spectral features are clustered using the aforementioned agglomerative hierarchical clustering method. Furthermore, by combining the properties of the multidimensional sample spectral features of each cluster, engine fault type labels for different clusters are determined. This allows for the construction of different categories of engine fault models based on these labels.

[0079] After obtaining models for diagnosing different types of engine faults, when it is necessary to identify engine faults, the multidimensional spectral features corresponding to the engine exhaust flame are sequentially input into the above engine fault diagnosis models to obtain the engine fault categories output by each engine fault diagnosis model. Finally, the engine fault categories output by each model are combined to obtain the final diagnosis result.

[0080] Furthermore, for multidimensional sample spectral features that cannot be identified, or multidimensional sample spectral features with contradictory fault category identification results, the method further includes the following S104.

[0081] S104, Combining the historical multidimensional spectral features of known engine fault types, perform neighbor classification and identification on the multidimensional spectral features to determine the fault diagnosis result corresponding to the multidimensional spectral features.

[0082] The neighbor classification and recognition is specifically implemented through a k-Nearest Neighbor (kNN) classifier. In this embodiment, for multidimensional sample spectral features that cannot be identified or for multidimensional sample spectral features with contradictory fault category identification results, firstly, the k closest historical multidimensional spectral features are determined, and then the engine fault type with the most common type among the k historical multidimensional spectral features is taken as the fault diagnosis result of the multidimensional sample spectral feature.

[0083] Please refer to Figure 5 , Figure 5 This is a schematic diagram of k-nearest neighbor classification provided in an embodiment of this application.

[0084] like Figure 5 As shown, Figure 5 The triangles and squares represent two different data categories, while the circle represents data of an unknown category. When k equals 3, the relationship between... Figure 5 If the known data adjacent to the circle includes two data points with the category of square and one data point with the category of circle, then the category of the data can be determined to be square.

[0085] In this embodiment of the application, the distance between the spectral features of the multidimensional samples that cannot be identified by the historical multidimensional spectral features can be represented by Euclidean distance, Manhattan distance, and Minkowski distance.

[0086] For example, suppose that any two multidimensional spectral features in the embodiments of this application are A( B () The Euclidean distance can be calculated using the following formula (11): (11); in, This represents the Euclidean distance between multidimensional spectral features A and B.

[0087] In summary, this application provides an engine fault diagnosis method. Based on the analysis wavelengths of different engine materials, the method extracts the spectrum from a hyperspectral image of the engine exhaust flame to determine the spectral values ​​corresponding to the analysis wavelengths of different engine materials in the hyperspectral image. The analysis wavelengths of different engine materials correspond to different engine fault types. Based on the spectral values ​​corresponding to the analysis wavelengths of different engine materials, the method determines the multidimensional spectral features of the hyperspectral image. Using a pre-trained engine fault diagnosis model, the method performs fault diagnosis and identification on the multidimensional spectral features to determine the engine fault diagnosis result. This method diagnoses engine faults by analyzing the engine exhaust flame spectrum, effectively identifying faults such as engine wear and ablation.

[0088] This application also provides an engine fault diagnosis device; please refer to... Figure 6 , Figure 6 This is a structural diagram of the engine fault diagnosis device provided in an embodiment of this application.

[0089] like Figure 6 As shown, the engine fault diagnosis device includes: The spectral value determination unit 601 is used to extract the spectrum of the hyperspectral image of the engine exhaust flame based on the analysis wavelength of different engine materials, and determine the spectral values ​​corresponding to the analysis wavelengths of different engine materials in the hyperspectral image; wherein, the analysis wavelengths of different engine analysis materials correspond to different engine fault types.

[0090] The multidimensional feature determination unit 602 is used to determine the multidimensional spectral features of the hyperspectral image based on the spectral values ​​corresponding to the analysis wavelengths of different engine materials.

[0091] The model diagnostic unit 603 is used to perform fault diagnosis and identification on the multidimensional spectral features using a pre-trained engine fault diagnosis model, and to determine the fault diagnosis result of the engine.

[0092] In one optional embodiment of this application, the step of extracting the spectrum from the hyperspectral image of the engine exhaust based on the analysis wavelength of different engine materials, and determining the spectral values ​​corresponding to the analysis wavelengths of different engine materials in the hyperspectral image, includes: dividing the hyperspectral image of the engine exhaust into regions and determining the average spectrum of each region in the hyperspectral image of the engine exhaust; combining the average spectrum of each region in the hyperspectral image of the engine exhaust corresponding to different bands, and using interpolation to determine the spectral values ​​corresponding to the analysis wavelengths according to the analysis wavelengths of different engine materials.

[0093] In one optional embodiment of this application, the process of region division of the engine exhaust hyperspectral image includes: determining a reference hyperspectral image and performing contrast enhancement processing on the exhaust portion and background portion of the reference hyperspectral image; the engine exhaust radiation intensity in the corresponding band of the reference hyperspectral image is greater than the engine exhaust radiation intensity in other bands; obtaining a grayscale histogram of the contrast-enhanced reference hyperspectral image and performing maximum inter-class variance calculation on the grayscale histogram to determine a binarization threshold for the grayscale histogram; performing binarization processing on the grayscale histogram according to the binarization threshold to obtain a binary image of the grayscale histogram; performing morphological processing on the binary image to obtain an engine exhaust image of the reference hyperspectral image; determining the region division position of the engine exhaust hyperspectral image based on the engine exhaust image; and performing region division on the engine exhaust hyperspectral image according to the region division position.

[0094] In one optional embodiment of this application, the multidimensional image features of the hyperspectral image are determined by the following formula: ; ; ; ; ; in, , , , , These represent the spectral features of different dimensions of the hyperspectral image; j represents the analysis wavelength type of the hyperspectral image, j=p or h or q, where p represents the injection disk material, h represents the coating material, and q represents the inner wall material; n represents the total number of wavelengths of the injection disk material; m represents the total number of analysis wavelengths of the coating material; and r represents the total number of analysis wavelengths of the inner wall material. This represents the i-th analysis wavelength corresponding to the analysis wavelength type j in the hyperspectral image; represents the spectral value corresponding to the i-th analytical wavelength when the analytical wavelength type is j; c represents the spectral characteristic coefficient, c=1 when j=p, c=2 when j=h, and c=3 when j=q.

[0095] In one optional embodiment of this application, the engine fault diagnosis model is a support vector machine model, and different engine fault diagnosis models correspond to different engine fault types. The engine fault diagnosis model is trained in the following way: for an engine fault diagnosis model used to identify any fault type, based on the first sample spectral features corresponding to the fault type and the second sample spectral features corresponding to other fault types, determine the hyperplane coefficients that maximize the spatial distance between the first sample spectral features and the second sample spectral features; and construct the fault diagnosis model corresponding to the fault type based on the hyperplane coefficients.

[0096] In one optional embodiment of this application, different engine fault types correspond to different engine fault diagnosis models; the step of using pre-trained engine fault diagnosis models to perform fault diagnosis and identification on the multidimensional spectral features and determine the engine fault diagnosis result includes: sequentially inputting the multidimensional spectral features into each of the engine fault diagnosis models to obtain the engine fault type output by each engine fault diagnosis model; and determining the engine fault diagnosis result based on the engine fault diagnosis type output by each engine fault diagnosis model.

[0097] In one optional embodiment of this application, if the engine diagnostic model cannot identify the multidimensional spectral features, the method further includes: combining historical multidimensional spectral features of known engine fault types to perform neighbor classification identification on the multidimensional spectral features, and determining the fault diagnosis result corresponding to the multidimensional spectral features.

[0098] The device embodiments provided in this embodiment and the method embodiments of this application belong to the same application concept. For technical details not described in detail in this embodiment, please refer to the specific processing content of the engine fault diagnosis method provided in the above embodiments of this application, which will not be repeated here.

[0099] This application also provides an electronic device, such as... Figure 7 As shown, Figure 7 This is a schematic diagram of an electronic device structure provided in an embodiment of this application.

[0100] like Figure 7 As shown, the electronic device includes: Processor 210; Memory 200 for storing executable instructions of the processor 210; The processor 210 is used to execute the engine fault diagnosis method disclosed in any of the above embodiments by running instructions in the memory 200.

[0101] The processor 210, memory 200, communication interface 220, input device 230, and output device 240 are interconnected via a bus. Among them: A bus can include a pathway for transmitting information between various components of a computer system.

[0102] Processor 210 can be a general-purpose processor, such as a general-purpose central processing unit (CPU), a microprocessor, etc., or an application-specific integrated circuit (ASIC), or one or more integrated circuits used to control the execution of the program of the present invention. It can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0103] Processor 210 may include a main processor, as well as a baseband chip, modem, etc.

[0104] The memory 200 stores a program that executes the technical solution of this invention, and may also store an operating system and other key business functions. Specifically, the program may include program code, which includes computer operation instructions. More specifically, the memory 200 may include read-only memory (ROM), other types of static storage devices capable of storing static information and instructions, random access memory (RAM), other types of dynamic storage devices capable of storing information and instructions, disk storage, flash memory, etc.

[0105] Input device 230 may include a device for receiving user input data and information, such as a keyboard, mouse, camera, scanner, touch screen, etc.

[0106] Output device 240 may include devices that allow information to be output to a user, such as a display screen, printer, speaker, etc.

[0107] The communication interface 220 may include a device that uses any transceiver to communicate with other devices or communication networks, such as Ethernet, Radio Access Network (RAN), Wireless Local Area Network (WLAN), etc.

[0108] The processor 210 executes the program stored in the memory 200 and calls other devices, and can be used to implement each step of any of the engine material erosion rate determination methods provided in the above embodiments of this application.

[0109] In addition to the methods and apparatus described above, embodiments of this application may also be computer program products, comprising computer program instructions that, when executed by a processor, cause the processor to perform the steps in the methods for determining the erosion rate of engine materials according to various embodiments of this application.

[0110] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this application. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0111] Furthermore, embodiments of this application may also be storage media storing a computer program, which is executed by a processor in the steps of the method for determining the erosion rate of engine materials in various embodiments of this application.

[0112] For the foregoing method embodiments, in order to simplify the description, they are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, because according to this application, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0113] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For apparatus embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0114] The steps in the methods of the various embodiments of this application can be adjusted, merged, or deleted in order according to actual needs, and the technical features described in each embodiment can be replaced or combined.

[0115] The modules and sub-modules in the apparatus and terminal in the various embodiments of this application can be merged, divided, and deleted according to actual needs.

[0116] It should be understood that the disclosed terminals, devices, and methods can be implemented in other ways, given the several embodiments provided in this application. For example, the terminal embodiments described above are merely illustrative. For instance, the division of modules or sub-modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple sub-modules or modules may be combined or integrated into another module, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or modules, and may be electrical, mechanical, or other forms.

[0117] The modules or submodules described as separate components may or may not be physically separate. The components that constitute a module or submodule may or may not be physical modules or submodules; that is, they may be located in one place or distributed across multiple network modules or submodules. Some or all of the modules or submodules can be selected to achieve the purpose of this embodiment's solution, depending on actual needs.

[0118] Furthermore, the functional modules or sub-modules in the various embodiments of this application can be integrated into one processing module, or each module or sub-module can exist physically separately, or two or more modules or sub-modules can be integrated into one module. The integrated modules or sub-modules described above can be implemented in hardware or in the form of software functional modules or sub-modules.

[0119] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0120] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software unit executed by a processor, or a combination of both. The software unit can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0121] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0122] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for diagnosing engine faults, characterized in that, include: Based on the analysis wavelengths of different engine materials, the hyperspectral image of the engine exhaust flame is subjected to spectral extraction to determine the spectral values ​​corresponding to the analysis wavelengths of different engine materials in the hyperspectral image; wherein, the analysis wavelengths of different engine analysis materials correspond to different engine fault types. The multidimensional spectral features of the hyperspectral image are determined based on the spectral values ​​corresponding to the analysis wavelengths of different engine materials. Using a pre-trained engine fault diagnosis model, fault diagnosis and identification are performed on the multidimensional spectral features to determine the engine fault diagnosis result; The multidimensional spectral features of the hyperspectral image are determined by the following formula: ； ； ； ； ； in, , , , , These represent the spectral features of different dimensions of the hyperspectral image; j represents the analysis wavelength type of the hyperspectral image, j=p or h or q, where p represents the injection disk material, h represents the coating material, and q represents the inner wall material; n represents the total number of wavelengths of the injection disk material; m represents the total number of analysis wavelengths of the coating material; and r represents the total number of analysis wavelengths of the inner wall material. This represents the i-th analysis wavelength corresponding to the analysis wavelength type j in the hyperspectral image; represents the spectral value corresponding to the i-th analytical wavelength when the analytical wavelength type is j; c represents the spectral characteristic coefficient, c=1 when j=p, c=2 when j=h, and c=3 when j=q.

2. The method according to claim 1, characterized in that, The process of extracting the spectrum from the hyperspectral image of the engine exhaust flame based on the analysis wavelengths of different engine materials, and determining the spectral values ​​corresponding to the analysis wavelengths of different engine materials in the hyperspectral image, includes: The hyperspectral image of the engine exhaust flame is divided into regions, and the average spectrum of each region in the hyperspectral image of the engine exhaust flame is determined. By combining the average spectrum of each region in the hyperspectral image of the engine exhaust flame corresponding to different wavebands, and based on the analysis wavelength of different engine materials, the spectral value corresponding to the analysis wavelength is determined by interpolation.

3. The method according to claim 2, characterized in that, The process of dividing the hyperspectral image of the engine exhaust plume into regions includes: A reference hyperspectral image is determined, and the exhaust flame portion and background portion in the reference hyperspectral image are subjected to contrast enhancement processing; the exhaust flame radiation intensity of the engine in the corresponding band of the reference hyperspectral image is greater than the exhaust flame radiation intensity of the engine in other bands. Obtain the gray-level histogram of the reference hyperspectral image after contrast enhancement processing, and perform maximum inter-class variance calculation on the gray-level histogram to determine the binarization threshold of the gray-level histogram; The gray-level histogram is binarized according to the binarization threshold to obtain a binary image of the gray-level histogram. Morphological processing is performed on the binary image to obtain the engine exhaust plume image of the reference hyperspectral image; Based on the engine exhaust plume image, determine the region division position of the engine exhaust plume hyperspectral image; The hyperspectral image of the engine exhaust flame is divided into regions based on the region division location.

4. The method according to claim 1, characterized in that, The engine fault diagnosis model is a support vector machine model, and different engine fault types correspond to different engine fault diagnosis models. The engine fault diagnosis model is trained in the following way: For an engine fault diagnosis model used to identify any fault type, the hyperplane coefficient that maximizes the spatial distance between the first sample spectral feature and the second sample spectral feature is determined based on the first sample spectral feature corresponding to the fault type and the second sample spectral feature corresponding to other fault types. Based on the hyperplane coefficients, a fault diagnosis model corresponding to this fault type is constructed.

5. The method according to claim 1, characterized in that, Different engine fault types correspond to different engine fault diagnosis models; The step of using a pre-trained engine fault diagnosis model to perform fault diagnosis and identification on the multidimensional spectral features, and determining the engine fault diagnosis result, includes: The multidimensional spectral features are sequentially input into each of the engine fault diagnosis models to obtain the engine fault types output by each engine fault diagnosis model. Based on the engine fault diagnosis type output by each engine fault diagnosis model, the fault diagnosis result of the engine is determined.

6. The method according to claim 1, characterized in that, If the engine diagnostic model cannot identify the multidimensional spectral features, the method further includes: By combining the historical multidimensional spectral features of known engine fault types, the multidimensional spectral features are classified and identified by proximity to determine the fault diagnosis result corresponding to the multidimensional spectral features.

7. An engine fault diagnosis device, characterized in that, The apparatus used in the method according to any one of claims 1-6 comprises: The spectral value determination unit is used to extract the spectrum from the hyperspectral image of the engine exhaust flame based on the analysis wavelength of different engine materials, and determine the spectral values ​​corresponding to the analysis wavelengths of different engine materials in the hyperspectral image; wherein, the analysis wavelengths of different engine analysis materials correspond to different engine fault types. A multidimensional feature determination unit is used to determine the multidimensional spectral features of the hyperspectral image based on the spectral values ​​corresponding to the analysis wavelengths of different engine materials. The model diagnostic unit is used to perform fault diagnosis and identification on the multidimensional spectral features using a pre-trained engine fault diagnosis model, and to determine the engine fault diagnosis result.

8. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the engine fault diagnosis method according to any one of claims 1 to 6 by running instructions in the memory.

9. A computer storage medium, characterized in that, The computer storage medium stores instructions, and when the instructions are executed, the engine fault diagnosis method according to any one of claims 1 to 6 is performed by the processor.

Images