Metal coating crack detection model construction and crack detection method, device and equipment

Abstract

The invention relates to the technical field of electronics, and discloses a metal coating crack detection model construction method, a metal coating crack detection method, a metal coating crack detection device and metal coating crack detection equipment. The metal coating crack detection model construction method comprises the following steps: performing feature extraction on voiceprint data of a plurality of metal coatings in a preparation process to obtain feature vectors of the voiceprint data; and associating the feature vectors of the voiceprint data with the corresponding crack state data to obtain an associated data set, and training a preset model by using the associated data set to obtain a metal coating crack detection model. The metal coating crack detection model can output the crack state data of the corresponding metal coating according to the feature vector of the voiceprint data in the metal coating preparation process, so that crack detection in the metal coating preparation process is realized, and the problem that crack monitoring is difficult to carry out in the coating preparation process in the prior art is solved. And the crack monitoring result is not comprehensive enough.

Description

Technical Field

[0001] The present invention relates to the field of electronic technology, and in particular to metal coating crack detection model construction and crack detection methods, devices and equipment. Background Art

[0002] Metal coatings generally refer to coatings that are applied to the surface of components using metal materials as raw materials through thermal spraying, cold spraying, surfacing, laser cladding, and other methods. Most of them have the characteristics of wear resistance and corrosion resistance. The preparation of metal coatings on the surfaces of some key structural parts is mainly to give full play to the wear resistance and corrosion resistance of the metal coating, thereby protecting the integrity of key structural parts and extending the service life of the structural parts. However, during the preparation process of metal coatings, cracks often appear due to thermal stress, external forces, etc. Tiny cracks or cracks inside the coating are difficult to observe with the naked eye and cannot be repaired in time, thus affecting the subsequent wear resistance and corrosion resistance of the coating during use. Therefore, crack monitoring of the coating is of great significance in ensuring the effectiveness of the coating.

[0003] Related technologies typically use ultrasonic, electromagnetic, radiographic, magnetic powder, microwave, and millimeter wave methods to assess coating integrity without damaging the material. These methods only detect early cracks after the coating is prepared, based on changes in the metal's crystal structure and magnetic properties. This makes it difficult to monitor cracks during the coating preparation process, resulting in incomplete crack monitoring results. Summary of the Invention

[0004] In view of this, the present invention provides a metal coating crack detection model construction and crack detection method, device and equipment to solve the problem in the related art that early cracks are only detected based on changes in metal crystal structure and magnetism after the coating is prepared, and it is difficult to monitor cracks during the coating preparation process, resulting in incomplete crack monitoring results.

[0005] In a first aspect, the present invention provides a method for constructing a metal coating crack detection model, the method comprising: obtaining multiple voiceprint data and crack status data of each voiceprint data, the voiceprint data being used to characterize the voiceprint information of the metal coating during the preparation process; performing feature extraction on each voiceprint data to obtain a feature vector of the corresponding voiceprint data; associating the feature vector of each voiceprint data with the crack status data to obtain an associated data set; and using the associated data set to train a preset model until the accuracy of the preset model meets preset conditions, thereby obtaining a metal coating crack detection model.

[0006] The present invention provides a method for constructing a metal coating crack detection model. The method extracts features from the voiceprint data of multiple metal coatings during the preparation process to obtain a feature vector of each voiceprint data, associates the feature vector of each voiceprint data with the corresponding crack status data to obtain a correlated data set, and uses the correlated data set to train a preset model to obtain a metal coating crack detection model. The metal coating crack detection model can output the crack status data of the corresponding metal coating based on the feature vector of the voiceprint data during the preparation process of the metal coating, thereby realizing crack detection during the preparation process of the metal coating. This solves the problem in the related art that early cracks are detected based on changes in the metal crystal structure and magnetism after the coating is prepared, which makes it difficult to monitor cracks during the preparation process of the coating, resulting in incomplete crack monitoring results.

[0007] In an optional embodiment, the step of extracting features from each voiceprint data to obtain a feature vector of the corresponding voiceprint data includes: processing each voiceprint data using a preset modal decomposition algorithm to obtain n intrinsic mode function components of the corresponding voiceprint data; processing the n intrinsic mode function components of each voiceprint data using a continuous mean square error algorithm to obtain a demarcation point m of the corresponding voiceprint data; based on the demarcation point m of each voiceprint data, truncating the first m intrinsic mode function components of the n intrinsic mode function components of the corresponding voiceprint data to obtain a first signal of the corresponding voiceprint data; and extracting features from each voiceprint data. The first signal and the second signal are fused to obtain a reconstructed signal of the corresponding voiceprint data, where the second signal is used to represent the remaining nm intrinsic mode function components of the n intrinsic mode function components of the corresponding voiceprint data except the first m intrinsic mode function components; principal component analysis is performed on the reconstructed signal of each voiceprint data to obtain the number k of source signals of the corresponding voiceprint data; the first k-1 intrinsic mode function components in the reconstructed signal of each voiceprint data are synthesized with the reconstructed signal to obtain a k-dimensional observation signal of the corresponding voiceprint data; blind separation processing is performed on the k-dimensional observation signal of each voiceprint data to obtain a eigenvector of the corresponding voiceprint data.

[0008] The method provided in this optional embodiment obtains a feature vector suitable for voiceprint data under complex interference signal conditions, which facilitates the subsequent construction of a metal coating crack detection model.

[0009] In an optional embodiment, based on the dividing point m of each voiceprint data, the first m intrinsic mode function components of the n intrinsic mode function components of the corresponding voiceprint data are truncated to obtain the first signal of the corresponding voiceprint data. The method includes: obtaining a preset frequency threshold; eliminating part of the signals whose frequency is greater than the preset frequency threshold in the first m intrinsic mode function components of each voiceprint data to obtain the first signal of the corresponding voiceprint data.

[0010] In an optional embodiment, the k-dimensional observation signal of each voiceprint data is blindly separated to obtain a feature vector of the corresponding voiceprint data, including: using an independent component analysis algorithm to blindly separate the k-dimensional observation signal of each voiceprint signal to obtain a feature vector of the corresponding voiceprint data.

[0011] In a second aspect, the present invention provides a method for detecting cracks in a metal coating, the method comprising: obtaining target voiceprint data during the metal coating welding process; performing feature extraction on the target voiceprint data to obtain a target feature vector; inputting the target feature vector into a pre-constructed metal coating crack detection model so that the metal coating crack detection model outputs target crack status data, and the metal coating crack detection model is constructed by the metal coating crack detection model construction method of the first aspect or any corresponding embodiment thereof.

[0012] The metal coating crack detection method provided by the present invention inputs the target feature vector of the target voiceprint data into the metal coating crack detection model, so that the model outputs the crack status data of the corresponding metal coating according to the target feature vector, thereby realizing crack detection during the preparation process of the metal coating, solving the problem in the related technology that early cracks are detected based on changes in the metal crystal structure and magnetism after the coating is prepared, making it difficult to monitor cracks during the preparation process of the coating, resulting in incomplete crack monitoring results.

[0013] In the third aspect, the present invention provides a device for constructing a metal coating crack detection model, which includes: a first acquisition module, used to acquire multiple voiceprint data and crack status data of each voiceprint data, the voiceprint data is used to characterize the voiceprint information of the metal coating during the preparation process; a first extraction module, used to extract features from each voiceprint data to obtain a feature vector of the corresponding voiceprint data; an association module, used to associate the feature vector of each voiceprint data with the crack status data to obtain an associated data set; a training module, used to train a preset model using the associated data set until the accuracy of the preset model meets the preset conditions, thereby obtaining a metal coating crack detection model.

[0014] In a fourth aspect, the present invention provides a metal coating crack detection device, which includes: a second acquisition module for acquiring target voiceprint data during the metal coating welding process; a second extraction module for performing feature extraction on the target voiceprint data to obtain a target feature vector; a detection module for inputting the target feature vector into a pre-constructed metal coating crack detection model so that the metal coating crack detection model outputs target crack status data, and the metal coating crack detection model is constructed by the metal coating crack detection model construction method of the above-mentioned first aspect or any corresponding embodiment thereof.

[0015] In a fifth aspect, the present invention provides a computer device comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, computer instructions being stored in the memory, and the processor executing the computer instructions to thereby execute the metal coating crack detection model construction method of the above-mentioned first aspect or any corresponding embodiment thereof, or execute the metal coating crack detection method of the above-mentioned second aspect.

[0016] In a sixth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the metal coating crack detection model construction method of the above-mentioned first aspect or any corresponding embodiment thereof, or to execute the metal coating crack detection method of the above-mentioned second aspect.

[0017] In the seventh aspect, the present invention provides a computer program product comprising computer instructions, the computer instructions being used to enable a computer to execute the metal coating crack detection model construction method of the above-mentioned first aspect or any corresponding embodiment thereof, or to execute the metal coating crack detection method of the above-mentioned second aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 1 is a flow chart of a method for constructing a metal coating crack detection model according to an embodiment of the present invention;

[0020] Figure 2 is a flow chart of another method for constructing a metal coating crack detection model according to an embodiment of the present invention;

[0021] Figure 3 1 is a schematic flow chart of a metal coating crack detection method according to an embodiment of the present invention;

[0022] Figure 4 is a structural block diagram of a device for constructing a metal coating crack detection model according to an embodiment of the present invention;

[0023] Figure 5 is a structural block diagram of a metal coating crack detection device according to an embodiment of the present invention;

[0024] Figure 6 Schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0025] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings 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 those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.

[0026] For metal coatings, crack monitoring is also necessary during the fabrication process. For example, cracking in metal coatings fabricated by overlay welding can be caused by a variety of factors, including material properties, welding process parameters, and operator technique. First, from a material perspective, overlay welding typically involves two or more metals with different properties, such as thermal expansion coefficient, thermal conductivity, and melting point. During the overlay welding process, these differing thermal behaviors can easily lead to stress concentration between the overlay layer and the base material, which can cause cracks. Second, the selection of welding process parameters has a significant impact on crack initiation. For example, excessive welding current can lead to excessive heat input, causing overheating and coarsening of the overlay layer, reducing its crack resistance. Excessive welding speeds can lead to rapid cooling of the weld pool, generating residual stresses and increasing crack susceptibility. Furthermore, process measures such as welding sequence, preheat temperature, and post-heat treatment can also influence cracking. Finally, welding technique is also a significant factor in crack initiation. During the welding process, the operator's proficiency, stability, and the quality of the weld joint assembly will have a direct impact on the welding quality. If the weld joint is poorly assembled, with defects such as gaps or misalignments, welding stress concentration is likely to occur, which in turn causes cracks.

[0027] Related technologies typically use ultrasonic, electromagnetic, radiographic, magnetic powder, microwave, and millimeter wave methods to assess coating integrity without damaging the material. These methods only detect early cracks after the coating is prepared, based on changes in the metal's crystal structure and magnetic properties. This makes it difficult to monitor cracks during the coating preparation process, resulting in incomplete crack monitoring results.

[0028] In view of this, a method for constructing a metal coating crack detection model provided in an embodiment of the present application can be applied to a server to realize the construction of a metal coating crack detection model. The method provided in an embodiment of the present application performs feature extraction on the voiceprint data of multiple metal coatings during the preparation process to obtain a feature vector of each voiceprint data, associates the feature vector of each voiceprint data with the corresponding crack status data to obtain a related data set, and uses the related data set to train a preset model to obtain a metal coating crack detection model. The metal coating crack detection model can output the crack status data of the corresponding metal coating according to the feature vector of the voiceprint data during the preparation process of the metal coating, thereby realizing crack detection during the preparation process of the metal coating, solving the problem in the related art that early cracks are detected based on changes in the metal crystal structure and magnetism after the coating is prepared, making it difficult to monitor cracks during the preparation process of the coating, resulting in incomplete crack monitoring results.

[0029] According to an embodiment of the present invention, an embodiment of a method for constructing a metal coating crack detection model is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0030] In this embodiment, a method for constructing a metal coating crack detection model is provided, which can be used in the above-mentioned server. Figure 1 FIG. 1 is a flow chart of a method for constructing a metal coating crack detection model according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0031] Step S101 , obtaining a plurality of voiceprint data and crack status data of each voiceprint data, wherein the voiceprint data is used to characterize the voiceprint information of the metal coating during the preparation process.

[0032] Exemplarily, voiceprint data is collected by a voiceprint signal acquisition system. The hardware of the voiceprint signal acquisition system mainly includes acoustic emission sensors, preamplifiers, acoustic emission cards, and computers for data storage and processing. The acoustic emission sensors can be set in locations including, but not limited to, the heat source of the surfacing welding and the substrate to which the metal coating is to be welded. Crack status data is used to characterize the severity of cracks in the metal coating. The specific content of the crack status data is not limited in the embodiments of this application, and those skilled in the art can determine it according to needs. In the embodiments of this application, multiple voiceprint data and the crack status data of each voiceprint data can be stored in a pre-built voiceprint signal data center. It mainly collects fatigue, wear, vibration, impact voiceprint signals of different materials and various environmental noise signals, and uniformly manages the voiceprint data collected in actual working conditions, uploads them to the cloud, and continuously updates them. This is used as a sample set for voiceprint signal denoising and blind separation and pattern recognition intelligent algorithms. By continuously training the intelligent algorithms, crack extension voiceprint signals can be better identified. At the same time, a metal material crack voiceprint data center is formed to fill the gap in the industry.

[0033] Step S102: extract features from each voiceprint data to obtain feature vectors corresponding to the voiceprint data.

[0034] For example, in an embodiment of the present application, feature extraction of voiceprint data can be achieved by blind separation of fatigue crack voiceprint signals under complex signal sources. The embodiment of the present application does not limit the specific feature extraction method, and those skilled in the art can determine it according to needs.

[0035] Step S103: Associating the feature vectors of each voiceprint data with the crack state data to obtain an associated data set.

[0036] For example, in the embodiments of the present application, there is no limitation on the manner of associating the feature vector with the crack state data, as long as the association can be achieved.

[0037] Step S104: training the preset model using the associated data set until the accuracy of the preset model meets the preset conditions, thereby obtaining a metal coating crack detection model.

[0038] For example, the embodiments of the present application do not limit the specific content of the preset conditions, and those skilled in the art may determine them according to their needs. The preset model may include but is not limited to a least squares support vector machine (Particle Swarm Optimization-Least Squares Support Vector Machine, PSO-LS-SVM) model.

[0039] The method for constructing a metal coating crack detection model provided in this embodiment performs feature extraction on the voiceprint data of multiple metal coatings during the preparation process to obtain a feature vector of each voiceprint data, associates the feature vector of each voiceprint data with the corresponding crack status data to obtain a correlated data set, and uses the correlated data set to train a preset model to obtain a metal coating crack detection model. The metal coating crack detection model can output the crack status data of the corresponding metal coating based on the feature vector of the voiceprint data during the preparation process of the metal coating, thereby realizing crack detection during the preparation process of the metal coating, solving the problem in the related art that early cracks are detected based on changes in the metal crystal structure and magnetism after the coating is prepared, making it difficult to monitor cracks during the preparation process of the coating, resulting in incomplete crack monitoring results.

[0040] In this embodiment, a method for constructing a metal coating crack detection model is provided, which can be used in the above-mentioned server. Figure 2 FIG. 1 is a flow chart of a method for constructing a metal coating crack detection model according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps:

[0041] Step S201: Acquire multiple voiceprint data and crack status data of each voiceprint data. The voiceprint data is used to characterize the voiceprint information of the metal coating during the preparation process. Figure 1 Step S101 of the illustrated embodiment will not be described in detail here.

[0042] Step S202: extract features from each voiceprint data to obtain feature vectors corresponding to the voiceprint data.

[0043] Specifically, the above step S202 includes:

[0044] Step S2021: Process each voiceprint data using a preset modal decomposition algorithm to obtain n intrinsic mode function components corresponding to the voiceprint data.

[0045] For example, in embodiments of the present application, the preset modal decomposition algorithm may include, but is not limited to, Ensemble Derivative-Enhanced Empirical Mode Decomposition (EDEMD). EDEMD is an improved EMD algorithm specifically designed for processing nonlinear and non-stationary signals. By introducing ensemble mean and derivative information, it addresses problems such as modal mixing and endpoint effects present in traditional EMD methods. EDEMD decomposition of voiceprint data is performed to obtain its n intrinsic mode function (IMF) components.

[0046] Step S2022: Process the n intrinsic mode function components of each voiceprint data using a continuous mean square error algorithm to obtain a demarcation point m of the corresponding voiceprint data.

[0047] For example, in the embodiment of the present application, the dividing point between the noise-dominated component and the signal-dominated component is found by calculating the continuous mean square error (MSE) of adjacent IMF components. i and the IMF i+1 Calculate the continuous mean square error MSE i , where i is equal to 1 to n-1, based on the calculated MSE i Construct the MSE curve. If the MSE curve is at point MSE i The sudden increase at i and the IMF i+1 If the properties of the two are significantly different, the first significantly increased MSE i i is used as the dividing point m.

[0048] Step S2023: Based on the demarcation point m of each voiceprint data, the first m intrinsic mode function components of the n intrinsic mode function components corresponding to the voiceprint data are truncated to obtain a first signal corresponding to the voiceprint data.

[0049] For example, in the embodiment of the present application, the first m intrinsic mode function components of each voiceprint data may be removed from the parts with higher frequencies according to a preset rule to obtain a first signal.

[0050] In some optional implementations, step S2023 includes:

[0051] Step a1: Obtain a preset frequency threshold.

[0052] Exemplarily, the preset frequency threshold may be determined based on experience. In the embodiment of the present application, the preset frequency threshold may be determined by the following formula:

[0053]

[0054] Among them, μ represents the preset frequency threshold; N represents the sampling length, which is set based on experience. Here, the sampling length is selected as 1000; σ represents the noise standard deviation.

[0055] Step a2: remove the signals whose frequencies are greater than a preset frequency threshold from the first m intrinsic mode function components of each voiceprint data to obtain the first signal corresponding to the voiceprint data.

[0056] For example, in an embodiment of the present application, for any voiceprint data, among the first m IMF components, the part of each IMF component that exceeds a preset frequency threshold is removed to obtain the corresponding processed IMF component, and the different processed IMF components constitute the first signal of the corresponding voiceprint data.

[0057] Step S2024: fuse the first signal and the second signal of each voiceprint data to obtain a reconstructed signal of the corresponding voiceprint data, wherein the second signal is used to represent the remaining nm intrinsic mode function components of the n intrinsic mode function components of the corresponding voiceprint data except the first m intrinsic mode function components.

[0058] Exemplarily, in the embodiment of the present application, the first signal and the second signal of each voiceprint data may be summed to obtain a reconstructed signal of the corresponding voiceprint data.

[0059] Step S2025: Perform principal component analysis on the reconstructed signal of each voiceprint data to obtain the number k of source signals corresponding to the voiceprint data.

[0060] For example, in the embodiment of the present application, the purpose of performing principal component analysis (PCA) on the reconstructed signal of each voiceprint data is to reduce dimensionality. PCA is a commonly used multivariate statistical analysis method, which is used to reduce the dimensionality of high-dimensional data and extract low-dimensional features (principal components) that best represent the original data information. Its core idea is to find a set of new orthogonal coordinate axes (principal components) so that the projection variance of the data on these axes is maximized. The first principal component corresponds to the direction with the largest data variance, the second principal component is orthogonal to the first principal component and has the second largest variance, and so on. Usually, the first k principal components are retained and the remaining components are discarded to achieve dimensionality reduction. Among them, the number of source signals k is determined based on the first k principal components retained.

[0061] Step S2026, synthesizing the first k-1 intrinsic mode function components in the reconstructed signal of each voiceprint data with the reconstructed signal to obtain a k-dimensional observation signal of the corresponding voiceprint data;

[0062] Exemplarily, in an embodiment of the present application, the first k-1 intrinsic mode function components in the reconstructed signal of each voiceprint data are superimposed on the reconstructed signal to obtain a k-dimensional observation signal, that is, the dimension of the synthetic signal is equal to the number of sources.

[0063] Step S2027: perform blind separation processing on the k-dimensional observation signal of each voiceprint data to obtain a feature vector corresponding to the voiceprint data.

[0064] In some optional implementations, the above step S2027 includes:

[0065] Step b1: Use the independent component analysis algorithm to perform blind separation processing on the k-dimensional observation signal of each voiceprint signal to obtain the feature vector of the corresponding voiceprint data.

[0066] For example, in the embodiment of the present application, the independent component analysis algorithm (ICA) is used to perform blind separation processing on the k-dimensional observation signal of each voiceprint signal to obtain the feature vector of each voiceprint data. ICA analysis is used to decouple the measurement signal and obtain an approximate estimate of the source signal s(t) (voiceprint data) from the k-dimensional observation signal x(t). That is, the output feature vector can be described as:

[0067]

[0068] Where A is the signal mixing matrix; W is the separation matrix; It is the independent signal vector corresponding to the source signal obtained by separation, which is the approximate approximation of s(t).

[0069] Step S203: Associate the feature vectors of each voiceprint data with the crack state data to obtain an associated data set. Figure 1 Step S103 of the illustrated embodiment will not be described in detail here.

[0070] Step S204: Use the associated data set to train the preset model until the accuracy of the preset model meets the preset conditions, thereby obtaining a metal coating crack detection model. Figure 1 Step S104 of the illustrated embodiment will not be described in detail here.

[0071] In this embodiment, a metal coating crack detection method is also provided, which can be used in a server to realize crack detection of metal coating. Figure 3 Flowchart of a metal coating crack detection method according to an embodiment of the present invention. Figure 3 As shown, the process includes the following steps:

[0072] Step S301: Acquire target voiceprint data during the metal coating welding process.

[0073] Exemplarily, the target voiceprint data is voiceprint data of a metal coating that needs to be crack detected during the preparation process.

[0074] Step S302: extract features from the target voiceprint data to obtain a target feature vector.

[0075] For example, the specific feature extraction process is described in the above embodiment and will not be repeated here.

[0076] In step S303, the target feature vector is input into a pre-built metal coating crack detection model so that the metal coating crack detection model outputs target crack state data. The metal coating crack detection model is constructed by the metal coating crack detection model construction method in the above embodiment.

[0077] Exemplarily, inputting the target feature vector into a pre-built metal coating crack detection model will output target crack state data, which can represent crack information existing in the corresponding metal coating during the preparation process.

[0078] The metal coating crack detection method provided in this embodiment inputs the target feature vector of the target voiceprint data into the metal coating crack detection model, so that the model outputs the crack status data of the corresponding metal coating according to the target feature vector, thereby realizing crack detection during the preparation process of the metal coating. It solves the problem in the related art that early cracks are detected based on changes in the metal crystal structure and magnetism after the coating is prepared, which makes it difficult to monitor cracks during the preparation process of the coating, resulting in incomplete crack monitoring results.

[0079] In the embodiments of the present application, the metal coating crack detection method can be applied to a voiceprint monitoring device. This device is mainly used to monitor the initiation and expansion of cracks in the metal coating during the actual preparation process. A browser / server architecture (B / S) can be used, and it has functions such as voiceprint status monitoring, voiceprint map viewing, autonomous learning, abnormal alarm sending and viewing, voiceprint data annotation, voiceprint database query, and data upload. The main purpose is to integrate the various hardware and algorithms involved in the present invention together, combine them with cloud technology, and form a portable voiceprint monitoring device for further promotion.

[0080] This embodiment also provides a device for constructing a metal coating crack detection model. This device is used to implement the above-mentioned embodiments and preferred embodiments. Details already described will not be repeated here. As used below, the term "module" may refer to a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation using hardware, or a combination of software and hardware, is also possible and contemplated.

[0081] This embodiment provides a device for constructing a metal coating crack detection model. Figure 4 Shown, including:

[0082] A first acquisition module 401 is used to acquire a plurality of voiceprint data and crack status data of each voiceprint data, wherein the voiceprint data is used to represent the voiceprint information of the metal coating during the preparation process;

[0083] The first extraction module 402 is used to extract features from each voiceprint data to obtain a feature vector corresponding to the voiceprint data;

[0084] The association module 403 is used to associate the feature vector of each voiceprint data with the crack state data to obtain an associated data set;

[0085] The training module 404 is used to train the preset model using the associated data set until the accuracy of the preset model meets the preset conditions, thereby obtaining a metal coating crack detection model.

[0086] In some optional implementations, the first extraction module 402 includes:

[0087] The first processing submodule is used to process each voiceprint data using a preset modal decomposition algorithm to obtain n intrinsic mode function components corresponding to the voiceprint data;

[0088] The calculation submodule is used to process the n intrinsic mode function components of each voiceprint data using the continuous mean square error algorithm to obtain the demarcation point m of the corresponding voiceprint data;

[0089] The second processing submodule is configured to truncate the first m intrinsic mode function components of the n intrinsic mode function components of the corresponding voiceprint data based on the demarcation point m of each voiceprint data to obtain a first signal of the corresponding voiceprint data;

[0090] A fusion submodule is configured to fuse the first signal and the second signal of each voiceprint data to obtain a reconstructed signal of the corresponding voiceprint data, wherein the second signal is used to represent the remaining nm intrinsic mode function components of the n intrinsic mode function components of the corresponding voiceprint data, excluding the first m intrinsic mode function components;

[0091] The analysis submodule is used to perform principal component analysis on the reconstructed signal of each voiceprint data to obtain the number k of source signals corresponding to the voiceprint data;

[0092] A synthesis submodule, configured to synthesize the first k-1 intrinsic mode function components in the reconstructed signal of each voiceprint data with the reconstructed signal to form a k-dimensional observation signal of the corresponding voiceprint data;

[0093] The third processing submodule is used to perform blind separation processing on the k-dimensional observation signal of each voiceprint data to obtain a feature vector corresponding to the voiceprint data.

[0094] In some optional implementations, the second processing submodule includes:

[0095] An acquiring unit, configured to acquire a preset frequency threshold;

[0096] The elimination unit is used to eliminate the signals whose frequencies are greater than a preset frequency threshold in the first m intrinsic mode function components of each voiceprint data to obtain the first signal corresponding to the voiceprint data.

[0097] In an optional embodiment, the third processing submodule includes:

[0098] The processing unit is used to perform blind separation processing on the k-dimensional observation signal of each voiceprint signal using an independent component analysis algorithm to obtain a feature vector of the corresponding voiceprint data.

[0099] This embodiment provides a metal coating crack detection device, such as Figure 5 As shown, including:

[0100] The second acquisition module 501 is used to acquire target voiceprint data during the metal coating welding process;

[0101] The second extraction module 502 is used to extract features from the target voiceprint data to obtain a target feature vector;

[0102] The detection module 503 is used to input the target feature vector into a pre-built metal coating crack detection model so that the metal coating crack detection model outputs target crack state data. The metal coating crack detection model is constructed using the metal coating crack detection model construction method of the above embodiment.

[0103] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0104] The metal coating crack detection model construction device and the metal coating crack detection device in this embodiment are presented in the form of functional units, where the units refer to ASIC (Application Specific Integrated Circuit) circuits, processors and memories that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0105] The embodiment of the present invention also provides a computer device having the above Figure 4 The metal coating crack detection model building device shown, or having the above Figure 5 The metal coating crack detection device shown.

[0106] See also Figure 6 , Figure 6 is a structural diagram of a computer device provided by an optional embodiment of the present invention, such as Figure 6As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components utilize different buses to communicate with each other and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of a GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories. Equally, multiple computer devices can be connected, and each device provides part of the necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 6 A processor 10 is taken as an example.

[0107] The processor 10 may be a central processing unit, a network processor, or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic, or any combination thereof.

[0108] The memory 20 stores instructions that can be executed by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0109] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created based on the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0110] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0111] The computer device further includes a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0112] The embodiment of the present invention also provides a computer-readable storage medium. The above-mentioned method according to the embodiment of the present invention can be implemented in hardware, firmware, or implemented as a computer code that can be recorded in a storage medium, or implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that a computer, a processor, a microprocessor controller or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor or hardware, the method shown in the above embodiment is implemented.

[0113] A portion of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the form in which the computer program instruction exists in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc. Accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium that can be accessed by the computer.

[0114] Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention. Such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for constructing a metal coating crack detection model, characterized in that: The method comprises: Acquire multiple voiceprint data and crack status data of each voiceprint data, where the voiceprint data is used to characterize the voiceprint information of the metal coating during the preparation process; Perform feature extraction on each voiceprint data to obtain the feature vector of the corresponding voiceprint data; Associating the feature vectors of each voiceprint data with the crack state data to obtain an associated data set; The preset model is trained using the associated data set until the accuracy of the preset model meets the preset conditions, thereby obtaining a metal coating crack detection model.

2. The method according to claim 1, characterized in that The step of extracting features from each voiceprint data to obtain a feature vector corresponding to the voiceprint data includes: Using a preset modal decomposition algorithm to process each voiceprint data, n intrinsic mode function components corresponding to the voiceprint data are obtained; The continuous mean square error algorithm is used to process the n intrinsic mode function components of each voiceprint data to obtain the demarcation point m of the corresponding voiceprint data; Based on the demarcation point m of each voiceprint data, truncating the first m intrinsic mode function components of the n intrinsic mode function components of the corresponding voiceprint data to obtain a first signal of the corresponding voiceprint data; fusing the first signal and the second signal of each voiceprint data to obtain a reconstructed signal of the corresponding voiceprint data, wherein the second signal is used to represent the remaining nm intrinsic mode function components of the n intrinsic mode function components of the corresponding voiceprint data, excluding the first m intrinsic mode function components; Perform principal component analysis on the reconstructed signal of each voiceprint data to obtain the number k of source signals corresponding to the voiceprint data; synthesizing the first k-1 intrinsic mode function components in the reconstructed signal of each voiceprint data with the reconstructed signal to obtain a k-dimensional observation signal of the corresponding voiceprint data; Blind separation processing is performed on the k-dimensional observation signal of each voiceprint data to obtain the feature vector corresponding to the voiceprint data.

3. The method according to claim 2, characterized in that Based on the demarcation point m of each voiceprint data, the first m intrinsic mode function components of the n intrinsic mode function components of the corresponding voiceprint data are truncated to obtain a first signal of the corresponding voiceprint data, the method comprising: Get the preset frequency threshold; Part of the signals whose frequencies are greater than the preset frequency threshold in the first m intrinsic mode function components of each voiceprint data are removed to obtain the first signal corresponding to the voiceprint data.

4. The method according to claim 2, characterized in that The blind separation process is performed on the k-dimensional observation signal of each voiceprint data to obtain the feature vector corresponding to the voiceprint data, including: An independent component analysis algorithm is used to perform blind separation processing on the k-dimensional observation signal of each voiceprint signal to obtain the feature vector corresponding to the voiceprint data.

5. A method for detecting cracks in metal coatings, characterized in that: The method comprises: Acquire target voiceprint data during metal coating welding; Performing feature extraction on the target voiceprint data to obtain a target feature vector; The target feature vector is input into a pre-constructed metal coating crack detection model so that the metal coating crack detection model outputs target crack state data, and the metal coating crack detection model is constructed by the metal coating crack detection model construction method according to any one of claims 1 to 4.

6. A device for constructing a metal coating crack detection model, characterized in that: The device comprises: A first acquisition module is used to acquire a plurality of voiceprint data and crack status data of each voiceprint data, wherein the voiceprint data is used to characterize the voiceprint information of the metal coating during the preparation process; The first extraction module is used to extract features from each voiceprint data to obtain a feature vector corresponding to the voiceprint data; An association module is used to associate the feature vectors of each voiceprint data with the crack state data to obtain an associated data set; The training module is used to train the preset model using the associated data set until the accuracy of the preset model meets the preset conditions, thereby obtaining a metal coating crack detection model.

7. A metal coating crack detection device, characterized in that: The device comprises: The second acquisition module is used to obtain the target voiceprint data during the metal coating welding process; A second extraction module is used to extract features from the target voiceprint data to obtain a target feature vector; A detection module is used to input the target feature vector into a pre-constructed metal coating crack detection model so that the metal coating crack detection model outputs target crack state data, and the metal coating crack detection model is constructed by the metal coating crack detection model construction method according to any one of claims 1 to 4.

8. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the metal coating crack detection model construction method according to any one of claims 1 to 4, or the metal coating crack detection method according to claim 5, by executing the computer instructions.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, which are used to enable a computer to execute the metal coating crack detection model construction method described in any one of claims 1 to 4, or the metal coating crack detection method described in claim 5.

10. A computer program product, characterized in that The method comprises computer instructions for causing a computer to execute the method for constructing a metal coating crack detection model according to any one of claims 1 to 4, or the method for detecting cracks in a metal coating according to claim 5.

Images