LDED dilution rate monitoring method based on Markov transfer field and related equipment

Abstract

The invention belongs to the technical field of additive manufacturing, and discloses an LDED dilution rate monitoring method based on a Markov transfer field and related equipment.The method comprises the steps that one-dimensional spectral signals of excited atoms in a molten pool are obtained in real time; preprocessing the one-dimensional spectral signal to obtain a preprocessed one-dimensional spectral signal; converting the preprocessed one-dimensional spectral signal into a two-dimensional feature map by adopting a Markov transfer field method; inputting the two-dimensional feature map into a trained ResNet18 model, and outputting a prediction range of the dilution rate; the training data of the ResNet18 model comprises a two-dimensional feature map with different dilution rate range labels. The objective of the invention is to solve the limitation of a one-dimensional spectral signal analysis method in processing a complex data structure and realize real-time monitoring and accurate classification of the dilution rate.

Description

Technical Field

[0001] The present invention belongs to the technical field of additive manufacturing, and in particular relates to a LDED dilution rate monitoring method based on Markov transfer field and related equipment. Background Art

[0002] Additive Manufacturing (AM), as an important part of advanced manufacturing technology, integrates multiple fields such as metallurgy, robotics, heat transfer, and computer science, and achieves the goal of preparing various complex parts based on the principles of discreteness and accumulation. Among them, additive manufacturing technology using metal powder as raw material, especially laser directed energy deposition (LDED) technology, has shown broad application prospects in aerospace, automobile manufacturing, biomedicine and other fields due to its advantages of high precision, high efficiency and high material utilization.

[0003] In the fields of aerospace, medical equipment, etc., the quality requirements for parts are extremely high. In the LDED process, the dilution rate is an important indicator for evaluating the quality of laser cladding forming and the bonding strength of the formed parts. It describes the proportion of substrate metal integrated into the cladding layer, and directly reflects the degree of mixing of substrate metal and cladding powder material in the cladding layer. Traditional dilution rate monitoring methods often rely on empirical formulas or offline detection methods. These methods are not only time-consuming and labor-intensive, but also difficult to achieve real-time monitoring and precise control of the dilution rate.

[0004] In order to achieve real-time monitoring of the dilution rate, researchers began to explore methods to analyze the dilution rate by extracting the spectral signal characteristics inside the molten pool. The spectral signal inside the molten pool contains rich physical and chemical information, which is closely related to the dilution rate. However, the traditional one-dimensional spectral signal analysis method has the limitations of information loss, unclear features or difficult interpretation when processing complex data structures, which makes it difficult to meet the needs of real-time monitoring of the dilution rate in the LDED process. Summary of the invention

[0005] In view of the problems existing in the prior art, the present invention provides a LDED dilution rate monitoring method and related equipment based on Markov transition field, which aims to solve the limitations of one-dimensional spectral signal analysis method when processing complex data structures and realize real-time monitoring and accurate classification of dilution rate.

[0006] In order to solve the above technical problems, the present invention is implemented by the following technical solutions:

[0007] According to a first aspect of the present invention, there is provided a method for monitoring LDED dilution rate based on a Markov transition field, comprising:

[0008] Obtain the one-dimensional spectrum signal of excited atoms in the molten pool in real time;

[0009] Preprocessing the one-dimensional spectral signal to obtain a preprocessed one-dimensional spectral signal;

[0010] Using the Markov transfer field method, the preprocessed one-dimensional spectral signal is converted into a two-dimensional feature map;

[0011] The two-dimensional feature map is input into a trained ResNet18 model, and a predicted range of dilution rate is output; the training data of the ResNet18 model includes two-dimensional feature maps with different dilution rate range labels.

[0012] In a possible implementation manner of the first aspect, a method for acquiring a one-dimensional spectrum signal of excited-state atoms in the molten pool is:

[0013] A photomultiplier tube or a spectrometer is used to collect the one-dimensional spectrum signal of excited atoms in the molten pool.

[0014] In a possible implementation manner of the first aspect, preprocessing the one-dimensional spectral signal includes:

[0015] The one-dimensional spectral signal is divided into several intervals according to the set length, and several segments of fine-grained one-dimensional spectral signals are obtained, which are expressed as:

[0016] Xi(k)={x t ∣t∈[(k-1)·m+1,k·m]},k=1,2,…,n;

[0017] Where Xi(k) is the kth segment fine-grained one-dimensional spectral signal; x t is the spectral intensity of the one-dimensional spectral signal; m is the length of each interval; n is the number of segments that divide the interval.

[0018] In a possible implementation manner of the first aspect, preprocessing the one-dimensional spectral signal includes:

[0019] The one-dimensional spectral signal is averaged to obtain a coarse-grained one-dimensional spectral signal, which is expressed as:

[0020]

[0021] In the formula, y k is a coarse-grained one-dimensional spectral signal; N is the scale factor, and the length of each coarse-grained time series is equal to the length of the time series sample divided by the scale factor N; x t is the spectral intensity of the one-dimensional spectral signal; L is the total length of the signal; k is the number of signal segments.

[0022] In a possible implementation manner of the first aspect, the Markov transition field method is used to convert the preprocessed one-dimensional spectral signal into a two-dimensional feature map, specifically:

[0023] The preprocessed one-dimensional spectral signal is divided into a plurality of subsequences according to a preset window length;

[0024] Discretize the divided subsequences so that each point in the one-dimensional spectral signal is mapped into a corresponding value range;

[0025] Calculate the transition probability for each point after mapping to obtain a state transition probability matrix;

[0026] The state transition probability matrix is ​​visualized through a heat map to obtain the MTF image.

[0027] In a possible implementation manner of the first aspect, the ResNet18 model includes an initial convolutional layer, a residual block, a global average pooling layer, and a fully connected layer;

[0028] The initial convolution layer uses a 7×7 convolution kernel, a step size of 2, a padding of 3, and an output channel of 64, and outputs the original image as a feature map of size 56×56×64;

[0029] The residual block includes 4 residual block sequences. The image processed by the initial convolution layer is input. Each sequence contains a different number of residual blocks. Each residual block has two 3x3 convolution layers with channel numbers of 64, 128, 256 and 512 respectively. The last residual block outputs a 7×7×512 feature map.

[0030] The feature map output by the last residual block is input into the global average pooling layer, which takes the average of the feature map and inputs the 512-dimensional feature vector output by the global average pooling layer into the fully connected layer, where the Softmax function outputs the dilution rate range.

[0031] According to a second aspect of the present invention, there is provided a LDED dilution rate monitoring device based on a Markov transition field, comprising:

[0032] An acquisition module is used to acquire one-dimensional spectrum signals of excited atoms in the molten pool in real time;

[0033] A preprocessing module, used for preprocessing the one-dimensional spectral signal to obtain a preprocessed one-dimensional spectral signal;

[0034] A conversion module, used for converting the preprocessed one-dimensional spectral signal into a two-dimensional feature map by using a Markov transfer field method;

[0035] The monitoring module is used to input the two-dimensional feature map into a trained ResNet18 model and output a predicted range of dilution rate; the training data of the ResNet18 model includes two-dimensional feature maps with different dilution rate range labels.

[0036] According to a third aspect of the present invention, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the LDED dilution rate monitoring method based on the Markov transition field is implemented.

[0037] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the LDED dilution rate monitoring method based on a Markov transition field is implemented.

[0038] According to a fifth aspect of the present invention, a computer program product is provided, which, when executed by a processor, implements the LDED dilution rate monitoring method based on Markov transition field.

[0039] Compared with the prior art, the present invention has at least the following beneficial effects:

[0040] The present invention provides a LDED dilution rate monitoring method based on Markov transfer field. Traditional spectral monitoring methods rely on empirical formulas or offline detection, which is difficult to achieve real-time monitoring and has limited accuracy. The present invention obtains the one-dimensional spectral signal of excited atoms in the molten pool in real time, and combines the Markov transfer field (MTF) method to convert it into a two-dimensional feature map, thereby capturing the dynamics of the molten pool in real time and improving the real-time performance and prediction accuracy of monitoring. At the same time, the MTF method can retain the spatiotemporal characteristics of the signal, solving the problems of information loss and difficulty in interpretation when processing complex data in traditional methods. In addition, the method of the present invention combines the trained and optimized ResNet18 model to achieve automatic classification and prediction of the dilution rate, improve the intelligent level of monitoring, and extract multi-scale spectral features through fine-grained and coarse-grained preprocessing, thereby enhancing the feature characterization capability. Compared with the problem of poor adaptability of traditional methods, the present invention can adapt to different dilution rate ranges, process parameters and material properties, and has strong generalization ability. At the same time, the two-dimensional feature map generated by the MTF method is intuitive and visualized, which is convenient for operators to understand and analyze the molten pool state. Through high-precision real-time monitoring, the present invention can effectively reduce material waste, reduce production costs, and improve production efficiency.

[0041] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the specific implementation modes of the present invention, the drawings required for use in the description of the specific implementation modes will be briefly introduced below. Obviously, the drawings described below are some implementation modes of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0043] Figure 1 The present invention is a flow chart of a method for monitoring LDED dilution rate based on Markov transition field.

[0044] Figure 2 A flow chart of a spectral signal passing through MTF and deriving a thermal map according to an embodiment of the present invention.

[0045] Figure 3 The original spectral signal provided for the embodiment of the present invention is preprocessed, converted into a two-dimensional thermal map and input into the Resnet18 training flow chart.

[0046] Figure 4 MTF heat map generated by data from experiments with different dilution rates provided in an embodiment of the present invention, wherein (a) is a two-dimensional feature map corresponding to label 0; (b) is a two-dimensional feature map corresponding to label 1; (c) is a two-dimensional feature map corresponding to label 2;

[0047] Figure 5 A schematic diagram of the Resnet18 network structure provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. 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.

[0049] like Figure 1 As shown, a LDED dilution rate monitoring method based on Markov transition field specifically includes the following steps:

[0050] S1. Obtain the one-dimensional spectrum signal of excited atoms in the molten pool in real time.

[0051] Specifically, in the LDED process, the metal substrate absorbs laser energy to generate plasma, and the plasma radiates a spectrum due to electron transitions. Therefore, using a photomultiplier tube or a spectrometer to collect the one-dimensional spectral signal of excited atoms in the substrate molten pool can effectively reflect parameters such as dilution rate, defects, and mechanical properties, and reflect the degree of fusion between the substrate and the powder.

[0052] S2. Preprocess the one-dimensional spectral signal to obtain a preprocessed one-dimensional spectral signal.

[0053] Due to the instability of the deposition process, the signal-to-noise ratio of the collected spectral data is low, and there is noise aliasing. For a spectral signal, it contains a lot of information. If the unprocessed spectral data is used as input, it will increase the amount of model calculation. Therefore, the following method is used for preprocessing to reduce the amount of computer calculation. In one feasible method, the one-dimensional spectral signal is divided into several intervals according to the set length to obtain several segments of fine-grained one-dimensional spectral signals. This method ignores the overall change trend of the signal, but retains the internal detail components of the signal. It can be expressed as:

[0054] Xi(k)={x t ∣t∈[(k-1)·m+1,k·m]},k=1,2,…,n;

[0055] Where Xi(k) is the kth fine-grained one-dimensional spectral signal; x t is the spectral intensity in the one-dimensional spectral signal; m is the length of each interval; n is the number of segments that divide the interval.

[0056] In another possible implementation, the one-dimensional spectral signal is averaged to obtain a coarse-grained one-dimensional spectral signal. This method ignores the internal detail components of the signal but retains the overall change trend. It can be expressed as:

[0057]

[0058] In the formula, y k is a coarse-grained one-dimensional spectral signal; N is the scale factor, and the length of each coarse-grained time series is equal to the length of the time series sample divided by the scale factor N; x t is the spectral intensity of the one-dimensional spectral signal; L is the total length of the signal; k is the number of signal segments.

[0059] S3. Using the Markov transfer field method, the preprocessed one-dimensional spectral signal is converted into a two-dimensional feature map.

[0060] The preprocessed one-dimensional spectral signal is divided into Q subsequences according to the preset window length, and the length of each subsequence is the same; for example, for X = {0.1, 0.3, 0.5, 0.7, 0.9, 0.2, 0.4, 0.6, 0.8, 1.0}, the window length is 5, and the number of discretization intervals is 4, which can be divided into two subsequences X1 = {0.1, 0.3, 0.5, 0.7, 0.9} and X2 = {0.2, 0.4, 0.6, 0.8, 1.0}.

[0061] Discretize the divided subsequences so that each point in Xi(k) is mapped to a corresponding value range;

[0062] For example, the X value range [0,1] is divided into 4 intervals [0,0.25), [0.25,0.5), [0.5,0.75), [0.75,1]. D1 = {1,2,3,4,4}, D2 = {1,2,3,4,4}. D1 and D2 are the discretized sequences.

[0063] The transition probability is calculated for each point after mapping to obtain the Markov transition matrix W. The state transition probability of two adjacent time steps is calculated to obtain a state transition probability matrix of size [Q,Q], where Q is the number of subsequences, and the value of each element in the matrix is ​​determined by the frequency of the data in X.

[0064] Construct a Markov transition field and output it as a heat map to obtain the image after MTF.

[0065] S4. Input the two-dimensional feature map into a trained ResNet18 model, and output a predicted range of dilution rate; the training data of the ResNet18 model includes two-dimensional feature maps with different dilution rate range labels.

[0066] Specifically, if Figure 5 As shown, the ResNet18 model includes an initial convolutional layer, a residual block, a global average pooling layer, and a fully connected layer.

[0067] Specifically, the initial convolution layer uses a 7×7 convolution kernel, a stride of 2, a padding of 3, and an output channel of 64, and outputs the original image as a feature map of size 56×56×64. The residual block contains 4 residual block sequences, and the image processed by the initial convolution layer is input. Each sequence contains a different number of residual blocks. Each residual block has two 3x3 convolution layers with channels of 64, 128, 256, and 512, respectively. The last residual block outputs a 7×7×512 feature map. The feature map (7×7×512) output by the last residual block is input into the global average pooling layer, which takes the average of the feature map to reduce the spatial dimension of the feature map. The 512-dimensional feature vector output by the global average pooling layer is input into the fully connected layer, where the Softmax function outputs the dilution rate range.

[0068] In order to explain the present invention more clearly, a specific case is described below.

[0069] In this case, a photomultiplier tube (PMT) was used to extract excited-state atomic signals in the molten pool. The sampling frequency was 44KHz and the sampling time was ≥10s. The substrate and powder materials were composed of elements such as iron and cobalt, respectively. The number of channels was adjusted to 2 to obtain the spectral signals of the two elements. Data covering different dilution rate ranges were obtained by controlling different process parameters (laser power, scanning speed, printing speed), and the data were exported as multiple sets of one-dimensional spectral signals.

[0070] As shown in Table 1, the experimental data of the present invention can be divided into three categories according to the dilution rate interval, and labels 0 to 2 are set.

[0071] First, the experimental data are divided and organized according to the dilution rate range [0-0.25), [0.25-0.35), [0.35-0.5], and the data volume is calculated. The experimental data of labels 0, 1, and 2 are averaged or segmented to obtain coarse-grained time series or fine-grained time series, such as Figure 3 As shown; and using the Markov transfer field method, the preprocessed one-dimensional spectral signal is converted into a two-dimensional feature map, such as Figure 2 As shown in the figure, the experimental data of each dilution rate interval are subjected to MTF in turn to obtain the thermal diagram of different dilution rate intervals, as shown in the figure. Figure 4 As shown; finally, the two-dimensional feature map is input into the trained ResNet18 model, and the model structure is as follows Figure 5 As shown, the predicted range of the dilution rate is output by Softmax, and the process parameters are dynamically adjusted to complete the real-time monitoring of the dilution rate.

[0072] Table 1 Experimental data

[0073]

[0074] The embodiment of the present invention provides a LDED dilution rate monitoring device based on Markov transition field, which is used to implement the aforementioned LDED dilution rate monitoring method based on Markov transition field, and specifically includes the following modules:

[0075] The acquisition module is used to acquire the one-dimensional spectrum signal of excited atoms in the molten pool in real time.

[0076] The preprocessing module is used to preprocess the one-dimensional spectral signal to obtain a preprocessed one-dimensional spectral signal.

[0077] The conversion module is used to convert the preprocessed one-dimensional spectral signal into a two-dimensional feature map by adopting the Markov transfer field method.

[0078] The monitoring module is used to input the two-dimensional feature map into a trained ResNet18 model and output a predicted range of dilution rate; the training data of the ResNet18 model includes two-dimensional feature maps with different dilution rate range labels.

[0079] All relevant contents of each step involved in the aforementioned embodiment of a LDED dilution rate monitoring method based on a Markov transfer field can be referred to the functional description of the functional module corresponding to a LDED dilution rate monitoring device based on a Markov transfer field in the embodiment of the present invention, and will not be repeated here. The division of modules in the embodiment of the present invention is schematic, which is only a logical function division. There may be other division methods in actual implementation. In addition, each functional module in each embodiment of the present invention can be integrated into a processor, or it can exist physically separately, or two or more modules can be integrated into one module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of software functional modules.

[0080] In another embodiment of the present invention, a computer device is provided, the computer device comprising a processor and a memory, the memory being used to store a computer program, the computer program comprising program instructions, and the processor being used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or may be 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., which are the computing core and control core of the terminal, which are suitable for implementing one or more instructions, and are specifically suitable for loading and executing one or more instructions in a computer storage medium to implement a corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of a LDED dilution rate monitoring method based on a Markov transition field.

[0081] 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 computer 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 computer device and the extended storage medium supported by the computer device. 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 the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the above-mentioned embodiment regarding a method for monitoring the LDED dilution rate based on a Markov transition field.

[0082] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0083] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks 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 produce 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.

[0084] 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 1A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0085] 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 steps for the functions specified in one or more boxes.

[0086] The present invention also provides a computer program product, which is used to execute any of the above-mentioned methods for monitoring LDED dilution rate based on Markov transfer field. Since the computer program product provided by the present invention and the above-mentioned method for monitoring LDED dilution rate based on Markov transfer field belong to the same inventive concept, the computer program product provided by the present invention has all the advantages of the above-mentioned method for monitoring LDED dilution rate based on Markov transfer field, so the beneficial effects of the computer program product provided by the present invention will not be described one by one here.

[0087] In the present invention, the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" etc. mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the described specific features, structures, materials or characteristics may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0088] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for monitoring LDED dilution rate based on Markov transition field, characterized in that: include: Obtain the one-dimensional spectrum signal of excited atoms in the molten pool in real time; Preprocessing the one-dimensional spectral signal to obtain a preprocessed one-dimensional spectral signal; Using the Markov transfer field method, the preprocessed one-dimensional spectral signal is converted into a two-dimensional feature map; The two-dimensional feature map is input into a trained ResNet18 model, and a predicted range of dilution rate is output; the training data of the ResNet18 model includes two-dimensional feature maps with different dilution rate range labels.

2. The LDED dilution rate monitoring method based on Markov transition field according to claim 1 is characterized in that: The method for acquiring the one-dimensional spectrum signal of excited atoms in the molten pool is: A photomultiplier tube or a spectrometer is used to collect the one-dimensional spectrum signal of excited atoms in the molten pool.

3. The LDED dilution rate monitoring method based on Markov transition field according to claim 1 is characterized in that: The preprocessing of the one-dimensional spectral signal comprises: The one-dimensional spectral signal is divided into several intervals according to the set length, and several segments of fine-grained one-dimensional spectral signals are obtained, which are expressed as: Xi(k)={x t ∣t∈[(k-1)·m+1,k·m]},k=1,2,…,n; Where Xi(k) is the kth segment fine-grained one-dimensional spectral signal; x t is the spectral intensity of the one-dimensional spectral signal; m is the length of each interval; n is the number of segments that divide the interval.

4. The LDED dilution rate monitoring method based on Markov transition field according to claim 1 is characterized in that: The preprocessing of the one-dimensional spectral signal comprises: The one-dimensional spectral signal is averaged to obtain a coarse-grained one-dimensional spectral signal, which is expressed as: In the formula, y k is a coarse-grained one-dimensional spectral signal; N is the scale factor, and the length of each coarse-grained time series is equal to the length of the time series sample divided by the scale factor N; x t is the spectral intensity of the one-dimensional spectral signal; L is the total length of the signal; k is the number of signal segments.

5. The LDED dilution rate monitoring method based on Markov transition field according to claim 1, characterized in that: The Markov transfer field method is used to convert the preprocessed one-dimensional spectral signal into a two-dimensional feature map, specifically: The preprocessed one-dimensional spectral signal is divided into a plurality of subsequences according to a preset window length; Discretize the divided subsequences so that each point in the one-dimensional spectral signal is mapped into a corresponding value range; Calculate the transition probability for each point after mapping to obtain a state transition probability matrix; The state transition probability matrix is ​​visualized through a heat map to obtain the MTF image.

6. The method for monitoring LDED dilution rate based on Markov transition field according to claim 1, characterized in that: The ResNet18 model includes an initial convolutional layer, a residual block, a global average pooling layer, and a fully connected layer; The initial convolution layer uses a 7×7 convolution kernel, a step size of 2, a padding of 3, and an output channel of 64, and outputs the original image as a feature map of size 56×56×64; The residual block includes 4 residual block sequences. The image processed by the initial convolution layer is input. Each sequence contains a different number of residual blocks. Each residual block has two 3x3 convolution layers with channel numbers of 64, 128, 256 and 512 respectively. The last residual block outputs a 7×7×512 feature map. The feature map output by the last residual block is input into the global average pooling layer, which takes the average of the feature map and inputs the 512-dimensional feature vector output by the global average pooling layer into the fully connected layer, where the Softmax function outputs the dilution rate range.

7. A LDED dilution rate monitoring device based on Markov transition field, characterized in that: include: An acquisition module is used to acquire one-dimensional spectrum signals of excited atoms in the molten pool in real time; A preprocessing module, used for preprocessing the one-dimensional spectral signal to obtain a preprocessed one-dimensional spectral signal; A conversion module, used for converting the preprocessed one-dimensional spectral signal into a two-dimensional feature map by using a Markov transfer field method; The monitoring module is used to input the two-dimensional feature map into a trained ResNet18 model and output a predicted range of dilution rate; the training data of the ResNet18 model includes two-dimensional feature maps with different dilution rate range labels.

8. A computer 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 LDED dilution rate monitoring method based on Markov transition field as described in any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for monitoring LDED dilution rate based on Markov transition field as claimed in any one of claims 1 to 6 is implemented.

10. A computer program product, characterized in that When the computer program product is executed by a processor, the method for monitoring the LDED dilution rate based on the Markov transition field as claimed in any one of claims 1 to 6 is implemented.

Images