Search

CN-122016765-A - Coal quality rapid online detection and analysis method based on multi-mode spectroscopy information

CN122016765ACN 122016765 ACN122016765 ACN 122016765ACN-122016765-A

Abstract

The invention discloses a coal quality rapid online detection and analysis method based on multi-mode spectroscopy information, and belongs to the technical field of coal quality detection. The method comprises the steps of 1, synchronously collecting a spectrum, a near infrared spectrum and a sound spectrum signal on the coal flow surface by using a time resolution laser induced plasma spectrum, a diffuse reflection probe, an acousto-optic adjustable filter and a microphone, 2, sequentially carrying out abnormal elimination, background subtraction, normalization and peak position translation complementation on the collected TRLIBS spectrum, carrying out dark background subtraction on the collected near infrared spectrum, extracting sound spectrum characteristics and correcting corrected spectrum data, 3, carrying out multi-mode weighted fusion on an NIRS coal quality parameter predicted value and a TRLIBS coal quality parameter predicted value which are respectively obtained, and obtaining a coal quality industrial index predicted result.

Inventors

  • LIU RUIBIN
  • Ding Caihao
  • CHEN GUANNAN

Assignees

  • 北京理工大学

Dates

Publication Date
20260512
Application Date
20251205

Claims (6)

  1. 1. A coal quality rapid on-line detection and analysis method based on multi-mode spectroscopy information is characterized by comprising the following steps, Step 1, synchronously collecting spectrum, near infrared spectrum and sound spectrum signals on the surface of coal flow by using a time resolution laser induced plasma spectrum, a diffuse reflection probe, an acousto-optic adjustable filter and a microphone respectively; Step 2, carrying out abnormal elimination, background subtraction, normalization and peak position translation complementation on the collected TRLIBS spectrums in sequence, carrying out dark background subtraction on the collected near infrared spectrums, extracting sound spectrum characteristics and correcting corrected spectrum data; Step 3, carrying out multi-mode weighted fusion on the NIRS coal quality parameter predicted value and TRLIBS coal quality parameter predicted value which are respectively obtained to obtain a coal quality industrial index predicted result; extracting a near infrared spectrum characteristic matrix from the near infrared spectrum subtracted from the dark bottom, and obtaining an NIRS coal quality parameter predicted value by adopting a partial least squares regression method; extracting a spectrum characteristic matrix from the corrected spectrum data, and acquiring TRLIBS coal quality parameter predicted values by adopting multiple linear regression; And 3.3, carrying out weighted fusion on the NIRS coal quality parameter predicted value and the TRLIBS coal quality parameter predicted value to obtain a coal quality industrial index predicted result.
  2. 2. The method for rapidly detecting and analyzing coal quality on line based on multi-mode spectroscopy information according to claim 1, wherein the method for realizing step 1 is as follows, Step 1.1, performing spectrum acquisition on the coal flow surface by utilizing time-resolved laser-induced plasma spectrum; Step 1.2, near infrared spectrum acquisition is carried out on the coal flow surface by utilizing a diffuse reflection probe and an acousto-optic adjustable filter; and 1.3, collecting an acoustic wave signal generated by laser-induced plasma by using a microphone.
  3. 3. The method for rapidly detecting and analyzing the coal quality on line based on the multi-mode spectroscopy information according to claim 2, wherein the method for realizing the step 1.1 is as follows, When coal flow is transported to the surface of a laser focusing point through a belt, a laser emits laser with the wavelength of 1064nm, and the laser is focused on the surface of the coal flow through a coaxial zoom system consisting of a concave mirror and a convex mirror after passing through a dichroic mirror to generate laser-induced plasma; step 1.1.2, the wide-wavelength light beam emitted by the laser induced plasma is reflected to a focusing lens by a dichroic mirror after passing through a coaxial zoom system consisting of a concave mirror and a convex mirror; Step 1.1.3 the light beam passing through the focusing lens is focused to an optical fiber and conducted to a spectrometer to form spectral data.
  4. 4. The method for rapidly detecting and analyzing the coal quality on line based on the multi-mode spectroscopy information according to claim 2, wherein the method for realizing the step 1.2 is as follows, Step 1.2.1, utilizing a diffuse reflection probe to emit detection light to irradiate the coal flow surface, and receiving the diffuse reflection light of the coal flow surface by a near infrared spectrometer; And 1.2.2, performing wavelength selection on the diffuse reflection light by utilizing an acousto-optic tunable filter to obtain a near infrared spectrum.
  5. 5. The method for rapidly detecting and analyzing the coal quality on line based on the multi-mode spectroscopy information according to claim 1, wherein the implementation method of the step2 is as follows, Step 2.1, on the preprocessing of spectrum data, sequentially adopting abnormal rejection, background subtraction, normalization and peak position translation completion to obtain corrected spectrum data; Step 2.2, subtracting dark bottom from the near infrared spectrum by subtracting instrument dark current noise from spectrum intensity; And 2.3, extracting positive peak values of the sound spectrum and signal energy as sound spectrum characteristics, and correcting the corrected spectrum data by utilizing the sound spectrum characteristics.
  6. 6. The method for rapidly detecting and analyzing coal quality on line based on multi-mode spectroscopy information according to claim 5, wherein the implementation method of step 2.1 is as follows, Step 2.1.1, eliminating abnormal spectrums of the spectrum data by utilizing a Laida criterion; step 2.1.2, removing the spectral background of the channel by utilizing a window translation minimum method and combining polynomial fitting; Step 2.1.3, carrying out channel-division background intensity normalization on the spectrum after background interference is removed; and 2.1.4, marking characteristic peaks of the normalized spectrum, translating drift peaks, establishing a multiple linear regression relation by using peak columns of correlation coefficients among the characteristic peaks, and complementing peak position translation by using the multiple linear regression relation to form corrected spectrum data.

Description

Coal quality rapid online detection and analysis method based on multi-mode spectroscopy information Technical Field The invention relates to a coal quality rapid online detection and analysis method based on multi-mode spectroscopy information, which belongs to the technical field of coal quality detection and is suitable for rapid and accurate online detection of coal quality industrial indexes. Background Energy is the foundation of social development, china is used as the major country for energy consumption, and mainly depends on fossil energy, especially coal. Coal is a major source of electricity supply in china, and although clean energy is being developed, coal power generation will still play a key role before new energy is replaced. Indexes such as calorific value, ash content, volatile matters and sulfur content of coal are important to the combustion of a power plant boiler, and the traditional artificial coal quality assay method cannot guide production rapidly and in real time. The on-line detection method (such as neutron activation method, gamma ray and X ray radiation method) based on the traditional radiation source has the problem of environmental unfriendly, and the substance detection technology based on the spectral information is green and safe, which is the necessary trend of future industry development. Although the single spectrum technology can detect various elements and indexes in coal in real time, due to the technical limitations of the spectrum detection equipment and detection content at the present stage, the accurate determination requirement of complex coal quality parameters is difficult to meet. Therefore, how to solve the problem that the accuracy of single LIBS for detecting moisture index is insufficient and the spectrum fluctuation is caused by the fluctuation of the coal flow surface in the coal quality detection is urgent to be solved. Disclosure of Invention The invention aims to provide a rapid online detection and analysis method for coal quality based on multi-mode spectroscopy information, aiming at the technical problems of insufficient accuracy of a single LIBS (laser induced breakdown spectroscopy) detection moisture index and spectrum fluctuation caused by coal flow surface fluctuation in coal quality detection. By fusing various kinds of spectroscopy information, spectra, sound spectrum pretreatment and PLS modeling, the accuracy of coal moisture index detection is further improved, and fluctuation of spectrum data is reduced. The invention aims at realizing the following technical scheme: the invention discloses a coal quality rapid online detection and analysis method based on multi-mode spectroscopy information, which comprises the following steps: Step 1, synchronously collecting spectrum, near infrared spectrum and sound spectrum signals on the surface of coal flow by using a time resolution laser induced plasma spectrum, a diffuse reflection probe, an acousto-optic adjustable filter and a microphone respectively; step 1.1, performing spectrum acquisition on the surface of the coal flow by utilizing a Time-Resolved Laser induced plasma spectrum (TRLIBS, time-Resolved Laser-Induced Breakdown Spectroscopy); When coal flow is transported to the surface of a laser focusing point through a belt, a laser emits laser with the wavelength of 1064nm, and the laser is focused on the surface of the coal flow through a coaxial zoom system consisting of a concave mirror and a convex mirror after passing through a dichroic mirror to generate laser-induced plasma; step 1.1.2, the wide-wavelength light beam emitted by the laser induced plasma is reflected to a focusing lens by a dichroic mirror after passing through a coaxial zoom system consisting of a concave mirror and a convex mirror; Step 1.1.3, the light beam passing through the focusing lens is focused to an optical fiber and conducted to a spectrometer to form spectral data; Step 1.2, near infrared spectrum acquisition is carried out on the coal flow surface by utilizing a diffuse reflection probe and an acousto-optic adjustable filter; step 1.2.1, utilizing a diffuse reflection probe to emit detection light to irradiate the coal flow surface, and receiving the diffuse reflection light of the coal flow surface by a near infrared spectrometer; step 1.2.2, performing wavelength selection on the diffuse reflection light by utilizing an acousto-optic tunable filter to obtain a near infrared spectrum; Step 1.3, collecting sound wave signals generated by laser-induced plasmas by utilizing a microphone; Step 2, sequentially carrying out abnormal elimination, background subtraction, normalization and peak position translation complementation on the collected TRLIBS spectrums, carrying out dark background subtraction on the collected Near infrared spectrums (NIRS, near-Infrared Spectroscopy), extracting sound spectrum characteristics and correcting corrected spectrum data; Step 2.1, on the preprocessing of spectrum