WO2023236387A1 - 元素信息的预测方法、装置、设备及介质 - Google Patents

元素信息的预测方法、装置、设备及介质 Download PDF

Info

Publication number
WO2023236387A1
WO2023236387A1 PCT/CN2022/119121 CN2022119121W WO2023236387A1 WO 2023236387 A1 WO2023236387 A1 WO 2023236387A1 CN 2022119121 W CN2022119121 W CN 2022119121W WO 2023236387 A1 WO2023236387 A1 WO 2023236387A1
Authority
WO
WIPO (PCT)
Prior art keywords
spectral
spectral intensities
spectra
element information
spectral intensity
Prior art date
Application number
PCT/CN2022/119121
Other languages
English (en)
French (fr)
Inventor
潘从元
王彬
张兵
Original Assignee
合肥金星智控科技股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 合肥金星智控科技股份有限公司 filed Critical 合肥金星智控科技股份有限公司
Publication of WO2023236387A1 publication Critical patent/WO2023236387A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/62Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light
    • G01N21/71Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light thermally excited
    • G01N21/718Laser microanalysis, i.e. with formation of sample plasma
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/24Classification techniques
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02PCLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30Computing systems specially adapted for manufacturing

Definitions

  • the present application relates to the technical field of metallurgical material composition detection, and in particular to a method, device, equipment and medium for predicting element information.
  • composition of metallurgical materials is one of the core parameters of metallurgical process control, and its detection timeliness is of great significance for optimizing process parameters and energy conservation and emission reduction.
  • detection of metallurgical material components is mostly done through offline detection after manual sampling and sample preparation, which requires sample preprocessing such as grinding, polishing or dissolving. The detection is time-consuming and has serious lag.
  • This application aims to solve at least one of the technical problems existing in the prior art. To this end, this application proposes a method, device, equipment and medium for predicting element information.
  • a method for predicting element information including:
  • each original characteristic spectrum at least includes spectral intensities at multiple locations of the metallurgical material
  • the spectral intensities of multiple positions in the standard spectrum are used as columns and the multiple standard spectra are used as rows to form a matrix.
  • the matrix is calculated through the first preset algorithm to obtain a preliminary set of spectral intensities of the multiple positions. ;
  • the preliminary set is screened based on shape distance analysis to obtain an effective set of spectral intensities.
  • the effective set of spectral intensities is calculated through a second preset algorithm to obtain reconstructed and normalized spectral intensity data.
  • the reconstructed normalized The spectral intensity data at least include the area, maximum value, kurtosis, and skewness of each peak;
  • the reconstructed and standardized spectral intensity data is input into an element prediction model to obtain element information in the metallurgical material, wherein the element prediction model uses the reconstructed and standardized spectral intensity data of the metallurgical material sample as a data sample,
  • the element information of the metallurgical material sample is trained as a label, wherein the element information includes the element type and the content of each element;
  • the preliminary set is screened based on shape distance analysis to obtain an effective set of spectral intensities.
  • the specific steps are as follows:
  • X c represents the peak center
  • A is the peak area
  • represents the half width and height
  • x represents the spectral intensity
  • y max represents the peak value
  • y min represents the valley value
  • x max represents the maximum value of the wavelength fitting this interval
  • x min represents the minimum value of the wavelength fitting this interval
  • D3 Obtain the number of vector sums greater than or equal to 0 and the number less than 0 in the preliminary set, determine the size between the number of vector sums greater than or equal to 0 and the number less than 0, and filter out the ones with the largest number as multiple positions.
  • a second preliminary set of spectral intensities If the vector sum is greater than or equal to 0, it is an upward trend. If the vector sum is less than 0, it is a downward trend.
  • the second preliminary set of spectral intensities at the multiple locations is The spectral intensity of each position in the set is compared with ⁇ T. If it is greater than or equal to ⁇ T, it is recorded as the second vector 1. If it is less than ⁇ T, it is recorded as the second vector -1. All second vectors are added.
  • a device for predicting element information is proposed.
  • the device for predicting element information is configured to perform the prediction method of element information described in the embodiment of the first aspect of the present application.
  • an electronic device including a processor and a memory.
  • the memory stores at least one instruction, at least a program, a code set or an instruction set.
  • the instruction, the program, The code set or the instruction set is loaded and executed by the processor to implement the prediction method of element information described in the embodiment of the first aspect of the present application.
  • An embodiment according to the fourth aspect of the present application provides a non-transitory computer-readable storage medium, which when instructions in the storage medium are executed by a processor of a mobile terminal, enables the mobile terminal to execute the first aspect of the present application.
  • the technical solutions provided by the embodiments of the present application can include the following beneficial effects: detecting the components of metallurgical materials through element prediction models, avoiding the need for manual grinding, polishing or dissolving operations, greatly improving the detection efficiency and accuracy.
  • FIG. 1 is a flowchart of a method for predicting element information according to an exemplary embodiment.
  • FIG. 2 is a block diagram of a device for predicting element information according to an exemplary embodiment.
  • FIG. 3 is an internal structure diagram of an electronic device according to an exemplary embodiment.
  • FIG. 4 is a schematic matrix diagram according to an exemplary embodiment.
  • Figure 5 is a schematic diagram of a second matrix according to an exemplary embodiment.
  • FIG. 6 is a schematic diagram of a second matrix with vectors according to an exemplary embodiment.
  • Figure 7 is a schematic diagram showing a preliminary collection of spectral intensities at multiple locations according to an exemplary embodiment.
  • Figure 8 is a second preliminary assembly schematic diagram according to an exemplary embodiment.
  • FIG. 9 is a schematic diagram illustrating an effective set of spectral intensities according to an exemplary embodiment.
  • Figure 1 is a flow chart of a method for predicting element information according to an exemplary embodiment. As shown in Figure 1, it includes the following steps:
  • step S101 multiple original characteristic spectra of the metallurgical material are obtained.
  • the multiple original characteristic spectra are obtained through multiple measurements.
  • Each original characteristic spectrum at least includes spectral intensities of multiple locations of the metallurgical material.
  • multiple original characteristic spectra of the metallurgical material are obtained.
  • the multiple original characteristic spectra are obtained through multiple measurements.
  • Each original characteristic spectrum at least includes spectral intensities of multiple locations of the metallurgical material. Specifically, include:
  • the original characteristic spectrum of the metallurgical material is obtained based on LIBS equipment, and the multiple original characteristic spectra are obtained through multiple measurements, where each original characteristic spectrum at least includes spectral intensities of multiple locations of the metallurgical material.
  • LIBS equipment is used to detect metallurgical materials using laser-induced breakdown spectroscopy to obtain the original characteristic spectrum of metallurgical materials.
  • step S102 the plurality of original characteristic spectra are standardized to obtain a plurality of standard spectra.
  • the multiple original characteristic spectra are standardized to obtain multiple standard spectra, including:
  • the plurality of original characteristic spectra are subjected to sequential processing of background subtraction, peak search, spectrum similarity screening at peaks, and bias correction at peaks to obtain the plurality of standard spectra.
  • B 1 is an adjustable parameter
  • ⁇ E is the channel width (energy step)
  • I(j) is the measured spectrum
  • C represents the fitting constant term
  • j and i represent different initial values of the number of spectral recording times
  • X i and Y i are the spectral intensity measured for different times respectively, and The average value of the measured spectral intensity, n represents the number of measurements;
  • step S103 the spectral intensities of multiple positions in the standard spectrum are used as columns and the multiple standard spectra are used as rows to form a matrix.
  • the matrix is calculated through the first preset algorithm to obtain the intensity of the multiple positions. Preliminary collection of spectral intensities.
  • a matrix is formed by using the spectral intensities at multiple locations in each standard spectrum as columns and multiple standard spectra as rows. , the matrix is calculated through the first preset algorithm, thereby obtaining a preliminary set of spectral intensities at multiple locations.
  • the spectral intensities of multiple positions in the standard spectrum are used as columns and the multiple standard spectra are used as rows to form a matrix.
  • the matrix is calculated through a first preset algorithm to obtain the multiple A preliminary collection of spectral intensities at locations, specifically including the following steps:
  • D1 use the spectral intensities of multiple positions in the standard spectrum as columns and the multiple standard spectra as rows to form a matrix
  • D3 subtract cyclically by subtracting the first row from the second row and subtracting the second row from the third row. If it is greater than or equal to 0, it is recorded as vector 1. If it is less than 0, it is recorded as vector -1;
  • D4 add the vectors at the same position in each row. If it is greater than or equal to 0, filter out the spectral intensity with a vector of 1 at that position. Add and average it to obtain the spectral intensity at that position. If it is less than 0, filter out that position. The spectral intensities with vectors of -1 are summed and averaged to obtain the spectral intensity at that location, thereby obtaining a preliminary set of spectral intensities at the multiple locations.
  • the spectral intensities of multiple positions in the standard spectrum are used as columns and multiple standard spectra are used as rows to form a matrix.
  • a row of 0 is added above the matrix to form a second matrix, to Subtract the first row from the second row, and subtract the second row from the third row. If it is greater than or equal to 0, it is recorded as vector 1. If it is less than 0, it is recorded as vector -1, as shown in Figure 6. Add up the vectors at the same position in each row. If it is greater than or equal to 0, filter out the spectral intensity of the position vector of 1 and average it to obtain the spectral intensity of the position. If it is less than 0, filter out the position vector of 1.
  • the spectral intensities of -1 are summed and averaged to obtain the spectral intensity at that position, thereby obtaining a preliminary set of spectral intensities at the multiple positions, as shown in Figure 7.
  • step S104 the preliminary set is screened based on shape distance analysis to obtain an effective set of spectral intensities, and the effective set of spectral intensities is calculated through a second preset algorithm to obtain reconstructed and standardized spectral intensity data.
  • the reconstructed normalized spectral intensity data at least includes the area, maximum value, kurtosis, and skewness of each peak.
  • the preliminary set is screened based on shape distance analysis to obtain an effective set of spectral intensities, and the effective set of spectral intensities is calculated through a second preset algorithm to obtain the reconstructed and standardized Spectral intensity data.
  • the preliminary set is screened based on shape distance analysis to obtain an effective set of spectral intensities, with the following steps:
  • X c represents the peak center
  • A is the peak area
  • represents the half width and height
  • x represents the spectral intensity
  • y max represents the peak value
  • y min represents the valley value
  • x max represents the maximum value of the wavelength fitting this interval
  • x min represents the minimum value of the wavelength fitting this interval
  • D3 Obtain the number of vector sums greater than or equal to 0 and the number less than 0 in the preliminary set, determine the size between the number of vector sums greater than or equal to 0 and the number less than 0, and filter out the ones with the largest number as multiple positions.
  • a second preliminary set of spectral intensities If the vector sum is greater than or equal to 0, it is an upward trend. If the vector sum is less than 0, it is a downward trend.
  • the second preliminary set of spectral intensities at the multiple locations is The spectral intensity of each position in the set is compared with ⁇ T. If it is greater than or equal to ⁇ T, it is recorded as the second vector 1. If it is less than ⁇ T, it is recorded as the second vector -1. All second vectors are added.
  • the first step is to obtain the number of vector sums greater than or equal to 0 and the number less than 0 in the preliminary set, and determine the size between the number of vector sums greater than or equal to 0 and the number less than 0. , screening out a large number of second preliminary sets as spectral intensities at multiple locations.
  • the screened out second preliminary set is shown in Figure 8.
  • the effective set of spectral intensity is shown in Figure 9 shown. At this time, the effective set of spectral intensity is an accelerated upward trend.
  • a second preset algorithm is used to calculate the effective set of spectral intensities to obtain reconstructed and normalized spectral intensity data.
  • the reconstructed and normalized spectral intensity data at least include the area of each peak, the maximum value, kurtosis, and skewness, specifically including:
  • (x 1 , y 1 ), (x 2 , y 2 ), (x 3 , y 3 ), (x 4 , y 4 ) are the four maximum continuous points of a peak;
  • step S105 the reconstructed and standardized spectral intensity data is input into an element prediction model to obtain element information in the metallurgical material, wherein the element prediction model is based on the reconstructed and standardized spectral intensity data of the metallurgical material sample.
  • the element information of the metallurgical material sample is trained as a label, where the element information includes the element type and the content of each element.
  • the reconstructed and standardized spectral intensity data is input into the element prediction model to obtain element information in the metallurgical material.
  • the element information includes element types and the content of each element.
  • the element prediction model specifically includes:
  • Each original characteristic spectrum at least includes spectral intensities at multiple locations of the metallurgical material sample;
  • the spectral intensities of multiple positions in the standard spectrum are used as columns and the multiple standard spectra are used as rows to form a matrix.
  • the matrix is calculated through a first preset algorithm to obtain a preliminary estimate of the spectral intensities of the multiple positions. gather;
  • the preliminary set is screened based on shape distance analysis to obtain an effective set of spectral intensities.
  • the effective set of spectral intensities is calculated through a second preset algorithm to obtain reconstructed and normalized spectral intensity data.
  • the reconstructed normalized The spectral intensity data at least include the area, maximum value, kurtosis, and skewness of each peak;
  • y represents the corresponding spectral intensity
  • x represents the corresponding spectral wavelength
  • t represents the number of repeated measurements.
  • regression fitting is performed, thereby obtaining a preliminary model, and the reconstructed and standardized spectral intensity data The fit corrects the reconstructed normalized spectral intensity data in the preliminary model, and finally obtains the element prediction model.
  • FIG. 2 is a block diagram of an apparatus for predicting element information according to an exemplary embodiment.
  • the device includes an acquisition module 201, a processing module 202, a first calculation module 203, a second calculation module 204 and a prediction module 205.
  • the acquisition module 201 acquires multiple original characteristic spectra of metallurgical materials.
  • the multiple original characteristic spectra are obtained through multiple measurements.
  • Each original characteristic spectrum at least includes spectral intensities at multiple locations of the metallurgical material;
  • the processing module 202 performs standardization processing on the plurality of original characteristic spectra to obtain a plurality of standard spectra;
  • the first calculation module 203 uses the spectral intensities of multiple positions in the standard spectrum as columns and multiple standard spectra as rows to form a matrix, calculates the matrix through a first preset algorithm, and obtains the multiple A preliminary collection of spectral intensities at locations;
  • the second calculation module 204 screens the preliminary set based on shape distance analysis to obtain a valid set of spectral intensities, calculates the valid set of spectral intensities through a second preset algorithm, and obtains reconstructed standardized spectral intensity data.
  • the reconstructed normalized spectral intensity data at least includes the area, maximum value, kurtosis, and skewness of each peak;
  • the prediction module 205 inputs the reconstructed and standardized spectral intensity data into an element prediction model to obtain element information in the metallurgical material, wherein the element prediction model is based on the reconstructed and standardized spectral intensity data of the metallurgical material sample.
  • the element information of the metallurgical material sample is trained as a label, where the element information includes the element type and the content of each element.
  • an electronic device is provided.
  • the electronic device may be a terminal, and its internal structure diagram may be as shown in FIG. 3 .
  • the electronic device includes a processor, memory, communication interface, display screen and input device connected through a system bus.
  • the processor of the electronic device is used to provide computing and control capabilities.
  • the memory of the electronic device includes non-volatile storage media and internal memory.
  • the non-volatile storage medium stores operating systems and computer programs. This internal memory provides an environment for the execution of operating systems and computer programs in non-volatile storage media.
  • the communication interface of the computer device is used for wired or wireless communication with external terminals.
  • the wireless mode can be implemented through WIFI, operator network, near field communication (NFC) or other technologies.
  • the computer program when executed by the processor, implements a prediction method of element information.
  • the display screen of the computer device may be a liquid crystal display or an electronic ink display.
  • the input device of the computer device may be a touch layer covered on the display screen, or may be a button, trackball or touch pad provided on the computer device shell. , it can also be an external keyboard, trackpad or mouse, etc.
  • FIG. 3 is only a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer equipment to which the solution of the present application is applied.
  • the specific computer equipment can May include more or fewer parts than shown, or combine certain parts, or have a different arrangement of parts.
  • the element information prediction device provided by the present application can be implemented in the form of a computer program, and the computer program can be run on the electronic device as shown in Figure 3.
  • the memory of the electronic device may store various program modules that constitute the prediction device of the element information, such as the acquisition module 201, the processing module 202, the first calculation module 203, the second calculation module 204 and the prediction module 205 shown in Figure 2.
  • the computer program composed of each program module causes the processor to execute the steps in a method for predicting element information in various embodiments of the present application described in this specification.
  • the electronic device shown in Figure 3 can obtain multiple original characteristic spectra of metallurgical materials through the acquisition module 201 of the element information prediction device shown in Figure 2.
  • the multiple original characteristic spectra are obtained through multiple measurements.
  • each original characteristic spectrum at least includes spectral intensities at multiple locations of the metallurgical material;
  • the processing module 202 performs standardization processing on the multiple original characteristic spectra to obtain multiple standard spectra;
  • the first calculation module 203 calculates the The spectral intensities of multiple positions in the standard spectrum are used as columns, and the multiple standard spectra are used as rows to form a matrix.
  • the matrix is calculated through a first preset algorithm to obtain a preliminary set of spectral intensities of the multiple positions;
  • the second calculation module 204 screens the preliminary set based on shape distance analysis to obtain an effective set of spectral intensities, calculates the effective set of spectral intensities through a second preset algorithm, and obtains reconstructed and standardized spectral intensity data,
  • the reconstructed standardized spectral intensity data at least includes the area, maximum value, kurtosis, and skewness of each peak;
  • the prediction module 205 inputs the reconstructed standardized spectral intensity data into the element prediction model to obtain the metallurgical Element information in materials, wherein the element prediction model is trained using the reconstructed and standardized spectral intensity data of metallurgical material samples as data samples and the element information of the metallurgical material samples as labels, where the element information includes Types of elements and the amount of each element.
  • the electronic device provided by this application can open each module in the element information prediction device through the memory and processor, and detect the composition of metallurgical materials through the element prediction model, avoiding the need for manual grinding, polishing or dissolution, etc., which is extremely Greatly improve the efficiency and accuracy of detection.
  • a computer-readable storage medium is provided with a computer program stored thereon.
  • the computer program is executed by a processor, the following steps are implemented: acquiring a plurality of original characteristic spectra of metallurgical materials, the plurality of original characteristics The spectrum is obtained through multiple measurements, and each original characteristic spectrum at least includes the spectral intensity of multiple locations of the metallurgical material; the multiple original characteristic spectra are standardized to obtain multiple standard spectra; the standard spectrum is The spectral intensities of multiple positions are used as columns, and the multiple standard spectra are used as rows to form a matrix.
  • the matrix is calculated through the first preset algorithm to obtain a preliminary set of spectral intensities of the multiple positions; based on shape distance analysis, the The preliminary set is screened to obtain a valid set of spectral intensities.
  • the valid set of spectral intensities is calculated through a second preset algorithm to obtain reconstructed and standardized spectral intensity data.
  • the reconstructed and standardized spectral intensity data is at least Including the area, maximum value, kurtosis, and skewness of each peak; input the reconstructed standardized spectral intensity data into the element prediction model to obtain the element information in the metallurgical material, wherein the element prediction model is
  • the reconstructed and standardized spectral intensity data of the metallurgical material sample is used as the data sample, and the element information of the metallurgical material sample is used as the label for training, where the element information includes the element type and the content of each element.
  • the non-transitory computer-readable storage medium provided by this application can execute one of the element information prediction methods in the above embodiments through instructions in the storage medium, and detect the composition of metallurgical materials through the element prediction model, avoiding the need for manual labor. Carrying out operations such as grinding, polishing or dissolving greatly improves the efficiency and accuracy of detection.
  • Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc.
  • Volatile memory may include random access memory (Random Access Memory, RAM) or external cache memory.
  • RAM Random Access Memory
  • SRAM Static Random Access Memory
  • DRAM Dynamic Random Access Memory

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Evolutionary Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Optics & Photonics (AREA)
  • Plasma & Fusion (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Investigating Or Analysing Materials By Optical Means (AREA)
  • Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)

Abstract

一种元素信息的预测方法、装置、设备及介质,包括:获取冶金物料的多个原始特征光谱,对多个原始特征光谱进行标准化处理得到多个标准光谱(S102);将标准光谱中多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵,通过第一预设算法对矩阵进行计算,获取多个位置的光谱强度的初步集合(S103);基于形状距离分析对初步集合进行筛选,获取光谱强度的有效集合,通过第二预设算法对光谱强度的有效集合进行计算,获取重构标准化的光谱强度数据;将重构标准化的光谱强度数据输入元素预测模型中,获取冶金物料中的元素信息。通过元素预测模型来预测冶金物料中元素信息的稳定性和准确性,达到工厂选取对应的最优处理方法的目的。

Description

元素信息的预测方法、装置、设备及介质
本申请要求于2022年06月10日提交中国专利局、申请号为202210649577.X,申请名称为“元素信息的预测方法、装置、设备及介质”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及冶金物料成分检测技术领域,尤其涉及一种元素信息的预测方法、装置、设备及介质。
背景技术
冶金物料的成分是冶金过程控制的核心参数之一,其检测时效性对优化工艺参数和节能减排等具有重要意义。目前,冶金物料成分检测多通过人工取样制样后的离线检测,需要研磨、抛光或溶解等进行样品预处理,检测耗时,具有严重的滞后性。
申请内容
本申请旨在至少解决现有技术中存在的技术问题之一。为此,本申请提出一种元素信息的预测方法、装置、设备及介质。
根据本申请实施例的第一方面,提供一种元素信息的预测方法,包括:
获取冶金物料的多个原始特征光谱,所述多个原始特征光谱是经过多次测量得到的,每个原始特征光谱至少包括所述冶金物料多个位置的光谱强度;
对所述多个原始特征光谱进行标准化处理得到多个标准光谱;
将所述标准光谱中多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵,通过第一预设算法对所述矩阵进行计算,获取所述多个位置的光谱强度的初步集合;
基于形状距离分析对所述初步集合进行筛选,获取光谱强度的有效集合,通过第二预设算法对所述光谱强度的有效集合进行计算,获取重构标准化的光谱强度数据,所述重构标准化的光谱强度数据至少包括每个峰的面积、最大值、峰度、偏度;
将所述重构标准化的光谱强度数据输入元素预测模型中,获取所述冶金物料中的元素信息,其中,所述元素预测模型是以冶金物料样品的重构标准化的光谱强度数据作为数据样本、所述冶金物料样品的元素信息作为标签训练的,其中,所述元素信息包括元素种类以及每种元素的含量;
其中,基于形状距离分析对所述初步集合进行筛选,获取光谱强度的有效集合,具体如下步骤:
D1,计算所述初步集合的平均数,并用洛伦茨函数拟合;
Figure PCTCN2022119121-appb-000001
其中X c表示峰中心,A为峰面积,ω表示半宽高,x表示光谱强度;
D2,通过如下分类界限公式获取界限值ΔT,
Figure PCTCN2022119121-appb-000002
其中,y max表示峰值,y min表示谷值,x max表示拟合这段区间的波长最大值,x min表示拟合这段区间的波长最小值;
D3,获取所述初步集合中的向量和大于等于0的数量及小于0的数量,判断向量和大于等于0的数量与小于0的数量之间的大小,筛选出数量多的作为多个位置的光谱强度的第二初步集合,其中,若向量和大于等于0的数量多,则为上升趋势,若小于0的数量多,则为下降趋势,将所述多个位置的光谱强度的第二初步集合中的各个位置的光谱强度与ΔT进行比较,若大于等于ΔT,记为第二向量1,若小于ΔT,记为第二向量-1,将所有第二向量进行加和,若大于等于0,则为加速状态,筛选出大于等于ΔT的光谱强度,形成所述光谱强度的有效集合,若小于0,则为平速状态,筛选出小于ΔT的光谱强度,形成所述光谱强度的有效集合。
根据本申请的第二方面的实施例提出一种元素信息的预测装置,所述元素信息的预测装置用于执行本申请的第一方面的实施例所述的元素信息的预测方法。
根据本申请的第三方面的实施例提出一种电子设备,包括处理器和存储器,所述存储器中存储有至少一条指令、至少一段程序、代码集或指令集,所述指令、所述程序、所述代码集或所述指令集由所述处理器加载并执行以实现本申请的第一方面的实施例所述的元素信息的预测方法。
根据本申请的第四方面的实施例提出一种非临时性计算机可读存储介质,当所述存储介质中的指令由移动终端的处理器执行时,使得移动终端能够执行本申请的第一方面的实施例所述的元素信息的预测方法。
本申请的实施例提供的技术方案可以包括以下有益效果:通过元素预测模型来对冶金 物料的成分进行检测,避免了需要人工进行研磨、抛光或溶解等操作,极大的提高了检测的效率及准确性。
本申请的附加方面和优点将在下面的描述中部分给出,部分将从下面的描述中变得明显,或通过本申请的实践了解到。
附图说明
本申请的上述和/或附加的方面和优点从结合下面附图对实施例的描述中将变得明显和容易理解,其中:
图1是根据一示例性实施例示出的一种元素信息的预测方法的流程图。
图2是根据一示例性实施例示出的一种元素信息的预测装置的框图。
图3是根据一示例性实施例示出的一种电子设备的内部结构图。
图4是根据一示例性实施例示出的一种矩阵示意图。
图5是根据一示例性实施例示出的一种第二矩阵示意图。
图6是根据一示例性实施例示出的一种带向量的第二矩阵示意图。
图7是根据一示例性实施例示出的一种多个位置光谱强度的初步集合示意图。
图8是根据一示例性实施例示出的一种第二初步集合示意图。
图9是根据一示例性实施例示出的一种光谱强度的有效集合示意图。
具体实施方式
这里将详细地对示例性实施例进行说明,其示例表示在附图中。下面的描述涉及附图时,除非另有表示,不同附图中的相同数字表示相同或相似的要素。以下示例性实施例中所描述的实施方式并不代表与本发明相一致的所有实施方式。相反,它们仅是与如所附权利要求书中所详述的、本发明的一些方面相一致的装置和方法的例子。
图1是根据一示例性实施例示出的一种元素信息的预测方法的流程图,如图1所示,包括以下步骤:
在步骤S101中,获取冶金物料的多个原始特征光谱,所述多个原始特征光谱是经过多次测量得到的,每个原始特征光谱至少包括所述冶金物料多个位置的光谱强度。
通过对需要待检冶金物料进行多次测量,从而获取待检冶金物料的多个原始特征光谱,而每次测量待检冶金物料时,获取的是待检冶金物料多个位置的光谱强度。
在一些实施例中,获取冶金物料的多个原始特征光谱,所述多个原始特征光谱是经过多次测量得到的,每个原始特征光谱至少包括所述冶金物料多个位置的光谱强度,具体包括:
基于LIBS设备获取所述冶金物料的原始特征光谱,多次测量获取所述多个原始特征光谱,其中,每个原始特征光谱至少包括所述冶金物料多个位置的光谱强度。
采用LIBS设备,利用激光诱导击穿光谱向冶金物料进行检测,从而获取冶金物料的原始特征光谱。
在步骤S102中,对所述多个原始特征光谱进行标准化处理得到多个标准光谱。
由于多个原始特征光谱存在一些不合理的数据,因此通过对多个原始特征光谱进行标准化处理得到多个标准光谱。
在一些实施例中,对所述多个原始特征光谱进行标准化处理得到多个标准光谱,包括:
对所述多个原始特征光谱进行扣背景、寻峰、峰处光谱相似性筛查、峰处偏正矫正顺序处理获取所述多个标准光谱。
为了对原始特征光谱进行矫正,通过对原始光谱进行扣背景、寻峰、峰处光谱相似性筛查、峰处偏正矫正顺序处理获取多个标准光谱。其中,
扣背景公式:
Figure PCTCN2022119121-appb-000003
其中,B 1为可调参数,ΔE为道宽(能量步长),I(j)为测量谱,C表示拟合常数项,j和i分别代表不同的采集光谱记录次数初始值;
相关系数公式:
Figure PCTCN2022119121-appb-000004
其中,X i和Y i分别为测量不同次数的光谱强度,
Figure PCTCN2022119121-appb-000005
Figure PCTCN2022119121-appb-000006
该次测量光谱强度的平均值,n代表测量的次数;
峰偏移矫正公式:
Figure PCTCN2022119121-appb-000007
其中,X c为矫正后的峰中心,A为峰面积,ω为半宽高,x表示光谱强度,A 0表述为拟合的常数参数。
在步骤S103中,将所述标准光谱中多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵,通过第一预设算法对所述矩阵进行计算,获取所述多个位置的光谱强度的 初步集合。
为了将对多个标准光谱进行统一化,获取精准的多个位置的光谱强度的初步集合,因此通过将每个标准光谱中多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵,通过第一预设算法对所述矩阵进行计算,从而获取多个位置的光谱强度的初步集合。
在一些实施例中,将所述标准光谱中的多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵,通过第一预设算法对所述矩阵进行计算,获取所述多个位置的光谱强度的初步集合,具体包括如下步骤:
D1,将所述标准光谱中多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵;
D2,在矩阵上方添加一行0,形成第二矩阵;
D3,以第二行减去第一行,第三行减去第二行的方式循环相减,大于等于0,记为向量1,如果小于0,记为向量-1;
D4,将每行同位置的向量进行加和,若大于等于0,则筛选出该位置向量为1的光谱强度加和求平均,获取该位置的光谱强度,若小于0,则筛选出该位置向量为-1的光谱强度加和求平均,获取该位置的光谱强度,从而获取所述多个位置的光谱强度的初步集合。
如图4所示,将所述标准光谱中多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵,如图5所示,在矩阵上方添加一行0,形成第二矩阵,以第二行减去第一行,第三行减去第二行的方式循环相减,大于等于0,记为向量1,如果小于0,记为向量-1,具体如图6所示。将每行同位置的向量进行加和,若大于等于0,则筛选出该位置向量为1的光谱强度加和求平均,获取该位置的光谱强度,若小于0,则筛选出该位置向量为-1的光谱强度加和求平均,获取该位置的光谱强度,从而获取所述多个位置的光谱强度的初步集合,具体如图7所示。
步骤S104中,基于形状距离分析对所述初步集合进行筛选,获取光谱强度的有效集合,通过第二预设算法对所述光谱强度的有效集合进行计算,获取重构标准化的光谱强度数据,所述重构标准化的光谱强度数据至少包括每个峰的面积、最大值、峰度、偏度。
为了获取更准确的冶金物料的光谱强度,基于形状距离分析对初步集合进行筛选,获取光谱强度的有效集合,通过第二预设算法对所述光谱强度的有效集合进行计算,获取重构标准化的光谱强度数据。
在一些实施例中,基于形状距离分析对所述初步集合进行筛选,获取光谱强度的有效集合,具体如下步骤:
D1,计算所述初步集合的平均数,并用洛伦茨函数拟合;
Figure PCTCN2022119121-appb-000008
其中X c表示峰中心,A为峰面积,ω表示半宽高,x表示光谱强度;
D2,通过如下分类界限公式获取界限值ΔT,
Figure PCTCN2022119121-appb-000009
其中,y max表示峰值,y min表示谷值,x max表示拟合这段区间的波长最大值,x min表示拟合这段区间的波长最小值;
D3,获取所述初步集合中的向量和大于等于0的数量及小于0的数量,判断向量和大于等于0的数量与小于0的数量之间的大小,筛选出数量多的作为多个位置的光谱强度的第二初步集合,其中,若向量和大于等于0的数量多,则为上升趋势,若小于0的数量多,则为下降趋势,将所述多个位置的光谱强度的第二初步集合中的各个位置的光谱强度与ΔT进行比较,若大于等于ΔT,记为第二向量1,若小于ΔT,记为第二向量-1,将所有第二向量进行加和,若大于等于0,则为加速状态,筛选出大于等于ΔT的光谱强度,形成所述光谱强度的有效集合,若小于0,则为平速状态,筛选出小于ΔT的光谱强度,形成所述光谱强度的有效集合。
假设计算获取的界限值ΔT为20,第一步获取所述初步集合中的向量和大于等于0的数量及小于0的数量,判断向量和大于等于0的数量与小于0的数量之间的大小,筛选出数量多的作为多个位置的光谱强度的第二初步集合,以图7为例,则筛选出来的第二初步集合就如图8所示,同时又因为计算获取的界限值ΔT为20,所以位置1的第二向量为-1,位置3和位置4的第二向量为1,将所有向量相加,由于是大于0的,因此再次筛选后,光谱强度的有效集合如图9所示。此时光谱强度的有效集合为加速上升趋势。
在一些实施例中,通过第二预设算法对所述光谱强度的有效集合进行计算,获取重构标准化的光谱强度数据,所述重构标准化的光谱强度数据至少包括每个峰的面积、最大值、峰度、偏度,具体包括:
通过如下公式计算各个峰的面积:
Figure PCTCN2022119121-appb-000010
其中,(x 1,y 1),(x 2,y 2),(x 3,y 3),(x 4,y 4)为一个峰的最大连续的四个点;
通过如下公式计算峰度:
Figure PCTCN2022119121-appb-000011
其中,
Figure PCTCN2022119121-appb-000012
表述光谱均值,s表述光谱平均数,X表示光谱强度;
通过如下公式计算偏度:
Figure PCTCN2022119121-appb-000013
其中,
Figure PCTCN2022119121-appb-000014
表述光谱均值,s表述光谱平均数,X表示光谱强度。
步骤S105中,将所述重构标准化的光谱强度数据输入元素预测模型中,获取所述冶金物料中的元素信息,其中,所述元素预测模型是以冶金物料样品的重构标准化的光谱强度数据作为数据样本、所述冶金物料样品的元素信息作为标签训练的,其中,所述元素信息包括元素种类以及每种元素的含量。
将重构标准化的光谱强度数据输入元素预测模型中,获取所述冶金物料中的元素信息,元素信息包括元素种类以及每种元素的含量。
在一些实施例中,所述元素预测模型,具体包括:
获取已知元素信息冶金物料样品的多个原始特征光谱,所述多个原始特征光谱是经过多次测量得到的,每个原始特征光谱至少包括所述冶金物料样品的多个位置的光谱强度;
对所述多个原始特征光谱进行标准化处理得到多个标准光谱;
将所述标准光谱中的多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵,通过第一预设算法对所述矩阵进行计算,获取所述多个位置的光谱强度的初步集合;
基于形状距离分析对所述初步集合进行筛选,获取光谱强度的有效集合,通过第二预设算法对所述光谱强度的有效集合进行计算,获取重构标准化的光谱强度数据,所述重构标准化的光谱强度数据至少包括每个峰的面积、最大值、峰度、偏度;
对所述重构标准化的光谱强度数据和所述元素信息数据进行回归拟合获取初步模型:
重构标准化的光谱强度数据和所述元素信息数据进行回归拟合公式:
Figure PCTCN2022119121-appb-000015
其中,X i和Y i分别为测量不同次数的光谱强度,X表示光谱强度,Y表示元素信息,β 0表示拟合的常数项,β 1表示拟合的系数,n表示测量的次数;
通过重构标准化的光谱强度数据拟合公式对所述初步模型中的所述重构标准化的光谱强度数据进行修正,获取所述元素预测模型;
重构标准化的光谱强度数据拟合公式:
Figure PCTCN2022119121-appb-000016
其中,y表示对应的光谱强度,x表示对应的光谱波长,t表示重复测量的次数。
具体的,通过已知元素信息冶金物料样品,同时通过获取该已知元素信息冶金物料样品的重构标准化的光谱强度数据,进行回归拟合,从而获取初步模型,经过重构标准化的光谱强度数据拟合对初步模型中的重构标准化的光谱强度数据进行修正,最终获取元素预测模型。
图2是根据一示例性实施例示出的一种元素信息的预测装置框图。参照图2,该装置包括获取模块201,处理模块202、第一计算模块203、第二计算模块204和预测模块205。
获取模块201,获取冶金物料的多个原始特征光谱,所述多个原始特征光谱是经过多次测量得到的,每个原始特征光谱至少包括所述冶金物料多个位置的光谱强度;
处理模块202,对所述多个原始特征光谱进行标准化处理得到多个标准光谱;
第一计算模块203,将所述标准光谱中的多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵,通过第一预设算法对所述矩阵进行计算,获取所述多个位置的光谱强度的初步集合;
第二计算模块204,基于形状距离分析对所述初步集合进行筛选,获取光谱强度的有效集合,通过第二预设算法对所述光谱强度的有效集合进行计算,获取重构标准化的光谱强度数据,所述重构标准化的光谱强度数据至少包括每个峰的面积、最大值、峰度、偏度;
预测模块205,将所述重构标准化的光谱强度数据输入元素预测模型中,获取所述冶金物料中的元素信息,其中,所述元素预测模型是以冶金物料样品的重构标准化的光谱强 度数据作为数据样本、所述冶金物料样品的元素信息作为标签训练的,其中,所述元素信息包括元素种类以及每种元素的含量。
关于上述实施例中的装置,其中各个模块执行操作的具体方式已经在有关该方法的实施例中进行了详细描述,此处将不做详细阐述说明。
在一个实施例中,提供了一种电子设备,该电子设备可以是终端,其内部结构图可以如图3所示。该电子设备包括通过系统总线连接的处理器、存储器、通信接口、显示屏和输入装置。其中,该电子设备的处理器用于提供计算和控制能力。该电子设备的存储器包括非易失性存储介质、内存储器。该非易失性存储介质存储有操作系统和计算机程序。该内存储器为非易失性存储介质中的操作系统和计算机程序的运行提供环境。该计算机设备的通信接口用于与外部的终端进行有线或无线方式的通信,无线方式可通过WIFI、运营商网络、近场通信(NFC)或其他技术实现。该计算机程序被处理器执行时以实现一种元素信息的预测方法。该计算机设备的显示屏可以是液晶显示屏或者电子墨水显示屏,该计算机设备的输入装置可以是显示屏上覆盖的触摸层,也可以是计算机设备外壳上设置的按键、轨迹球或触控板,还可以是外接的键盘、触控板或鼠标等。
本领域技术人员可以理解,图3中示出的结构,仅仅是与本申请方案相关的部分结构的框图,并不构成对本申请方案所应用于其上的计算机设备的限定,具体的计算机设备可以包括比图中所示更多或更少的部件,或者组合某些部件,或者具有不同的部件布置。
在一个实施例中,本申请提供的元素信息的预测装置可以实现为一种计算机程序的形式,计算机程序可在如图3所示的电子设备上运行。电子设备的存储器中可存储组成该元素信息的预测装置的各个程序模块,比如,图2所示的获取模块201,处理模块202、第一计算模块203、第二计算模块204和预测模块205。各个程序模块构成的计算机程序使得处理器执行本说明书中描述的本申请各个实施例的一种元素信息的预测方法中的步骤。
例如,图3所示的电子设备可以通过如图2所示的元素信息的预测装置的获取模块201,获取冶金物料的多个原始特征光谱,所述多个原始特征光谱是经过多次测量得到的,每个原始特征光谱至少包括所述冶金物料多个位置的光谱强度;处理模块202,对所述多个原始特征光谱进行标准化处理得到多个标准光谱;第一计算模块203,将所述标准光谱中的多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵,通过第一预设算法对所述矩阵进行计算,获取所述多个位置的光谱强度的初步集合;第二计算模块204,基于形状距离分析对所述初步集合进行筛选,获取光谱强度的有效集合,通过第二预设算法对所述光谱强度的有效集合进行计算,获取重构标准化的光谱强度数据,所述重构标准化的光谱强度数据至少包括每个峰的面积、最大值、峰度、偏度;预测模块205,将所述重构标准化的光谱强度数据输入元素预测模型中,获取所述冶金物料中的元素信息,其中,所 述元素预测模型是以冶金物料样品的重构标准化的光谱强度数据作为数据样本、所述冶金物料样品的元素信息作为标签训练的,其中,所述元素信息包括元素种类以及每种元素的含量。
本申请提供的电子设备通过存储器和处理器可以打开元素信息的预测装置中的各个模块,通过元素预测模型来对冶金物料的成分进行检测,避免了需要人工进行研磨、抛光或溶解等操作,极大的提高了检测的效率及准确性。
在一个实施例中,提供了一种计算机可读存储介质,其上存储有计算机程序,计算机程序被处理器执行时实现以下步骤:获取冶金物料的多个原始特征光谱,所述多个原始特征光谱是经过多次测量得到的,每个原始特征光谱至少包括所述冶金物料多个位置的光谱强度;对所述多个原始特征光谱进行标准化处理得到多个标准光谱;将所述标准光谱中多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵,通过第一预设算法对所述矩阵进行计算,获取所述多个位置的光谱强度的初步集合;基于形状距离分析对所述初步集合进行筛选,获取光谱强度的有效集合,通过第二预设算法对所述光谱强度的有效集合进行计算,获取重构标准化的光谱强度数据,所述重构标准化的光谱强度数据至少包括每个峰的面积、最大值、峰度、偏度;将所述重构标准化的光谱强度数据输入元素预测模型中,获取所述冶金物料中的元素信息,其中,所述元素预测模型是以冶金物料样品的重构标准化的光谱强度数据作为数据样本、所述冶金物料样品的元素信息作为标签训练的,其中,所述元素信息包括元素种类以及每种元素的含量。
本申请提供的非临时性计算机可读存储介质通过存储介质中的指令可以执行上述实施例中的一种元素信息的预测方法,通过元素预测模型来对冶金物料的成分进行检测,避免了需要人工进行研磨、抛光或溶解等操作,极大的提高了检测的效率及准确性。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程,是可以通过计算机程序来指令相关的硬件来完成的,计算机程序可存储于一非易失性计算机可读取存储介质中,该计算机程序在执行时,可包括如上述各方法的实施例的流程。其中,本申请所提供的各实施例中所使用的对存储器、数据库或其它介质的任何引用,均可包括非易失性和易失性存储器中的至少一种。非易失性存储器可包括只读存储器(Read-Only Memory,ROM)、磁带、软盘、闪存或光存储器等。易失性存储器可包括随机存取存储器(Random Access Memory,RAM)或者外部高速缓冲存储器。作为说明而非局限,RAM以多种形式可得,比如静态随机存取存储器(Static Random Access Memory,SRAM)和动态随机存取存储器(Dynamic Random Access Memory,DRAM)等。
以上实施例的各技术特征可以进行任意的组合,为使描述简洁,未对上述实施例中的各个的技术特征所有可能的组合都进行描述,然而,只要这些技术特征的组合不存在矛盾, 都应当认为是本说明书记载的范围。
以上实施例仅表达了本申请的几种实施方式,其描述较为具体和详细,但并不能因此而理解为对发明专利范围的限制。应当指出的是,对于本领域的普通技术人员来说,在不脱离本申请构思的前提下,还可以做出若干变形和改进,这些都属于本申请的保护范围。因此,本申请专利的保护范围应以所附权利要求为准。

Claims (9)

  1. 一种元素信息的预测方法,其特征在于,包括:
    获取冶金物料的多个原始特征光谱,所述多个原始特征光谱是经过多次测量得到的,每个原始特征光谱至少包括所述冶金物料多个位置的光谱强度;
    对所述多个原始特征光谱进行标准化处理得到多个标准光谱;
    将所述标准光谱中多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵,通过第一预设算法对所述矩阵进行计算,获取所述多个位置的光谱强度的初步集合;
    基于形状距离分析对所述初步集合进行筛选,获取光谱强度的有效集合,通过第二预设算法对所述光谱强度的有效集合进行计算,获取重构标准化的光谱强度数据,所述重构标准化的光谱强度数据至少包括每个峰的面积、最大值、峰度、偏度;
    将所述重构标准化的光谱强度数据输入元素预测模型中,获取所述冶金物料中的元素信息,其中,所述元素预测模型是以冶金物料样品的重构标准化的光谱强度数据作为数据样本、所述冶金物料样品的元素信息作为标签训练的,其中,所述元素信息包括元素种类以及每种元素的含量;
    基于形状距离分析对所述初步集合进行筛选,获取光谱强度的有效集合,具体如下步骤:
    D1,计算所述初步集合的平均数,并用洛伦茨函数拟合;
    Figure PCTCN2022119121-appb-100001
    其中X c表示峰中心,A为峰面积,ω表示半宽高,x表示光谱强度;
    D2,通过如下分类界限公式获取界限值ΔT,
    Figure PCTCN2022119121-appb-100002
    其中,y max表示峰值,y min表示谷值,x max表示拟合这段区间的波长最大值,x min表示拟合这段区间的波长最小值;
    D3,获取所述初步集合中的向量和大于等于0的数量及小于0的数量,判断向量和大 于等于0的数量与小于0的数量之间的大小,筛选出数量多的作为多个位置的光谱强度的第二初步集合,其中,若向量和大于等于0的数量多,则为上升趋势,若小于0的数量多,则为下降趋势,将所述多个位置的光谱强度的第二初步集合中的各个位置的光谱强度与ΔT进行比较,若大于等于ΔT,记为第二向量1,若小于ΔT,记为第二向量-1,将所有第二向量进行加和,若大于等于0,则为加速状态,筛选出大于等于ΔT的光谱强度,形成所述光谱强度的有效集合,若小于0,则为平速状态,筛选出小于ΔT的光谱强度,形成所述光谱强度的有效集合。
  2. 根据权利要求1所述的元素信息的预测方法,其特征在于,获取冶金物料的多个原始特征光谱,所述多个原始特征光谱是经过多次测量得到的,每个原始特征光谱至少包括所述冶金物料多个位置的光谱强度,具体包括:
    基于LIBS设备获取所述冶金物料的原始特征光谱,多次测量获取所述多个原始特征光谱,其中,每个原始特征光谱至少包括所述冶金物料多个位置的光谱强度。
  3. 根据权利要求1-2任一项所述的元素信息的预测方法,其特征在于,对所述多个原始特征光谱进行标准化处理得到多个标准光谱,具体包括:
    对所述多个原始特征光谱进行扣背景、寻峰、峰处光谱相似性筛查、峰处偏正矫正顺序处理获取所述多个标准光谱。
  4. 根据权利要求1-3任一项所述的元素信息的预测方法,其特征在于,将所述标准光谱中的多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵,通过第一预设算法对所述矩阵进行计算,获取所述多个位置的光谱强度的初步集合,具体包括如下步骤:
    D1,将所述标准光谱中多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵;
    D2,在矩阵上方添加一行0,形成第二矩阵;
    D3,以第二行减去第一行,第三行减去第二行的方式循环相减,大于等于0,记为向量1,如果小于0,记为向量-1;
    D4,将每行同位置的向量进行加和,若大于等于0,则筛选出该位置向量为1的光谱强度加和求平均,获取该位置的光谱强度,若小于0,则筛选出该位置向量为-1的光谱强度加和求平均,获取该位置的光谱强度,从而获取所述多个位置的光谱强度的初步集合。
  5. 根据权利要求1-4任一项所述的元素信息的预测方法,其特征在于,通过第二预设算法对所述光谱强度的有效集合进行计算,获取重构标准化的光谱强度数据,所述重构标准化的光谱强度数据至少包括每个峰的面积、最大值、峰度、偏度,具体包括:
    通过如下公式计算各个峰的面积:
    Figure PCTCN2022119121-appb-100003
    其中,(x 1,y 1),(x 2,y 2),(x 3,y 3),(x 4,y 4)为一个峰的最大连续的四个点;
    通过如下公式计算峰度:
    Figure PCTCN2022119121-appb-100004
    其中,
    Figure PCTCN2022119121-appb-100005
    表述光谱均值,s表述光谱平均数,X表示光谱强度;
    通过如下公式计算偏度:
    Figure PCTCN2022119121-appb-100006
    其中,
    Figure PCTCN2022119121-appb-100007
    表述光谱均值,s表述光谱平均数,X表示光谱强度。
  6. 根据权利要求1-5任一项所述的元素信息的预测方法,其特征在于,所述元素预测模型,具体包括:
    获取已知元素信息冶金物料样品的多个原始特征光谱,所述多个原始特征光谱是经过多次测量得到的,每个原始特征光谱至少包括所述冶金物料样品的多个位置的光谱强度;
    对所述多个原始特征光谱进行标准化处理得到多个标准光谱;
    将所述标准光谱中的多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵,通过第一预设算法对所述矩阵进行计算,获取所述多个位置的光谱强度的初步集合;
    基于形状距离分析对所述初步集合进行筛选,获取光谱强度的有效集合,通过第二预设算法对所述光谱强度的有效集合进行计算,获取重构标准化的光谱强度数据,所述重构标准化的光谱强度数据至少包括每个峰的面积、最大值、峰度、偏度;
    对所述重构标准化的光谱强度数据和所述元素信息数据进行回归拟合获取初步模型:
    重构标准化的光谱强度数据和所述元素信息数据进行回归拟合公式:
    Figure PCTCN2022119121-appb-100008
    其中,X i和Y i分别为测量不同次数的光谱强度,X表示光谱强度,Y表示元素信 息,β 0表示拟合的常数项,β 1表示拟合的系数,n表示测量的次数;
    通过重构标准化的光谱强度数据拟合公式对所述初步模型中的所述重构标准化的光谱强度数据进行修正,获取所述元素预测模型;
    重构标准化的光谱强度数据拟合公式:
    Figure PCTCN2022119121-appb-100009
    其中,y表示对应的光谱强度,x表示对应的光谱波长,t表示重复测量的次数。
  7. 一种元素信息的预测装置,其用于执行如权利要求1-6任一项所述的元素信息的预测方法,其特征在于,包括:
    获取模块,获取冶金物料的多个原始特征光谱,所述多个原始特征光谱是经过多次测量得到的,每个原始特征光谱至少包括所述冶金物料多个位置的光谱强度;
    处理模块,对所述多个原始特征光谱进行标准化处理得到多个标准光谱;
    第一计算模块,将所述标准光谱中的多个位置的光谱强度作为列,多个标准光谱作为行,形成矩阵,通过第一预设算法对所述矩阵进行计算,获取所述多个位置的光谱强度的初步集合;
    第二计算模块,基于形状距离分析对所述初步集合进行筛选,获取光谱强度的有效集合,通过第二预设算法对所述光谱强度的有效集合进行计算,获取重构标准化的光谱强度数据,所述重构标准化的光谱强度数据至少包括每个峰的面积、最大值、峰度、偏度;
    预测模块,将所述重构标准化的光谱强度数据输入元素预测模型中,获取所述冶金物料中的元素信息,其中,所述元素预测模型是以冶金物料样品的重构标准化的光谱强度数据作为数据样本、所述冶金物料样品的元素信息作为标签训练的,其中,所述元素信息包括元素种类以及每种元素的含量。
  8. 一种电子设备,其特征在于,包括处理器和存储器,所述存储器中存储有至少一条指令、至少一段程序、代码集或指令集,所述指令、所述程序、所述代码集或所述指令集由所述处理器加载并执行以实现根据权利要求1-6中任一项所述的元素信息的预测方法。
  9. 一种非临时性计算机可读存储介质,其特征在于,当所述存储介质中的指令由移动终端的处理器执行时,使得移动终端能够执行根据权利要求1-6中任一项所述的元素信息的预测方法。
PCT/CN2022/119121 2022-06-10 2022-09-15 元素信息的预测方法、装置、设备及介质 WO2023236387A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210649577.X 2022-06-10
CN202210649577.XA CN114739980B (zh) 2022-06-10 2022-06-10 元素信息的预测方法、装置、设备及介质

Publications (1)

Publication Number Publication Date
WO2023236387A1 true WO2023236387A1 (zh) 2023-12-14

Family

ID=82287156

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/119121 WO2023236387A1 (zh) 2022-06-10 2022-09-15 元素信息的预测方法、装置、设备及介质

Country Status (2)

Country Link
CN (1) CN114739980B (zh)
WO (1) WO2023236387A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118246818A (zh) * 2024-05-28 2024-06-25 东北大学 一种铁矿粉动态综合评价方法、装置、介质以及电子设备

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114739980B (zh) * 2022-06-10 2022-09-06 合肥金星智控科技股份有限公司 元素信息的预测方法、装置、设备及介质
CN118624589A (zh) * 2024-08-02 2024-09-10 中国科学院近代物理研究所 基于libs确定痕量元素浓度的方法和设备

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120029836A1 (en) * 2008-11-06 2012-02-02 Centre National De La Recherche Scientifique - Cnrs System and method for quantitative analysis of the elemental composition of a material by laser-induced breakdown spectroscopy (libs)
US20160349174A1 (en) * 2015-06-01 2016-12-01 Ingrain, Inc. Combined Vibrational Spectroscopy And Laser Induced Breakdown Spectroscopy For Improved Mineralogical And Geochemical Characterization Of Petroleum Source Or Reservoir Rocks
CN110208252A (zh) * 2019-06-30 2019-09-06 华中科技大学 一种基于激光诱导击穿光谱分析的煤灰熔点预测方法
CN111351778A (zh) * 2020-03-27 2020-06-30 合肥金星机电科技发展有限公司 基于libs光谱的成分分析方法
CN111693513A (zh) * 2020-07-27 2020-09-22 中国科学院空天信息创新研究院 一种基于激光诱导击穿光谱的在线检测方法
CN112304924A (zh) * 2020-10-28 2021-02-02 武汉纺织大学 一种利用光谱特征扩充提高激光探针分类精度的方法
CN114002204A (zh) * 2021-10-15 2022-02-01 华中科技大学 一种基于光谱抖动的激光诱导击穿光谱分析方法
CN114235785A (zh) * 2021-11-09 2022-03-25 合肥金星智控科技股份有限公司 一种激光诱导击穿光谱质量确定方法及系统
CN114486821A (zh) * 2022-04-06 2022-05-13 合肥金星智控科技股份有限公司 冶金光谱特征回归方法、装置、电子设备和存储介质
CN114739980A (zh) * 2022-06-10 2022-07-12 合肥金星智控科技股份有限公司 元素信息的预测方法、装置、设备及介质

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102410993B (zh) * 2011-08-01 2014-06-18 清华大学 基于激光诱导等离子体发射光谱标准化的元素测量方法
RU2677417C2 (ru) * 2014-11-25 2019-01-16 Хэллибертон Энерджи Сервисиз, Инк. Получение флюидальных характеристик пористых материалов с использованием спектроскопии возбуждения лазерным пробоем
CN104807787B (zh) * 2015-05-20 2017-05-31 重庆邮电大学 一种基于激光诱导击穿光谱技术的茶叶分类鉴别方法
CN107941758B (zh) * 2017-10-16 2020-08-18 华南理工大学 一种基于光谱预处理的煤质热值测量方法
CN109540941A (zh) * 2018-12-21 2019-03-29 四川聚创石墨烯科技有限公司 一种检测碳材料类型的方法

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120029836A1 (en) * 2008-11-06 2012-02-02 Centre National De La Recherche Scientifique - Cnrs System and method for quantitative analysis of the elemental composition of a material by laser-induced breakdown spectroscopy (libs)
US20160349174A1 (en) * 2015-06-01 2016-12-01 Ingrain, Inc. Combined Vibrational Spectroscopy And Laser Induced Breakdown Spectroscopy For Improved Mineralogical And Geochemical Characterization Of Petroleum Source Or Reservoir Rocks
CN110208252A (zh) * 2019-06-30 2019-09-06 华中科技大学 一种基于激光诱导击穿光谱分析的煤灰熔点预测方法
CN111351778A (zh) * 2020-03-27 2020-06-30 合肥金星机电科技发展有限公司 基于libs光谱的成分分析方法
CN111693513A (zh) * 2020-07-27 2020-09-22 中国科学院空天信息创新研究院 一种基于激光诱导击穿光谱的在线检测方法
CN112304924A (zh) * 2020-10-28 2021-02-02 武汉纺织大学 一种利用光谱特征扩充提高激光探针分类精度的方法
CN114002204A (zh) * 2021-10-15 2022-02-01 华中科技大学 一种基于光谱抖动的激光诱导击穿光谱分析方法
CN114235785A (zh) * 2021-11-09 2022-03-25 合肥金星智控科技股份有限公司 一种激光诱导击穿光谱质量确定方法及系统
CN114486821A (zh) * 2022-04-06 2022-05-13 合肥金星智控科技股份有限公司 冶金光谱特征回归方法、装置、电子设备和存储介质
CN114739980A (zh) * 2022-06-10 2022-07-12 合肥金星智控科技股份有限公司 元素信息的预测方法、装置、设备及介质

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
SUN, LANXIANG; WANG, WEI; ZHANG, PENG; XIN, YONG; QI, LIFENG: "Research Progress of Laser-induced Breakdown Spectroscopy in Metallurgical Online Analysis Application", YEJIN FENXI, YEJIN FENXI BIANJIBU, vol. 41, no. 12, 15 December 2021 (2021-12-15), pages 58 - 67, XP009551110, ISSN: 1000-7571, DOI: 10.13228/j.boyuan.issn1000-7571.011612 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118246818A (zh) * 2024-05-28 2024-06-25 东北大学 一种铁矿粉动态综合评价方法、装置、介质以及电子设备

Also Published As

Publication number Publication date
CN114739980B (zh) 2022-09-06
CN114739980A (zh) 2022-07-12

Similar Documents

Publication Publication Date Title
WO2023236387A1 (zh) 元素信息的预测方法、装置、设备及介质
WO2021000524A1 (zh) 孔位保护门检测方法、装置、计算机设备和存储介质
JP2000162135A (ja) 検査方法、検査システムおよび電子デバイスの製造方法
JP2018529074A5 (zh)
CN117312968A (zh) 一种盐碱农田土壤有机质含量的预测方法
CN114611582B (zh) 一种基于近红外光谱技术分析物质浓度的方法及系统
CN116793979A (zh) 一种水样中化学需氧量的检测方法
US7865322B2 (en) Relative noise
CN110793920B (zh) 一种化学成像与高光谱联用的大气遥测方法
CN111105417A (zh) 影像噪声定位方法及系统
US20220381832A1 (en) Production of a Quality Test System
CN110688934A (zh) 一种空间采样主动学习分类方法、电子设备及存储介质
CN114004147B (zh) 一种土壤湿润状态下同时预测多种土壤属性的方法
CN115909324A (zh) 一种广陈皮陈化年份高光谱鉴别方法及系统
CN111797742B (zh) 一种小麦胚乳蛋白体面积的测定方法
CN111291020A (zh) 基于局部加权线性动态系统的动态过程软测量建模方法
CN114398228A (zh) 一种设备资源使用情况的预测方法、装置及电子设备
CN115931828B (zh) 适应复杂土壤基体的成分分析与预测方法、单元及系统
CN116108368B (zh) 一种基于深度学习混合模型的沉积微相识别方法和装置
CN116754467B (zh) 一种天然橡胶老化性能的测评方法
CN116166969A (zh) 一种基于谱线强度相关性匹配的光谱标准化方法
CN115049025B (zh) 一种基于弹性分段标准化算法的模型迁移方法及系统
CN115510097A (zh) 周期性指标判定方法、装置、设备及计算机可读存储介质
CN117783885B (zh) 一种基于电化学机理的储能电池热失控预测方法和系统
CN117235150A (zh) 面向复杂场景的小样本时序数据外推分析方法及系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22945520

Country of ref document: EP

Kind code of ref document: A1