An online monitoring platform and adaptive manufacturing method for underwater additive manufacturing

Abstract

The present invention discloses an online monitoring platform and an adaptive manufacturing method for underwater additive manufacturing, wherein the monitoring platform comprises an underwater sealing cover installed on a substrate for sealing a laser head, wherein an inner hollow shaft motor and an outer hollow shaft motor are installed inside the underwater sealing cover, wherein the inner hollow shaft motor is installed above the laser head, and its inner ring is fixedly connected to the laser head, and a laser beam splitter is installed on the outer ring; the outer ring of the outer hollow shaft motor forms an axial concentric fit with the outer wall of the small hole at the lower end of the inner ring of the underwater sealing cover, and a side monitoring CCD, a line laser generator, a line laser monitoring CCD, a temperature sensor, a high-precision spectrometer, and a line laser generator are installed on the inner ring of the outer hollow shaft motor; and an ultrasonic detector is installed on the back of the substrate. The present invention takes into account the relevant issues of underwater additive manufacturing, and has the characteristics of stable monitoring structure, flexible monitoring process, fast calculation speed, and real-time and accurate results, thereby realizing online monitoring of underwater additive processes.

Description

Technical Field

[0001] The present invention belongs to the technical field of laser additive manufacturing, and in particular relates to an online monitoring platform and an adaptive manufacturing method for underwater additive manufacturing. Background Art

[0002] Underwater additive manufacturing technology is a manufacturing method that adds materials layer by layer to a working surface underwater to build a three-dimensional object. In the marine field, underwater additive manufacturing technology is widely used to repair damaged marine facilities and build marine facilities. Underwater, once the equipment is damaged, repairing it in situ underwater can greatly improve the repair efficiency and reduce the repair cost. However, in the process of underwater additive manufacturing, affected by the complex underwater environment, the quality of the manufactured components needs to consider many factors. Therefore, it is very important to monitor the performance and structure of the components in the underwater additive manufacturing process in real time. However, there are many problems with the current underwater additive manufacturing monitoring technology. Due to the complex underwater environment, online monitoring needs to consider many factors. Considering the limited space for online monitoring of underwater additive manufacturing, the impact of the drainage device on the monitoring device when discharging water during underwater additive manufacturing cannot be ignored. In addition, the flexible control of the monitoring device in such a limited space is also a key issue in the development of underwater additive manufacturing technology. Therefore, the realization of online monitoring of underwater laser additive manufacturing requires a new online monitoring technology to solve these problems. Summary of the invention

[0003] To solve the above problems, the present invention discloses an online monitoring platform and an adaptive manufacturing method for underwater additive manufacturing. The online monitoring platform monitors the forming status of components in real time through sensors and image processing technology, and combines the adaptive manufacturing method to effectively predict the upcoming molten pool, thereby effectively improving the quality and efficiency of underwater additive manufacturing.

[0004] The present invention first provides an online monitoring platform for underwater additive manufacturing, comprising an underwater sealing cover installed on a substrate for sealing a laser head, wherein an inner hollow shaft motor and an outer hollow shaft motor are installed inside the underwater sealing cover, wherein the inner hollow shaft motor is installed above the laser head, and its inner ring is fixedly connected to the laser head, and a laser beam splitter is installed on the outer ring; the outer ring of the outer hollow shaft motor forms an axial concentric fit with the outer wall of a small hole at the lower end of the inner ring of the underwater sealing cover, and a lateral monitoring CCD, a line laser generator, a line laser monitoring CCD, a temperature sensor, and a high-precision spectrometer are installed on the inner ring of the outer hollow shaft motor. The line laser generator and the line laser monitoring CCD are symmetrically installed about a plane perpendicular to the scanning direction where a test point on the deposition layer is located, and the line laser generator is aligned with the laser beam splitter, and the lateral monitoring CCD is installed and arranged at 90 degrees on the horizontal plane with the laser generator and the line laser monitoring CCD; an ultrasonic detector is installed on the back of the substrate.

[0005] Furthermore, the upper end of the laser beam splitter is installed on any annular positioning hole of the outer ring of the inner hollow shaft motor by bolt connection; the temperature sensor and the high-precision spectrometer are installed on the inner ring positioning hole of the outer hollow shaft motor by bolt connection, and are aligned with the point required to be tested in the deposition layer.

[0006] A method for realizing underwater additive manufacturing adaptive manufacturing using the above-mentioned online monitoring platform for underwater additive manufacturing comprises the following steps:

[0007] S1. The laser head emits laser light to melt the powder and form a molten pool on the substrate;

[0008] S2. The side monitoring CCD takes a picture of the side of the molten pool from the side to obtain a side view of the molten pool surface; the coaxial monitoring CCD takes a picture from above the molten pool to obtain a top view of the molten pool surface; the ultrasonic measuring instrument detects the morphology of the remelting molten pool from the bottom of the substrate to obtain the morphology of the remelting molten pool inside the deposition layer formed on the substrate;

[0009] S3. The computer constructs a three-dimensional geometric morphology of the molten pool, including a three-dimensional model of the molten pool, through views obtained by the side monitoring CCD and the coaxial monitoring CCD;

[0010] S4. In parallel with step S3, the line laser generator emits a line laser which is refracted by a laser beam splitter behind the scanning point, and the line laser profile is acquired by a 305-line laser monitoring CCD, and transmitted to a computer to construct a three-dimensional model of the deposition layer;

[0011] S5. In parallel with step S3, the computer constructs a three-dimensional model of the remelting molten pool through the data fed back by the ultrasonic measuring instrument. The computer calculates the surface displacement and volume change caused by the cooling, solidification and shrinkage of the molten pool to the deposited layer based on the size difference between the three-dimensional model of the molten pool and the three-dimensional model of the deposited layer. Combined with the microstructure, stress state and temperature parameters obtained by the temperature sensor and the high-precision spectrometer monitoring device, a real-time three-dimensional numerical model of the molten pool-deposited layer is constructed to realize online monitoring of underwater additive manufacturing.

[0012] Furthermore, in the process of online monitoring of underwater additive manufacturing, for the laser unmelted part of the underwater additive manufacturing process, the computer summarizes a series of model parameters such as the shape geometric parameters of the molten pool-deposition layer, temperature and material stress state, microstructure, and element content obtained by online monitoring to generate input and output data sets; by selecting deep learning models and parameters, the deep learning model is trained using the TensorFlow deep learning framework to accurately learn the relationship between input parameters and output results, as well as the relationship between output results and processing parameters; the real-time data of the model is input into the trained deep learning model, and the deep learning model will output the model parameters to be deposited, and reflect them in the modified processing parameters, so as to compensate for the output results of the deep learning model, and finally realize the real-time prediction of the underwater additive manufacturing process.

[0013] Furthermore, the stress state is measured by calculating the surface displacement difference between the molten pool and the deposited layer. The specific method is: through the generated three-dimensional geometric model of the molten pool and the geometric dimensions of the deposited layer, the surface displacement and volume shrinkage values ​​caused by the shrinkage of the material from the molten state to the solidified state are obtained, and the surface displacement and volume shrinkage values ​​are added as loads to the real-time generated model to calculate the stress value caused by the cooling of the molten pool; in addition, the surface shrinkage of the remelted molten pool is obtained by an ultrasonic measuring instrument, and based on this, it is added to the three-dimensional model as a load to obtain the stress state after the stress superposition caused by the increase in the number of deposited layers; with the advancement of the laser heat source, each formation of a new molten pool will be calculated through the above process, and the calculation results will be continuously iterated, thereby reflecting the stress changes inside the deposited layer.

[0014] Furthermore, the monitoring of element content is achieved by calibrating the spectral lines of plasma excitation measured by a high-precision spectrometer, thereby establishing a plasma spectrum database of different elements. In the real-time monitoring process, the element type and distribution are determined by analyzing the peak wavelength of the spectral line and imported into the three-dimensional model.

[0015] The microstructure monitoring method is to establish a connection between the plasma excitation spectrum signal and the microstructure by building a deep learning network before testing. During real-time monitoring, the spectral line characteristics are analyzed through the deep learning network to determine the microstructure of the test point and import it into the three-dimensional model. Beneficial Effects

[0016] The invention proposes an underwater additive manufacturing online monitoring platform, which solves the problem of difficulty in building an underwater online monitoring platform due to the complex underwater additive manufacturing environment and narrow space.

[0017] The present invention proposes an online monitoring platform for underwater additive manufacturing. Its monitoring method is different from that of additive manufacturing in the air. It adopts an adjustable rotating monitoring method, which can realize online monitoring of multi-directional scanning conditions.

[0018] The monitoring method proposed in the present invention can capture the surface molten pool morphology, form two line lasers under the refraction of the line laser beam splitter, obtain the geometric morphology of the deposition layer at different positions, and obtain more accurate stress distribution through interpolation calculation.

[0019] The underwater additive manufacturing online monitoring platform proposed in the present invention not only realizes online monitoring of the molten pool through a high-speed camera CCD in the narrow space provided by the water-gas mixed drainage, but also monitors the elemental composition and microstructure through the plasma excitation spectrum induced by the heat source laser. It is low in cost and easy to operate.

[0020] The adaptive manufacturing method of the present invention is based on establishing a three-dimensional numerical model, which on the one hand reflects the real-time state of the deposition process, and on the other hand predicts the relevant parameters of the molten pool to be generated and is reflected in adjusting specific processing parameters to obtain components with higher precision and better mechanical properties.

[0021] The prediction-compensation mechanism involved in the adaptive method proposed in the present invention cannot establish complex nonlinear relationships using traditional finite element methods, and requires a lot of experiments and adjustments to obtain accurate results. However, deep learning models can learn complex nonlinear relationships by learning a large amount of data, thereby enabling accurate prediction and compensation based on online monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a 1 / 4 bottom sectional view of the underwater additive manufacturing online monitoring platform of the present invention;

[0023] Figure 2 It is a partial enlarged view of the monitoring area of ​​the underwater additive manufacturing online monitoring platform of the present invention;

[0024] Figure 3 It is a 1 / 4 cross-sectional view of the underwater additive manufacturing device combined with online monitoring of the present invention;

[0025] Figure 4 It is a schematic diagram of the principle of the laser beam splitter for the beam splitting light source in the present invention;

[0026] Figure 5 It is a flow chart of the adaptive manufacturing method applicable to online monitoring of underwater additive manufacturing platform in the present invention;

[0027] Description of Reference Numerals

[0028] 100-underwater sealing cover, 200-base, 101-inner hollow shaft motor, 102-laser head, 201-laser, 201-laser beam splitter, 301-temperature sensor, 302-outer hollow shaft motor, 303-line laser generator, 304-high-precision spectrometer, 305-line laser monitoring CCD, 306-side monitoring CCD, 307-coaxial monitoring CCD, 401-ultrasonic detector.

[0029] Depend on Figure 1-5 As shown, the components required for the implementation of the underwater laser additive manufacturing online monitoring platform include an underwater sealing cover 100, an inner hollow shaft motor 101, a laser head 102, a laser beam splitter 201, a temperature sensor 301, an outer hollow shaft motor 302, a line laser generator 303, a high-precision spectrometer 304, a line laser monitoring CCD 305, a side monitoring CCD 306, a coaxial monitoring CCD 307, an ultrasonic detector 401, and a high-performance multi-core computer. The inner hollow shaft motor 101 is installed above the laser head 102, and its inner ring cooperates with the laser head 102, and the outer ring cooperates with the device for installing the laser beam splitter 201. The inner hollow shaft motor 101 and the device fixed thereon can realize circumferential rotation. The side monitoring CCD306, line laser generator 303, line laser monitoring CCD305, temperature sensor 301, and high-precision spectrometer 304 are all installed on the inner ring of the outer hollow shaft motor 302, and the outer ring thereof forms an axial concentric fit with the outer wall of the small hole at the lower end of the inner ring of the underwater sealing cover 100 to achieve circumferential rotation. Among them, the line laser generator 303 and the line laser monitoring CCD are symmetrically installed about the plane perpendicular to the scanning direction where the test point on the sedimentary layer is located, so that the line laser monitoring CCD305 can capture the image formed by the laser emitted by the line laser generator 303 in the sedimentary layer. It should be noted that the upper end of the laser beam splitter 201 is installed on any annular positioning hole on the outer ring of the inner hollow shaft motor 101 by bolt connection: Figure 4 As shown, the laser emitted by the line laser generator 303 is split by two beam splitters inside the laser beam splitter 201. The line laser first contacts the near-end beam splitter lens (angle α=45°), part of which is refracted to the rear end of the deposition layer, and part of which is transmitted to the far-end beam splitter lens (angle α=60°) for further refraction, and reaches a position slightly in front of the deposition layer. The side monitoring CCD306 is installed on the inner ring of the outer hollow shaft motor 302, and is arranged at 90° on the horizontal plane with the laser generator and the line laser monitoring CCD305. The temperature sensor 301 and the high-precision spectrometer 304 are installed on the inner ring positioning hole of the outer hollow shaft motor 302 by bolt connection, and are aligned with the points to be tested in the deposition layer. The ultrasonic detector is located on the back of the substrate 200, and is used to detect the geometric morphology of the remelting pool and feed it back to the high-performance multi-core computer.

[0030] During the manufacturing process, the monitoring equipment is fixed on the annular positioning holes of the inner and outer hollow shaft motors by bolting on land according to the scanning monitoring requirements. The inner and outer hollow shaft motors, including the side monitoring CCD306, line laser generator 303, line laser monitoring CCD305, temperature sensor 301, and high-precision spectrometer 304 arranged thereon, are controlled by a computer to achieve 360° free rotation around the laser head 102. Based on the scanning path, the line laser generator 303 placed on the inner ring of the outer hollow shaft motor 302 and the laser beam splitter 201 placed on the outer ring of the inner hollow shaft motor 101 are located in the same plane in the scanning direction to ensure that the line laser is always perpendicular to the deposition layer. The high-performance multi-core computer is used to receive the data collected by each monitoring device, integrate and process them, analyze and calculate them, and generate a numerical three-dimensional model of the molten pool-deposition layer in real time.

[0031] The method of implementing adaptive manufacturing using the above-mentioned underwater additive manufacturing online monitoring platform mainly includes the following steps:

[0032] High energy laser is emitted from the laser head 102 to melt the coaxially output powder and form a molten pool on the substrate 200;

[0033] The side monitoring CCD306 takes a picture of the side of the molten pool from the side to obtain a side view of the molten pool surface; the coaxial monitoring CCD307 takes a picture from above the molten pool to obtain a top view of the molten pool surface; the ultrasonic measuring instrument detects the morphology of the remelting molten pool from the bottom of the substrate 200 to obtain the morphology of the remelting molten pool inside the deposition layer formed on the substrate 200;

[0034] The high-performance multi-core computer constructs the three-dimensional geometric morphology of the molten pool, including the three-dimensional model of the molten pool, through the views obtained by the side monitoring CCD306 and the coaxial monitoring CCD307;

[0035] In parallel with step (3), the line laser generator 303 emits a line laser which is refracted by the laser beam splitter 201 at the rear of the scanning point. The line laser profile is acquired by the line laser monitoring CCD 305 and transmitted to a high-performance multi-core computer to construct a three-dimensional model of the sedimentary layer.

[0036] In parallel with step (3), the computer constructs a three-dimensional model of the remelting pool through the data fed back by the ultrasonic measuring instrument.

[0037] The high-performance multi-core computer calculates the surface displacement and volume change caused by the cooling, solidification and shrinkage of the molten pool to the deposited layer based on the size difference between the three-dimensional model of the molten pool and the three-dimensional model of the deposited layer. Combined with the microstructure, stress condition and temperature parameters obtained by the temperature sensor and the high-precision spectrometer monitoring device, a real-time three-dimensional numerical model of the molten pool-deposited layer is built to realize the online monitoring of underwater additive manufacturing. For the laser unmelted part of the underwater additive manufacturing process, the computer summarizes the shape geometric parameters, temperature and material stress state, microstructure, element content and a series of model parameters of the molten pool-deposited layer obtained by online monitoring to generate input and output data sets. By selecting appropriate deep learning models and parameters, the model is trained using the TensorFlow deep learning framework to accurately learn the relationship between input parameters and output results, as well as the relationship between output results and processing parameters. Furthermore, the real-time data of the model is input into the trained deep learning model, and the deep learning model will output the model parameters to be deposited and reflect them in the modified processing parameters to achieve compensation for the output results of the deep learning model, and finally obtain a processed component with higher precision and better mechanical properties, realizing real-time prediction of the underwater additive manufacturing process.

[0038] Through the analysis of data before and after the laser scanning time point, this adaptive manufacturing process mainly includes two aspects: on the one hand, the data of the deposited and deposited molten pools are recorded, that is, the online monitoring of underwater additive manufacturing; on the other hand, for the molten pool to be generated, based on the numerical model analysis of the deposited layer and the molten pool generated in real time, the laser parameters to be deposited, including scanning speed, laser power, etc., are optimized to achieve real-time prediction of the underwater additive manufacturing process.

[0039] The size and shape monitoring of the molten pool-deposition layer includes the morphology monitoring of the molten pool and the morphology monitoring of the deposition layer.

[0040] The morphology monitoring of the molten pool includes morphology monitoring of the molten pool surface and morphology monitoring of the remelting molten pool.

[0041] The morphology monitoring of the molten pool surface is performed by using the side monitoring CCD306 and the coaxial monitoring CCD307 to obtain the side view and the top view of the molten pool respectively, and the surface molten pool size in different views is obtained by taking the surface image of the molten pool, and the surface three-dimensional geometric size of the molten pool is synthesized to generate a three-dimensional model;

[0042] The morphology of the remelting pool is monitored by ultrasonic monitoring. The morphology of the remelting pool refers to the part of the upper molten pool that is re-melted by the laser when another deposition layer is further generated on the upper surface of the deposited deposition layer. The boundary of the laser remelting pool is detected by an ultrasonic probe set at the bottom of the additive manufacturing 200-base, and the three-dimensional geometric dimensions of the remelting pool are obtained and a three-dimensional model is generated.

[0043] The monitoring of the deposited layer morphology adopts the line laser method. The line laser generator 303 emits a line laser which is refracted by the laser beam splitter 201 behind the scanning point to form two line lasers. The near-end line laser is about 3 mm away from the molten pool. With the help of the line laser monitoring CCD 305, the line laser irradiated on the deposited layer is collected to form the geometric morphology of the deposited layer and generate a three-dimensional model.

[0044] The stress state is measured by calculating the surface displacement difference between the molten pool and the deposited layer. The specific method is as follows: the surface displacement and volume shrinkage value generated by the material shrinking from the molten state to the solidified state are obtained through the three-dimensional geometric model of the molten pool and the geometric dimensions of the deposited layer generated above. The surface displacement and volume shrinkage values ​​are added as loads to the model generated in real time to calculate the stress value generated by the cooling of the molten pool. In addition, the calculation of the stress state of the remelted molten pool is similar to this, except that the surface shrinkage of the remelted molten pool is obtained by an ultrasonic measuring instrument, and based on this, it is added to the three-dimensional model as a load to obtain the stress state after the stress superposition caused by the increase in the number of deposited layers. With the advancement of the laser heat source, each formation of a new molten pool will be calculated through the above process, and the calculation results will be continuously iterated to reflect the stress changes inside the deposited layer.

[0045] The monitoring of element content is achieved by calibrating the spectral lines of plasma excitation measured by the high-precision spectrometer 304, and then establishing a plasma spectrum database of different elements. In the real-time monitoring process, the element type and distribution are determined by analyzing the peak wavelength of the spectral line and imported into the three-dimensional model.

[0046] The microstructure monitoring method is to establish a connection between the plasma excitation spectrum signal and the microstructure by building a deep learning network before testing. During real-time monitoring, the spectral line characteristics are analyzed through the deep learning network to determine the microstructure of the test point and import it into the three-dimensional model.

[0047] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. An online monitoring platform for underwater additive manufacturing, characterized in that: It comprises an underwater sealing cover installed on a base for sealing a laser head, wherein an inner hollow shaft motor and an outer hollow shaft motor are installed inside the underwater sealing cover, wherein the inner hollow shaft motor is installed above the laser head, and its inner ring is fixedly connected to the laser head, and a laser beam splitter is installed on the outer ring; the outer ring of the outer hollow shaft motor forms an axial concentric fit with the outer wall of the small hole at the lower end of the inner ring of the underwater sealing cover, and a side monitoring CCD, a line laser generator, a line laser monitoring CCD, a temperature sensor and a high-precision spectrometer are installed on the inner ring of the outer hollow shaft motor, the line laser generator and the line laser monitoring CCD are symmetrically installed about a plane perpendicular to the scanning direction where a test point on the sedimentary layer is located, and the line laser generator is aligned with the laser beam splitter, and the side monitoring CCD, the laser generator and the line laser monitoring CCD are arranged at 90 degrees on the horizontal plane; an ultrasonic detector is installed on the back of the base.

2. The online monitoring platform for underwater additive manufacturing according to claim 1, characterized in that: The upper end of the laser beam splitter is installed on any annular positioning hole of the outer ring of the inner hollow shaft motor through bolt connection; the temperature sensor and the high-precision spectrometer are installed on the inner ring positioning hole of the outer hollow shaft motor through bolt connection and are aligned with the test point in the deposition layer.

3. A method for realizing underwater additive manufacturing adaptive manufacturing using the online monitoring platform for underwater additive manufacturing as claimed in claim 1 or 2, characterized in that: The method comprises the following steps: S1. The laser head emits laser light to melt the powder and form a molten pool on the substrate; S2. The side monitoring CCD takes a picture of the side of the molten pool from the side to obtain a side view of the molten pool surface; the coaxial monitoring CCD takes a picture from above the molten pool to obtain a top view of the molten pool surface, and the ultrasonic measuring instrument detects the morphology of the remelted molten pool from the bottom of the substrate to obtain the morphology of the remelted molten pool inside the deposition layer formed above the substrate; S3. The computer constructs a three-dimensional geometric morphology of the molten pool through the views obtained by the side monitoring CCD and the coaxial monitoring CCD, including a three-dimensional model of the molten pool; S4. In parallel with step S3, the line laser generator emits a line laser, which is refracted by a laser beam splitter and the part behind the scanning point is obtained by the line laser monitoring CCD to obtain the line laser profile, and transmits it to the computer to construct a three-dimensional model of the deposition layer; S5. In parallel with step S3, the computer constructs a three-dimensional model of the remelting molten pool through the data fed back by the ultrasonic measuring instrument. The computer calculates the surface displacement and volume change of the deposited layer caused by the cooling and solidification shrinkage of the molten pool based on the size difference between the three-dimensional model of the molten pool and the three-dimensional model of the deposited layer. Combined with the microstructure, stress state and temperature parameters obtained by the temperature sensor and the high-precision spectrometer monitoring device, a real-time three-dimensional numerical model of the molten pool-deposited layer is constructed to realize online monitoring of underwater additive manufacturing.

4. The method according to claim 3, characterized in that During the online monitoring of underwater additive manufacturing, for the laser unmelted part of the underwater additive manufacturing process, the computer summarizes the geometric parameters of the molten pool-deposition layer, the temperature and material stress state, microstructure, and element content model parameters obtained by online monitoring to generate input and output data sets; by selecting deep learning models and parameters, the deep learning model is trained using the TensorFlow deep learning framework to accurately learn the relationship between input parameters and output results, as well as the relationship between output results and processing parameters; the real-time data of the model is input into the trained deep learning model, and the deep learning model will output the model parameters to be deposited and reflect them in the modified processing parameters to achieve compensation for the output results of the deep learning model, and finally achieve real-time prediction of the underwater additive manufacturing process.

5. The method according to claim 3, characterized in that: The stress state is measured by calculating the surface displacement difference between the molten pool and the deposited layer. The specific method is: the surface displacement and volume shrinkage value caused by the shrinkage of the material from the molten state to the solidified state are obtained through the generated three-dimensional geometric model of the molten pool and the geometric dimensions of the deposited layer, and the surface displacement and volume shrinkage values ​​are added as loads to the real-time generated model to calculate the stress value generated by the cooling of the molten pool; in addition, the surface shrinkage of the remelted molten pool is obtained by an ultrasonic measuring instrument, and the surface shrinkage data obtained is added as a load to the three-dimensional model to obtain the stress state after the stress superposition caused by the increase in the number of deposited layers; with the advancement of the laser heat source, each formation of a new molten pool will undergo the above-mentioned calculation of the stress state, and the calculation results will be continuously iterated to reflect the stress changes inside the deposited layer.

6. The method according to claim 4, characterized in that The monitoring of element content is achieved by calibrating the spectral lines of plasma excitation measured by a high-precision spectrometer, and then establishing a plasma spectral database of different elements; in the real-time monitoring process, the type and distribution of elements are determined by analyzing the peak wavelength of the spectral lines and imported into the three-dimensional model.

7. The method according to claim 3, characterized in that The microstructure monitoring method is to establish a connection between the plasma excitation spectrum signal and the microstructure by building a deep learning network before testing. During real-time monitoring, the spectral line characteristics are analyzed through the deep learning network to determine the microstructure of the test point and import it into the three-dimensional model.

Images