Defective abrasive particle type migration identification method and related equipment
By combining the abrasive integrated feature extractor and the Gaussian kernel function, a transfer learning model for image transfer of defective abrasive image is constructed with improved domain adaptation, which solves the problem of low image recognition accuracy of defective abrasive abrasive type recognition .
Patent Information
- Application Number
- CN202411045858.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-31
- Publication Date
- 2024-11-08
AI Technical Summary
In the prior art, the incomplete image of the broken abrasive particles and the lack of significant identification features lead to low abrasive particle recognition accuracy, which is difficult to meet the requirements of industrial applications for high identification accuracy of wear states.
Using abrasive particle comprehensive feature extractor, abrasive particle features are extracted from the spatial domain and the frequency domain. Combined with Gaussian kernel function and attention mechanism, a residual abrasive particle image transfer learning model driven by improved domain adaptation is constructed, and multi-channel feature screening is performed. With the fusion strategy, we realize intelligent identification of abrasive type.
Improve the feature extraction ability and identification accuracy of the image of broken abrasive particles, enhance the weighting of multivariate features of abrasive particles and the positioning accuracy of prominent feature areas, solve the problems of incomplete image of broken abrasive particles and lack of identification features, and achieve more accurate identification of abrasive particles.
Smart Images

Figure BDA0004973480310000094 
Figure BDA0004973480310000096 
Figure BDA0004973480310000104
Abstract
Description
Background Art
[0002] During the long-term operation of mechanical equipment, due to the complexity of operating conditions and the harshness of the environment, wear phenomena gradually emerge, which not only leads to a reduction in system reliability, but also may cause serious accidents, resulting in immeasurable economic losses and casualties. Therefore, accurate and effective monitoring of the wear state of mechanical equipment has always been a top priority in industrial research. As a direct product of friction pair wear, abrasive particles carry a large amount of key wear information during the operation of the equipment, making abrasive particle analysis occupy a pivotal position in the evaluation of equipment wear state. Wear Debris Analysis (WDA), which takes abrasive particles as the research object, can not only monitor the wear state of mechanical equipment, but also accurately analyze the wear mechanism by identifying different types of abrasive particles, thereby finding out the causes of induced wear failures. In particular, the microscopic morphology of the surface of different types of abrasive particles can accurately reveal the different wear mechanisms of mechanical equipment, so the identification of typical failure abrasive particle types can provide strong support for the wear state monitoring of mechanical equipment. Therefore, studying an engineering-oriented abrasive particle identification technology is of great significance to improving the level of wear state monitoring of mechanical equipment.
[0003] As a direct product of friction pair wear, abrasive particles theoretically carry all the information of the entire wear process. Different types of abrasive particles have different morphological characteristics: spherical abrasive particles have a smooth surface and a spherical shape; cutting abrasive particles are slender and curved; severe sliding abrasive particles have obvious parallel scratches and irregular contours; fatigue spalling abrasive particles have pits and pits on the surface, and their shape contours are also very irregular. According to the above characteristics of abrasive particles, existing scholars have studied two types of abrasive particle identification methods: feature parameter method and deep learning method. The feature parameter method relies on manually summarized identification feature parameters, such as abrasive particle aspect ratio, area, roundness, roughness, etc. Although these methods have been fully studied in laboratory environments and have achieved high abrasive particle identification accuracy, they are affected by factors such as abrasive particle chain adhesion and abrasive particles blocking each other. When these methods are applied to the incomplete abrasive particle images collected in engineering scenarios, the correct feature parameters cannot be extracted, resulting in a decrease in their identification accuracy. The deep learning method, due to its superior feature extraction ability, can avoid the pain points of the feature parameter method to a certain extent, and therefore has more engineering application prospects. However, the models constructed by deep learning methods often require a large number of sample training to support the improvement of their feature extraction capabilities, and the models are not very interpretable, which makes it difficult for existing deep learning models to accurately and completely extract the discriminative features of wear particles.
[0004] In summary, due to the incomplete image information and the lack of significant identification features of the incomplete wear particle images collected in engineering scenarios, the existing feature parameter method and deep learning method are difficult to meet the requirements of high identification accuracy of wear status in actual industrial applications. Summary of the invention
[0005] The purpose of the present invention is to provide a method and related equipment for identifying the migration of defective abrasive particles, so as to solve the technical problem in the prior art that the defective abrasive particle images are incomplete and significant identification features are missing, resulting in low abrasive particle identification accuracy.
[0006] In order to achieve the above object, the present invention adopts the following technical scheme:
[0007] A method for identifying the migration of defective abrasive particles comprises the following steps:
[0008] Combined with the generation mechanism of wear particles, a comprehensive feature extractor of wear particles is established from two dimensions: spatial domain and frequency domain.
[0009] Based on the wear particle comprehensive feature extractor, a parameter-adaptive Gaussian kernel function is designed for the multivariate features of wear particles, and an improved domain adaptation-driven transfer learning model for incomplete wear particle images is established.
[0010] The transfer learning model of incomplete wear particle images is used to guide the wear particle comprehensive feature extractor to extract effective identification features of wear particle images. The spatial and channel attention mechanisms are combined to establish a multi-feature screening and fusion strategy for incomplete wear particles, and then a multi-channel feature intelligent identification model for incomplete wear particle types is constructed. The incomplete wear particles are identified using the incomplete wear particle type intelligent identification model.
[0011] Furthermore, the steps for establishing the wear particle comprehensive feature extractor are as follows:
[0012] A shape feature encoder is built based on the EfficientNet model, and convolution operations are designed for different shape feature layers of wear particles to extract three-layer wear particle shape features. An encoder with three convolutional layers is designed for wear particle texture features to extract single-layer wear particle texture features.
[0013] Combine three layers of shape features and one layer of texture features, group the features along the channel dimension in multiple subspaces, and use three parallel sub-branches to periodically arrange the features for each pair of shape features and texture features that match the dimensions.
[0014] The grouped convolution method is used to achieve feature fusion by adding the arranged features;
[0015] Three trainable adaptive filters are designed for incomplete wear particle images, and the discrete cosine transform method is optimized in the frequency dimension to achieve the extraction of morphological features of wear particle images after filtering in different frequency domains.
[0016] The three feature extraction branches of shape, texture and filtering are combined to complete the construction of the wear particle comprehensive feature extractor through feature splicing.
[0017] Furthermore, the three feature extraction branches of shape, texture and filtering are respectively: based on the three-layer wear shape features, a wear shape feature map is generated through a domain connection decoder; based on the single-layer wear texture features, a bilinear interpolation method is used to generate a wear texture feature map; based on the image after frequency domain filtering, a wear frequency domain feature map is generated through an inverse discrete cosine transform method.
[0018] Furthermore, the specific steps of designing a parameter-adaptive Gaussian kernel function for multivariate features of wear particles based on a comprehensive feature extractor of wear particles and establishing an improved domain adaptation-driven transfer learning model for incomplete wear particle images are as follows:
[0019] Perform dimension reduction on the multivariate features extracted by the wear particle comprehensive feature extractor and calculate the joint distribution of the multivariate features of the wear particles;
[0020] The domain adaptation method of joint maximum mean difference is used to calculate the average embedding distance between the feature distribution P of the complete wear particle image and the feature distribution Q of the incomplete wear particle image in Hilbert space, and the calculation result JMMD that measures their difference is obtained;
[0021] The Gaussian kernel function is selected as the kernel function for calculating the joint maximum mean difference, and a parameter-trainable kernel function is set for each wear particle image feature map to achieve adaptive adjustment of the key parameters of the kernel function in JMMD.
[0022] The feature extractor is guided to extract effective identification features of wear particle images by aligning feature distribution, and a transfer learning model for incomplete wear particle images is constructed.
[0023] Furthermore, the specific process of combining the spatial and channel attention mechanisms to establish a multi-feature screening and fusion strategy for defective wear particles, and then constructing a multi-channel feature intelligent identification model for defective wear particle types is as follows:
[0024] The attention mechanism is used to locate the significant feature areas of the wear particle image from the spatial dimension, and the multi-channel features of the wear particles are weighted from the channel dimension to achieve the screening of effective wear particle identification information;
[0025] The features filtered by the attention mechanism are used as input, a classifier is designed for the multi-channel feature map of wear particles, and an intelligent identification model for the type of incomplete wear particles is established.
[0026] Furthermore, the training set of the incomplete wear particle type intelligent identification model is wear particle images with different degrees of incompleteness generated based on complete wear particle image samples.
[0027] Furthermore, the loss function of the incomplete wear particle type intelligent identification model is a weighted sum of spatial domain feature map generation loss, domain adaptation joint distribution loss and wear particle image classification loss.
[0028] In a second aspect, the present invention provides a defective wear particle type migration identification system, comprising a wear particle comprehensive feature extractor construction module, a defective wear particle image migration learning model construction module and a defective wear particle identification module, wherein:
[0029] Wear particle comprehensive feature extractor construction module: used to combine the generation mechanism of wear particles and establish a wear particle comprehensive feature extractor from two dimensions: spatial domain and frequency domain;
[0030] The module for constructing the transfer learning model of incomplete wear particle images: Based on the wear particle comprehensive feature extractor, a parameter-adaptive Gaussian kernel function is designed for the multivariate features of wear particles, and an improved domain adaptation-driven transfer learning model of incomplete wear particle images is established;
[0031] Incomplete wear particle identification module: It is used to guide the wear particle comprehensive feature extractor to extract effective identification features of wear particle images through the incomplete wear particle image transfer learning model, and combine the space and channel attention mechanism to establish a multi-feature screening and fusion strategy for incomplete wear particles, and then construct a multi-channel feature incomplete wear particle type intelligent identification model, and use the incomplete wear particle type intelligent identification model to realize the identification of incomplete wear particles.
[0032] According to a third aspect, a terminal device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.
[0033] According to a fourth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and the computer program implements the steps of the above method when executed by a processor.
[0034] Compared with the prior art, the present invention has the following beneficial technical effects:
[0035] The present invention provides a method for migration and identification of defective wear particle types. Based on a comprehensive feature extractor for wear particles, an improved domain adaptation-driven transfer learning model for defective wear particle images is established. A multi-feature screening and fusion strategy for defective wear particles is established in combination with a space and channel attention mechanism. Then, a multi-channel feature intelligent identification model for defective wear particle types is constructed. The defective wear particles are identified by using the intelligent identification model for defective wear particle types. By introducing a complete wear particle image data set, the model is guided to extract effective identification features of the defective wear particle images, and a migration learning model for defective wear particle images is established, thereby improving the model's feature extraction capability for the defective wear particle images. The multi-features of the wear particles are weighted from two dimensions, space and channel, and a classifier is designed for the multi-channel feature map of the wear particles, thereby enhancing the model's identification capability for the defective wear particle images, thereby solving the technical problem that the defective wear particle images are incomplete and the significant identification features are missing, resulting in low wear particle identification accuracy.
[0036] Furthermore, three sub-branches were designed from the spatial domain and frequency domain to extract the shape features, texture features, and image filtering features of the wear particle image, which improved the interpretability of the model feature extractor. By adopting a soft grouping strategy, the model's accuracy in locating the significant feature areas of the wear particle image was improved.
[0037] Furthermore, the Gaussian kernel function is selected as the kernel function for calculating the joint maximum mean difference, and the adaptive adjustment of the key parameters of the kernel function in JMMD is achieved by setting a parameter-trainable kernel function for each wear particle image feature map.
[0038] Furthermore, the attention mechanism is used to locate the significant feature areas of the wear particle image from the spatial dimension, and the multi-channel features of the wear particles are weighted from the channel dimension to achieve the fusion and screening of effective wear particle identification information.
[0039] Furthermore, loss functions are designed for the feature extractor, transfer learning model, and type identification model respectively, and the overall loss function of the model is constructed in a weighted manner. The model parameters are trained and optimized through typical failed wear particle samples to achieve accurate identification of typical failed wear particles.
[0040] Furthermore, the number of iterations and learning rate are set, and the Adam optimizer is used to optimize the weight parameters of the wear particle identification model. The recognition accuracy of the model is tested using a partial incomplete wear particle image dataset as a test set. The incomplete wear particle image dataset is generated using complete wear particle image samples. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] Figure 1 The figure is a flow chart of a method for identifying migration of defective abrasive particles in an embodiment of the present invention.
[0042] Figure 2This is the overall framework diagram of the incomplete wear particle identification model constructed in the present invention.
[0043] Figure 3 This is a structural diagram of the shape feature encoder used in the present invention.
[0044] Figure 4 This is a structural diagram of the soft grouping strategy used in the present invention.
[0045] Figure 5 Schematic diagram of spatial domain feature extraction of the present invention.
[0046] Figure 6 It is a schematic diagram of frequency domain feature extraction of the present invention.
[0047] Figure 7 Schematic diagram of the improved domain adaptation method based on the joint maximum mean difference of the present invention.
[0048] Figure 8 This is a flow chart of the incomplete wear particle image expansion strategy used in the present invention.
[0049] Fig. 9 This is an iterative result diagram of Loss and Accuracy during the model training process of the present invention.
[0050] Fig.10 Schematic diagram of the confusion matrix of the present invention compared with other image recognition methods.
[0051] Fig.11 This is a visualization of the output features of the fully connected layer of the model of the present invention.
[0052] Fig.12 A schematic diagram of a computer device.
[0053] Fig.13 A block diagram of an electronic device. DETAILED DESCRIPTION
[0054] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0055] In the description of the present invention, it should be understood that the terms “include” and “comprises” indicate the presence of described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.
[0056] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.
[0057] It should be further understood that the term "and / or" used in the present specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes these combinations. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present invention generally indicates that the associated objects are in an "or" relationship.
[0058] It should be understood that, although the terms first, second, third, etc. may be used to describe preset ranges, etc. in the embodiments of the present invention, these preset ranges should not be limited to these terms. These terms are only used to distinguish preset ranges from each other. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be referred to as the second preset range, and similarly, the second preset range may also be referred to as the first preset range.
[0059] The word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrases "if it is determined" or "if (stated condition or event) is detected" may be interpreted as "when it is determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)", depending on the context.
[0060] Various structural schematic diagrams of the embodiments disclosed in the present invention are shown in the accompanying drawings. These figures are not drawn to scale, and some details are magnified and some details may be omitted for the purpose of clear expression. The shapes of various regions and layers shown in the figures and the relative sizes and positional relationships therebetween are only exemplary, and may deviate in practice due to manufacturing tolerances or technical limitations, and those skilled in the art may additionally design regions / layers with different shapes, sizes, and relative positions according to actual needs.
[0061] The present invention is further described in detail below in conjunction with the accompanying drawings:
[0062] like Figure 1 As shown, a method for identifying the migration of defective abrasive particles includes the following steps:
[0063] Step 1: Combined with the generation mechanism of wear particles, a comprehensive feature extractor for wear particles is established from two dimensions: spatial domain and frequency domain:
[0064] Detailed process as Figure 2 As shown in FIG. 1 , specifically, s101 builds a shape feature encoder based on the EfficientNet model, and designs convolution operations for different shape feature layers of wear particles to extract three layers of wear particle shape features, such as Figure 3 As shown in the figure, an encoder with three convolutional layers is designed for wear particle texture features to extract single-layer wear particle texture features.
[0065] s102. Design a soft grouping strategy for the fusion of abrasive particle shape and texture features. Combine three layers of shape features and one layer of texture features to group features along the channel dimension in multiple subspaces. For each pair of shape features that match the dimensions, and texture features Three parallel sub-branches are used to periodically arrange the features to obtain G i ,in:
[0066]
[0067] Where: M is the number of groups; K = 32 / M is the number of channels for each group of features; H i ×W i is the feature dimension.
[0068] S103, such as Figure 4 As shown, the soft-grouped feature G is processed by group convolution. i Processing, in three sub-branches for G i According to {N 1 , N 2 , N 3} are grouped for convolution, and a residual structure is introduced to achieve feature fusion by adding features:
[0069]
[0070] Where: N j is the number of groups during convolution, taking {j=1, 2, 3}, {N 1 =4,N 2 =8,N 3 =16}, corresponding to three sub-branches respectively; H i ×W i is the feature dimension; It is a group convolution operation, and the convolution kernel shape is (64 / N j )×1×1;Y i is the output feature.
[0071] S104. Design three trainable adaptive filters for incomplete wear particle images, optimize the discrete cosine transform method in the frequency dimension, and realize the extraction of morphological features of wear particle images after filtering in different frequency domains:
[0072]
[0073] Among them: According to the distribution of image energy, the initial frequency domain range of the three filters is divided to obtain is the trainable weight matrix; σ(·) is the normalization operation.
[0074] s105, such as Figure 5 As shown in FIG, based on the three-layer wear particle shape features, the wear particle shape feature map is generated by connecting the decoder to the field; based on the single-layer wear particle texture features, the wear particle texture feature map is generated by bilinear interpolation; Figure 6 As shown, based on the image after frequency domain filtering, the wear particle frequency domain feature map is generated by inverse discrete cosine transform.
[0075] s106. Combine the three feature extraction branches of shape, texture and filtering to complete the construction of the wear particle comprehensive feature extractor through feature splicing.
[0076] Step 2: Based on the wear particle comprehensive feature extractor, a parameter-adaptive Gaussian kernel function is designed for the multivariate features of wear particles, and an improved domain adaptation-driven transfer learning model for incomplete wear particle images is established:
[0077] like Figure 2 As shown, specifically, S201, the multivariate features extracted by the wear particle comprehensive feature extractor are reduced in dimension by using feature flattening, and converted into an m-order tensor inner product space To calculate the multivariate characteristics of wear particles X 1 , ..., X m The joint distribution of:
[0078]
[0079] Where: X 1:m For the characteristics in the domain m variables on ; Calculates the inner product.
[0080] S202, using the domain adaptation method of the joint maximum mean difference, the average embedding distance between the feature distribution P of the complete wear particle image and the feature distribution Q of the incomplete wear particle image in the Hilbert space is calculated to obtain the calculation result JMMD that measures their difference:
[0081]
[0082] Where: n s is the number of complete wear particle image samples in each group; n t is the number of samples of each group of incomplete wear particle images; k l (·) is the kernel function calculation of samples on different domains l.
[0083] S203, selecting a Gaussian kernel function as the kernel function for calculating the joint maximum mean difference, and setting a parameter-trainable kernel function k(x, x′) for each wear particle image feature map to achieve adaptive adjustment of key parameters of the kernel function in JMMD:
[0084] k(x, x′)=exp(-||xx′|| 2 / (2α·σ 2 ))
[0085] Where: x, x′ are the sample data to be calculated; α is the model trainable parameter; σ is the standard deviation of the sample data.
[0086] S204, such as Figure 7 As shown in the figure, the feature extractor is guided to extract effective identification features of the wear particle image by aligning the feature distribution, and a transfer learning model for incomplete wear particle images is constructed.
[0087] Step 3: Guide the feature extractor to extract effective identification features of the wear particle image through the wear particle salient feature transfer learning model, and establish a multi-feature screening and fusion strategy for incomplete wear particles by combining the spatial and channel attention mechanisms, and then build a multi-channel feature incomplete wear particle type intelligent identification model, and use the incomplete wear particle type intelligent identification model to realize the identification of incomplete wear particles:
[0088] like Figure 2 As shown, specifically, S301, the attention mechanism is used to locate the significant feature area of the wear particle image from the spatial dimension, and the multi-channel features of the wear particles are weighted from the channel dimension to achieve the screening of effective wear particle identification information.
[0089] S302. Using the above features filtered by the attention mechanism as input, a classifier is designed for the multi-channel feature map of wear particles, and an intelligent identification model for the type of incomplete wear particles is established.
[0090] The training and optimization process of the intelligent identification model of defective wear particle types is as follows:
[0091] Loss functions are designed for the feature extractor, transfer learning model, and type identification model respectively. The overall loss function of the model is constructed in a weighted manner. The model parameters are trained and optimized through typical failed wear particle samples to achieve accurate identification of typical failed wear particles.
[0092] Specifically, S401, such as Figure 8As shown in the figure, a sample expansion strategy for incomplete wear particle images is established, and wear particle images with different degrees of incompleteness are generated based on complete wear particle image samples, which serve as the data set for model training.
[0093] S402, the model loss function is divided into three parts: spatial domain feature map generation loss, domain adaptation joint distribution loss and wear particle image classification loss. The loss functions of different modules are weighted summed to form the total loss function Loss of the model:
[0094]
[0095] Among them: The spatial domain features include shape features L S and texture feature L T , the former uses the intersection-over-union loss function L IOU and the binary cross entropy loss function L BCE Calculate, the latter uses the mean square error loss function L MSE ; The joint distribution loss of domain adaptation is measured by using JMMD; The wear particle image classification loss L C Using the cross entropy loss function L CE Calculation is performed; α, β, γ are the coefficients of the objective function loss term; considering that image classification is the main task of the model, α=0.3, β=1.0, y=1.0.
[0096] S403, set the number of iterations and learning rate, use Adam optimizer to optimize the weight parameters of the wear particle identification model, and the iteration results are as follows: Fig. 9 The recognition accuracy of the model is tested by taking some incomplete wear particle image datasets as the test set. Compared with other existing wear particle type classification methods, the recognition accuracy results of different types of wear particles are shown in the figure. Fig.10 As shown in the figure, an ablation experiment is designed to test the improvement of recognition accuracy by different modules.
[0097] The combined modules are:
[0098] Combination 1: spatial domain feature module; Combination 2: spatial domain feature + frequency domain feature + attention mechanism module; Combination 3: spatial domain feature + domain adaptation module; Combination 4: spatial domain feature + frequency domain feature + domain adaptation module; Combination 5: spatial domain feature + attention mechanism + domain adaptation module; Combination 6: spatial domain feature + frequency domain feature + attention mechanism + domain adaptation module; The visualization results of feature distribution are as follows Fig.11 shown.
[0099] It will be appreciated by those skilled in the art that various aspects of the present invention may be implemented as systems, methods or program products. Therefore, various aspects of the present invention may be specifically implemented in the following forms, namely: complete hardware implementation, complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software implementations, which may be collectively referred to herein as "circuits", "modules" or "platforms".
[0100] In yet another embodiment of the present invention, a defective wear particle type migration identification system is provided, comprising a wear particle comprehensive feature extractor construction module, a defective wear particle image migration learning model construction module and a defective wear particle identification module, wherein:
[0101] Wear particle comprehensive feature extractor construction module: used to combine the generation mechanism of wear particles and establish a wear particle comprehensive feature extractor from two dimensions: spatial domain and frequency domain;
[0102] The module for constructing the transfer learning model of incomplete wear particle images: Based on the wear particle comprehensive feature extractor, a parameter-adaptive Gaussian kernel function is designed for the multivariate features of wear particles, and an improved domain adaptation-driven transfer learning model of incomplete wear particle images is established;
[0103] Incomplete wear particle identification module: It is used to guide the wear particle comprehensive feature extractor to extract effective identification features of wear particle images through the incomplete wear particle image transfer learning model, and combine the space and channel attention mechanism to establish a multi-feature screening and fusion strategy for incomplete wear particles, and then construct a multi-channel feature incomplete wear particle type intelligent identification model, and use the incomplete wear particle type intelligent identification model to realize the identification of incomplete wear particles.
[0104] In another embodiment of the present invention, a terminal device is provided, which includes a processor and a memory, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, and is specifically suitable for loading and executing one or more instructions to implement corresponding method processes or corresponding functions; the processor described in the embodiment of the present invention can be used for the operation of the defective wear particle type migration identification method, including:
[0105] Combined with the generation mechanism of wear particles, a comprehensive feature extractor of wear particles is established from two dimensions: spatial domain and frequency domain.
[0106] Based on the wear particle comprehensive feature extractor, a parameter-adaptive Gaussian kernel function is designed for the multivariate features of wear particles, and an improved domain adaptation-driven transfer learning model for incomplete wear particle images is established.
[0107] The transfer learning model of incomplete wear particle images is used to guide the wear particle comprehensive feature extractor to extract effective identification features of wear particle images. The spatial and channel attention mechanisms are combined to establish a multi-feature screening and fusion strategy for incomplete wear particles, and then a multi-channel feature intelligent identification model for incomplete wear particle types is constructed. The incomplete wear particles are identified using the incomplete wear particle type intelligent identification model.
[0108] In another embodiment of the present invention, the present invention also provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a terminal device for storing programs and data. It is understandable that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and the extended storage medium supported by the terminal device, and can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that more specific examples (non-exhaustive list) of the computer-readable storage medium here include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.
[0109] Computer readable storage media also include data signals propagated in baseband or as part of a carrier wave, which carry readable program codes. Such propagated data signals can take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The readable storage medium can also be any readable medium other than a readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, device, or device. The program code contained on the readable storage medium can be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the above.
[0110] Program code for performing the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages such as Java, C++, and the like, as well as conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., via the Internet using an Internet service provider).
[0111] The processor may load and execute one or more instructions stored in a computer-readable storage medium to implement the corresponding steps of the defective abrasive particle type migration identification method in the above embodiment; the processor may load and execute the following steps of one or more instructions in the computer-readable storage medium:
[0112] Combined with the generation mechanism of wear particles, a comprehensive feature extractor of wear particles is established from two dimensions: spatial domain and frequency domain.
[0113] Based on the wear particle comprehensive feature extractor, a parameter-adaptive Gaussian kernel function is designed for the multivariate features of wear particles, and an improved domain adaptation-driven transfer learning model for incomplete wear particle images is established.
[0114] The transfer learning model of incomplete wear particle images is used to guide the wear particle comprehensive feature extractor to extract effective identification features of wear particle images. The spatial and channel attention mechanisms are combined to establish a multi-feature screening and fusion strategy for incomplete wear particles, and then a multi-channel feature intelligent identification model for incomplete wear particle types is constructed. The incomplete wear particles are identified using the incomplete wear particle type intelligent identification model.
[0115] Fig.12 A schematic diagram of a computer device provided by an embodiment of the present invention.
[0116] See also Fig.12, the terminal device is a computer device, and the computer device 60 of this embodiment includes: a processor 61, a memory 62, and a computer program 63 stored in the memory 62 and executable on the processor 61. When the computer program 63 is executed by the processor 61, the method for calculating the composition of fluid in the reservoir transformation wellbore in the embodiment is realized. To avoid repetition, it is not described one by one here. Alternatively, when the computer program 63 is executed by the processor 61, the functions of each model / unit in the calculation system for the composition of fluid in the reservoir transformation wellbore in the embodiment are realized. To avoid repetition, it is not described one by one here.
[0117] The computer device 60 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The computer device 60 may include, but is not limited to, a processor 61 and a memory 62. Those skilled in the art will appreciate that Fig.12 This is only an example of the computer device 60 and does not constitute a limitation of the computer device 60. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device may also include input and output devices, network access devices, buses, etc.
[0118] The processor 61 may be a central processing unit (CPU), or other general-purpose processors, central processing units, graphics processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, data processing logic based on quantum computing, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc.
[0119] The memory 62 may be an internal storage unit of the computer device 60, such as a hard disk or memory of the computer device 60. The memory 62 may also be an external storage device of the computer device 60, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc., equipped on the computer device 60.
[0120] Furthermore, the memory 62 may include both an internal storage unit of the computer device 60 and an external storage device. The memory 62 is used to store computer programs and other programs and data required by the computer device. The memory 62 may also be used to temporarily store data that has been output or is to be output.
[0121] Any reference to memory, database or other media used in the embodiments provided in the present application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.
[0122] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.
[0123] Fig.13 The present invention is a block diagram of an electronic device provided according to an embodiment.
[0124] See also Fig.13 The terminal device 600 is an electronic device, and the electronic device is in the form of a general computing device. The components of the electronic device may include but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different platform components (including the storage unit 620 and the processing unit 610), a display unit 640, etc.
[0125] The storage unit stores program codes, which can be executed by the processing unit 610, so that the processing unit 610 performs the steps according to various exemplary embodiments of the present invention described in the above method section of this specification. For example, the processing unit 610 can perform the following steps: Figure 1 Follow the steps shown in .
[0126] The storage unit 620 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 6201 and / or a cache memory unit 6202 , and may further include a read-only memory unit (ROM) 6203 .
[0127] The storage unit 620 may also include a program / utility 6204 having a set (at least one) of program modules 6205, such program modules 6205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.
[0128] Bus 630 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.
[0129] The electronic device 600 may also communicate with one or more external devices 700 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 600, and / or communicate with any device that enables the electronic device 600 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 650. Furthermore, the electronic device 600 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 660. The network adapter 660 may communicate with other modules of the electronic device 600 via a bus 630. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms, etc.
[0130] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention described and shown in the drawings here can usually be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.
[0131] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.
[0132] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.
[0133] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.
[0134] In the embodiments provided by the present invention, it should be understood that the disclosed devices / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.
[0135] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0136] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.
[0137] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.
[0138] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0139] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0140] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The above content is only for explaining the technical idea of the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution according to the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.
Claims
1. A method for identifying the migration of defective abrasive particles, characterized in that: The following steps are involved: Combined with the generation mechanism of wear particles, a comprehensive feature extractor of wear particles is established from two dimensions: spatial domain and frequency domain. Based on the wear particle comprehensive feature extractor, a parameter-adaptive Gaussian kernel function is designed for the multivariate features of wear particles, and an improved domain adaptation-driven transfer learning model for incomplete wear particle images is established. The transfer learning model of incomplete wear particle images is used to guide the wear particle comprehensive feature extractor to extract effective identification features of wear particle images. The spatial and channel attention mechanisms are combined to establish a multi-feature screening and fusion strategy for incomplete wear particles, and then a multi-channel feature intelligent identification model for incomplete wear particle types is constructed. The incomplete wear particles are identified using the incomplete wear particle type intelligent identification model.
2. The method for identifying the migration of defective abrasive particles according to claim 1, characterized in that: The steps for establishing the wear particle comprehensive feature extractor are as follows: A shape feature encoder is built based on the EfficientNet model, and convolution operations are designed for different shape feature layers of wear particles to extract three layers of wear particle shape features. An encoder with three convolutional layers is designed for wear particle texture features to extract single-layer wear particle texture features. Combine three layers of shape features and one layer of texture features, group the features along the channel dimension in multiple subspaces, and use three parallel sub-branches to periodically arrange the features for each pair of shape features and texture features that match the dimensions. The grouped convolution method is used to achieve feature fusion by adding the arranged features; Three trainable adaptive filters are designed for incomplete wear particle images, and the discrete cosine transform method is optimized in the frequency dimension to achieve the extraction of morphological features of wear particle images after filtering in different frequency domains. The three feature extraction branches of shape, texture and filtering are combined to complete the construction of the wear particle comprehensive feature extractor through feature splicing.
3. A method for identifying the migration of defective abrasive particles according to claim 2, characterized in that: The three feature extraction branches of shape, texture and filtering are: based on the three-layer wear shape features, a wear shape feature map is generated through a domain connection decoder; based on the single-layer wear texture features, a wear texture feature map is generated by bilinear interpolation; based on the image after frequency domain filtering, a wear frequency domain feature map is generated by discrete cosine inverse transform.
4. The method for identifying the migration of defective abrasive particles according to claim 1, characterized in that: The specific steps of designing a parameter-adaptive Gaussian kernel function for multivariate features of wear particles based on a comprehensive feature extractor of wear particles and establishing an improved domain adaptation-driven transfer learning model for incomplete wear particle images are as follows: Perform dimension reduction on the multivariate features extracted by the wear particle comprehensive feature extractor and calculate the joint distribution of the multivariate features of the wear particles; The domain adaptation method of joint maximum mean difference is used to calculate the average embedding distance between the feature distribution P of the complete wear particle image and the feature distribution Q of the incomplete wear particle image in Hilbert space, and the calculation result JMMD that measures their difference is obtained; The Gaussian kernel function is selected as the kernel function for calculating the joint maximum mean difference, and a parameter-trainable kernel function is set for each wear particle image feature map to achieve adaptive adjustment of the key parameters of the kernel function in JMMD. The feature extractor is guided to extract effective identification features of wear particle images by aligning feature distribution, and a transfer learning model for incomplete wear particle images is constructed.
5. The method for identifying the migration of defective abrasive particles according to claim 1, characterized in that: The specific process of combining the spatial and channel attention mechanisms to establish a multi-feature screening and fusion strategy for defective wear particles, and then constructing a multi-channel feature intelligent identification model for defective wear particle types is as follows: The attention mechanism is used to locate the significant feature areas of the wear particle image from the spatial dimension, and the multi-channel features of the wear particles are weighted from the channel dimension to achieve the screening of effective wear particle identification information; The features filtered by the attention mechanism are used as input, a classifier is designed for the multi-channel feature map of wear particles, and an intelligent identification model for the type of incomplete wear particles is established.
6. A method for identifying the migration of defective abrasive particles according to claim 1 or 5, characterized in that: The training set of the incomplete wear particle type intelligent identification model is wear particle images with different degrees of incompleteness generated based on complete wear particle image samples.
7. A method for identifying the migration of defective abrasive particles according to claim 1 or 5, characterized in that: The loss function of the incomplete wear particle type intelligent identification model is the weighted sum of the spatial domain feature map generation loss, the domain adaptation joint distribution loss and the wear particle image classification loss.
8. A defective abrasive particle type migration identification system, characterized in that: It includes a wear particle comprehensive feature extractor construction module, a defective wear particle image transfer learning model construction module and a defective wear particle identification module, wherein: Wear particle comprehensive feature extractor construction module: used to combine the generation mechanism of wear particles and establish a wear particle comprehensive feature extractor from two dimensions: spatial domain and frequency domain; The module for constructing the transfer learning model of incomplete wear particle images: Based on the wear particle comprehensive feature extractor, a parameter-adaptive Gaussian kernel function is designed for the multivariate features of wear particles, and an improved domain adaptation-driven transfer learning model of incomplete wear particle images is established; Incomplete wear particle identification module: It is used to guide the wear particle comprehensive feature extractor to extract effective identification features of wear particle images through the incomplete wear particle image transfer learning model, and combine the space and channel attention mechanism to establish a multi-feature screening and fusion strategy for incomplete wear particles, and then construct a multi-channel feature incomplete wear particle type intelligent identification model, and use the incomplete wear particle type intelligent identification model to realize the identification of incomplete wear particles.
9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.
