A substance identification method and system based on fiber grating biochemical sensing

Abstract

The application is suitable for biochemical sensing technology, and discloses a substance identification method and system based on fiber grating biochemical sensing. The method sequentially collects a tilted grating fiber transmission spectrum, analyzes a harmonic resonance sequence, generates an envelope feature, performs adaptive filtering and noise reduction, extracts a multi-dimensional optical feature, inverts a refractive index with temperature correction, matches a target substance by combining a refractive index with a dispersion relationship curve, quantifies a concentration based on molecular optical gradient, and outputs a standardized detection data package. The identification logic covers spectrum preprocessing, feature extraction, optical parameter inversion, substance matching, and concentration quantification. The method solves the problems of insufficient refractive index measurement accuracy, low substance identification accuracy, and large concentration detection error of traditional schemes, and can simultaneously improve the refractive index solving accuracy, biochemical substance discrimination reliability, and target substance concentration quantification accuracy in a complex biochemical medium environment.

Description

Technical Field

[0001] This application pertains to biochemical sensing technology, and particularly relates to a method and system for substance identification based on fiber Bragg grating biochemical sensing. Background Technology

[0002] Traditional fiber Bragg grating biochemical identification relies on vertical gratings with the cladding removed to collect a single wavelength offset for detection.

[0003] Traditional fiber optic biochemical identification relies solely on a single parameter to roughly calculate the refractive index, making it difficult to distinguish minute differences in refractive index between trace biochemical substances. This results in technical problems such as low accuracy in measuring the refractive index of trace biochemical substances, easy misjudgment of substances, and large errors in concentration quantification. It is impossible to simultaneously complete accurate substance identification and concentration quantification output.

[0004] The information disclosed in the background section is only for enhancing the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] Therefore, embodiments of this application provide a substance identification method and system based on fiber optic grating biochemical sensing, which can solve the problem in related technologies that it is impossible to simultaneously complete accurate substance identification and quantitative concentration output.

[0006] The first aspect of this application provides a method for substance identification based on fiber Bragg grating biochemical sensing, the method comprising: Acquire the initial transmission spectrum data output after the optical fiber containing the tilted grating is transmitted in the medium under test; An extreme value analysis operation is performed on the initial transmission spectrum data to obtain the cladding resonant wavelength sequence and the corresponding transmittance numerical sequence of the cladding resonant wavelength sequence. Based on the cladding resonant wavelength sequence and the transmittance numerical sequence, an envelope generation operation is performed to obtain envelope feature data; Based on the envelope feature data, an adaptive filtering operation is performed to obtain a noise-suppressed envelope data sequence. Based on the envelope data sequence, geometric morphological features and energy attenuation features within a specific wavelength range are extracted to obtain a multidimensional feature parameter set characterizing the optical properties of the medium. Based on the multidimensional feature parameter set, a refractive index inversion operation is performed to obtain the refractive index value of the medium under test at a specific wavelength, and a temperature dependence correction coefficient is calculated based on the refractive index value. Based on the refractive index value and a preset material refractive index database, the target biochemical substance identifier is determined by comparing the refractive index value and the similarity of the dispersion relation curve. The material refractive index database is obtained by offline collection of refractive index values ​​and full-band dispersion relation curve data of various known biochemical substances at standard temperature and standard concentration. Based on the target biochemical substance identifier and the temperature-dependent correction coefficient, a concentration quantization operation is performed to correct the refractive index value and invert the concentration distribution range of the target substance in the test medium, thereby obtaining a detection result data package containing the substance identifier, concentration value, and confidence level.

[0007] Optionally, in a first implementation of the first aspect of this application, the step of performing an extremum analysis operation on the initial transmission spectrum data to obtain the cladding resonant wavelength sequence and the corresponding transmittance numerical sequence includes: Discrete sampling is performed on the initial transmission spectrum data to obtain a spectral data matrix; A sliding window is set to traverse the spectral data matrix, and the first and second derivatives of the transmittance within the window are calculated. When the first derivative changes from a negative value to a positive value after crossing the zero point and the second derivative is greater than a preset threshold, the zero point position is determined to be a candidate extreme point. According to the preset noise floor, the target extreme point is determined among the candidate extreme points, wherein the preset noise floor is set based on the statistical average of the background spectral transmittance collected multiple times by the optical fiber under no-load conditions; The target extreme points are sorted in ascending order of wavelength to obtain the cladding resonant wavelength sequence and the corresponding transmittance numerical sequence.

[0008] Optionally, in a second implementation of the first aspect of this application, the step of performing an envelope generation operation based on the cladding resonant wavelength sequence and the transmittance numerical sequence to obtain envelope feature data includes: Calculate the transmittance change of adjacent wavelength points in the cladding resonant wavelength sequence, remove invalid sampling points whose change is not greater than a preset change threshold, and obtain the retained sampling points. Based on the retained sampling points, curve fitting is performed to obtain a continuous spectral curve; Second derivative analysis is performed on the continuous spectral curve to locate the set of maxima with negative curvature and the set of minima with positive curvature. By fitting the set of maximum points and the set of minimum points, we obtain the set of peak coordinates of the upper envelope and the set of valley coordinates of the lower envelope. The envelope feature data is determined based on the set of peak coordinates of the upper envelope and the set of valley coordinates of the lower envelope.

[0009] Optionally, in the fourth implementation of the first aspect of this application, determining the envelope feature data based on the upper envelope peak coordinate set and the lower envelope valley coordinate set includes: Extract the transmittance peak fluctuation range corresponding to the upper envelope peak coordinate set, and extract the transmittance valley fluctuation range corresponding to the lower envelope valley coordinate set; The set of upper envelope peak coordinates, the set of lower envelope valley coordinates, the transmittance peak fluctuation range, and the transmittance valley fluctuation range are determined as the envelope feature data.

[0010] Optionally, in the fifth implementation of the first aspect of this application, the step of performing a refractive index inversion operation based on the multidimensional feature parameter set to obtain the refractive index value of the medium under test at a specific wavelength includes: Based on the multidimensional feature parameter set, the wavelength offset feature and the peak attenuation coefficient feature are separated. The wavelength offset feature and the peak attenuation coefficient feature are matched with a preset dispersion relation model to invert the refractive index value of the medium under test at a specific wavelength. The model coefficients of the dispersion relation model are obtained by fitting a standard refractive index solution to a tilted fiber grating of the same specification.

[0011] Optionally, in the sixth implementation of the first aspect of this application, the step of calculating the temperature-dependent correction coefficient based on the refractive index value includes: The refractive index value, the wavelength offset feature and the peak attenuation coefficient feature in the multidimensional feature parameter group are matched with the preset full-range calibration database to select the target refractive index temperature calibration curve with the highest matching degree. The full-range calibration database stores the refractive index temperature mapping relationship of different biochemical substances and the temperature and refractive index cross-sensitive calibration parameters of the same specification tilted fiber grating. The partial derivative of the refractive index with temperature is calculated based on the target refractive index temperature calibration curve, and the partial derivative is determined as the temperature-dependent correction coefficient.

[0012] Optionally, in the seventh implementation of the first aspect of this application, the step of determining the target biochemical substance identifier by comparing the refractive index value and the similarity of the dispersion relation curve based on the refractive index value and a preset material refractive index database includes: Based on the refractive index value and the material refractive index database, a refractive index tolerance matching operation is performed to obtain candidate materials that fall within the preset tolerance range of the refractive index value and the pre-stored full-wavelength dispersion relationship curves corresponding to each candidate material in the candidate material set. Based on the pre-stored full-wavelength dispersion relation curve, a curve similarity calculation operation is performed to obtain the matching confidence score of each candidate substance; The candidate substances are sorted in descending order according to the matching confidence score, and the identifier corresponding to the top-ranked target candidate substance is determined as the target biochemical substance identifier.

[0013] Optionally, in the eighth implementation of the first aspect of this application, the step of extracting geometric morphological features and energy attenuation features within a specific wavelength range based on the envelope data sequence to obtain a multidimensional feature parameter set characterizing the optical properties of the medium includes: Based on the envelope data sequence, the slope change rate and curvature distribution extreme values ​​of the envelope curve are extracted within the preset core response wavelength range to obtain geometric morphological features. The core response wavelength range is set to be the range of ±n nanometers of the Bragg wavelength center value of the tilted grating, where n is a constant determined according to the grating tilt angle and demodulation resolution calibration. The integral area value enclosed by the envelope curve and the horizontal axis is calculated for the envelope data sequence within a preset integral wavelength range. The resonance peak parameters in the current envelope data sequence are compared with the corresponding resonance peak parameters of the pre-stored reference spectrum to obtain the peak attenuation coefficient and wavelength shift. The integral area value, the peak attenuation coefficient, and the wavelength shift are used as energy attenuation characteristics. The integral wavelength range is set to the full spectrum range from the start wavelength to the end wavelength of the cladding resonance wavelength sequence. The geometric features and the energy decay features are normalized and integrated to obtain the multidimensional feature parameter set that includes geometric dimension, energy dimension and offset dimension.

[0014] Optionally, in the ninth implementation of the first aspect of this application, the step of performing an adaptive filtering operation based on the envelope feature data to obtain a noise-suppressed envelope data sequence includes: Based on the transmittance peak fluctuation range and transmittance valley fluctuation range contained in the envelope feature data, calculate the overall fluctuation amplitude of the envelope and determine whether the proportion of the overall fluctuation amplitude to the envelope mean exceeds a preset proportion threshold. If the preset ratio threshold is exceeded, the wavelength interval between adjacent resonance peaks is calculated based on the cladding resonance wavelength sequence, and the filter window size is adaptively adjusted to obtain filter window parameters that match the current spectral sampling resolution. Perform a smoothing filter operation on the envelope feature data according to the filter window parameters, and record the wavelength offset and smoothing quantification index before and after filtering to obtain the noise-suppressed envelope data sequence.

[0015] Secondly, embodiments of this application provide a substance identification system based on fiber Bragg grating biochemical sensing, the system comprising: The acquisition module is used to acquire the initial transmission spectrum data output by the optical fiber containing the tilted grating after transmission in the medium under test. The analysis module is used to perform extreme value analysis on the initial transmission spectrum data to obtain the cladding resonant wavelength sequence and the corresponding transmittance value sequence. The generation module is used to perform an envelope generation operation based on the cladding resonant wavelength sequence and the transmittance value sequence to obtain envelope feature data; The filtering module is used to perform adaptive filtering operations based on the envelope feature data to obtain a noise-suppressed envelope data sequence. The extraction module is used to extract geometric morphological features and energy attenuation features within a specific wavelength range based on the envelope data sequence, so as to obtain a multidimensional feature parameter set characterizing the optical properties of the medium. The inversion module is used to perform a refractive index inversion operation based on the multidimensional feature parameter set, obtain the refractive index value of the medium under test at a specific wavelength, and calculate the temperature dependence correction coefficient based on the refractive index value. The comparison module is used to determine the target biochemical substance identifier by comparing the refractive index value and the similarity of the dispersion relation curve based on the refractive index value and the preset material refractive index database. The material refractive index database is obtained by offline collection of refractive index values ​​and full-band dispersion relation curve data of various known biochemical substances at standard temperature and standard concentration. The quantization module is used to perform concentration quantization operations based on the target biochemical substance identifier and the temperature-dependent correction coefficient, so as to correct the refractive index value and invert the concentration distribution range of the target substance in the test medium, and obtain a detection result data package containing the substance identifier, concentration value and confidence level.

[0016] The beneficial effects of this application embodiment compared with traditional technology are as follows: by sequentially acquiring the transmission spectrum of tilted grating fiber, analyzing the resonance sequence, generating envelope features, adaptive filtering and noise reduction, extracting multi-dimensional optical features, inverting the refractive index with temperature correction, matching the target substance by combining the refractive index and dispersion relationship curve, quantifying the concentration based on the molecular optical gradient, and outputting a standardized detection data package, an identification logic covering spectral preprocessing, feature extraction, optical parameter inversion, substance matching, and concentration quantification is generated. This solves the problems of insufficient accuracy in measuring the refractive index of trace biochemical substances, low accuracy in substance identification, and large concentration detection error in traditional schemes. It can simultaneously improve the accuracy of refractive index solution, the reliability of biochemical substance discrimination, and the accuracy of target substance concentration quantification in complex biochemical media environments. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or conventional technology will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of an embodiment of the substance identification method based on fiber optic grating biochemical sensing in this application. Figure 2 This is a schematic diagram of Embodiment 2 of the substance identification method based on fiber Bragg grating biochemical sensing in this application. Figure 3 This is a schematic diagram of Embodiment 3 of the substance identification method based on fiber optic grating biochemical sensing in this application. Figure 4 This is a schematic diagram of Embodiment 4 of the substance identification method based on fiber optic grating biochemical sensing in this application. Figure 5 This is a schematic diagram of Embodiment 5 of the substance identification method based on fiber Bragg grating biochemical sensing in this application. Figure 6 This is a schematic diagram of Embodiment Six of the Substance Identification Method Based on Fiber Bragg Grating Biochemical Sensing in this application; Figure 7 This is a schematic diagram of Embodiment Seven of the Substance Identification Method Based on Fiber Bragg Grating Biochemical Sensing in this application. Figure 8 This is a schematic diagram of Embodiment 8 of the substance identification method based on fiber optic grating biochemical sensing in this application. Figure 9 This is a schematic diagram of Embodiment Nine of the Substance Identification Method Based on Fiber Bragg Grating Biochemical Sensing in this application. Figure 10 This is a schematic diagram of one embodiment of the substance identification system based on fiber optic grating biochemical sensing in this application. Detailed Implementation

[0019] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that this disclosure will be more comprehensive and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art. The described features, structures, or characteristics may be integrated in any suitable manner into one or more embodiments. In the following description, numerous specific details are set forth to give a full understanding of embodiments of this disclosure.

[0020] The terms “a,” “one,” and “” are used to indicate the existence of one or more elements / components / etc.; the terms “including” and “having” are used to indicate an open-ended meaning of inclusion and that other elements / components / etc. may exist in addition to those listed. The terms “first” and “second” are used only as markers and are not a limitation on the number of objects.

[0021] Traditional fiber optic biochemical identification relies solely on a single parameter to roughly calculate the refractive index, making it difficult to distinguish minute differences in refractive index between trace biochemical substances. This results in technical problems such as low accuracy in measuring the refractive index of trace biochemical substances, easy misjudgment of substances, and large errors in concentration quantification. It is impossible to simultaneously complete accurate substance identification and concentration quantification output.

[0022] In view of this, this application provides a substance identification method and system based on fiber optic grating biochemical sensing. By sequentially acquiring the transmission spectrum of a tilted grating fiber, analyzing the resonance sequence, generating envelope features, adaptively filtering and denoising, extracting multi-dimensional optical features, inverting the temperature-corrected refractive index, matching the target substance by combining the refractive index and dispersion relationship curve, quantifying the concentration based on the molecular optical gradient, and outputting a standardized detection data package, an identification logic covering spectral preprocessing, feature extraction, optical parameter inversion, substance matching, and concentration quantification is generated. This solves the problems of insufficient refractive index measurement accuracy, low substance identification accuracy, and large concentration detection error in traditional schemes for trace biochemical substances. It can simultaneously improve the accuracy of refractive index calculation, the reliability of biochemical substance discrimination, and the accuracy of target substance concentration quantification in complex biochemical media environments.

[0023] To illustrate the technical solution of this application, specific embodiments are described below.

[0024] This disclosure provides a substance identification method based on fiber Bragg grating biochemical sensing, such as... Figure 1 The illustrated method is a substance identification method based on fiber Bragg grating biochemical sensing. This method may include the following steps: Step S101: Obtain the initial transmission spectrum data output after the optical fiber containing the tilted grating is transmitted in the medium under test.

[0025] In the embodiments of this application, the substance identification system based on fiber optic grating biochemical sensing is equipped with an intelligent sensor with an integrated tilted grating. The system controls a broadband light source to output a continuous broadband light signal, which is incident on the tilted fiber grating inside the intelligent sensor that is immersed in the biochemical medium to be tested. The light undergoes core-cladding mode coupling inside the grating, and the transmitted light carrying the optical information of the medium is connected to a photodetector to complete photoelectric conversion.

[0026] The original digital spectral signal is uniformly sampled and output through the analog-to-digital conversion module. Basic electrical noise is filtered out from the original digital spectral signal, and the standardized initial transmission spectral data is reconstructed. The output band of the broadband light source can cover the C-band communication range, and the test medium can include biochemical samples such as blood, water quality reagents, and food extracts.

[0027] Step S102: Perform extreme value analysis on the initial transmission spectrum data to obtain the cladding resonant wavelength sequence and the corresponding transmittance numerical sequence.

[0028] In the implementation of this application, the extreme value analysis submodule is called to load the complete initial transmission spectrum data. All effective resonant minimum points are screened through sliding window differential peak finding logic, and false extreme points caused by background noise are eliminated. The effective extreme points are sorted according to the wavelength ascending order rule, and the corresponding wavelength coordinates and transmittance values ​​are bound one by one to generate two sets of one-dimensional sequences with consistent length and strict mapping, namely the cladding resonant wavelength sequence and the transmittance value sequence.

[0029] Step S103: Based on the cladding resonant wavelength sequence and transmittance numerical sequence, perform an envelope generation operation to obtain envelope feature data.

[0030] In the embodiments of this application, discrete resonant wavelength and transmittance pairing data are read, invalid sampling points with too small fluctuation amplitude are filtered out, and a piecewise interpolation algorithm is used to fit and generate a continuous spectral curve without discontinuities.

[0031] The upper and lower envelope feature coordinates are located by using the second derivative extrema, and the peak and valley coordinates are integrated with the transmittance fluctuation index and encapsulated into complete envelope feature data.

[0032] Step S104: Based on the envelope feature data, perform an adaptive filtering operation to obtain the noise-suppressed envelope data sequence.

[0033] In the embodiments of this application, the peak and valley transmittance fluctuation ranges within the envelope feature data are analyzed, the overall envelope volatility is calculated, and compared with a preset proportional threshold.

[0034] When the volatility exceeds the limit, the filter window size is adaptively matched according to the wavelength interval of adjacent resonance peaks, and the envelope curve is smoothed using a five-point moving average algorithm. The wavelength shift and smoothness index before and after filtering are recorded simultaneously, and the noise-reduced envelope data sequence is output.

[0035] Step S105: Based on the envelope data sequence, extract the geometric morphological features and energy attenuation features within a specific wavelength range to obtain a multidimensional feature parameter set characterizing the optical properties of the medium.

[0036] In the embodiments of this application, the core response wavelength range corresponding to the tilted grating is locked, and the curve slope change rate and curvature extreme value are extracted as geometric features through differential operation.

[0037] The area of ​​integration is obtained by performing a definite integral operation on the envelope curve of the entire spectrum. The attenuation coefficient of the resonance peak and the wavelength shift are extracted as energy attenuation characteristics by comparing with the standard reference spectrum.

[0038] The geometric morphological features and energy attenuation features are normalized and then spliced ​​and fused into a multi-dimensional medium optical feature parameter set.

[0039] Step S106: Perform refractive index inversion operation based on the multidimensional characteristic parameter group to obtain the refractive index value of the medium under test at a specific wavelength, and calculate the temperature dependence correction coefficient based on the refractive index value.

[0040] In the embodiments of this application, two types of features, wavelength shift and peak attenuation, are separated from the multidimensional feature parameter group, and then iteratively fitted into a pre-calibrated dispersion relation model to invert the high-precision refractive index at the target reference wavelength.

[0041] Refractive index, wavelength shift, and peak attenuation are simultaneously input into a full-range temperature calibration database. The optimal temperature calibration curve is matched, and the temperature-dependent correction coefficient for refractive index is obtained through differentiation, thus completing the extraction of temperature-sensitive parameters. Specifically, the measured refractive index value, wavelength shift characteristics, and peak attenuation coefficient are used as input vectors. The full-range calibration database is traversed and searched, and the Euclidean distance between each existing calibration curve and the input vector is calculated. The curve with the smallest distance is selected as the target refractive index and temperature calibration curve. The calibration curve characterizes the evolution of the refractive index of a specific biochemical substance with temperature under known grating coupling conditions. The first-order partial derivative of this curve along the temperature dimension is calculated. The resulting partial derivative is the temperature-dependent correction coefficient. It is used for refractive index temperature drift compensation during subsequent concentration inversion.

[0042] Step S107: Based on the refractive index value and the preset material refractive index database, the target biochemical substance identifier is determined by comparing the refractive index value and the similarity of the dispersion relation curve. The material refractive index database is obtained by offline collection of refractive index values ​​and full-band dispersion relation curve data of various known biochemical substances at standard temperature and standard concentration.

[0043] In the embodiments of this application, a matching tolerance range is set based on the inverted refractive index, and all candidate materials that meet the tolerance range are screened in the pre-built material refractive index database. The full-band standard dispersion relation curves of each type of candidate material are retrieved.

[0044] The similarity between the measured dispersion relation curve of the test medium and each standard curve is calculated using a time-series matching algorithm. This is then converted into a matching confidence score. Candidate substances are arranged in descending order of the score, and the unique identifier of the substance ranked first is used as the identifier of the target biochemical substance.

[0045] Step S108: Based on the target biochemical substance identifier and temperature-dependent correction coefficient, perform concentration quantification to correct the refractive index value and invert the concentration distribution range of the target substance in the test medium, and obtain a detection result data package containing the substance identifier, concentration value and confidence level.

[0046] In the embodiments of this application, the refractive index change gradient curve of the target biochemical substance at different concentrations is retrieved according to the target biochemical substance identifier, and the original refractive index is compensated and corrected by combining the real-time ambient temperature and the temperature dependence correction coefficient.

[0047] The corrected refractive index is input into the trained concentration inversion model to solve for the concentration distribution range of the target substance inside the medium.

[0048] The confidence level of the test results was calculated using Monte Carlo simulation. The substance identification, concentration range, and confidence index were uniformly encapsulated and encrypted to generate a standardized test result data packet for output.

[0049] The beneficial effects of this application embodiment compared with traditional technology are as follows: by sequentially acquiring the transmission spectrum of tilted grating fiber, analyzing the resonance sequence, generating envelope features, adaptive filtering and noise reduction, extracting multi-dimensional optical features, inverting the refractive index with temperature correction, matching the target substance by combining the refractive index and dispersion relationship curve, quantifying the concentration based on the molecular optical gradient, and outputting a standardized detection data package, an identification logic covering spectral preprocessing, feature extraction, optical parameter inversion, substance matching, and concentration quantification is generated. This solves the problems of insufficient accuracy in measuring the refractive index of trace biochemical substances, low accuracy in substance identification, and large concentration detection error in traditional schemes. It can simultaneously improve the accuracy of refractive index solution, the reliability of biochemical substance discrimination, and the accuracy of target substance concentration quantification in complex biochemical media environments.

[0050] In the optional embodiments of this application, refer to Figure 2 , Figure 2 This is a schematic diagram of Embodiment 2 of the material identification method based on fiber optic grating biochemical sensing in this application. An extreme value analysis operation is performed on the initial transmission spectrum data to obtain the cladding resonant wavelength sequence and the corresponding transmittance numerical sequence. Specifically, this may include the following steps: Step S201: Discrete sampling is performed on the initial transmission spectrum data to obtain the spectral data matrix.

[0051] In the embodiments of this application, the continuous initial transmission spectrum is discretely digitally sampled at a uniform fixed sampling interval, and the wavelength and corresponding transmittance values ​​are stored in a row-column regularized manner to generate a two-dimensional spectral data matrix.

[0052] The sampling resolution can be adaptively adjusted according to the performance of the demodulation equipment, and the matrix storage format can be a floating-point array or a structured data table.

[0053] Step S202: Set a sliding window to traverse the spectral data matrix and calculate the first and second derivatives of the transmittance within the window.

[0054] In the embodiments of this application, a sliding window of a preset size is loaded, and all sampled data are traversed point by point along the wavelength dimension of the spectral matrix. The first and second derivatives are calculated for each set of transmittance values ​​within the window. The derivative calculation adopts the central difference algorithm to determine the local concavity and convexity change trend of the spectral curve.

[0055] Step S203: When the first derivative changes from a negative value to a positive value after crossing the zero point and the second derivative is greater than a preset threshold, the zero point position is determined to be a candidate extreme point.

[0056] In the embodiments of this application, the trend of derivative changes is monitored in real time. When the first derivative changes from negative to positive and crosses zero, and the value of the second derivative at that position is higher than the system's preset discrimination threshold, the wavelength coordinate is marked as a candidate minimum extreme point and temporarily stored in the cache list.

[0057] This judgment logic can filter out resonant absorption peaks and spurious zeros generated in the flat region of the curve.

[0058] Step S204: Determine the target extreme point among the candidate extreme points according to the preset noise floor, wherein the preset noise floor is set based on the statistical average of the background spectral transmittance collected multiple times by the optical fiber under no-load conditions.

[0059] In the embodiments of this application, the pre-stored statistical mean of the background spectrum of the unloaded optical fiber is retrieved, and a uniform noise threshold is generated based on the mean.

[0060] Traverse all candidate extreme points, compare the transmittance of each point with the noise limit, and retain only the valid points with transmittance below the limit as the target extreme points, eliminating false extreme points caused by environmental electrical noise and optical path loss.

[0061] Step S205: Sort the target extreme points in ascending order of wavelength to obtain the cladding resonant wavelength sequence and the corresponding transmittance numerical sequence.

[0062] In the embodiments of this application, the wavelength and transmittance binary data bound to all target extreme points are extracted, and sorted in ascending order along the wavelength from smallest to largest using a sorting algorithm. The resulting arrays are split into two ordered one-dimensional arrays, which serve as the cladding resonant wavelength sequence and the transmittance numerical sequence, respectively, with the indices of the two arrays corresponding one-to-one.

[0063] In this embodiment, a standardized spectral matrix is ​​generated by discrete sampling, and the resonant minimum value is located by using the double derivative peak-judgment logic. The noise limit is set according to the unloaded background spectrum to eliminate false extreme values, and the resonant sequence is generated in an orderly manner, thereby reducing the interference of optical path and circuit noise on the extraction of resonant peaks.

[0064] In the optional embodiments of this application, refer to Figure 3 , Figure 3 This is a schematic diagram of Embodiment 3 of the material identification method based on fiber optic grating biochemical sensing in this application. Based on the cladding resonant wavelength sequence and transmittance numerical sequence, an envelope generation operation is performed to obtain envelope feature data. Specifically, this may include the following steps: Step S301: Calculate the transmittance change of adjacent wavelength points in the cladding resonant wavelength sequence, remove invalid sampling points whose change is not greater than a preset change threshold, and obtain the retained sampling points.

[0065] In the embodiments of this application, the transmittance difference between each group of adjacent wavelength sampling points in the calculation sequence is used as the transmittance change. The change is compared with a preset change threshold, and smooth invalid sampling points with a difference less than or equal to the threshold are deleted. The remaining data points with obvious fluctuation characteristics are summarized into a set of retained sampling points to reduce the interference of invalid data on curve fitting.

[0066] Step S302: Perform curve fitting based on the retained sampling points to obtain a continuous spectral curve.

[0067] In the embodiments of this application, a cubic spline interpolation algorithm is used to perform piecewise curve fitting on the discrete retained sampling points to ensure the continuity of the first and second derivatives at each segment node, eliminate the quantization step error caused by discrete sampling, and generate a smooth and uninterrupted continuous spectral curve across the entire wavelength range to improve the accuracy of subsequent envelope extraction.

[0068] Step S303: Perform second derivative analysis on the continuous spectral curve to locate the set of maxima with negative curvature and the set of minima with positive curvature.

[0069] In the embodiments of this application, the second derivative is solved point by point for the complete continuous spectral curve, and the value of the second derivative corresponds to the curvature of the curve.

[0070] Curvature extrema with negative second derivatives are grouped into a set of upper envelope maxima, while curvature extrema with positive second derivatives are grouped into a set of lower envelope minima, in order to distinguish between the two types of envelope feature coordinates.

[0071] Step S304: Fit the set of maximum points and the set of minimum points to obtain the set of peak coordinates of the upper envelope and the set of valley coordinates of the lower envelope.

[0072] In the embodiments of this application, polynomial fitting is performed again on the set of maximum points and the set of minimum points to generate two independent smooth curves. Feature coordinates are extracted uniformly along the two curves and encapsulated to form the upper envelope peak coordinate set and the lower envelope valley coordinate set, so as to completely record the envelope fluctuation changes of the entire spectrum.

[0073] Step S305: Determine the envelope feature data based on the set of peak coordinates of the upper envelope and the set of valley coordinates of the lower envelope.

[0074] In the embodiments of this application, the entire coordinate set of the upper and lower envelopes is integrated, the transmittance fluctuation range index is extracted synchronously, and the coordinate set and fluctuation index are combined and encapsulated into a unified structured data package as complete envelope feature data.

[0075] In this embodiment, by filtering out smooth and invalid sampling points and generating a continuous spectrum, the upper and lower envelope coordinates are distinguished according to the curvature of the second derivative, and the overall spectral fluctuation features are completely extracted. This can transform the discrete resonance sequence into a continuous envelope feature that can be quantified and analyzed, providing complete spectral fluctuation information support for subsequent adaptive noise discrimination and multidimensional feature extraction.

[0076] In the optional embodiments of this application, refer to Figure 4 , Figure 4 This is a schematic diagram of Embodiment 4 of the substance identification method based on fiber optic grating biochemical sensing in this application. The envelope feature data is determined based on the upper envelope peak coordinate set and the lower envelope valley coordinate set, which may specifically include the following steps: Step S401: Extract the transmittance peak fluctuation range corresponding to the upper envelope peak coordinate set, and extract the transmittance valley fluctuation range corresponding to the lower envelope valley coordinate set.

[0077] In the embodiments of this application, the transmittance values ​​corresponding to all peak coordinates of the upper envelope are traversed, and the difference between the maximum and minimum values ​​is calculated to obtain the transmittance peak fluctuation range; the transmittance values ​​corresponding to all valley coordinates of the lower envelope are traversed simultaneously, and the difference between the values ​​is solved to obtain the transmittance valley fluctuation range. The above two fluctuation ranges are used to characterize the overall noise fluctuation intensity of the spectrum.

[0078] Step S402: The set of upper envelope peak coordinates, the set of lower envelope valley coordinates, the transmittance peak fluctuation range, and the transmittance valley fluctuation range are determined as envelope feature data.

[0079] In the embodiments of this application, two sets of coordinates and two types of transmittance fluctuation range indicators are uniformly stored in a structured data container. The fields in the container are fixed to correspond to four types of data, forming standardized envelope feature data, which can be directly used for subsequent filtering pre-operations such as fluctuation amplitude calculation and noise threshold determination.

[0080] In this embodiment, the upper and lower envelope spatial coordinates and transmittance fluctuation amplitude are stored synchronously, and the dual information of spectral spatial distribution and intensity fluctuation is fully preserved. This enables the noise level to be determined using only a single data packet, simplifies the adaptive filtering pre-data reading logic, and reduces the number of data interactions between modules.

[0081] In the optional embodiments of this application, refer to Figure 5 , Figure 5 This is a schematic diagram of Embodiment 5 of the material identification method based on fiber optic grating biochemical sensing in this application. Based on a multi-dimensional feature parameter set, a refractive index inversion operation is performed to obtain the refractive index value of the medium under test at a specific wavelength. Specifically, this may include the following steps: Step S501: Based on the multidimensional feature parameter group, separate the wavelength offset feature and the peak attenuation coefficient feature.

[0082] In the embodiments of this application, the fields of each dimension within the multidimensional feature parameter group are analyzed, and the wavelength offset and peak attenuation coefficient are extracted separately as optical features directly related to the refractive index of the medium. Auxiliary features such as geometric integral and curvature are stripped off, and the feature vector is output separately for dispersion model fitting calculation.

[0083] Step S502: Match the wavelength offset feature and peak attenuation coefficient feature with the preset dispersion relation model to invert the refractive index value of the medium under test at a specific wavelength. The model coefficients of the dispersion relation model are obtained by fitting the standard refractive index solution to the tilted fiber grating of the same specification.

[0084] In the embodiments of this application, a Cauchy dispersion relation model that has been pre-calibrated by a gradient standard refractive index solution is loaded, and the wavelength shift and peak attenuation characteristics are substituted into the model for nonlinear iterative solution. After the iteration converges, a high-precision refractive index value at a specific reference wavelength is output, and the numerical accuracy is retained to four decimal places.

[0085] The dispersion model coefficients are determined by fitting multiple sets of standard solutions of the same type of tilted grating through calibration experiments, thus adapting to the inherent coupling characteristics of the grating.

[0086] In this embodiment, two types of optical features directly related to the refractive index are separated, and nonlinear iterative inversion is completed based on the dispersion model calibrated by the standard solution. This overcomes the limitations of traditional single-wavelength offset fitting. At the same time, combined with the information on light energy attenuation and loss, the accuracy of solving the refractive index of trace biochemical media is significantly improved.

[0087] In the optional embodiments of this application, refer to Figure 6 , Figure 6 This is a schematic diagram of Embodiment Six of the material identification method based on fiber optic grating biochemical sensing in this application. The calculation of the temperature-dependent correction coefficient based on the refractive index value may specifically include the following steps: Step S601: Match the refractive index value, wavelength offset feature and peak attenuation coefficient feature in the multidimensional feature parameter group with the preset full-range calibration database, and select the target refractive index temperature calibration curve with the highest matching degree. The full-range calibration database stores the refractive index temperature mapping relationship of different biochemical substances and the temperature and refractive index cross-sensitive calibration parameters of the same specification tilted fiber grating.

[0088] In the embodiments of this application, a built-in full-range calibration database is retrieved. The database stores the refractive index temperature mapping curves of various biochemical substances in multiple temperature and concentration ranges, as well as the temperature-light cross-sensitivity calibration parameters of the tilted grating itself.

[0089] Using measured refractive index, wavelength shift, and peak attenuation characteristics as joint search criteria, all calibration curves in the database are traversed, the matching degree between each set of curves and the measured characteristics is calculated, and the curve with the highest matching degree is selected as the target refractive index temperature calibration curve.

[0090] Step S602: Calculate the partial derivative of the refractive index with temperature based on the target refractive index temperature calibration curve, and determine the partial derivative as the temperature-dependent correction coefficient.

[0091] In the embodiments of this application, the first partial derivative of the refractive index with respect to temperature is calculated along the temperature dimension of the target calibration curve. This partial derivative represents the amount of refractive index change corresponding to a unit temperature change and is directly used as a temperature-dependent correction coefficient for refractive index temperature compensation calculation in the subsequent concentration quantization stage.

[0092] In this embodiment, the optimal temperature calibration curve is matched by multi-feature joint retrieval, and the quantitative temperature-dependent correction coefficient is obtained by differentiation. This can completely compensate for the cross-sensitivity interference of the grating and biochemical medium, and eliminate the refractive index measurement offset caused by environmental temperature fluctuations.

[0093] In the optional embodiments of this application, refer to Figure 7 , Figure 7This is a schematic diagram of Embodiment Seven of the substance identification method based on fiber optic grating biochemical sensing in this application. The method determines the target biochemical substance identifier by comparing the refractive index value and the similarity of the dispersion curve, based on the refractive index value and a preset substance refractive index database. Specifically, this may include the following steps: Step S701: Based on the refractive index values ​​and the material refractive index database, perform a refractive index tolerance matching operation to obtain candidate materials that fall within the preset tolerance range of the refractive index values, as well as the pre-stored full-wavelength dispersion relationship curves corresponding to each candidate material in the candidate material set.

[0094] In the embodiments of this application, a fixed matching tolerance range is set with the measured refractive index as the center. All biochemical substances whose refractive index values ​​fall within the tolerance range are retrieved from the offline generated material refractive index database and collected to form a candidate material set. Simultaneously, the full-band standard dispersion relation curves of each candidate material are retrieved in advance to establish a one-to-one mapping relationship between the material identifier and the standard dispersion relation curve.

[0095] Step S702: Based on the pre-stored full-wavelength dispersion relation curve, perform curve similarity calculation to obtain the matching confidence score of each candidate substance.

[0096] In the embodiments of this application, the measured full-band dispersion relation curve of the medium under test is extracted, and the cumulative distance between the measured curve and the standard dispersion relation curve of each candidate material is calculated using a dynamic time warping algorithm. The distance value is converted into a matching confidence score in the range of 0 to 100 through an exponential decay mapping rule. The higher the score, the closer the optical properties of the medium are to those of the candidate material.

[0097] Step S703: Sort each candidate substance in descending order according to the matching confidence score, and determine the identifier corresponding to the target candidate substance at the top of the ranking as the target biochemical substance identifier.

[0098] In the embodiments of this application, the matching confidence score is used as the sorting key to sort all candidate substances in descending order, and the unique molecular or substance identifier bound to the first candidate substance in the sort is read. This identifier is then output as the final target biochemical substance identifier to the concentration quantification module.

[0099] In this embodiment, the range of candidate substances is first narrowed down quickly by refractive index tolerance, and then the similarity of the full-band dispersion relation curve is used to complete the discrimination, which solves the problem of confusion between biochemical substances with similar refractive indices when relying solely on a single refractive index value for identification.

[0100] In the optional embodiments of this application, refer to Figure 8 , Figure 8This is a schematic diagram of Embodiment 8 of the material identification method based on fiber optic grating biochemical sensing in this application. Based on the envelope data sequence, geometric morphological features and energy attenuation features within a specific wavelength range are extracted to obtain a multidimensional feature parameter set characterizing the optical properties of the medium. Specifically, this may include the following steps: Step S801: Based on the envelope data sequence, extract the slope change rate and curvature distribution extreme values ​​of the envelope curve within the preset core response wavelength range to obtain geometric morphological features. The core response wavelength range is set as the range of ±n nanometers of the Bragg wavelength center value of the tilted grating, where n is a constant determined based on the grating tilt angle and demodulation resolution calibration.

[0101] In the embodiments of this application, the system defines the core response wavelength range based on the optical coupling characteristics of the smart sensor equipped with a tilted grating. The Bragg wavelength center of the tilted grating is the core fundamental mode resonance reference position. Within the symmetrical range to the left and right of this center wavelength, the interaction intensity between the higher-order cladding evanescent field and the biochemical medium to be measured is the highest, and the spectral curvature and slope are most sensitive to minute changes in refractive index.

[0102] n is an adjustable range width parameter that adapts to the grating tilt angle, fiber specifications, and demodulation source bandwidth. The range boundary is determined by offline calibration experiments. The cladding resonance peak response outside this range is weak and has a high noise ratio, and is not included in the geometric feature calculation.

[0103] After locking the wavelength range of the symmetrical core response, the slope change rate of the envelope curve is calculated point by point using the second-order central difference algorithm. At the same time, the maximum extreme value of the curvature distribution in the whole range is extracted. The slope change rate and curvature extreme value are uniformly collected into geometric morphological features to reflect the curve deformation caused by the local abrupt change in the refractive index of the medium.

[0104] Step S802: Calculate the integral area value enclosed by the envelope curve and the horizontal axis within the preset integral wavelength range of the envelope data sequence, and compare the resonance peak parameters in the current envelope data sequence with the corresponding resonance peak parameters of the pre-stored reference spectrum to obtain the peak attenuation coefficient and wavelength shift. Use the integral area value, peak attenuation coefficient, and wavelength shift as energy attenuation characteristics. The integral wavelength range is set to the full spectrum range from the start wavelength to the end wavelength of the cladding resonance wavelength sequence.

[0105] In the embodiments of this application, the system pre-stores the factory reference spectral envelope collected by the same specification tilted fiber grating under an empty clean medium, as a disturbance-free reference template throughout the entire life cycle.

[0106] Extract the starting wavelength of the cladding resonant wavelength sequence With termination wavelength To define the integration boundary across the entire spectrum, the composite Simpson numerical integration algorithm is used to perform definite integration on the measured envelope function T(λ) within the interval: ; The obtained definite integral value This is the integral area value, which physically represents the total absorption loss of transmitted optical energy in the medium under test within the complete cladding mode coupling passband.

[0107] Secondly, cross-state comparisons were performed by identifying the main resonance peak of the same order as the reference spectrum in the measured envelope: wavelength offset : Locating and measuring the center wavelength of the main resonance peak Subtract the center wavelength of the corresponding order resonance peak in the reference spectrum. ,Right now It is used to characterize the overall spectral shift caused by changes in the refractive index of the medium; Peak attenuation coefficient Extract the transmittance valley value of the measured main resonance peak. Transmittance valleys corresponding to the order of the reference spectrum Perform normalized decay calculation, i.e. The larger the value, the stronger the energy confinement and dissipation of the medium under test on the higher-order cladding modes.

[0108] Integral area value Wavelength offset Peak attenuation coefficient The data is concatenated according to predefined field widths and encapsulated into a structured energy attenuation feature. This feature complements the aforementioned local geometric features, eliminating the risk of missed detections due to the influence of a single wavelength drift on the response of complex media.

[0109] Step S803: Normalize and integrate the geometric morphological features and energy decay features to obtain a multi-dimensional feature parameter set including geometric dimension, energy dimension and offset dimension.

[0110] In the embodiments of this application, all indicators within the geometric morphology characteristics and energy decay characteristics are linearly normalized to eliminate the dimensional differences of each indicator.

[0111] The normalized slope, curvature, integral area, attenuation coefficient, and wavelength shift are concatenated into a unified tensor to form a multidimensional feature parameter set covering three dimensions: geometric deformation, energy loss, and spectral drift.

[0112] In this embodiment, multiple features, including local curve geometric deformation and full-spectrum light energy loss, are extracted simultaneously. It is no longer limited to a single wavelength shift parameter and can fully analyze all spectral change information generated by the interaction between the biochemical medium and the evanescent field of the cladding.

[0113] In the optional embodiments of this application, refer to Figure 9 , Figure 9 This is a schematic diagram of Embodiment Nine of the substance identification method based on fiber Bragg grating biochemical sensing in this application. Based on the envelope feature data, an adaptive filtering operation is performed to obtain a noise-suppressed envelope data sequence. Specifically, this may include the following steps: Step S901: Based on the transmittance peak fluctuation range and transmittance valley fluctuation range contained in the envelope feature data, calculate the overall fluctuation amplitude of the envelope and determine whether the proportion of the overall fluctuation amplitude to the envelope mean exceeds the preset proportion threshold.

[0114] In the embodiments of this application, two sets of structured fields are deconstructed from the envelope feature data: the upper envelope transmittance peak fluctuation range. Fluctuation range of lower envelope transmittance trough .

[0115] First, calculate the range of the upper and lower envelopes separately: ; ; The sum of the upper and lower envelope ranges is used as the overall fluctuation range of the envelope. ,Right now Simultaneously extract the arithmetic mean of transmittance of the current envelope data sequence across the entire spectrum. We use this as the mean of the envelope.

[0116] Further calculation of relative volatility : ; The relative volatility Compared with the preset ratio threshold .

[0117] Perform a logical comparison: like If the current spectrum is significantly affected by ambient stray light or electrical noise from the demodulation circuit, the subsequent adaptive filter window reconstruction logic will be triggered. like If the current envelope is in a low-noise steady state, the smoothing filtering step is skipped and the envelope data sequence is directly transmitted to reduce the processing latency on the embedded side.

[0118] Among them, the preset ratio threshold Instead of being limited to a fixed constant, it is specified during the factory calibration stage based on the type of the medium being tested: for low-turbidity, transparent biochemical media such as water, The typical calibration is 10%; for high-scattering media such as blood and fermentation broth, The typical calibration is 20%, which balances noise reduction intensity and spectral morphology fidelity.

[0119] In step S902, if it is determined that the preset ratio threshold is exceeded, the wavelength interval between adjacent resonance peaks is calculated based on the cladding resonance wavelength sequence, and the filter window size is adaptively adjusted to obtain the filter window parameters that match the current spectral sampling resolution.

[0120] In the embodiments of this application, when the volatility exceeds a preset threshold, the cladding resonant wavelength sequence is traversed, the wavelength interval between all adjacent resonant peaks is solved and a representative interval value is taken; the physical wavelength size of the filter window is converted according to a fixed ratio based on the interval value, and the number of data points contained in the window is converted in conjunction with the original spectral sampling resolution to generate filter window parameters that are adapted to the current spectral characteristics.

[0121] Step S903: Perform a smoothing filter operation on the envelope feature data according to the filter window parameters, and record the wavelength offset and smoothing quantification index before and after filtering to obtain the noise-suppressed envelope data sequence.

[0122] In the embodiments of this application, a five-point moving window mean filtering algorithm is used to traverse the entire envelope curve with adaptively generated window parameters to complete smooth noise reduction.

[0123] The difference in the center wavelength of the resonance peak before and after filtering is recorded as the wavelength offset, and the smoothness index is quantified by the variance of the first derivative of adjacent data.

[0124] The smoothed complete transmittance and wavelength data are encapsulated to generate a noise-suppressed envelope data sequence.

[0125] In this embodiment, the filtering logic is adaptively triggered based on the amplitude of the spectrum fluctuations. The size of the filtering window is dynamically matched according to the spacing between the resonance peaks to adapt to the spectral data under different test media and different noise intensities. This fully filters out high-frequency noise while retaining the effective resonance peak deformation characteristics.

[0126] Figure 10 This illustration shows a structural schematic diagram of a substance identification system based on fiber Bragg grating biochemical sensing according to an embodiment of this application. Specifically, the aforementioned substance identification system 1000 based on fiber Bragg grating biochemical sensing may include: The acquisition module 1001 is used to acquire the initial transmission spectrum data output by the optical fiber containing the tilted grating after transmission in the medium under test. The analysis module 1002 is used to perform extreme value analysis on the initial transmission spectrum data to obtain the cladding resonant wavelength sequence and the corresponding transmittance numerical sequence. The generation module 1003 is used to perform an envelope generation operation based on the cladding resonant wavelength sequence and the transmittance numerical sequence to obtain envelope feature data; The filtering module 1004 is used to perform adaptive filtering based on the envelope feature data to obtain the noise-suppressed envelope data sequence. The extraction module 1005 is used to extract geometric morphological features and energy attenuation features within a specific wavelength range based on the envelope data sequence, so as to obtain a multi-dimensional feature parameter set characterizing the optical properties of the medium. The inversion module 1006 is used to perform a refractive index inversion operation based on a multidimensional characteristic parameter set, obtain the refractive index value of the medium under test at a specific wavelength, and calculate the temperature dependence correction coefficient based on the refractive index value. The comparison module 1007 is used to determine the target biochemical substance identifier by comparing the refractive index value and the similarity of the dispersion curve based on the refractive index value and the preset material refractive index database. The material refractive index database is obtained by offline collection of refractive index values ​​and full-band dispersion curve data of various known biochemical substances at standard temperature and standard concentration. The quantization module 1008 is used to perform concentration quantification operations based on the target biochemical substance identifier and temperature-dependent correction coefficient, so as to correct the refractive index value and invert the concentration distribution range of the target substance in the test medium, and obtain a detection result data package containing the substance identifier, concentration value and confidence level.

[0127] The beneficial effects of this application embodiment compared with traditional technology are as follows: by sequentially acquiring the transmission spectrum of tilted grating fiber, analyzing the resonance sequence, generating envelope features, adaptive filtering and noise reduction, extracting multi-dimensional optical features, inverting the refractive index with temperature correction, matching the target substance by combining the refractive index and dispersion relationship curve, quantifying the concentration based on the molecular optical gradient, and outputting a standardized detection data package, an identification logic covering spectral preprocessing, feature extraction, optical parameter inversion, substance matching, and concentration quantification is generated. This solves the problems of insufficient accuracy in measuring the refractive index of trace biochemical substances, low accuracy in substance identification, and large concentration detection error in traditional schemes. It can simultaneously improve the accuracy of refractive index solution, the reliability of biochemical substance discrimination, and the accuracy of target substance concentration quantification in complex biochemical media environments.

[0128] The parsing module 1002 can also be specifically used for: performing discrete sampling on the initial transmission spectrum data to obtain a spectral data matrix; setting a sliding window to traverse the spectral data matrix and calculating the first and second derivatives of the transmittance within the window; when the first derivative changes from a negative value to a positive value after crossing zero and the second derivative is greater than a preset threshold, determining the zero point as a candidate extreme point; determining the target extreme point among the candidate extreme points according to a preset noise floor, wherein the preset noise floor is set based on the statistical average of the background spectral transmittance collected multiple times in an unloaded state; and sorting the target extreme points in ascending order of wavelength to obtain the cladding resonant wavelength sequence and the corresponding transmittance numerical sequence.

[0129] The generation module 1003 can also be specifically used for: calculating the transmittance change of adjacent wavelength points in the cladding resonant wavelength sequence, removing invalid sampling points whose change is not greater than a preset change threshold, and obtaining retained sampling points; performing curve fitting based on the retained sampling points to obtain a continuous spectral curve; performing second derivative analysis on the continuous spectral curve to locate the set of maxima with negative curvature and the set of minima with positive curvature; fitting based on the set of maxima and the set of minima to obtain the set of upper envelope peak coordinates and the set of lower envelope valley coordinates; and determining the envelope feature data based on the set of upper envelope peak coordinates and the set of lower envelope valley coordinates.

[0130] The generation module 1003 can also be specifically used to: extract the transmittance peak fluctuation range corresponding to the upper envelope peak coordinate set and extract the transmittance valley fluctuation range corresponding to the lower envelope valley coordinate set; and determine the upper envelope peak coordinate set, the lower envelope valley coordinate set, the transmittance peak fluctuation range, and the transmittance valley fluctuation range as envelope feature data.

[0131] The inversion module 1006 can also be specifically used to: separate the wavelength offset feature and the peak attenuation coefficient feature based on the multidimensional feature parameter group; match the wavelength offset feature and the peak attenuation coefficient feature with the preset dispersion relation model to invert the refractive index value of the medium under test at a specific wavelength, wherein the model coefficients of the dispersion relation model are obtained by fitting the standard refractive index solution to the tilted fiber grating of the same specification through calibration experiments.

[0132] The inversion module 1006 can also be specifically used to: match the refractive index value, the wavelength shift feature and the peak attenuation coefficient feature in the multidimensional feature parameter group with the preset full-range calibration database, and select the target refractive index temperature calibration curve with the highest matching degree. The full-range calibration database stores the refractive index temperature mapping relationship of different biochemical substances and the temperature and refractive index cross-sensitive calibration parameters of the same specification tilted fiber grating; calculate the partial derivative of the refractive index with temperature according to the target refractive index temperature calibration curve, and determine the partial derivative as the temperature-dependent correction coefficient.

[0133] The comparison module 1007 can also be specifically used for: performing a refractive index tolerance matching operation based on the refractive index value and the material refractive index database to obtain candidate substances that fall within the preset tolerance range of the refractive index value and the pre-stored full-wavelength dispersion relationship curves corresponding to each candidate substance in the candidate substance set; performing a curve similarity calculation operation based on the pre-stored full-wavelength dispersion relationship curves to obtain the matching confidence score of each candidate substance; and sorting each candidate substance in descending order based on the matching confidence score, and determining the identifier corresponding to the target candidate substance at the top of the sort as the target biochemical substance identifier.

[0134] The extraction module 1005 can also be specifically used for: extracting the slope change rate and curvature distribution extreme values ​​of the envelope curve within a preset core response wavelength range based on the envelope data sequence to obtain geometric morphological features, wherein the core response wavelength range is set to the range of ±n nanometers of the Bragg wavelength center value of the tilted grating; calculating the integral area value enclosed by the envelope curve and the horizontal axis within a preset integral wavelength range for the envelope data sequence, and comparing the resonance peak parameters in the current envelope data sequence with the corresponding resonance peak parameters of the pre-stored reference spectrum to obtain the peak attenuation coefficient and wavelength offset, and using the integral area value, peak attenuation coefficient, and wavelength offset as energy attenuation features, wherein the integral wavelength range is set to the full spectrum range from the start wavelength to the end wavelength of the cladding resonance wavelength sequence; and normalizing and integrating the geometric morphological features and energy attenuation features to obtain a multi-dimensional feature parameter set containing geometric, energy, and offset dimensions.

[0135] The filtering module 1004 can also be specifically used for: calculating the overall fluctuation amplitude of the envelope based on the peak and valley fluctuation ranges of transmittance contained in the envelope feature data, and determining whether the proportion of the overall fluctuation amplitude to the envelope mean exceeds a preset threshold; if it is determined to exceed the preset threshold, calculating the wavelength interval between adjacent resonance peaks based on the cladding resonance wavelength sequence, performing an adaptive adjustment operation of the filtering window size to obtain filtering window parameters that match the current spectral sampling resolution; performing a smoothing filtering operation on the envelope feature data based on the filtering window parameters, and recording the wavelength offset and smoothing quantification index before and after filtering to obtain the noise-suppressed envelope data sequence.

[0136] It should be noted that, in this document, the terms "comprising," "including," and any other variations are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Specific examples have been used in this document to illustrate the principles and implementation methods of the technical solutions of this application. The above examples are only for the purpose of helping to understand the methods and core ideas of this application. The above are merely preferred embodiments of this application. It should be pointed out that, due to the limitations of written expression and the objective existence of infinite specific structures, those skilled in the art can make several improvements, modifications, or changes without departing from the principles of this application, and can also combine the above technical features in an appropriate manner; these improvements, modifications, changes, or combinations, or the direct application of the concept and technical solutions of this application to other situations without modification, should all be considered within the scope of protection of this application.

Claims

1. A method for substance identification based on fiber Bragg grating biochemical sensing, characterized in that, The method includes: Acquire the initial transmission spectrum data output after the optical fiber containing the tilted grating is transmitted in the medium under test; An extreme value analysis operation is performed on the initial transmission spectrum data to obtain the cladding resonant wavelength sequence and the corresponding transmittance numerical sequence of the cladding resonant wavelength sequence. Based on the cladding resonant wavelength sequence and the transmittance numerical sequence, an envelope generation operation is performed to obtain envelope feature data; Based on the envelope feature data, an adaptive filtering operation is performed to obtain a noise-suppressed envelope data sequence. Based on the envelope data sequence, geometric morphological features and energy attenuation features within a specific wavelength range are extracted to obtain a multidimensional feature parameter set characterizing the optical properties of the medium. Based on the multidimensional feature parameter set, a refractive index inversion operation is performed to obtain the refractive index value of the medium under test at a specific wavelength, and a temperature dependence correction coefficient is calculated based on the refractive index value. Based on the refractive index value and a preset material refractive index database, the target biochemical substance identifier is determined by comparing the refractive index value and the similarity of the dispersion relation curve. The material refractive index database is obtained by offline collection of refractive index values ​​and full-band dispersion relation curve data of various known biochemical substances at standard temperature and standard concentration. Based on the target biochemical substance identifier and the temperature-dependent correction coefficient, a concentration quantization operation is performed to correct the refractive index value and invert the concentration distribution range of the target substance in the test medium, thereby obtaining a detection result data package containing the substance identifier, concentration value, and confidence level.

2. The substance identification method based on fiber optic grating biochemical sensing as described in claim 1, characterized in that, The extreme value analysis operation performed on the initial transmission spectrum data to obtain the cladding resonant wavelength sequence and the corresponding transmittance numerical sequence includes: Discrete sampling is performed on the initial transmission spectrum data to obtain a spectral data matrix; A sliding window is set to traverse the spectral data matrix, and the first and second derivatives of the transmittance within the window are calculated. When the first derivative changes from a negative value to a positive value after crossing the zero point and the second derivative is greater than a preset threshold, the zero point position is determined to be a candidate extreme point. According to the preset noise floor, the target extreme point is determined among the candidate extreme points, wherein the preset noise floor is set based on the statistical average of the background spectral transmittance collected multiple times by the optical fiber under no-load conditions; The target extreme points are sorted in ascending order of wavelength to obtain the cladding resonant wavelength sequence and the corresponding transmittance numerical sequence.

3. The substance identification method based on fiber optic grating biochemical sensing as described in claim 1, characterized in that, The step of performing an envelope generation operation based on the cladding resonant wavelength sequence and the transmittance numerical sequence to obtain envelope feature data includes: Calculate the transmittance change of adjacent wavelength points in the cladding resonant wavelength sequence, remove invalid sampling points whose change is not greater than a preset change threshold, and obtain the retained sampling points. Based on the retained sampling points, curve fitting is performed to obtain a continuous spectral curve; Second derivative analysis is performed on the continuous spectral curve to locate the set of maxima with negative curvature and the set of minima with positive curvature. By fitting the set of maximum points and the set of minimum points, we obtain the set of peak coordinates of the upper envelope and the set of valley coordinates of the lower envelope. The envelope feature data is determined based on the set of peak coordinates of the upper envelope and the set of valley coordinates of the lower envelope.

4. The substance identification method based on fiber optic grating biochemical sensing as described in claim 3, characterized in that, The step of determining the envelope feature data based on the upper envelope peak coordinate set and the lower envelope valley coordinate set includes: Extract the transmittance peak fluctuation range corresponding to the upper envelope peak coordinate set, and extract the transmittance valley fluctuation range corresponding to the lower envelope valley coordinate set; The set of upper envelope peak coordinates, the set of lower envelope valley coordinates, the transmittance peak fluctuation range, and the transmittance valley fluctuation range are determined as the envelope feature data.

5. The substance identification method based on fiber optic grating biochemical sensing as described in claim 1, characterized in that, The step of performing a refractive index inversion operation based on the multidimensional feature parameter set to obtain the refractive index value of the medium under test at a specific wavelength includes: Based on the multidimensional feature parameter set, the wavelength offset feature and the peak attenuation coefficient feature are separated. The wavelength offset feature and the peak attenuation coefficient feature are matched with a preset dispersion relation model to invert the refractive index value of the medium under test at a specific wavelength. The model coefficients of the dispersion relation model are obtained by fitting a standard refractive index solution to a tilted fiber grating of the same specification.

6. The substance identification method based on fiber Bragg grating biochemical sensing as described in claim 1, characterized in that, The calculation of the temperature-dependent correction factor based on the refractive index value includes: The refractive index value, the wavelength offset feature and the peak attenuation coefficient feature in the multidimensional feature parameter group are matched with the preset full-range calibration database to select the target refractive index temperature calibration curve with the highest matching degree. The full-range calibration database stores the refractive index temperature mapping relationship of different biochemical substances and the temperature and refractive index cross-sensitive calibration parameters of the same specification tilted fiber grating. The partial derivative of the refractive index with temperature is calculated based on the target refractive index temperature calibration curve, and the partial derivative is determined as the temperature-dependent correction coefficient.

7. The substance identification method based on fiber optic grating biochemical sensing as described in claim 1, characterized in that, The step of determining the target biochemical substance identifier based on the refractive index value and a preset material refractive index database by comparing the refractive index values ​​and the similarity of dispersion curves includes: Based on the refractive index value and the material refractive index database, a refractive index tolerance matching operation is performed to obtain candidate materials that fall within the preset tolerance range of the refractive index value and the pre-stored full-wavelength dispersion relationship curves corresponding to each candidate material in the candidate material set. Based on the pre-stored full-wavelength dispersion relation curve, a curve similarity calculation operation is performed to obtain the matching confidence score of each candidate substance; The candidate substances are sorted in descending order according to the matching confidence score, and the identifier corresponding to the top-ranked target candidate substance is determined as the target biochemical substance identifier.

8. The substance identification method based on fiber Bragg grating biochemical sensing as described in claim 1, characterized in that, The step involves extracting geometric morphological features and energy attenuation features within a specific wavelength range based on the envelope data sequence to obtain a multidimensional feature parameter set characterizing the optical properties of the medium, including: Based on the envelope data sequence, the slope change rate and curvature distribution extreme values ​​of the envelope curve are extracted within the preset core response wavelength range to obtain geometric morphological features. The core response wavelength range is set to be the range of ±n nanometers of the Bragg wavelength center value of the tilted grating, where n is a constant determined according to the grating tilt angle and demodulation resolution calibration. The integral area value enclosed by the envelope curve and the horizontal axis is calculated for the envelope data sequence within a preset integral wavelength range. The resonance peak parameters in the current envelope data sequence are compared with the corresponding resonance peak parameters of the pre-stored reference spectrum to obtain the peak attenuation coefficient and wavelength shift. The integral area value, the peak attenuation coefficient, and the wavelength shift are used as energy attenuation characteristics. The integral wavelength range is set to the full spectrum range from the start wavelength to the end wavelength of the cladding resonance wavelength sequence. The geometric features and the energy decay features are normalized and integrated to obtain the multidimensional feature parameter set that includes geometric dimension, energy dimension and offset dimension.

9. The substance identification method based on fiber optic grating biochemical sensing as described in claim 1, characterized in that, The step of performing an adaptive filtering operation based on the envelope feature data to obtain a noise-suppressed envelope data sequence includes: Based on the transmittance peak fluctuation range and transmittance valley fluctuation range contained in the envelope feature data, calculate the overall fluctuation amplitude of the envelope and determine whether the proportion of the overall fluctuation amplitude to the envelope mean exceeds a preset proportion threshold. If the preset ratio threshold is exceeded, the wavelength interval between adjacent resonance peaks is calculated based on the cladding resonance wavelength sequence, and the filter window size is adaptively adjusted to obtain the filter window parameters that match the current spectral sampling resolution. Perform a smoothing filter operation on the envelope feature data according to the filter window parameters, and record the wavelength offset and smoothing quantification index before and after filtering to obtain the noise-suppressed envelope data sequence.

10. A substance identification system based on fiber Bragg grating biochemical sensing, characterized in that, include: The acquisition module is used to acquire the initial transmission spectrum data output by the optical fiber containing the tilted grating after transmission in the medium under test. The analysis module is used to perform extreme value analysis on the initial transmission spectrum data to obtain the cladding resonant wavelength sequence and the corresponding transmittance value sequence. The generation module is used to perform an envelope generation operation based on the cladding resonant wavelength sequence and the transmittance value sequence to obtain envelope feature data; The filtering module is used to perform adaptive filtering operations based on the envelope feature data to obtain a noise-suppressed envelope data sequence. The extraction module is used to extract geometric morphological features and energy attenuation features within a specific wavelength range based on the envelope data sequence, so as to obtain a multidimensional feature parameter set characterizing the optical properties of the medium. The inversion module is used to perform a refractive index inversion operation based on the multidimensional feature parameter set, obtain the refractive index value of the medium under test at a specific wavelength, and calculate the temperature dependence correction coefficient based on the refractive index value. The comparison module is used to determine the target biochemical substance identifier by comparing the refractive index value and the similarity of the dispersion relation curve based on the refractive index value and the preset material refractive index database. The material refractive index database is obtained by offline collection of refractive index values ​​and full-band dispersion relation curve data of various known biochemical substances at standard temperature and standard concentration. The quantization module is used to perform concentration quantization operations based on the target biochemical substance identifier and the temperature-dependent correction coefficient, so as to correct the refractive index value and invert the concentration distribution range of the target substance in the test medium, and obtain a detection result data package containing the substance identifier, concentration value and confidence level.

Images