Search

CN-121978083-A - LIBS and PLS-based aluminum alloy component detection method and system

CN121978083ACN 121978083 ACN121978083 ACN 121978083ACN-121978083-A

Abstract

The invention discloses an aluminum alloy component detection method and system based on LIBS and PLS, and belongs to the technical field of metal material component analysis. The method comprises the steps of 1, carrying out deoxidization leveling treatment on an aluminum alloy sample, focusing on the surface of the aluminum alloy sample by utilizing an aluminum alloy component detection system to obtain an LIBS spectrum of the aluminum alloy sample, 2, carrying out abnormal spectrum rejection on the LIBS spectrum of the aluminum alloy sample by a Laida method, carrying out background interference removal and spectrum normalization by adopting a window translation minimum method and channel background intensity normalization, extracting characteristic peaks of normalized spectrums strongly related to components of the aluminum alloy sample, 3, constructing and training an aluminum alloy component prediction model by adopting PLS, and 4, inputting the characteristic peaks of the normalized spectrums of the aluminum alloy sample into the trained aluminum alloy component prediction model to obtain component prediction results of the aluminum alloy sample. Compared with the prior art, the invention realizes nondestructive rapid high-precision second-level detection of the aluminum alloy components.

Inventors

  • LIU RUIBIN
  • Ding Caihao
  • CHEN GUANNAN
  • YANG XIAONING
  • SUN HAOHAN
  • LIU XIAODONG
  • GAO HAN

Assignees

  • 北京理工大学

Dates

Publication Date
20260505
Application Date
20251205

Claims (7)

  1. 1. An aluminum alloy component detection method based on LIBS and PLS is characterized by comprising the following steps, Step 1, deoxidizing and leveling an aluminum alloy sample, and focusing on the surface of the aluminum alloy sample by utilizing an aluminum alloy component detection system to obtain an LIBS spectrum of the aluminum alloy sample; Step 2, removing background interference and normalizing the spectrum by adopting a window translation minimum value method and sub-channel background intensity normalization after removing abnormal spectrum of LIBS spectrum of the aluminum alloy sample by a Laida method, and extracting characteristic peaks of normalized spectrum which are strongly related to components of the aluminum alloy sample; step 3, constructing and training an aluminum alloy component prediction model by adopting PLS; Step 3.1, dividing a training set, a verification set and a test set according to the proportion of the aluminum alloy sample; Step 3.2, taking a characteristic peak of a normalized spectrum of the aluminum alloy sample as an independent variable, and detecting a component true value of a national standard method by utilizing an aluminum alloy component as an independent variable; 3.3, constructing an aluminum alloy component prediction model by adopting a multiple linear regression relation of PLS of independent variables and dependent variables; Training an aluminum alloy component prediction model; Step 3.4.1, determining the main component number of the aluminum alloy component prediction model by using ten-fold cross validation; 3.4.2, setting iteration rounds, and obtaining the minimum value of root mean square error of the verification set in a cyclic iteration mode until reaching the iteration rounds to obtain a trained aluminum alloy component prediction model; And 4, inputting characteristic peaks of the normalized spectrum of the aluminum alloy sample into a trained aluminum alloy composition prediction model to obtain a composition prediction result of the aluminum alloy sample.
  2. 2. The method for detecting the components of the aluminum alloy based on LIBS and PLS according to claim 1, wherein the implementation method of the step 1 is as follows, Step 1.1, removing an oxide layer from an aluminum alloy sample and flattening the surface of the sample; and 1.2, utilizing an aluminum alloy component detection system to emit nanosecond pulse laser to focus on the surface of an aluminum alloy sample to generate plasma, and further obtaining LIBS spectrum.
  3. 3. The method for detecting the components of the aluminum alloy based on LIBS and PLS according to claim 1, wherein the implementation method of the step 2 is as follows, Step 2.1, obtaining the average spectrum intensity and standard deviation of the LIBS spectrum, carrying out abnormal spectrum elimination on the LIBS spectrum by adopting a Laida method, reserving the LIBS spectrum meeting the requirement shown in the formula (1), and removing the average value of the reserved spectrum to obtain the average spectrum; [ average spectral intensity.+ -. 3. Standard deviation ] (1) Step 2.2, removing the background interference of the sub-channels from the background curve obtained by using a window translation minimum method in the corresponding channel in the average spectrum; step 2.3, normalizing the spectrum from which the background interference is removed by using a multichannel background intensity normalization method; and 2.4, extracting characteristic peaks of the normalized spectrum of the aluminum alloy sample by utilizing a characteristic spectral line library of common elements of the aluminum alloy.
  4. 4. The method for detecting aluminum alloy components based on LIBS and PLS according to claim 3, wherein the step 2.2 is realized by the following steps, 2.2.1, Obtaining a background minimum value of a reserved spectrum by adopting a window translation minimum value method, and performing polynomial fitting on the extracted background minimum value to form a background curve; And 2.2.2, subtracting the background curve of the corresponding channel from the average spectrum to remove the background interference of the sub-channels.
  5. 5. The method for detecting aluminum alloy components based on LIBS and PLS according to claim 3, wherein the step 2.2 is realized by the following steps, Step 2.3.1, integrating the intensity of the background curve of each wavelength channel; Step 2.3.2, dividing the average spectrum intensity by the integral value of the background curve intensity to obtain a normalized spectrum.
  6. 6. A LIBS and PLS-based aluminum alloy composition rapid detection system for realizing the method as claimed in claim 1 is characterized by comprising a LIBS device and a calculation host; The LIBS device is used for collecting laser-induced plasma spectrums of aluminum alloy samples and unknown aluminum alloy samples to be tested, and obtaining the LIBS spectrums of the aluminum alloy samples with the wavelength range of 180-950nm and the resolution of not more than 0.05nm by focusing 1064nm nanosecond pulse laser with the emission energy range of 0-200mJ on the surfaces of the samples; the calculation host is used for obtaining the component prediction result of the aluminum alloy sample.
  7. 7. A rapid detection system for constituents of aluminum alloy based on LIBS and PLS according to claim 6, wherein the repetition frequency of the LIBS device is 5-15Hz.

Description

LIBS and PLS-based aluminum alloy component detection method and system Technical Field The invention relates to an aluminum alloy component detection method and system based on LIBS and PLS, belongs to the technical field of metal material component analysis, and is suitable for detecting the content of an aluminum alloy matrix and an alloy element in industrial sites and laboratory scenes. Background The aluminum alloy is widely applied to the fields of aerospace, automobiles, buildings and the like due to the characteristics of low density, high strength, corrosion resistance and the like, and the material performance and the application scene are directly determined by the influence of the components such as Cu content on the strength and the influence of the Mg content on the corrosion resistance. The existing aluminum alloy component detection method has the defects that 1, the detection period is long, the traditional method needs to carry out pretreatment such as digestion and dilution on samples, the whole process takes 1-4 hours, the rapid quality control or instant analysis requirement cannot be met, 2, the sample destructiveness is that part of the method (such as a chemical analysis method) needs to consume more samples (5-10 g), the samples cannot be reused after being processed and are not suitable for precious samples (such as failure analysis samples and research and development samples), 3, the scene limitation is that large-scale instruments such as ICP-MS (inductively coupled plasma-mass spectrometry) only can stably run in a laboratory and cannot be suitable for dust and vibration environments of an industrial field, the traditional LIBS detection is mainly used for modeling a single aluminum alloy series, the adaptability to different series of aluminum alloys is poor, the error rate is often more than 5%, 4, the operation is complicated, the professional is required to carry out sample pretreatment and instrument operation, the labor cost is high, and the automatic detection is difficult to realize. Therefore, how to realize nondestructive, rapid and high-precision second-level detection of aluminum alloy components in industrial and laboratory scenes has become a problem to be solved. Disclosure of Invention The invention aims to solve the technical problems of long period, destructiveness, scene limitation and poor adaptability of the existing aluminum alloy component detection, and provides an aluminum alloy component detection method and system based on LIBS and PLS, wherein the method and system are used for realizing nondestructive and high-precision second-level detection of different series of aluminum alloy components in industrial fields and laboratory scenes by adopting a Laser-induced breakdown spectroscopy (LIBS, laser-Induced Breakdown Spectro scopy) to acquire the modeling prediction of an emission spectrum of an aluminum alloy combined with a partial least Square method (PLS, partial Least Square s). The invention aims at realizing the following technical scheme: the invention discloses an aluminum alloy component detection method based on LIBS and PLS, which comprises the following steps: step 1, deoxidizing and leveling an aluminum alloy sample, and focusing on the surface of the aluminum alloy sample by utilizing an aluminum alloy component detection system to obtain an LIBS spectrum of the aluminum alloy sample; Step 1.1, removing an oxide layer from an aluminum alloy sample and flattening the surface of the sample; step 1.2, utilizing an aluminum alloy component detection system to emit nanosecond pulse laser to focus on the surface of an aluminum alloy sample to generate plasma so as to obtain LIBS spectrum; Step 2, removing background interference and normalizing the spectrum by adopting a window translation minimum value method and sub-channel background intensity normalization after removing abnormal spectrum of LIBS spectrum of the aluminum alloy sample by a Laida method, and extracting characteristic peaks of normalized spectrum which are strongly related to components of the aluminum alloy sample; Step 2.1, obtaining the average spectrum intensity and standard deviation of the LIBS spectrum, carrying out abnormal spectrum elimination on the LIBS spectrum by adopting a Laida method, reserving the LIBS spectrum meeting the requirement shown in the formula (1), and removing the average value of the reserved spectrum to obtain the average spectrum; [ average spectral intensity.+ -. 3. Standard deviation ] (1) Step 2.2, removing the background interference of the sub-channels from the background curve obtained by using a window translation minimum method in the corresponding channel in the average spectrum; 2.2.1, obtaining a background minimum value of a reserved spectrum by adopting a window translation minimum value method, and performing polynomial fitting on the extracted background minimum value to form a background curve; step 2.2.2, subtracting the backg