WO2023221282A1

WO2023221282A1 - Model correction method, spectroscopy device, computer device, and storage medium

Info

Publication number: WO2023221282A1
Application number: PCT/CN2022/106217
Authority: WO
Inventors: 潘从元; 张兵; 贾军伟
Original assignee: 合肥金星智控科技股份有限公司
Priority date: 2022-05-18
Filing date: 2022-07-18
Publication date: 2023-11-23
Also published as: CN114636688A; CN114636688B

Abstract

A model correction method, a spectroscopy device, a computer device, and a storage medium. The model correction method comprises: acquiring original spectral data of a plurality of focusing distances on the basis of a laser-induced breakdown spectroscopy technology (S10); establishing a prediction model by using the original spectral data of one of the focusing distances (S20); predicting predicted spectral data of the remaining focusing distances by using the prediction model (S30); and correcting the prediction model for the plurality of focusing distances on the basis of an error between the predicted spectral data and the original spectral data (S40).

Description

Model calibration method, spectroscopic equipment, computer equipment and storage media

priority information

This application requests the priority and rights of the patent application with patent application number 202210538088.7, which was submitted to the State Intellectual Property Office of China on May 18, 2022, and its full text is incorporated herein by reference.

Technical field

The present application relates to the technical field of laser-induced breakdown spectroscopy analysis, and in particular to a model correction method, spectral equipment, computer equipment and storage media.

Background technique

Current LIBS equipment application sites are subject to periodic equipment maintenance or stockpile replacement. Large changes in the liquid and material levels of the materials to be measured, as well as changes in the measurement distance after the equipment is moved, will result in the inability to collect effective spectra, and the focus needs to be adjusted. When the focus point is adjusted, the accuracy of the original model decreases, and recalibration requires a lot of manpower, material resources, and time.

Contents of the invention

Embodiments of the present application provide a model correction method, spectral equipment, computer equipment and storage media.

The model correction method in the embodiment of this application includes:

Based on laser-induced breakdown spectroscopy technology, original spectral data at multiple focusing distances are obtained;

Build a predictive model using raw spectral data at one of the focusing distances;

Using the prediction model, predict the predicted spectral data for the remaining focus distances;

Predictive model corrections for multiple focus distances are based on errors between the predicted spectral data and the original spectral data.

The model correction method of the embodiment of the present application can realize the migration correction of the prediction model of laser-induced breakdown spectroscopy equipment under different focusing distances, and improve the accuracy and stability of the prediction model. Laser-induced breakdown spectroscopy equipment can quickly adapt to changes in on-site test conditions, solving the problem of prediction model failure caused by changes in focus distance due to changes in the position of detected materials at the application site. There is no need to recalibrate, saving manpower, material resources and time.

In some embodiments, predictive model correction for multiple focus distances based on an error between the predicted spectral data and the original spectral data includes:

Establish a correction function between original spectral data and predicted spectral data;

Based on the correction function, establish a relationship between the loss function and the correction function;

Based on the relationship, determining a value of the correction function;

The prediction model of multiple focus distances is corrected using the determined value of the correction function.

In some embodiments, establishing a correction function between original spectral data and predicted spectral data includes:

determining an original concentration of an element based on the original spectral data;

determining a predicted concentration of an element based on the predicted spectral data;

The correction function is established based on the original concentration and the predicted concentration.

In some embodiments, the correction function adopts the following relationship:

C＝Z(l)c

Among them, c is the predicted concentration of the element at the focusing distance l of the prediction model, C is the original concentration of the element, and Z(l) is the corresponding correction function at the focusing distance l.

In some embodiments, determining a value of the correction function based on the relationship includes:

Determine the minimum value of the loss function using the least squares method;

The minimum value is used to determine the value of the correction function.

In some implementations, the loss function adopts the following relationship:

Among them, J is the loss function, i is the sample number, j is the focusing distance number, M is the number of samples, N is the number of focusing distances, C _i is the original concentration of the element of the i-th sample, Z(l _j ) is the j-th The correction function corresponding to the focusing distance l _j , c ^j _i is the predicted concentration of the element at the jth focusing distance l _j of the i-th sample of the prediction model.

In some embodiments, the model correction method includes:

The corrected prediction model is used to predict the elements in the sample at the corresponding focusing distance.

The spectroscopic equipment in the embodiment of the present application includes a sample stage, a laser and a spectrometer. The sample stage is used to carry the sample, the laser is used to emit laser light to the sample, and the spectrometer is used to receive the laser light reflected by the sample. The spectrometer includes a processor, and the processor is used to implement the model correction method in any of the above embodiments.

The spectrometer of the spectroscopic equipment in the embodiment of the present application has a processor that can implement the model correction method. When the spectroscopic equipment is measured at different measurement distances, the model correction method can be used to collect effective spectra without recalibration, saving manpower, material resources and time.

The computer device in the embodiment of the present application includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, it implements the model correction method in any of the above embodiments.

The non-volatile computer-readable storage medium of computer-executable instructions according to the embodiment of the present application, when the computer-executable instructions are executed by one or more processors, cause the processor to execute any one of the above-mentioned embodiments. The model calibration method.

Additional aspects and advantages of the application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the application.

Description of the drawings

The above and/or additional aspects and advantages of the present application will become apparent and readily understood from the description of the embodiments in conjunction with the following drawings, in which:

Figure 1 is a schematic flow chart of the model correction method in the embodiment of the present application;

Figure 2 is a schematic structural diagram of the spectroscopic equipment in the embodiment of the present application;

Figure 3 is a schematic diagram of the model correction method in the embodiment of the present application to obtain spectral data at multiple focusing distances;

Figure 4 is a diagram showing the results of the prediction model established at the focus distance of 152.5cm by the model correction method in the embodiment of the present application, respectively predicting different focus distance data before correction;

Figure 5 is a graph showing the relationship between the focusing distance and error of one of the samples in Figure 4;

Figure 6 is a schematic flow chart of the model correction method in the embodiment of the present application;

Figure 7 is a schematic flow chart of the model correction method in the embodiment of the present application;

Figure 8 is a schematic flow chart of the model correction method in the embodiment of the present application;

Figure 9 is a diagram showing the results of predicting different focus distance data after correcting the prediction model established by the model correction method in the embodiment of the present application at a focus distance of 152.5 cm.

Description of main component symbols and drawings:

Spectroscopic Instruments 100;

Sample stage 10, laser 20, spectrometer 30, beam expander 40, reflector 50, focusing lens 60, collection lens 70, optical fiber 80, processor 31.

Detailed ways

Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The embodiments described below with reference to the drawings are exemplary and are only used to explain the present application and cannot be understood as limiting the present application.

In the description of this application, it needs to be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", " The directions or positional relationships indicated by "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc. are based on the directions shown in the accompanying drawings or positional relationship is only for the convenience of describing the present application and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present application. In addition, the terms “first” and “second” are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity of indicated technical features. Thus, features defined as “first” and “second” may explicitly or implicitly include one or more of the described features. In the description of this application, "plurality" means two or more than two, unless otherwise explicitly and specifically limited.

In this application, unless otherwise explicitly stated and limited, the term "above" or "below" a first feature on a second feature may include direct contact between the first and second features, or may also include the first and second features. Not in direct contact but through additional characteristic contact between them. Furthermore, the terms "above", "above" and "above" a first feature on a second feature include the first feature being directly above and diagonally above the second feature, or simply mean that the first feature is higher in level than the second feature. “Below”, “under” and “under” the first feature is the second feature includes the first feature being directly below and diagonally below the second feature, or simply means that the first feature is less horizontally than the second feature.

The following disclosure provides many different embodiments or examples for implementing the various structures of the present application. To simplify the disclosure of the present application, the components and arrangements of specific examples are described below. Of course, they are merely examples and are not intended to limit the application. Furthermore, this application may repeat reference numbers and/or reference letters in different examples, such repetition being for the purposes of simplicity and clarity and does not by itself indicate a relationship between the various embodiments and/or arrangements discussed. In addition, this application provides examples of various specific processes and materials, but one of ordinary skill in the art will recognize the application of other processes and/or the use of other materials.

Please refer to Figure 1. The model correction method in the embodiment of this application includes:

Step S10: Based on laser-induced breakdown spectroscopy technology, obtain original spectral data at multiple focusing distances;

Step S20: Establish a prediction model using the original spectral data at one of the focusing distances;

Step S30: Use the prediction model to predict the predicted spectral data of the remaining focusing distances;

Step S40: Calibrate the prediction model of multiple focus distances based on the error between the predicted spectral data and the original spectral data.

Please refer to Figure 2. The spectroscopic equipment 100 in the embodiment of the present application includes a sample stage 10, a laser 20 and a spectrometer 30. The sample stage 10 is used to carry the sample, the laser 20 is used to emit laser light to the sample, and the spectrometer 30 is used to receive the laser light reflected by the sample. , the spectrometer 30 includes a processor 31, which is used to implement the above model correction method. In other words, the processor 31 is used to obtain original spectral data of multiple focusing distances based on laser-induced breakdown spectroscopy technology; and is used to establish a prediction model using the original spectral data of one focusing distance; and is used to use the prediction model to predict Predicted spectral data for the remaining focus distances; and prediction model correction for multiple focus distances based on errors between the predicted spectral data and the original spectral data.

The spectrometer 30 of the spectroscopic device 100 in the embodiment of the present application has a processor 31 that can implement the model correction method. When the spectroscopic device 100 is measured at different measurement distances, the model correction method can be used to collect effective spectra without recalibration, saving manpower and Material resources and time.

Specifically, the spectroscopy device 100 can be a device for spectrum analysis such as a laser-induced breakdown spectroscopy device. The laser 20 of the spectrometry device 100 can emit laser light. The laser can be a high-energy pulse laser. The laser light emitted by the laser 20 can pass through a beam expander. 40 and then hit the reflective mirror 50, which can reflect the laser to the sample stage 10. A focusing lens 60 can be provided between the sample stage 10 and the reflector 50. The laser light reflected to the sample stage 10 can be scattered after contacting the sample on the sample stage 10. The laser light scattered by the sample stage 10 can be The collection lens 70 collects, and then the collected laser light can be transmitted to the spectrometer 30 through the optical fiber 80 and analyzed by the processor 31, so that the spectroscopic device 100 can collect the original spectral data of the sample.

In the model correction method and the spectroscopic device 100 of the embodiment of the present application, the prediction model migration correction of the laser-induced breakdown spectroscopy device under different focusing distances can be realized, thereby improving the accuracy and stability of the prediction model. Laser-induced breakdown spectroscopy equipment can quickly adapt to changes in on-site test conditions, solving the problem of prediction model failure caused by changes in focus distance due to changes in the position of detected materials at the application site. There is no need to recalibrate, saving manpower, material resources and time.

Specifically, the laser-induced breakdown spectroscopy (LIBS) technology based on the model calibration method is a composition analysis technology based on the one-to-one correspondence between the wavelengths of atomic spectra and ion spectra and specific elements, and the spectral signal intensity is related to the corresponding element. The content also has a certain quantitative relationship. Plasma is formed by focusing the surface of the sample with a high-energy pulse laser. A spectrometer is used to record the spectral information emitted by the plasma. By analyzing the plasma spectrum, the elements in the sample are analyzed based on the position of the characteristic wavelength and the spectral intensity. Qualitative and quantitative analysis.

First, the model calibration method takes step S10, which can use LIBS technology equipment to acquire the original spectral data of the sample at multiple focusing distances.

The sample may be a metal sample, such as steel, alloy, etc., the number of samples may be multiple, and multiple samples may have multiple samples of the same element type. Multiple focusing distances may be the distance at which the optical system of the LIBS technology device focuses on the sample. For example, as shown in Figure 3, equipment using LIBS technology acquires raw spectral data from the sample at two focusing distances. When the sample is located at a, the focusing distance from the sample to the optical system of the LIBS technology equipment is l ₁ , and for The original spectral data is obtained from the sample at position a with a focusing distance of l ₁ , and then the sample can be placed at position b, and then the original spectral data is obtained from the sample at position b with a focusing distance of l ₂ .

Then, the model correction method takes step S20 to establish a prediction model using the original spectral data of one of the multiple focusing distances obtained in step S10.

Then, the model correction method takes step S30. The prediction model established in step S20 can predict the spectral data of the remaining focusing distances among the plurality of focusing distances based on the original spectral data of one of the focusing distances.

Then, the model correction method takes step S40 to correct the prediction model established in step S20 based on the error between the predicted spectral data of multiple focusing distances in step S30 and the original spectral data of multiple focusing distances in step S10.

In summary, the embodiment of using LIBS technology equipment to collect spectral data of samples at a certain focusing distance will be further explained:

The sample is carbon steel, the sample quantity is 10 pieces, and the sample number is 0 to 9. There are 11 focusing distances between 152.5cm and 376.5cm.

First, as in step S10, LIBS technology equipment can be used to obtain raw spectral data of 11 carbon steel samples at 11 focusing distances between 152.5cm and 376.5cm.

Then in step S20, the original spectral data of 10 carbon steel samples collected at a focusing distance of 152.5 cm can be used to establish a prediction model that can predict the spectral data.

Then in step S30, the predicted spectral data of 10 pieces of carbon steel with 11 focusing distances between 152.5cm and 376.5cm can be predicted according to the prediction model, and further the quantitative analysis data of the concentration of manganese (Mn) element in the carbon steel can be obtained. The results are shown in Figure 4.

Then in step S40, calculate the error between the predicted value and the original value of the 11 focusing distances of each piece of carbon steel at the focusing distance of 152.5cm to 376.5cm. As shown in Figure 5, one piece of carbon steel is at the focusing distance of 152.5cm. The error between the predicted value and the original value of Mn element concentration for 11 focusing distances to 376.5cm. The prediction model established on the original spectral data of 10 carbon steel samples collected at a focusing distance of 152.5cm can be corrected based on the calculated error.

Referring to Figure 6, in some embodiments, based on the error between the predicted spectral data and the original spectral data, the prediction model correction (step S40) for multiple focusing distances includes:

Step S41: Establish a correction function between the original spectral data and the predicted spectral data;

Step S42: Based on the correction function, establish the relationship between the loss function and the correction function;

Step S43: Based on the relationship, determine the value of the correction function;

Step S44: Use the determined values of the correction function to correct the prediction models of multiple focus distances.

In some embodiments, the processor 31 is configured to establish a correction function of the original spectral data and the predicted spectral data; and to establish a relationship between the loss function and the correction function based on the correction function; and to determine the correction based on the relationship. a numerical value of the function; and a predictive model correction for a plurality of focus distances using the determined numerical value of the correction function.

In this way, the functional relationship is used to establish the correction function, the loss function, and the relationship between the correction function and the loss function to correct the prediction model and improve the accuracy and stability of the prediction model.

Specifically, after obtaining the error data between the predicted spectral data and the original spectral data in the model correction method, the prediction model for multiple focusing distances can be corrected.

In order to implement step S40, the model correction method can first take step S41, using the error between the original spectrum data and the predicted spectrum data to establish a correction function of the original spectrum data and the predicted spectrum data, and then take step S42, based on the correction function established in step S41. , establish the relationship between the loss function and the correction function. The loss function can be a fitting function of the focusing distance and error established using the principle function and construct a loss function. The error is the error data between the original spectral data and the predicted spectral data.

Then step S43 can be taken, and the specific value of the correction function can be obtained based on the relationship between the loss function and the correction function established in step S42. The specific value of the correction function can be obtained by solving the functional relationship between the loss function and the correction function. Then step S44 is taken, and the specific values of the correction function obtained in step S43 are brought into corresponding prediction models of multiple focus distances to correct the prediction model.

Referring to Figure 7, in some embodiments, establishing a correction function between original spectral data and predicted spectral data (step S41) includes:

Step S411: Determine the original concentration of the element based on the original spectral data;

Step S412: Determine the predicted concentration of the element based on the predicted spectrum data;

Step S413: Establish a correction function based on the original concentration and the predicted concentration.

In certain embodiments, the processor 31 is configured to determine an original concentration of the element based on the original spectral data; and to determine a predicted concentration of the element based on the predicted spectral data; and to establish a correction function based on the original concentration and the predicted concentration.

In this way, using the relationship between spectral data and element concentration, the correction function established based on the original spectral data and predicted spectral data is converted into a correction function established based on the original concentration and predicted concentration, thereby improving the accuracy and stability of the prediction model.

Specifically, in order to implement step S41, the model calibration method may first take step S411 to determine the element in the sample detected at the corresponding focus distance based on the relationship between the original spectral data obtained in step S10 and a certain element of the sample detected at the corresponding focus distance. The original concentration of an element. Step S412 may then be taken to determine the predicted concentration of the element in the sample detected at the corresponding focus distance based on the relationship between the predicted spectrum data predicted in step S30 and a certain element in the sample detected at the corresponding focus distance. Then step S413 may be taken to establish a correction function corresponding to the focus distance based on the original concentration in step S411 and the predicted concentration in step S412.

In some embodiments, the correction function takes the following relationship:

C＝Z(l)c

In this way, the predicted concentration of an element is based on the relationship between the original concentration of the element and the correction function. The correction function can predict the predicted concentration of elements at other focusing distances through the original concentration of the element, realizing the migration correction of the prediction model at different focusing distances. .

Specifically, the relational expression of the correction function determines the relationship between the source, spectral data and element concentration. LIBS technology is based on the one-to-one correspondence between the wavelengths of atomic spectra and ion spectra and specific elements, and mainly involves spectral information, high-energy pulse lasers , plasma, characteristic wavelength and other data.

The relational expression used in the correction function can be derived from the specific formula:

Since the laser pulse power density formula is

In the formula, PD is the laser power density, E is the pulse energy, w is the pulse width, and d is the focused spot diameter;

The plasma temperature T is related to the pulse power density PD, then the plasma temperature can be expressed as T = f (E, w, d), where the pulse energy is E = g (l), which is brought into the plasma absolute intensity formula

in the middle, then

In the formula, w is the pulse width, d is the focused spot diameter, i represents the i energy level, j represents the j energy level, A _ij is the transition probability, g _i is the high energy level degeneracy, λ _ij is the radiation wavelength, and U is the current temperature. The matching function corresponding to the ion is the following, E _i is the high energy level energy, K _B is the Boltzmann constant, where F is the scaling factor related to system parameters, plasma temperature, element characteristics, etc., C is the original concentration of the element, I _ij is the element characteristic;

The current measurement parameters are consistent. When the influence of self-absorption is ignored, each coefficient before c in the above formula can be recorded as a constant a, so the formula can be rewritten as:

I＝aC

C＝I/a

Among them, I is the characteristic of the element, and C is the original concentration of the element;

However, the current experimental conditions have changes in focusing distance, so the formula is rewritten as

C＝Z(l)c

Referring to Figure 8, in some embodiments, based on the relationship, determining the value of the correction function (step S43) includes:

Step S431: Use the least squares method to determine the minimum value of the loss function;

Step S432: Use the minimum value to determine the value of the correction function.

In some embodiments, the processor 31 is configured to determine a minimum value of the loss function using a least squares method; and is configured to determine a value of the correction function using the minimum value.

In this way, when determining the minimum value of the loss function, the least squares method can be used to solve the minimum value of the loss function, thereby determining the value of the correction function.

Specifically, in order to implement step S43, the model correction method can first take step S431 based on the loss function constructed in step S42, bring the loss function into the data and use the least squares method to obtain the minimum value of the loss function, and then take the step S431. Step S432: Since there is a relationship between the loss function and the correction function, the specific value of the correction function can be further obtained by solving the minimum value of the loss function through the least squares method.

In some implementations, the loss function adopts the following relationship:

In this way, the loss function relationship includes the original concentration and predicted element concentration of multiple groups of samples at multiple focusing distances, so that the loss function relationship can achieve data fitting.

Specifically, it can be understood that the relational expression of the loss function can be a functional relational expression constructed from the square of the difference between the predicted concentration of the element obtained according to the correlation correction function of the prediction model and the original concentration of the element. The loss function can be solved by using the least squares method. By making the loss function reach the minimum value, the loss function can be solved.

For example, when the number of samples is 10 and the number of focusing distances obtained by the LIBS technology equipment for the samples is 11, the samples can be numbered from 1 to 10, and the different focusing distances can be numbered from 1 to 11. Therefore, the loss The function can take

formula, and then substitute the original concentration of the element, the predicted concentration of the element and other relevant data of the sample corresponding to the sample number in the loss function at the corresponding focusing distance number into the loss function, and solve the minimum value of the loss function through the least squares method. The specific values of the corresponding correction functions are obtained, and then the prediction models of the multiple focus distances can be corrected through the determined values of the correction functions of the multiple focus distances.

Referring to Figure 1, in some embodiments, the model correction method includes:

Step S50: Use the corrected prediction model to predict the elements in the sample at the corresponding focusing distance.

In some embodiments, the processor 31 is configured to use the corrected prediction model to predict elements in the sample at the corresponding focusing distance.

In this way, the prediction model corrected by the model calibration method is more accurate in predicting the spectral data and element concentration of the sample than the prediction model before correction.

Specifically, the model correction method can use step S50 to correct the prediction model. The prediction model corresponding to the focus distance established in step S20 is corrected based on the specific value of the correction function obtained in step S40. The corrected prediction model can be By predicting the elements in the sample at multiple focusing distances, more accurate element measurement data of the sample can be obtained.

The LIBS technology equipment in the above embodiment is still used to collect spectral data of samples at a certain focusing distance to further explain:

In step S30, the prediction model predicts the predicted spectrum data of 10 pieces of carbon steel at 11 focusing distances between 152.5cm and 376.5cm, and obtains the quantitative analysis data of the concentration of manganese (Mn) element in the carbon steel. The result is: Figure 4.

In step S50, the elements in the sample at the corresponding focusing distance are predicted according to the corrected prediction model, and 10 pieces of carbon steel at 11 focusing distances between 152.5cm and 376.5cm are predicted according to the corrected prediction model. The predicted spectral data were corrected to obtain the corrected quantitative analysis data of the concentration of manganese (Mn) element in carbon steel. The results are shown in Figure 9.

It can be understood that compared with Figure 4, the predicted concentration difference of each numbered sample predicted by the prediction model in Figure 9 at different focusing distances is greatly reduced after correction. Therefore, the model correction method can achieve different focusing distances. Calibrate the prediction model below.

The non-volatile computer-readable storage medium of computer-executable instructions in the embodiment of the present application, when the computer-executable instructions are executed by one or more processors, causes the processor to execute the model correction method in any of the above-mentioned embodiments.

Specifically, the computer device in the embodiment of the present application can be a calculator, a programmable controller, a desktop computer, a laptop computer, a tablet computer, a server, and other devices. The computer device can include a processor, a memory, and a computer connected through a system bus. Communication interfaces, displays and input devices.

The processor of the computer device may be a central processing unit (Central Processing Unit, CPU). The processor can also be other general-purpose processors, Digital Signal Processor (DSP), Application Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA) or other Chips such as programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or combinations of these types of chips.

The computer program can be stored in the memory. The memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer executable programs and modules, as described in the method in the above method embodiment. Corresponding program instructions/modules. The processor executes various functional applications and data processing of the processor by running non-transient software programs, instructions and modules stored in the memory, that is, implementing the method in the above method embodiment.

Any process or method descriptions in flowcharts or otherwise described herein may be understood to represent modules, segments, or portions of code that include one or more executable instructions for implementing the specified logical functions or steps of the process. , and the scope of the preferred embodiments of the present application includes additional implementations in which functions may be performed out of the order shown or discussed, including in a substantially simultaneous manner or in the reverse order, depending on the functionality involved, which shall It should be understood by those skilled in the technical field to which the embodiments of this application belong.

The logic and/or steps represented in the flowcharts or otherwise described herein, for example, may be considered a sequenced list of executable instructions for implementing the logical functions, and may be embodied in any computer-readable medium, For use by, or in combination with, instruction execution systems, devices or equipment (such as computer-based systems, systems including processing modules, or other systems that can fetch instructions from and execute instructions from the instruction execution system, device or equipment) or equipment. For the purposes of this specification, a "computer-readable medium" may be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer readable media include the following: electrical connections with one or more wires (electronic device), portable computer disk cartridges (magnetic device), random access memory (RAM), Read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium may even be paper or other suitable medium on which the program may be printed, as the paper or other medium may be optically scanned, for example, and subsequently edited, interpreted, or otherwise suitable as necessary. process to obtain the program electronically and then store it in computer memory.

The processor can be a Central Processing Unit (CPU), other general-purpose processors, Digital Signal Processor (DSP), Application Specific Integrated Circuit (ASIC), or off-the-shelf programmable processors. Gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc.

It should be understood that various parts of the embodiments of the present application can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if it is implemented in hardware, as in another embodiment, it can be implemented by any one or a combination of the following technologies known in the art: a logic gate circuit with a logic gate circuit for implementing a logic function on a data signal. Discrete logic circuits, application specific integrated circuits with suitable combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), etc.

Those of ordinary skill in the art can understand that all or part of the steps involved in implementing the methods of the above embodiments can be completed by instructing relevant hardware through a program. The program can be stored in a computer-readable storage medium. The program can be stored in a computer-readable storage medium. When executed, one of the steps of the method embodiment or a combination thereof is included.

In addition, each functional unit in various embodiments of the present application can be integrated into a processing module, each unit can exist physically alone, or two or more units can be integrated into one module. The above integrated modules can be implemented in the form of hardware or software function modules. If the integrated module is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

The storage media mentioned above can be read-only memory, magnetic disks or optical disks, etc.

Although the embodiments of the present application have been shown and described above, it can be understood that the above-mentioned embodiments are illustrative and cannot be understood as limitations of the present application. Those of ordinary skill in the art can make modifications to the above-mentioned embodiments within the scope of the present application. The embodiments are subject to changes, modifications, substitutions and variations.

Claims

A model correction method, characterized by including:

Based on laser-induced breakdown spectroscopy technology, original spectral data at multiple focusing distances are obtained;

Build a predictive model using raw spectral data at one of the focusing distances;

Using the prediction model, predict the predicted spectral data for the remaining focus distances;

Predictive model corrections for multiple focus distances are based on errors between the predicted spectral data and the original spectral data.
The model correction method according to claim 1, characterized in that, based on the error between the predicted spectral data and the original spectral data, the prediction model correction for multiple focusing distances includes:

Establish a correction function between original spectral data and predicted spectral data;

Based on the correction function, establish a relationship between the loss function and the correction function;

Based on the relationship, determining a value of the correction function;

The prediction model of multiple focus distances is corrected using the determined value of the correction function.
The model correction method according to claim 2, characterized in that establishing a correction function between original spectral data and predicted spectral data includes:

determining an original concentration of an element based on the original spectral data;

determining a predicted concentration of an element based on the predicted spectral data;

The correction function is established based on the original concentration and the predicted concentration.
The model correction method according to claim 3, characterized in that the correction function adopts the following relationship:

C＝Z(l)c

Among them, c is the predicted concentration of the element at the focusing distance l of the prediction model, C is the original concentration of the element, and Z(l) is the corresponding correction function at the focusing distance l.
The model correction method according to claim 2, characterized in that, based on the relationship, determining the value of the correction function includes:

Determine the minimum value of the loss function using the least squares method;

The minimum value is used to determine the value of the correction function.
The model correction method according to claim 2, characterized in that the loss function adopts the following relationship:

Among them, J is the loss function, i is the sample number, j is the focusing distance number, M is the number of samples, N is the number of focusing distances, C i is the original concentration of the element of the i-th sample, Z(l j ) is the j-th The correction function corresponding to the focusing distance l j , c j i is the predicted concentration of the element at the jth focusing distance l j of the i-th sample of the prediction model.
The model correction method according to claim 1, characterized in that the model correction method includes:

The corrected prediction model is used to predict the elements in the sample at the corresponding focusing distance.
A kind of spectroscopic equipment, characterized in that the spectroscopic equipment includes:

Sample stage, used to carry samples;

a laser for emitting laser light to the sample;

A spectrometer, the spectrometer is used to receive the laser light reflected by the sample, the spectrometer includes a processor, the processor is used to implement the model correction method according to any one of claims 1-7.
A computer device, characterized in that the computer device includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, it implements the method described in any one of claims 1-7 Model calibration method.
A non-volatile computer-readable storage medium of computer-executable instructions, characterized in that when the computer-executable instructions are executed by one or more processors, the processor is caused to execute claims 1-7 Any one of the model correction methods.