A multi-point calibration method and system for a gas analyzer
By using multi-point calibration of the reference analyzer and spectral curve calibration, combined with concentration inversion calculation, efficient and low-cost factory calibration of the gas analyzer was achieved, ensuring high-precision measurement results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NANJING ANRONX ELECTRONICS TECH CO LTD
- Filing Date
- 2023-04-11
- Publication Date
- 2026-05-19
AI Technical Summary
Traditional multi-point calibration methods require the collection of a large number of standard gases of different concentrations, resulting in long calibration times, high costs, and low efficiency for gas analyzers, which cannot meet the requirements for high precision.
Multi-point calibration was performed using a reference analyzer to obtain the reference calibration coefficient A. The target analyzer was then calibrated using spectral curves and concentration inversion with a small number of standard gases of different concentrations. The factory calibration coefficient B of the target analyzer was calculated by combining the reference calibration coefficient and the concentration calibration coefficient.
This shortens the factory calibration time of the gas analyzer, reduces calibration costs, and ensures high-precision measurement results, meeting the high-precision requirements of the analyzer.
Smart Images
Figure CN116337783B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a gas analyzer, specifically to a multi-point calibration method and system for a gas analyzer, belonging to the field of environmental monitoring technology. Background Technology
[0002] Before any gas analyzer can be put into operation, it needs to undergo factory calibration to obtain a calibration coefficient curve showing a linear relationship between concentration and response values. The response value of the sample to be tested is then substituted into this curve to obtain the concentration value of the sample. Common calibration methods include single-point calibration and multi-point calibration. Single-point calibration uses a reference gas with a concentration close to that of the standard sample being measured as the analytical standard. The reference gas is measured repeatedly two or more times to obtain a calibration coefficient. This method is simple to operate and has a short calibration time, but it has low accuracy and a small analyzer range. Multi-point calibration uses multiple reference gases of different concentrations as analytical standards. The reference gases are measured repeatedly two or more times, and the spectral values at different pixels of the obtained spectral curve are fitted with the corresponding concentration values to obtain a standard parameter curve. This method requires multiple concentrations of standard gas and is more complex to operate, but it offers high measurement accuracy and a wide analyzer range. With the increasing stringency of environmental protection standards in my country, the accuracy requirements for gas analyzers are also increasing. Therefore, more and more gas analyzers are choosing multi-point calibration methods to ensure the accuracy of gas analysis measurements.
[0003] However, the traditional multi-point calibration method requires the collection of a large amount of standard gas data of different concentrations to ensure high measurement accuracy. Therefore, the calibration time is long and a large amount of standard gas is consumed. In the actual production process, due to certain differences between different gas analyzers, each gas analyzer needs to be calibrated individually, which leads to reduced measurement efficiency and increased calibration cost. Summary of the Invention
[0004] To overcome the shortcomings of the prior art, the present invention provides a multi-point calibration method and system for gas analyzers, which can reduce the factory calibration time of each gas analyzer, improve measurement efficiency, and reduce calibration costs while ensuring measurement accuracy.
[0005] This invention discloses a multi-point calibration method for a gas analyzer, comprising the following steps:
[0006] S1, n C = [C1, C2, ..., C] are introduced into the reference analyzer. n Standard gases of different concentrations were used to obtain reference spectral curve data. The reference analyzer was then calibrated at the factory using a multi-point calibration method to obtain the reference calibration coefficient A.
[0007] S2, introduce m c = [c1, c2, ..., c] into the target analyzer. m]∈C Standard gases of different concentrations were used to obtain the target spectral curve data s=[s1,s2,…,s m ],3≤m <n;
[0008] S3, using the reference spectrum curve of the reference analyzer as the standard, calibrate the wavelength pixel position of the target spectrum curve obtained in the target analyzer under the same gas concentration.
[0009] S4. Input the calibrated target spectral curve data s' into the reference analyzer, and obtain the inversion concentration values c'=[c1',c2'…,c...] of each standard gas through the reference calibration coefficient A. m '];
[0010] S5, fit c and c' to obtain the concentration calibration coefficient P;
[0011] S6. Input the reference calibration coefficient A and the concentration calibration coefficient P into the target analyzer to calculate the factory calibration coefficient B of the target analyzer.
[0012] In step S1, the reference analyzer is calibrated at the factory using a multi-point calibration method to obtain the reference calibration coefficient A, as follows:
[0013] S11, n C = [C1, C2, ..., C] are introduced into the reference analyzer. n For standard gases of different concentrations, obtain the reference spectral curve data S=[S1,S2,…,S…] for each concentration of standard gas. n ] T ;
[0014] S12, select k wavelength pixels on the spectral curve of each concentration standard gas;
[0015] S13, based on the functional relationship between the standard gas concentration C and the spectral amplitude S, S=A*C, the reference calibration coefficient A=[a1,a2,…,a] is calculated using a fitting algorithm. k ].
[0016] Optionally, in step S3, the wavelength pixel positions of the target spectral curve obtained under the same gas concentration in the target analyzer are calibrated by an interpolation translation algorithm, a neural network deep learning algorithm, or a wavelength-pixel mapping method.
[0017] Specifically, the interpolation and translation algorithm steps are as follows:
[0018] S31, Select curve data at the same concentration from the reference spectral curve and the target spectral curve, and calculate the wavelength pixel offset Δλ of the target spectral curve.
[0019] S32, extract all pixel intervals in the target spectral curve that include the standard gas absorption band, select a wider interval than the above pixel interval as a wavelength window, and interpolate all pixels of the target spectrum within the wavelength window to obtain the interpolated target spectral curve.
[0020] S33, calculate the wavelength pixel offset between the interpolated target spectral curve and the reference spectral curve;
[0021] S34, calibrate the spectral curves obtained at various gas concentrations in the target analyzer by shifting the target spectral curve according to the wavelength offset.
[0022] Furthermore, the wavelength pixel offset is the sum of the wavelength pixel coarse offset and the wavelength pixel fine offset. The accuracy of the wavelength pixel coarse offset is 1 interpolation point, and the accuracy of the wavelength pixel fine offset is 0.0001 interpolation points.
[0023] In step S4, s' is substituted into the formula s'=A*c', and the inversion concentration value c'=[c1',c2'…,c m ').
[0024] In step S5, c and c' are fitted to obtain the concentration calibration coefficient P, as follows:
[0025] Let c be X and c′ be Y. Perform a second-order polynomial fitting on X and Y: Y = p²X 2 +p1X+p0,P=[p0,p1,p2].
[0026] In step S6, the factory calibration coefficient of the target analyzer is calculated based on the reference calibration coefficient A and the concentration calibration coefficient P as B = A * P = [a1 * P, a2 * P, ..., a k *P] T .
[0027] The present invention also discloses a multi-point calibration system for a gas analyzer, including a reference analyzer and a target analyzer. The reference analyzer includes a first measuring cell and a first spectrometer, and the target analyzer includes a second measuring cell and a second spectrometer.
[0028] The reference analyzer also includes a calibration calculation module, a storage module, and a concentration calibration module. Standard gases of different concentrations are introduced into the measurement cell. The spectrometer is used to acquire the reference spectral curve data of the standard gases of each concentration and send them to the calibration calculation module and the storage module respectively. The calibration calculation module calculates the reference calibration coefficient and sends it to the storage module.
[0029] The target analyzer also includes a pixel calibration module and a second calibration calculation module. Standard gases of different concentrations are introduced into the second measurement cell. The second spectrometer acquires the target spectral curve data of each concentration of standard gas and sends it to the second calibration calculation module and the second pixel calibration module, respectively. The reference spectral curve data is input into the second pixel calibration module. The second pixel calibration module is used to calibrate the position of the target spectral curve according to the reference spectral curve and sends the calibrated target spectral curve data to the concentration calibration module of the reference analyzer. The concentration calibration module calculates the inversion concentration value of each standard gas based on the calibrated target spectral curve and calculates the concentration calibration coefficient based on the standard concentration value and the inversion concentration value of each standard gas. The calibration calculation module is used to calculate the factory calibration coefficient of the target analyzer based on the concentration calibration coefficient and the reference calibration coefficient.
[0030] The present invention also discloses a non-transitory computer-readable storage medium, including any instruction for executing the multi-point calibration method of the gas analyzer.
[0031] The method and system provided by this invention first obtains the reference calibration coefficient of a reference analyzer using a traditional multi-point calibration method. Then, a small number of standard gases of different concentrations are introduced into the target analyzer to obtain a small number of target spectral curve data. The target spectral curve is then calibrated at the wavelength and pixel level based on the reference spectral curve. The calibrated target spectral curve is then input into the reference analyzer for concentration inversion calculation and concentration fitting calculation to obtain the concentration calibration coefficient P. Finally, matrix calculation is performed based on the reference calibration coefficient A and the concentration calibration coefficient P to obtain the factory calibration coefficient B of the target analyzer. At this point, the number of pixels in the factory calibration coefficient of the target analyzer is equal to the number of pixels in the reference calibration coefficient, ensuring the factory calibration accuracy of the target analyzer. Using this method to calibrate the target analyzer can not only greatly reduce the amount of standard gas introduced, saving production costs, but also shorten the factory calibration time. Furthermore, the indication error of the obtained factory calibration coefficient for concentration inversion calculation is within 1%, fully meeting the high precision requirements of the analyzer. Attached Figure Description
[0032] Figure 1 This is a schematic diagram of the multi-point calibration method for the gas analyzer provided by the present invention;
[0033] Figure 2 A schematic diagram of the method flow provided for a specific embodiment of the present invention;
[0034] Figure 3 This is a graph showing the benchmark calibration coefficients of the benchmark analyzer in a specific embodiment of the present invention;
[0035] Figure 4 This is a factory calibration coefficient curve of the target analyzer in a specific embodiment of the present invention;
[0036] Figure 5 This is a comparison chart of the reference spectrum curve of the reference analyzer and the target spectrum curve of the target analyzer at a concentration of 750 in a specific embodiment of the present invention;
[0037] Figure 6 This is a table showing the SO2 concentration error calculated by inversion after factory calibration using this invention;
[0038] Figure 7 This is a schematic diagram of the multi-point calibration system for the gas analyzer provided by the present invention. Detailed Implementation
[0039] To more clearly illustrate the present invention, the invention will be further described below with reference to the accompanying drawings and specific embodiments.
[0040] like Figure 1 As shown, a multi-point calibration method for a gas analyzer includes the following steps:
[0041] S1, n C = [C1, C2, ..., C] are introduced into the reference analyzer. n Standard gases of different concentrations were used to obtain reference spectral curve data. The reference analyzer was then calibrated at the factory using a multi-point calibration method to obtain the reference calibration coefficient A.
[0042] S2, introduce m c = [c1, c2, ..., c] into the target analyzer. m ]∈C Standard gases of different concentrations were used to obtain the target spectral curve data s=[s1,s2,…,s m ],3≤m <n;
[0043] S3, using the reference spectrum curve of the reference analyzer as the standard, calibrate the wavelength pixel position of the target spectrum curve obtained in the target analyzer under the same gas concentration.
[0044] S4. Input the calibrated target spectral curve data s' into the reference analyzer, and obtain the inversion concentration values c'=[c1',c2'…,c...] of each standard gas through the reference calibration coefficient A. m '];
[0045] S5, fit c and c' to obtain the concentration calibration coefficient P;
[0046] S6. Input the reference calibration coefficient A and the concentration calibration coefficient P into the target analyzer to calculate the factory calibration coefficient B of the target analyzer.
[0047] In step S1, the reference analyzer is calibrated at the factory using a multi-point calibration method to obtain the reference calibration coefficient A, as follows:
[0048] S11, n C = [C1, C2, ..., C] are introduced into the reference analyzer. n For standard gases of different concentrations, obtain the reference spectral curve data S=[S1,S2,…,S…] for each concentration of standard gas. n ] T ;
[0049] S12, select k wavelength pixels on the spectral curve of each concentration standard gas;
[0050] S13, based on the functional relationship between the standard gas concentration C and the spectral amplitude S, S=A*C, the reference calibration coefficient A=[a1,a2,…,a] is calculated using a fitting algorithm. k ].
[0051] In step S3, the wavelength pixel positions of the target spectral curve obtained in the target analyzer under the same gas concentration are calibrated by interpolation translation algorithm, neural network deep learning algorithm, or wavelength-pixel mapping method.
[0052] Specifically, the interpolation and translation algorithm steps are as follows:
[0053] S31, Select curve data at the same concentration from the reference spectral curve and the target spectral curve, and calculate the wavelength pixel offset Δλ of the target spectral curve.
[0054] S32, extract all pixel intervals in the target spectral curve that include the standard gas absorption band, select a wider interval than the above pixel interval as a wavelength window, and interpolate all pixels of the target spectrum within the wavelength window to obtain the interpolated target spectral curve.
[0055] S33, calculate the wavelength pixel offset between the interpolated target spectral curve and the reference spectral curve;
[0056] S34, calibrate the spectral curves obtained at various gas concentrations in the target analyzer by shifting the target spectral curve according to the wavelength offset.
[0057] Furthermore, the wavelength pixel offset is the sum of the wavelength pixel coarse offset and the wavelength pixel fine offset. The accuracy of the wavelength pixel coarse offset is 1 interpolation point, and the accuracy of the wavelength pixel fine offset is 0.0001 interpolation points.
[0058] In step S4, s' is substituted into the formula s'=A*c', and the inversion concentration value c'=[c1',c2'…,c m ').
[0059] In step S5, c and c' are fitted to obtain the concentration calibration coefficient P, as follows:
[0060] Let c be X and c′ be Y. Perform a second-order polynomial fitting on X and Y: Y = p²X 2 +p1X+p0,P=[p0,p1,p2].
[0061] In step S6, the factory calibration coefficient of the target analyzer is calculated based on the reference calibration coefficient A and the concentration calibration coefficient P as B = A * P = [a1 * P, a2 * P, ..., a k *P] T .
[0062] This invention first utilizes a reference analyzer to perform multi-point calibration to obtain reference calibration coefficients. To ensure calibration accuracy, sufficient spectral data needs to be acquired, thus requiring the introduction of standard gases of different concentrations for a sufficiently long time. Assuming the introduction time for a standard gas of one concentration is t, the calibration time required for n sets of standard gases of different concentrations is approximately n*t. Now, m sets of standard gases of different concentrations are introduced into the target analyzer, where m is much smaller than n. Then, the concentration calibration coefficient P is obtained by sequentially performing wavelength pixel point calibration, concentration inversion calculation, and fitting calculation on the spectral curve within the target analyzer. Finally, matrix calculation is performed based on the reference calibration coefficient A and the concentration calibration coefficient P to obtain the factory calibration coefficient B of the target analyzer. At this point, the number of pixels contained in the factory calibration coefficient of the target analyzer is equal to the number of pixels contained in the reference calibration coefficient, ensuring the factory calibration accuracy of the target analyzer. Furthermore, the calibration time required for each target analyzer is approximately m*t, which can significantly shorten the time when performing factory calibration on multiple analyzers.
[0063] To more clearly illustrate the implementation process of this method, we will now take the introduction of SO2 gas of different concentrations as an example to demonstrate the factory calibration of the target analyzer. Figure 2 As shown, a multi-point calibration method for a gas analyzer includes the following steps:
[0064] S101. Nitrogen gas and 11 sets of SO2 gas with concentrations C = [750, 600, 450, 300, 150, 100, 80, 60, 40, 20, 5] are introduced into the reference analyzer. The reference analyzer is calibrated at the factory using a multi-point calibration method, and the reference spectral curve data of SO2 gas at each concentration are obtained, denoted as S = [S1, S2, ..., S...]. 11 ] T ;
[0065] S102, the 300nm absorption band spectrum curve of SO2 gas at each concentration, select k wavelength pixels. In actual calculation, k is greater than or equal to 150, but in order to facilitate the specific explanation of the steps of this method, in this embodiment, 5 absorption peak pixels are selected for each spectrum curve for calculation.
[0066] S103, based on the functional relationship between the standard gas concentration C and the reference spectral curve S, S=A*C, the reference calibration coefficient of the reference analyzer is calculated using a fitting algorithm.
[0067] A=[13.50,-10.05,10.21,-11.02,12.76];
[0068] S104. Nitrogen gas and three sets of SO2 gas with concentration values of c = [750, 100, 5] are introduced into the target analyzer. The target spectral curve data of SO2 gas at each concentration are obtained and denoted as s = [s1, s2, s3].
[0069] S105, using the reference spectral curve as a standard, can calibrate the wavelength pixel position of the target spectral curve in the target analyzer through interpolation translation algorithm, neural network deep learning algorithm, or wavelength-pixel mapping method.
[0070] In this embodiment, an interpolation translation algorithm is preferably used, and the steps are as follows:
[0071] S1051, select the reference spectral curve S1 and the target spectral curve s1 to calculate the wavelength pixel offset Δλ of the target spectral curve. For example... Figure 5 As shown, a reference spectral curve and a target spectral curve with a concentration of 750 are selected to calculate the overall wavelength pixel offset of the target spectral curve.
[0072] S1052, extract all pixel intervals in the target spectral curve that include the standard gas absorption band, select a wider interval than the above pixel interval as a wavelength window, and interpolate all pixels of the target spectrum within the wavelength window to obtain the interpolated target spectral curve.
[0073] In this embodiment, the wavelength window pixel range is selected as 5-132, and FFT interpolation is used to interpolate 16 points between every two adjacent pixels.
[0074] S1053 calculates the wavelength pixel offset between the interpolated target spectral curve and the reference spectral curve using the correlation coefficient method or the least squares method.
[0075] This example uses the correlation coefficient method for calculation, and the steps include:
[0076] a. Starting from the leftmost pixel of the wavelength window and ending at the rightmost pixel of the wavelength window, extract one interpolation point every 16 pixels of the interpolated target spectral curve to obtain 321 new target spectral curves.
[0077] b. Calculate the correlation coefficient between each of the new target spectral curves and the reference spectral curves;
[0078] c. Compare all correlation coefficients and select the target spectral curve with the largest correlation coefficient with the reference spectral curve. In this embodiment, the target spectral curve with the largest correlation coefficient with the reference spectral curve is the 18th target spectral curve. Therefore, the coarse offset of the wavelength pixel is 18 interpolation points.
[0079] d. Using the polynomial fitting method, the origin and the points symmetrical to the left and right of the origin are defined as X values, and the maximum correlation coefficient between the reference spectrum curve and the target background spectrum and the correlation coefficients adjacent to the maximum correlation coefficient are taken as Y values. Polynomial fitting is performed on X and Y.
[0080] e. The x-coordinate of the extreme point of the multivariate fitting curve is 0.5225, then the fine offset of the wavelength pixel is 0.4854 interpolation points;
[0081] Therefore, the wavelength pixel offset Δλ of the target spectral curve = coarse offset + fine offset = 18.5225.
[0082] The target spectral curve achieves an accuracy of 0.0001 interpolation points through two processes: coarse offset calculation and fine offset calculation. This improves the correction accuracy of spectral wavelength pixels and further ensures the accuracy of subsequent concentration inversion calculation and concentration calibration coefficient calculation.
[0083] S1054. Based on Δλ, the wavelength pixel positions of the target spectral curve s=[s1,s2,s3] are calibrated to obtain a new set of spectral data, denoted as s'=[s1',s2',s3').
[0084] S106, input s'=[s1',s2',s3'] into the benchmark analyzer, substitute s' into the formula s'=A*c', and use an iterative optimization algorithm to calculate the inversion concentration values c'=[728.52,95.58,4.52] of each standard gas. In this embodiment, the iterative optimization algorithm uses the steepest descent method.
[0085] S107, perform binomial fitting on c and c', let c = [750, 100, 5] be X, and c' = [728.52, 95.58, 4.52] be Y, Y = p2X 2 +p1X+p0, we get the concentration calibration coefficient P=[-0.000023,1.0454,0.2774].
[0086] S108, the factory calibration coefficient of the target analyzer is calculated based on the reference calibration coefficient A and the concentration calibration coefficient P as B = A * P = [a1 * P, a2 * P, ..., a k *P] T
[0087] The functional relationship between the inverted gas concentration data c′ and the calibrated target spectral data s′ in the target analyzer is s′=A*c′;
[0088] Let c be X and c′ be Y. Perform a second-order polynomial fitting on X and Y: Y = p²X 2 +p1X+p0, P=[p0,p1,p2], therefore c′=P*[1,c,c 2 ] T ;
[0089] s′=A*(P*[1,c,c 2 ] T )=B*[1,c,c 2 ] T
[0090] B = A * P = [a1 * P, a2 * P, ..., a k *P] T
[0091] The factory calibration coefficients of the target analyzer were calculated as follows:
[0092]
[0093] At this point, the factory calibration coefficient B of the target analyzer is a given 3*5 matrix with 5 pixel coordinates. In actual calculations, when the pixel value k of each absorption peak in the spectral curve is greater than or equal to 150, the reference calibration coefficient A of the reference analyzer has k pixel coordinates. Figure 3 As shown, after matrix calculation, the pixel coordinates of the factory calibration coefficient B are also k, as follows. Figure 4 As shown.
[0094] To meet accuracy requirements, the traditional multi-point calibration method for the target analyzer requires calibrating with a sufficient number of standard gases of various concentrations, taking approximately 120 minutes. This new method, however, can obtain factory calibration coefficients with comparable accuracy to the traditional multi-point calibration method by collecting fewer standard gas spectral data, taking approximately 30 minutes. This significantly reduces calibration time when calibrating multiple target analyzers. To further verify the accuracy of this calibration method, SO2 gas with concentrations of [600, 450, 300, 150, 80, 60, 40, 20] was introduced into the target analyzer. After factory calibration using the above method, the analyzer performed inversion calculations to obtain the concentration value of SO2 gas at each concentration. The full-scale indication error for each concentration is as follows: Figure 6 As shown, it can be seen that using this method to calibrate the target analyzer can not only greatly reduce the amount of standard gas supplied and save production costs, but also shorten the factory calibration time. The error of the inversion concentration indication is also within 1%, which fully meets the high precision requirements of the analyzer.
[0095] like Figure 7 As shown, this application also discloses a multi-point calibration system for a gas analyzer, including a reference analyzer and a target analyzer. The reference analyzer includes a first measuring cell and a first spectrometer, and the target analyzer includes a second measuring cell and a second spectrometer.
[0096] The reference analyzer also includes a calibration calculation module, a storage module, and a concentration calibration module. Standard gases of different concentrations are introduced into the measurement cell. The spectrometer is used to acquire the reference spectral curve data of the standard gases of each concentration and send them to the calibration calculation module and the storage module respectively. The calibration calculation module calculates the reference calibration coefficient and sends it to the storage module.
[0097] The target analyzer also includes a pixel calibration module and a second calibration calculation module. Standard gases of different concentrations are introduced into the second measurement cell. The second spectrometer acquires the target spectral curve data of each concentration of standard gas and sends it to the second calibration calculation module and the second pixel calibration module, respectively. The reference spectral curve data is input into the second pixel calibration module. The second pixel calibration module is used to calibrate the position of the target spectral curve according to the reference spectral curve and sends the calibrated target spectral curve data to the concentration calibration module of the reference analyzer. The concentration calibration module calculates the inversion concentration value of each standard gas based on the calibrated target spectral curve and calculates the concentration calibration coefficient based on the standard concentration value and the inversion concentration value of each standard gas. The calibration calculation module is used to calculate the factory calibration coefficient of the target analyzer based on the concentration calibration coefficient and the reference calibration coefficient.
[0098] A non-transitory computer-readable storage medium includes instructions for performing a multi-point calibration method for a gas analyzer as described in any of the above embodiments.
[0099] An electronic device includes a non-transitory computer-readable storage medium; and one or more processors capable of executing the instructions of the non-transitory computer-readable storage medium.
[0100] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0101] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0102] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0103] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0104] Specific embodiments have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.
Claims
1. A multi-point calibration method for a gas analyzer, characterized in that, Includes the following steps: S1, n C = [C1, C2, ..., C] are introduced into the reference analyzer. n Standard gases of different concentrations were used to obtain reference spectral curve data. The reference analyzer was then calibrated at the factory using a multi-point calibration method to obtain the reference calibration coefficient A. S2, introduce m c = [c1, c2, ..., c] into the target analyzer. m ]∈C standard gases of different concentrations, to obtain target spectral curve data s=[s1,s2,…,s m ],3≤m <n; S3, using the reference spectrum curve of the reference analyzer as the standard, calibrate the wavelength pixel position of the target spectrum curve obtained in the target analyzer under the same gas concentration. S4. Input the calibrated target spectral curve data s' into the reference analyzer, and obtain the inversion concentration values c'=[c1',c2'…,c...] of each standard gas through the reference calibration coefficient A. m ']; S5, fit c and c' to obtain the concentration calibration coefficient P; S6. Input the reference calibration coefficient A and the concentration calibration coefficient P into the target analyzer to calculate the factory calibration coefficient B of the target analyzer.
2. The multi-point calibration method for a gas analyzer according to claim 1, characterized in that, In step S1, the reference analyzer is calibrated at the factory using a multi-point calibration method to obtain the reference calibration coefficient A, including the following steps: S11, n C = [C1, C2, ..., C] are introduced into the reference analyzer. n For standard gases of different concentrations, obtain the reference spectral curve data S=[S1,S2,…,S…] for each concentration of standard gas. n ] T ; S12, select k wavelength pixels on the spectral curve of each concentration standard gas; S13, based on the functional relationship between the standard gas concentration C and the spectral amplitude S, S=A*C, the reference calibration coefficient A=[a1,a2,…,a] is calculated using a fitting algorithm. k ].
3. The multi-point calibration method for a gas analyzer according to claim 1, characterized in that, In step S3, the wavelength pixel positions of the target spectral curve obtained in the target analyzer under the same gas concentration are calibrated by interpolation translation algorithm, neural network deep learning algorithm, or wavelength-pixel mapping method.
4. The multi-point calibration method for a gas analyzer according to claim 3, characterized in that, The interpolation translation algorithm steps are as follows: S31, Select curve data at the same concentration from the reference spectral curve and the target spectral curve, and calculate the wavelength pixel offset Δλ of the target spectral curve. S32, extract all pixel intervals in the target spectral curve that include the standard gas absorption band, select a wider interval than the above pixel interval as a wavelength window, and interpolate all pixels of the target spectrum within the wavelength window to obtain the interpolated target spectral curve. S33, calculate the wavelength pixel offset between the interpolated target spectral curve and the reference spectral curve; S34, calibrate the spectral curves obtained at various gas concentrations in the target analyzer by shifting the target spectral curve according to the wavelength offset.
5. The multi-point calibration method for a gas analyzer according to claim 4, characterized in that, The wavelength pixel offset is the sum of the wavelength pixel coarse offset and the wavelength pixel fine offset. The accuracy of the wavelength pixel coarse offset is 1 interpolation point, and the accuracy of the wavelength pixel fine offset is 0.0001 interpolation points.
6. The multi-point calibration method for a gas analyzer according to claim 1, characterized in that, In step S4, s' is substituted into the formula s'=A*c', and the inversion concentration value of the standard gas c'=[c1',c2'…,c m ').
7. The multi-point calibration method for a gas analyzer according to claim 1, characterized in that, In step S5, let c be X and c′ be Y, and perform a second-order polynomial fitting on X and Y: Y = p2X 2 +p1X+p0, we get the concentration calibration coefficient P=[p0,p1,p2].
8. The multi-point calibration method for a gas analyzer according to any one of claims 1 to 7, characterized in that, In step S6, the factory calibration coefficient of the target analyzer is calculated based on the reference calibration coefficient A and the concentration calibration coefficient P as B = A * P = [a1 * P, a2 * P, ..., a k *P] T .
9. A multi-point calibration system for a gas analyzer, comprising a reference analyzer and a target analyzer, wherein the reference analyzer includes a first measuring cell and a first spectrometer, and the target analyzer includes a second measuring cell and a second spectrometer; characterized in that, The reference analyzer also includes a calibration calculation module, a storage module, and a concentration calibration module. Standard gases of different concentrations are introduced into the measurement cell. The spectrometer is used to acquire the reference spectral curve data of the standard gases of each concentration and send them to the calibration calculation module and the storage module respectively. The calibration calculation module calculates the reference calibration coefficient and sends it to the storage module. The target analyzer also includes a pixel calibration module and a second calibration calculation module. Standard gases of different concentrations are introduced into the second measurement cell. The second spectrometer acquires the target spectral curve data of each concentration of standard gas and sends it to the second calibration calculation module and the second pixel calibration module, respectively. The reference spectral curve data is input into the second pixel calibration module. The second pixel calibration module is used to calibrate the position of the target spectral curve according to the reference spectral curve and sends the calibrated target spectral curve data to the concentration calibration module of the reference analyzer. The concentration calibration module calculates the inversion concentration value of each standard gas based on the calibrated target spectral curve and calculates the concentration calibration coefficient based on the standard concentration value and the inversion concentration value of each standard gas. The calibration calculation module is used to calculate the factory calibration coefficient of the target analyzer based on the concentration calibration coefficient and the reference calibration coefficient.