In-situ measurement method and system of ammonia escape based on laser spectroscopy technology

By analyzing the laser spectrum and flow data at historical moments, filtering the characteristic wavelength and time lag, adjusting the spectral data and filtering, the problems of environmental complexity and concentration in the measurement of ammonia escape amount are solved, and more accurate Ammonia escape measurement.

CN118518627BActive Publication Date: 2024-11-05山东创宇能源科技股份有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410575632.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-05-10
Publication Date
2024-11-05
Estimated Expiration
2044-05-10

AI Technical Summary

Technical Problem

In the measurement of ammonia escape amount, the measurement results are inaccurate due to the complex measurement environment and uneven distribution of ammonia concentration.

Method used

By obtaining laser spectral data and flow timing data at historical moments, filtering characteristic wavelengths, analyzing time lag and wavelength offsets, adjusting spectral data, obtaining final absorption timing data, and improving measurement accuracy through filtering and denoising processing.

Benefits of technology

More accurate measurement of ammonia escape amount is achieved, reducing the impact of environmental factors and wavelength shift on the measurement results, and improving data quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118518627B_ABST
    Figure CN118518627B_ABST
Patent Text Reader

Abstract

The present invention relates to the technical field of laser measuring instruments, and specifically to an in-situ measurement method and system for ammonia escape based on laser spectroscopy technology. Reductant flow time series data, spectral data and initial absorbance time series data within a preset historical period are obtained, and then characteristic wavelengths are screened out according to the change trend of the absorbance. Then, the changes between the initial absorbance time series data and the reductant flow time series data and related conditions are analyzed to determine the time lag and the optimal wavelength offset. After correcting the spectral data based on the optimal offset, the final absorbance time series data of the characteristic wavelength more accurately reflects the change in the absorbance of the material, and then analyzes the unevenness of the ammonia concentration in the pipeline. By analyzing the change fluctuations and data errors of the final absorbance time series data, determining the abnormality degree value, and performing filtering and denoising processing, the denoised time series data obtained provides a more accurate basis for the measurement of the ammonia escape amount at the current moment, thereby improving the accuracy of the measurement.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of laser measuring instruments, and in particular to an in-situ measurement method and system for ammonia escape based on laser spectroscopy technology. Background Art

[0002] Laser spectroscopy technology has been widely used and studied in the in-situ measurement of ammonia escape. The main advantages of this technology are its high sensitivity, rapid response, non-contact and real-time performance. In the denitrification process, the use of laser spectroscopy to accurately measure the amount of ammonia escape can achieve effective monitoring and avoid excessive ammonia producing more ammonium salts that clog the air preheater. Therefore, accurate measurement of ammonia escape is particularly important.

[0003] In the prior art, when measuring the amount of ammonia escape, regression analysis is usually performed using the ammonia absorption rate in the laser spectrum to obtain the amount of ammonia escape. However, since the measurement environment is inside a pipeline, the temperature and other factors will fluctuate to a certain extent, and the ammonia concentration distribution is not uniform, the final measurement result of the ammonia escape will be inaccurate. Summary of the invention

[0004] In order to solve the technical problem that the measurement environment is relatively complex when measuring ammonia slip, and the uneven distribution of ammonia concentration will lead to inaccurate measurement results of ammonia slip using the ammonia absorbance in laser spectroscopy data, the purpose of the present invention is to provide an ammonia slip in-situ measurement method and system based on laser spectroscopy technology, and the technical scheme adopted is as follows:

[0005] Obtain the flow time series data of the reducing agent in a preset historical period and the laser spectrum data at each historical moment; obtain the initial absorption rate time series data of each wavelength according to the absorption rate of each wavelength in each laser spectrum data;

[0006] Among all the laser spectrum data, according to the absorption rate change trend of the wavelength, the characteristic wavelength is screened out; according to the change of the initial absorption rate time series data and the flow time series data of all the characteristic wavelengths, the time lag is obtained; any historical moment is selected as the time to be measured; after the laser spectrum data at the time to be measured is offset at each preset wavelength offset, the updated absorption rate time series data of each characteristic wavelength is obtained; under the time lag, the correlation between all the updated absorption rate time series data and the flow time series data corresponding to different preset wavelength offsets in the preset time neighborhood corresponding to the time to be measured is analyzed, and the optimal offset of the laser spectrum data at the time to be measured is screened out;

[0007] After adjusting the corresponding laser spectrum data according to each optimal offset, the final absorption rate time series data of each characteristic wavelength is obtained; the abnormality degree value corresponding to each historical moment is obtained according to the change fluctuation of all the final absorption rate time series data; the final absorption rate time series data of each characteristic wavelength is filtered and denoised according to the abnormality degree value to obtain denoised time series data;

[0008] The ammonia slip amount at the current moment is detected based on all the denoised time series data.

[0009] Furthermore, the method for acquiring the characteristic wavelength includes:

[0010] For any wavelength, the absorption rates of the wavelength in the laser spectrum data at all historical moments are accumulated as the absorption rate and value of the wavelength;

[0011] A preset wavelength range is constructed with each wavelength as the center; in each preset wavelength range, other wavelengths except the central wavelength are used as comparison wavelengths, and when the absorbance and value of all comparison wavelengths are less than the absorbance and value of the central wavelength, the central wavelength is used as the characteristic wavelength.

[0012] Furthermore, the method for acquiring the initial absorption rate time series data includes:

[0013] The absorbance of each wavelength in the laser spectrum data at all historical moments is arranged in time series to obtain the initial absorbance time series data of each wavelength.

[0014] Furthermore, the method for obtaining the time lag includes:

[0015] Obtain the historical moment of the maximum value in the traffic time series data as the target moment, for any target moment, calculate the slope value of the fitting straight line of the data values ​​of the first number of historical moments preset before the target moment as the first slope, calculate the slope value of the fitting straight line of the data values ​​of the first number of historical moments preset after the target moment as the second slope, and normalize the difference between the first slope and the second slope as the convexity value corresponding to the target moment;

[0016] A preset hysteresis value range, for any hysteresis value in the preset hysteresis value range, calculate the time matching deviation value between the initial absorption rate time series data of all characteristic wavelengths and the flow time series data under the hysteresis value; among all the time matching deviation values, the hysteresis value corresponding to the smallest time matching deviation value is used as the time lag;

[0017] The method for obtaining the moment matching deviation value includes: selecting any characteristic wavelength as the wavelength to be measured; taking the historical moments corresponding to all the maximum values ​​of the initial absorption rate time series data of the wavelength to be measured as the comparison moments; for any target moment, calculating the numerical difference between the value after the target moment plus the hysteresis value and each comparison moment, multiplying the minimum numerical difference by the protrusion degree value corresponding to the target moment, and obtaining the deviation factor of the target moment; taking the value after accumulating the deviation factors of all target moments as the deviation sum value corresponding to the wavelength to be measured; and normalizing the value after accumulating the deviation sum values ​​of all characteristic wavelengths to obtain the moment matching deviation value.

[0018] Furthermore, the method for obtaining the optimal offset includes:

[0019] Under each preset wavelength offset, the updated absorbance time series data corresponding to each characteristic wavelength is subjected to STL time series decomposition to obtain the initial trend item time series data corresponding to each characteristic wavelength;

[0020] Each of the initial trend item time series data is offset by the time lag amount to obtain the final trend item time series data. The time lag amount offset process is: offset the moment axis of each initial trend item time series data in the direction where the moment becomes larger by the time lag amount units;

[0021] Calculate the Pearson correlation coefficient between each final trend item time series data and the flow time series data within the preset time neighborhood as a correlation value; and use the cumulative sum of the correlation values ​​corresponding to all characteristic wavelengths as a judgment value;

[0022] Among all the judgment values ​​corresponding to the preset wavelength offsets, the preset wavelength offset corresponding to the maximum judgment value is used as the optimal offset of the laser spectrum data at the time to be measured.

[0023] Furthermore, the preset time neighborhood is set to be centered at the time to be measured, and a preset second number of historical moments are taken on both sides of the front and back to form the time neighborhood.

[0024] Furthermore, the method for obtaining the abnormality degree value includes:

[0025] Perform STL time series decomposition on the final absorbance time series data of each characteristic wavelength to obtain seasonal term time series data and residual term time series data;

[0026] In the seasonal item time series data and the residual item time series data corresponding to each characteristic wavelength, a first abnormal factor at each historical moment is obtained according to the seasonal item data value and the residual item data value at each historical moment, wherein the first abnormal factor is positively correlated with the seasonal item data value, and the first abnormal factor is negatively correlated with the residual item data value;

[0027] The sum of the first anomaly factors of each historical moment at all characteristic wavelengths is taken as the second anomaly factor; within the preset time neighborhood centered on each historical moment, the sum of the second anomaly factors of all historical moments is normalized to a value that is the anomaly degree value of the central historical moment.

[0028] Furthermore, the method for obtaining denoised time series data includes:

[0029] Preset filter coefficients;

[0030] Obtaining a final filter coefficient at each historical moment according to the abnormality degree value at each historical moment and the preset filter coefficient, wherein both the abnormality degree value and the preset filter coefficient are positively correlated with the final filter coefficient;

[0031] The final absorption rate time series data of each characteristic wavelength is Gaussian filtered according to the final filtering coefficient at each historical moment to obtain the denoised time series data of each characteristic wavelength.

[0032] Further, the detecting of ammonia slip amount according to all the denoised time series data includes:

[0033] Get the actual concentration of ammonia at each historical moment;

[0034] The normalized value of the sum of the absorbances at the same historical moment in the denoised time series data of all characteristic wavelengths is used as the adjustment coefficient, the product of the sum of the absorbances at the same historical moment in the initial absorbance time series data of all characteristic wavelengths and the corresponding adjustment coefficient is used as the independent variable, and the actual concentration of ammonia at the same historical moment is used as the dependent variable, and PLS partial least squares is performed to obtain a regression model;

[0035] The product of the absorbance and value of all characteristic wavelengths in the laser spectrum data at the current moment and the adjustment coefficient is used as the input of the regression model, and the output is the actual concentration of ammonia at the current moment.

[0036] The present invention also proposes an in-situ measurement system for ammonia escape based on laser spectroscopy technology, the system comprising:

[0037] A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the steps of any one of the methods can be implemented when the processor executes the computer program.

[0038] The present invention has the following beneficial effects:

[0039] The present invention first obtains the flow time series data of the reducing agent within a preset time period, the spectral data at each historical moment, and extracts the initial absorption rate time series data of each wavelength. Based on the change of the absorption rate of the wavelength in the spectral data, the characteristic wavelength is screened out. Due to the influence of environmental factors, the change of the absorption rate in the spectral data caused by the reducing agent will have a time lag and the wavelength in the spectral data will have a certain offset. Therefore, the change between the initial absorption rate time series data of the characteristic wavelength and the flow time series data of the catalyst is analyzed to determine the time lag; at the same time, under different preset wavelength offsets, the updated absorption rate time series data of the characteristic wavelength is obtained, and combined with the obtained time lag, the correlation between the updated absorption rate time series data and the flow time series data is analyzed, so as to screen out the optimal offset of the wavelength offset of the spectral data at each historical moment. Furthermore, after the spectral data is corrected according to the optimal offset, the final absorbance time series data of the characteristic wavelength at this time can more accurately characterize the change in the absorbance of the substance. At this time, the uneven ammonia concentration inside the pipeline can be analyzed, which is specifically reflected in the change and fluctuation of the data. Therefore, the change and fluctuation of the final absorbance time series data are analyzed to obtain the abnormality degree value corresponding to each historical moment. Finally, the final absorbance time series data is filtered and denoised according to the abnormality degree value, and high-quality denoised time series data corresponding to the absorbance of the characteristic wavelength is obtained. At this time, the measurement of the ammonia escape amount at the current moment can be performed based on the denoised time series data to obtain a more accurate measurement result. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 A flow chart of an in-situ measurement method for ammonia escape based on laser spectroscopy technology provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0042] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of an in-situ measurement method and system of ammonia escape based on laser spectroscopy technology proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0043] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0044] The specific scheme of the in-situ measurement method and system of ammonia escape based on laser spectroscopy technology provided by the present invention is described in detail below with reference to the accompanying drawings.

[0045] See also Figure 1 , which shows a method flow chart of an in-situ measurement method of ammonia escape based on laser spectroscopy technology provided by an embodiment of the present invention, the method comprising the following steps:

[0046] Step S1: Obtain the flow time series data of the reducing agent in a preset historical period and the laser spectrum data at each historical moment; obtain the initial absorption rate time series data of each wavelength according to the absorption rate of each wavelength in each laser spectrum data.

[0047] In industrial production, especially in industries involving combustion processes, such as boiler operation, flue gas usually contains nitrogen oxides, which are a strong greenhouse gas that can absorb and re-radiate heat from the earth's surface, leading to global warming. In addition, nitrogen oxides are also a toxic substance that has certain hazards to human health and the environment. Therefore, flue gas needs to be denitrated. In essence, by adding ammonia (reducing agent), under the action of a catalyst, ammonia selectively reacts with nitrogen oxides in the flue gas to generate harmless nitrogen and water, thereby removing harmful substances in the flue gas.

[0048] In the denitrification process, the reducing agent usually uses anhydrous ammonia, ammonia water or urea. However, because the reaction is reversible, if too much ammonia is injected, the reverse reaction will be increased, and more ammonium salts will be produced, which will block the air preheater, reduce system efficiency, and may also cause equipment damage, increased maintenance costs and environmental pollution. Therefore, it is necessary to measure the amount of ammonia escape to facilitate the control of the ammonia concentration.

[0049] When measuring the amount of ammonia escape, in-situ semiconductor laser spectroscopy data is usually used to measure the ammonia escape concentration. In this embodiment of the present invention, the amount of ammonia escape in the tail flue of the denitrification process is detected, and the sampling probe is equipped with a high-precision filter. After filtering the flue gas, the influence of dust scattering on the measurement can be effectively avoided, and then the laser spectrum data at each historical moment in the preset historical period is obtained. Since the generation of ammonia is also related to the input amount of the reducing agent, the flow time series data of the reducing agent used in the preset time period is also obtained; wherein, the wavelength resolution of the laser spectrum data is 1nm, and the spectrum data is obtained every 1 second, and the sampling time interval of the flow time series data of the reducing agent is also 1 second. It should be noted that the data acquisition device can be adjusted according to the implementation scenario, which is not limited here; the preset historical period is set to the first 30 minutes of the current moment, and the specific duration can be changed according to the implementation scenario, which is not limited.

[0050] After obtaining the laser spectrum data at multiple historical moments, since the reaction is ongoing all the time, the absorption rates of various substances to different wavelengths will change at different historical moments. Therefore, here, based on the laser spectrum data at all historical moments, the time series data of the absorption rate of each wavelength changing with time can be obtained, that is, the initial absorption rate time series data of each wavelength, which can be used in subsequent analysis.

[0051] Preferably, in one embodiment of the present invention, the method for acquiring initial absorption rate time series data includes:

[0052] The initial absorption rate time series data of each wavelength can be obtained by arranging the absorption rates of the laser spectrum data of each wavelength at all sampling historical moments in time series.

[0053] At this point, the required data can be obtained, which can be analyzed in the subsequent process to measure the ammonia slip.

[0054] Step S2: Among all the laser spectrum data, the characteristic wavelength is screened out according to the absorption rate change trend of the wavelength; the time lag is obtained according to the change of the initial absorption rate time series data and the flow time series data of all the characteristic wavelengths; any historical moment is selected as the time to be measured; after the laser spectrum data at the time to be measured is offset at each preset wavelength offset, the updated absorption rate time series data of each characteristic wavelength is obtained; under the time lag, the correlation between all the updated absorption rate time series data and the flow time series data corresponding to different preset wavelength offsets in the preset time neighborhood corresponding to the time to be measured is analyzed, and the optimal offset of the laser spectrum data at the time to be measured is screened out.

[0055] Since the measurement environment is inside the flue, its temperature, pressure and other parameters will fluctuate to a certain extent, and this change will affect the intensity of the reaction between substances, and then affect the content of the final generated substance; and the ammonia concentration is not evenly distributed inside the pipeline, so the spectral data will have a certain deviation from the actual data. At this time, if the ammonia escape amount is directly determined by the absorbance at a single wavelength in the laser spectral data, it will cause inaccurate measurement results. Therefore, in an embodiment of the present invention, it is proposed to correct the data and analyze the changes in the absorbance at multiple wavelengths, so as to achieve accurate measurement of the ammonia escape amount.

[0056] Since the reaction is reversible, but the types of substances produced are relatively fixed, this means that during the reaction, certain products will continue to be produced and their concentrations will change. Therefore, this property can be used to select the wavelength with a more significant change in absorbance as the characteristic wavelength.

[0057] Preferably, in one embodiment of the present invention, the method for acquiring the characteristic wavelength includes:

[0058] In order to more accurately characterize the overall absorption effect of each wavelength, for each wavelength, the absorbance in the laser spectrum data at all sampling historical moments is accumulated as the absorbance and value of each wavelength.

[0059] Then, the changing trend of the absorbance and value is analyzed to screen out the characteristic wavelength. Specifically, a preset wavelength range is constructed with each wavelength as the center; in each preset wavelength range, other wavelengths except the central wavelength are used as comparison wavelengths. When the absorbance and value of all comparison wavelengths are less than the absorbance and value of the central wavelength, the central wavelength can be used as the characteristic wavelength, thereby ensuring that the selected characteristic wavelength has the most outstanding absorbance performance within the preset wavelength range.

[0060] It should be noted that the preset wavelength range is set to 11, that is, 5 wavelengths are taken on both sides of the central wavelength. If the number of wavelengths on one side is less than 5, the total number of wavelengths on the other side can be supplemented to 11. The setting of the preset wavelength range can be adjusted according to the implementation scenario and is not limited here.

[0061] At this point, the characteristic wavelengths are screened out by analyzing the comprehensive absorption rate of each wavelength at all historical moments. In the subsequent process, the characteristic wavelengths can be analyzed to measure the ammonia escape amount.

[0062] Due to the influence of measurement environment factors, the wavelength of laser spectrum data at different historical moments will be shifted, and this shift will also be affected by the flow rate of the reducing agent, but this impact may not occur immediately, because there is a certain time delay in the reaction, which leads to a certain time deviation between the change in the data value in the initial absorption rate time series data of the characteristic wavelength and the flow time series data of the reducing agent, which is specifically manifested as time lag. Therefore, before analyzing the wavelength offset of the laser spectrum, it is necessary to analyze the time lag first to eliminate the influence of reaction delay.

[0063] Preferably, in one embodiment of the present invention, the method for obtaining the time lag includes:

[0064] Because the maximum value of a set of data usually represents the local highest point of the data, it is usually more obvious and easier to analyze. Therefore, first obtain the historical moment of the maximum value in the flow time series data as the target moment. For any target moment, calculate the slope value of the fitting line of the data values ​​of the first number of historical moments preset before the target moment as the first slope, and calculate the slope value of the fitting line of the data values ​​of the first number of historical moments preset after the target moment as the second slope. The difference between the first slope and the second slope is normalized to the value obtained as the convexity value corresponding to the target moment. The convexity value mainly represents the flow change of the reducing agent at each target moment, and can be used as a credibility for determining the time lag in the subsequent process. The formula model of the convexity value is:

[0065] TC i =norm(|K1 i -K2 i |)

[0066] Among them, TC i K1 represents the convexity value corresponding to the i-th target moment; i K2 represents the first slope value corresponding to the i-th target moment; i represents the second slope value corresponding to the i-th target moment; norm() represents the normalization function.

[0067] In the formula model of the convexity value, when the difference between the first slope value and the second slope corresponding to each target moment | K1 i -K2 i| is larger, indicating that the data change trends on both sides of the target moment are quite different, which also indicates that at the target moment, the more obvious the flow change of the reducing agent is, the more violent the material reaction should be, which can be used as a credibility indicator in the subsequent determination of the time lag. It should be noted that the preset first number is set to 5, and the specific value can be adjusted according to the implementation scenario. If the number of historical moments on one side is less than 5, then there are several historical moments and the data values ​​of these historical moments can be linearly fitted. If the number of historical moments on one side is 0, the fitting line does not exist, and the slope can be set to 0; the method of linear fitting is a process well known to those skilled in the art, and will not be elaborated or limited here.

[0068] Since the specific value of the time lag is uncertain, the traversal method is used in this embodiment of the present invention to obtain the time lag. A preset lag value range is set. For any lag value in the preset lag value range, the moment matching deviation value between the initial absorbance time series data of all characteristic wavelengths and the flow time series data is calculated under the lag value; among the moment matching deviation values ​​corresponding to all lag values, the lag value corresponding to the smallest moment matching deviation value is used as the time lag. It should be noted that the preset lag value range is set to 0 to 10, and the specific range can be adjusted according to the actual situation, which is not limited here.

[0069] The method for obtaining the moment matching deviation value specifically includes: selecting any characteristic wavelength as the wavelength to be measured; similarly, analyzing the maximum value in the initial absorption rate time series data of the wavelength to be measured, because the flow rate of the reducing agent increases, correspondingly, the generated substances will also increase, so the absorption rate will increase accordingly. The historical moments corresponding to all the maximum values ​​of the initial absorption rate time series data of the wavelength to be measured are used as comparison moments; for any target moment, the numerical difference between the value after adding the hysteresis value to the target moment and each comparison moment is calculated, and the minimum numerical difference is multiplied by the convexity value of the target moment to obtain the deviation factor of the target moment; the value after the deviation factors of all target moments are accumulated is used as the deviation sum value. At this time, by analyzing the initial absorption rate time series data of a characteristic wavelength, the deviation sum value corresponding to the hysteresis value is obtained, and then the value after the deviation sum values ​​of all characteristic wavelengths under the hysteresis value are accumulated is normalized to obtain the moment matching deviation value under the hysteresis value.

[0070] The formula model of the time matching deviation value may be specifically, for example:

[0071]

[0072] Among them, Δt j represents the time matching deviation value under the jth hysteresis value; L represents the total number of characteristic wavelengths; I represents the total number of target moments; TC i represents the convexity value at the i-th target moment; Ti represents the historical moment value of the i-th target moment; Z j represents the jth lag value; Represents the historical time value corresponding to the rth maximum value in the initial absorption rate time series data of the lth characteristic wavelength; n l It represents the total number of maximum values ​​in the initial absorbance time series data of the lth characteristic wavelength; min() represents the minimum value; norm() represents the normalization function.

[0073] In the formula model of the time matching deviation value, for each hysteresis value, first, in any characteristic wavelength, the historical moment corresponding to each maximum value in the initial absorbance time series data of the characteristic wavelength is used as the comparison moment; at this time, the numerical difference between the value of each target moment in the flow time series data of the reducing agent plus the hysteresis value and the value of each comparison moment in the characteristic wavelength is analyzed. The smaller the numerical difference is, the closer the time distance between a maximum value with a more obvious change in the flow time series data of the reducing agent and a maximum value with a more obvious change in the initial absorption rate time series data of the characteristic wavelength is, that is, the higher the matching degree between the target time and the historical time corresponding to the maximum value in the initial absorption rate time series data; and the larger the convexity value of the target time, the more intense the material reaction should be, and the higher the confidence, so the minimum numerical difference is multiplied by the convexity value of the target time to obtain the deviation factor of the target time. Based on this method, the deviation factor of each target moment can be obtained, all deviation factors are accumulated, and the accumulated value is used as the deviation and value Finally, the deviations and values ​​corresponding to all characteristic wavelengths are accumulated and normalized to obtain the moment matching deviation value under the hysteresis value; and the smaller the moment matching deviation value, the better the match between the initial absorption rate time series data of the characteristic wavelength and the changing trend of the flow time series data of the reducing agent under the hysteresis value, and the hysteresis value is more likely to be an accurate time lag.

[0074] At this point, the analysis has obtained the time deviation between the change in the data value in the initial absorption rate time series data of the characteristic wavelength and the flow time series data of the reducing agent, that is, the time lag. Then, the influence of the reaction delay can be eliminated based on the time lag, and the wavelength offset of the laser spectrum data can be accurately analyzed.

[0075] The time lag is the overall time difference between the initial absorption rate time series data and the flow time series data, while the wavelength offset of the laser spectrum data is affected by environmental factors at different historical moments, which will produce different offsets. Therefore, it is necessary to analyze the wavelength offset of the laser spectrum data at each historical moment. For the sake of convenience of explanation and illustration, one historical moment is selected as the moment to be measured.

[0076] Similarly, since the specific value of the wavelength offset is uncertain, the traversal method is used in this embodiment of the present invention to obtain the optimal offset of each laser spectrum data. Different preset wavelength offsets are set, and then the laser spectrum data at the time to be measured is wavelength offset at each preset wavelength offset. The specific operation is: the absorbance value remains unchanged, and the wavelength is offset. For example: if the wavelength offset is 2, the wavelength dimension is moved two units in the direction of increasing wavelength as a whole, that is, before the offset, the absorbance corresponding to wavelength 1 is 5, and the absorbance corresponding to wavelength 3 is 6. After the offset, the absorbance corresponding to wavelength 1 is 6; if the wavelength offset is -1, the wavelength dimension is moved one unit in the direction of decreasing wavelength as a whole, that is, before the offset, the absorbance corresponding to wavelength 2 is 7, and the absorbance corresponding to wavelength 3 is 6. After the offset, the absorbance corresponding to wavelength 3 is 7. It should be noted that the preset wavelength offset is set to -3, -2, -1, 0, 1, 2, 3; the specific value can be adjusted according to the implementation scenario, and is not limited here.

[0077] After the laser spectrum data at the time to be measured completes the wavelength shift, the updated absorbance time series data of each characteristic wavelength is obtained. At this time, the difference between the updated absorbance time series data and the initial absorbance time series data is only the change in the absorbance at the time to be measured.

[0078] Since the length of the updated absorption rate time series data is relatively long, a preset time neighborhood of the time to be measured is determined for the time to be measured.

[0079] Preferably, in one embodiment of the present invention, the preset time neighborhood is set to be centered at the time to be measured, and a preset second number of historical moments are taken on both sides to form the time neighborhood. It should be noted that the preset second number is 10. If the number of historical moments on one side of the time to be measured is less than 10, the total number of 21 can be supplemented on the other side.

[0080] Then, combined with the time lag, the correlation between the updated absorbance time series data and the flow time series data under each preset wavelength offset is analyzed, that is, the influence of the reducing agent flow rate on the absorbance is analyzed, and the optimal offset of the laser spectrum data at the time to be measured is screened out.

[0081] Preferably, in one embodiment of the present invention, the method for obtaining the optimal offset includes:

[0082] Since the trend term can capture the long-term growth or decrease pattern in the time series, it is helpful to evaluate the similarity or difference. Therefore, at each preset wavelength offset, the updated absorbance time series data corresponding to each characteristic wavelength is decomposed by STL time series to obtain the initial trend term time series data corresponding to each characteristic wavelength.

[0083] Then each initial trend item time series data is offset by the time lag to obtain the final trend item time series data. The specific operation is to offset the moment axis in the initial trend item time series data in the direction of increasing time. The offset unit is equal to the time lag. For example: if the time lag is 1, the data value corresponding to historical moment 2 is 5, and the data value corresponding to historical moment 3 is 6, then in the final trend item time series data, the data value corresponding to historical moment 2 should be 6.

[0084] At this point, the final trend item time series data has eliminated the time delay, so the accuracy of the results is improved in the subsequent calculation process. Then calculate the Pearson correlation coefficient between each final trend item time series data and the flow time series data in the preset time neighborhood as the correlation value. The larger the value and the closer it is to 1, the better the performance of the absorbance for the actual substance content concentration within the local time range under the preset wavelength offset, that is, the more positively correlated the change between the absorbance and the flow rate of the reducing agent, indicating that the preset wavelength offset is more likely to be the offset of the real and accurate laser spectrum data, and then the cumulative sum of the correlation values ​​corresponding to all characteristic wavelengths is used as the judgment value. At this time, based on the above method, the corresponding judgment value under each preset wavelength offset can be obtained.

[0085] Finally, the preset wavelength offset corresponding to the maximum judgment value is taken as the optimal offset of the laser spectrum data at the time to be measured, denoted as PY u Wherein, u represents the time to be measured. It should be noted that the STL time series decomposition algorithm is a technical means well known to those skilled in the art, and will not be described in detail here.

[0086] At this point, the optimal offset of the laser spectrum data at each historical moment can be obtained through the above process. In the subsequent process, each laser spectrum data can be adjusted and then analyzed.

[0087] Step S3: After adjusting the corresponding laser spectrum data according to each optimal offset, obtain the final absorption rate time series data of each characteristic wavelength; obtain the abnormality degree value corresponding to each historical moment according to the change fluctuation of all the final absorption rate time series data; filter and denoise the final absorption rate time series data of each characteristic wavelength according to the abnormality degree value to obtain denoised time series data.

[0088] After obtaining the optimal offset corresponding to the laser spectrum data at each historical moment, each laser spectrum data can be corrected and adjusted. The specific operation is that the absorbance value remains unchanged. If the optimal offset is a positive number, the wavelength is offset by the optimal offset units in the direction of increasing wavelength, and the adjusted laser spectrum data can be obtained; if the optimal offset is a negative number, the wavelength is offset by the optimal offset units in the direction of decreasing wavelength. At this time, the absorbance time series data of the characteristic wavelength can be obtained again and used as the final absorbance time series data.

[0089] The final absorbance time series data has eliminated the effects of time lag and wavelength offset. However, considering that the location of ammonia concentration detection at the end of the flue of the denitrification process is generally in a flue section of 5×10m or even larger, there will be greater uncertainty in the flow of the detection surface in the flue, including catalyst blockage and other situations, which will cause a certain degree of unevenness in the concentration of ammonia. As a result, the obtained spectral absorbance data will also produce certain fluctuations within a smaller time period. This fluctuation is not caused by the actual change in ammonia concentration, but by the uneven distribution of ammonia concentration. Therefore, it is necessary to analyze this uneven situation and then filter and denoise the final absorbance data to eliminate the impact of this situation.

[0090] First, according to the fluctuation of the final absorption rate time series data and the data error, the abnormality degree value corresponding to each historical moment is obtained. The abnormality degree value can be used as a reference indicator for subsequent filtering and denoising of the final absorption rate time series data.

[0091] Preferably, in one embodiment of the present invention, the method for obtaining the abnormality degree value includes:

[0092] The trend term can reflect the overall trend of data changes over time, while the season term emphasizes more on periodicity, and the residual term is more used to reflect the abnormal situation of the data. Therefore, in this embodiment of the present invention, the season term and the residual term are selected to calculate the abnormality value.

[0093] The final absorbance time series data of each characteristic wavelength is decomposed by STL time series to obtain seasonal item time series data and residual item time series data; in the seasonal item time series data and residual item time series data corresponding to each characteristic wavelength, the first abnormal factor of each historical moment is obtained according to the seasonal item data value and the residual item data value at each historical moment, and the first abnormal factor is positively correlated with the seasonal item data value, and the first abnormal factor is negatively correlated with the residual item data value. Among them, the positive correlation relationship indicates that the dependent variable will increase as the independent variable increases, and the dependent variable will decrease as the independent variable decreases. The specific relationship can be a multiplication relationship, an addition relationship, and the power of an exponential function, which is determined by actual application; the negative correlation relationship indicates that the dependent variable will decrease as the independent variable increases, and the dependent variable will increase as the independent variable decreases. It can be a subtraction relationship, a division relationship, etc., which is determined by actual application.

[0094] The sum of the first anomaly factors of each historical moment at all characteristic wavelengths is taken as the second anomaly factor; within the preset time neighborhood centered on each historical moment, the sum of the second anomaly factors of all historical moments is normalized to the value as the anomaly degree value of the central historical moment. Taking historical moment a as an example of the central historical moment, the formula model of the anomaly degree value can be specifically, for example:

[0095]

[0096] Among them, YC a represents the abnormality value of the central historical moment a; D represents the total number of historical moments in the preset time neighborhood; L represents the total number of characteristic wavelengths; It represents the value of the central historical moment a in the seasonal time series data corresponding to the lth characteristic wavelength; It represents the value of the central historical moment a in the residual term time series data corresponding to the lth characteristic wavelength; γ represents the preset parameter; norm() represents the normalization function.

[0097] In the formula model of the abnormality degree value, since the seasonal term can represent the component of relatively high-frequency changes in the data, the larger the value of the seasonal term is, the greater the fluctuation of the data in a short period of time can be regarded as, and it can be used to represent the uneven distribution of ammonia concentration in the measurement environment, that is, the higher the abnormality degree is, and the residual term can represent the reliability of the STL decomposition structure. The smaller the value of the residual term is, the higher the reliability of the decomposition is. Therefore, in the seasonal term time series data and the residual term time series data corresponding to the characteristic wavelength, the seasonal term time series data value at each historical moment is used as the numerator, and the residual term time series data value at each historical moment is used as the denominator, so as to obtain the first abnormal factor at each historical moment. Then the first anomaly factors at all characteristic wavelengths at each historical moment are combined to obtain the sum as the second anomaly factor Finally, a local analysis is performed. In the preset time neighborhood centered on each historical moment, the second anomaly factors of all historical moments are accumulated and the sum is normalized to obtain the anomaly degree value of the central historical moment.

[0098] It should be noted that the preset parameter γ is used to prevent the denominator from being 0. At this moment, the value is 0.001. The specific value can be adjusted according to the implementation scenario and is not limited here.

[0099] In other embodiments of the present invention, the following method may also be used to obtain the abnormality level value at each historical moment:

[0100] In the final absorbance time series data of each characteristic wavelength, the variance of the data values ​​of all historical moments is calculated within a preset time neighborhood centered at each historical moment, and the variance is used as the fluctuation factor corresponding to the central historical moment.

[0101] Then the sum of the fluctuation factors at all characteristic wavelengths at each historical moment is taken as the abnormality degree value of that historical moment.

[0102] Based on the above process, the abnormality value at each historical moment can be obtained, and then the final absorbance time series data of each characteristic wavelength can be filtered and denoised to obtain denoised time series data.

[0103] Preferably, in one embodiment of the present invention, the method for acquiring denoised time series data includes:

[0104] The preset filter coefficient is set. Since the greater the abnormality value of the data at each historical moment, the greater the noise level of the data, and a stronger smoothing effect is required, the preset filter coefficient can be adjusted according to the abnormality value of each historical moment to obtain the final filter coefficient; and the preset filter coefficient and the abnormality value are both positively correlated with the final filter coefficient. The formula model of the final filter coefficient can be, for example:

[0105] σ t =(1+YC t )×σ0

[0106] Among them, σ t represents the final filter coefficient at the tth historical moment; YC t represents the abnormality value at the tth historical moment; σ0 represents the preset filter coefficient.

[0107] In the formula model of the final filter coefficient, the preset filter coefficient is adjusted by the abnormality value of each historical moment. When the abnormality value corresponding to a certain historical moment is larger, it means that the data at this time is more affected by noise, and a greater smoothing effect is required. Therefore, 1+YC tAs the adjustment value of the preset filter coefficient, the adjustment value is then multiplied by the preset filter coefficient to obtain the final filter coefficient at each historical moment.

[0108] After obtaining the final filter coefficient at each historical moment, the final absorbance time series data of each characteristic wavelength can be Gaussian filtered to obtain denoised time series data of each characteristic wavelength. It should be noted that Gaussian filtering is a technical means well known to those skilled in the art and will not be described in detail here; the preset filter coefficient is set to 1.6, and the specific value can be adjusted according to the implementation scenario and is not limited here.

[0109] At this point, the final absorbance time series data of each characteristic wavelength can be filtered and denoised, thereby improving the accuracy of subsequent ammonia escape measurement.

[0110] Step S4: Detect the ammonia slip amount at the current moment based on all the denoised time series data.

[0111] At this time, the denoised time series data of each characteristic wavelength can accurately characterize the change of the absorbance of the substance at the characteristic wavelength over time, so based on this, accurate measurement of ammonia escape can be achieved.

[0112] Preferably, in one embodiment of the present invention, the detection of ammonia slip is performed based on all denoised time series data, including:

[0113] In this embodiment of the present invention, a multivariate regression analysis method is adopted, and a regression model needs to be constructed.

[0114] First, the actual concentration of ammonia at each historical moment is obtained; the specific detection method can be chemical sampling and measurement.

[0115] Then, the normalized sum of the absorbances at the same historical moment in the denoised time series data of all characteristic wavelengths is used as the adjustment coefficient, and the product of the sum of the absorbances at the same historical moment in the initial absorbance time series data of all characteristic wavelengths and the corresponding adjustment coefficient is used as the independent variable, where the adjustment coefficient is a fixed value, and the actual concentration of ammonia at the same historical moment is used as the dependent variable. PLS partial least squares is performed to obtain the regression model.

[0116] Finally, the product of the absorbance and value of all characteristic wavelengths in the laser spectrum data at the current moment and the adjustment coefficient is used as the input of the regression model, and the output is the actual concentration of ammonia at the current moment. It should be noted that PLS partial least squares is a technical means well known to those skilled in the art and will not be described in detail here.

[0117] In other embodiments of the present invention, the following method may also be used to detect the ammonia slip amount at the current moment.

[0118] Obtain the actual concentration of ammonia at each historical moment; the specific detection method can be chemical sampling and measurement.

[0119] Then, the data values ​​in the denoised time series data of all characteristic wavelengths at each historical moment are taken as independent variables, and the actual concentration of ammonia at each historical moment is taken as the dependent variable, and the PLS partial least squares is used to obtain the regression model.

[0120] Then, the preset time period includes the preset historical time period and the current moment, and the flow time series data of the reducing agent in the preset time period, the laser spectrum data at each moment, and the initial absorption rate time series data of each wavelength are obtained. After the analysis in the above process, the denoised time series data of each characteristic wavelength at the current moment can be obtained, and then the data values ​​in the denoised time series data of all characteristic wavelengths at the current moment are used as the input of the regression model, and the output is the actual concentration of ammonia at the current moment.

[0121] It should be noted that, in order to facilitate calculation, all indicator data involved in the calculation in the embodiment of the present invention are preprocessed to eliminate the dimension effect. The specific means of eliminating the dimension effect are technical means well known to those skilled in the art and are not limited here.

[0122] An embodiment of the present invention also provides an in-situ measurement system for ammonia escape based on laser spectroscopy technology, the system comprising a memory, a processor and a computer program, wherein the memory is used to store the corresponding computer program, the processor is used to run the corresponding computer program, and when the computer program runs on the processor, it can implement the steps of an in-situ measurement method for ammonia escape based on laser spectroscopy technology.

[0123] In summary, the embodiment of the present invention first obtains the flow time series data of the reducing agent within a preset time period, the spectral data at each historical moment, and extracts the initial absorption rate time series data of each wavelength. Based on the change in the absorption rate of the wavelength in the spectral data, the characteristic wavelength is screened out. Due to the influence of environmental factors, the change in the absorption rate in the spectral data caused by the reducing agent will have a time lag and the wavelength in the spectral data will have a certain offset. Therefore, the change between the initial absorption rate time series data of the characteristic wavelength and the flow time series data of the catalyst is analyzed to determine the time lag; at the same time, under different preset wavelength offsets, the updated absorption rate time series data of the characteristic wavelength is obtained, and combined with the obtained time lag, the correlation between the updated absorption rate time series data and the flow time series data is analyzed, so as to screen out the optimal offset of the wavelength offset of the spectral data at each historical moment. Furthermore, after the spectral data is corrected according to the optimal offset, the final absorbance time series data of the characteristic wavelength at this time can more accurately characterize the change in the absorbance of the substance. At this time, the uneven ammonia concentration inside the pipeline can be analyzed, which is specifically reflected in the change and fluctuation of the data. Therefore, the change and fluctuation of the final absorbance time series data and the data error are analyzed to obtain the abnormality degree value corresponding to each historical moment. Finally, the final absorbance time series data is filtered and denoised according to the abnormality degree value, and high-quality denoised time series data corresponding to the absorbance of the characteristic wavelength is obtained. At this time, the measurement of the ammonia escape amount at the current moment can be performed based on the denoised time series data to obtain a more accurate measurement result.

[0124] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0125] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for in-situ measurement of ammonia escape based on laser spectroscopy technology, characterized in that: The method comprises: Obtain the flow time series data of the reducing agent in a preset historical period and the laser spectrum data of ammonia at each historical moment; obtain the initial absorption rate time series data of each wavelength according to the absorption rate of each wavelength in each laser spectrum data; Among all the laser spectrum data, according to the absorption rate change trend of the wavelength, the characteristic wavelength is screened out; according to the change of the initial absorption rate time series data and the flow time series data of all the characteristic wavelengths, the time lag is obtained; any historical moment is selected as the time to be measured; after the laser spectrum data at the time to be measured is offset at each preset wavelength offset, the updated absorption rate time series data of each characteristic wavelength is obtained; under the time lag, the correlation between all the updated absorption rate time series data and the flow time series data corresponding to different preset wavelength offsets in the preset time neighborhood corresponding to the time to be measured is analyzed, and the optimal offset of the laser spectrum data at the time to be measured is screened out; After adjusting the corresponding laser spectrum data according to each optimal offset, the final absorption rate time series data of each characteristic wavelength is obtained; the abnormality degree value corresponding to each historical moment is obtained according to the change fluctuation of all the final absorption rate time series data; the final absorption rate time series data of each characteristic wavelength is filtered and denoised according to the abnormality degree value to obtain denoised time series data; Detecting the ammonia slip amount at the current moment according to all the denoised time series data; The method for acquiring the characteristic wavelength comprises: For any wavelength, the absorption rates of the wavelength in the laser spectrum data at all historical moments are accumulated as the absorption rate and value of the wavelength; Taking each wavelength as the center, a preset wavelength range is constructed; in each preset wavelength range, other wavelengths except the central wavelength are used as comparison wavelengths, and when the absorbance and value of all comparison wavelengths are less than the absorbance and value of the central wavelength, the central wavelength is used as the characteristic wavelength; The method for acquiring the initial absorption rate time series data comprises: Arrange the absorbance of each wavelength in the laser spectrum data at all historical moments in time series to obtain the initial absorbance time series data of each wavelength; The method for obtaining the time lag comprises: The historical moment of the maximum value in the traffic time series data is obtained as the target moment. For any target moment, the slope value of the fitting straight line of the data values ​​of the first number of historical moments preset before the target moment is calculated as the first slope. The slope value of the fitting straight line of the data values ​​of the first number of historical moments preset after the target moment is calculated as the second slope. The difference between the first slope and the second slope is normalized and used as the convexity value corresponding to the target moment. The formula model of the convexity value is: in, Indicates The convexity value corresponding to the target moment; Indicates The first slope value corresponding to the target moment; Indicates The second slope value corresponding to the target moment; represents the normalization function; A preset hysteresis value range, for any hysteresis value in the preset hysteresis value range, calculate the time matching deviation value between the initial absorption rate time series data of all characteristic wavelengths and the flow time series data under the hysteresis value; among all the time matching deviation values, the hysteresis value corresponding to the smallest time matching deviation value is used as the time lag; The method for obtaining the moment matching deviation value comprises: selecting any characteristic wavelength as the wavelength to be measured; taking the historical moments corresponding to all the maximum values ​​of the initial absorption rate time series data of the wavelength to be measured as the comparison moments; for any target moment, calculating the numerical difference between the value after adding the hysteresis value to the target moment and each comparison moment, multiplying the minimum numerical difference by the protrusion degree value corresponding to the target moment, and obtaining the deviation factor of the target moment; taking the value after accumulating the deviation factors of all target moments as the deviation sum value corresponding to the wavelength to be measured; normalizing the value after accumulating the deviation sum values ​​of all characteristic wavelengths, and obtaining the moment matching deviation value; the formula model of the moment matching deviation value comprises: in, Indicates Moment matching deviation value under hysteresis value; Represents the total number of characteristic wavelengths; represents the total number of target moments; Indicates The convexity value of the target moment; Indicates The historical moment value of the target moment; Indicates hysteresis value; Indicates The initial absorption rate time series data of the characteristic wavelength The historical moment value corresponding to the maximum value; Indicates The total number of maxima in the initial absorbance time series data of the characteristic wavelength; Indicates finding the minimum value; represents the normalization function; The method for obtaining the optimal offset comprises: Under each preset wavelength offset, the updated absorbance time series data corresponding to each characteristic wavelength is subjected to STL time series decomposition to obtain the initial trend item time series data corresponding to each characteristic wavelength; Each of the initial trend item time series data is offset by the time lag amount to obtain the final trend item time series data. The time lag amount offset process is: offset the moment axis of each initial trend item time series data in the direction where the moment becomes larger by the time lag amount units; Calculate the Pearson correlation coefficient between each final trend item time series data and the flow time series data within the preset time neighborhood as a correlation value; and use the cumulative sum of the correlation values ​​corresponding to all characteristic wavelengths as a judgment value; Among all the judgment values ​​corresponding to the preset wavelength offsets, the preset wavelength offset corresponding to the maximum judgment value is used as the optimal offset of the laser spectrum data at the time to be measured; The preset time neighborhood is set to be centered at the time to be measured, and a preset second number of historical moments are taken on both sides to form a time neighborhood; The method for obtaining the abnormality degree value includes: Perform STL time series decomposition on the final absorbance time series data of each characteristic wavelength to obtain seasonal term time series data and residual term time series data; In the seasonal item time series data and the residual item time series data corresponding to each characteristic wavelength, a first abnormal factor at each historical moment is obtained according to the seasonal item data value and the residual item data value at each historical moment, wherein the first abnormal factor is positively correlated with the seasonal item data value, and the first abnormal factor is negatively correlated with the residual item data value; The sum of the first anomaly factors at all characteristic wavelengths at each historical moment is taken as the second anomaly factor; within the preset time neighborhood centered on each historical moment, the sum of the second anomaly factors of all historical moments is normalized and taken as the anomaly degree value of the central historical moment; the formula model of the anomaly degree value includes: in, Representing the historical moments of the center The abnormality value of Represents the total number of historical moments within a preset time neighborhood; Represents the total number of characteristic wavelengths; Indicated in In the seasonal time series data corresponding to the characteristic wavelength, the central historical moment The value of Indicated in In the residual time series data corresponding to the characteristic wavelength, the central historical moment The value of Indicates preset parameters; Represents the normalization function.

2. The in-situ measurement method of ammonia escape based on laser spectroscopy technology according to claim 1 is characterized in that: The method for acquiring denoised time series data comprises: Preset filter coefficients; Obtaining a final filter coefficient at each historical moment according to the abnormality degree value at each historical moment and the preset filter coefficient, wherein both the abnormality degree value and the preset filter coefficient are positively correlated with the final filter coefficient; The final absorption rate time series data of each characteristic wavelength is Gaussian filtered according to the final filtering coefficient at each historical moment to obtain the denoised time series data of each characteristic wavelength.

3. The in-situ measurement method of ammonia escape based on laser spectroscopy technology according to claim 1 is characterized in that:The detecting of ammonia slip amount according to all the denoised time series data includes: Get the actual concentration of ammonia at each historical moment; The normalized value of the sum of the absorbances at the same historical moment in the denoised time series data of all characteristic wavelengths is used as the adjustment coefficient, the product of the sum of the absorbances at the same historical moment in the initial absorbance time series data of all characteristic wavelengths and the corresponding adjustment coefficient is used as the independent variable, and the actual concentration of ammonia at the same historical moment is used as the dependent variable, and PLS partial least squares is performed to obtain a regression model; The product of the absorbance and value of all characteristic wavelengths in the laser spectrum data at the current moment and the adjustment coefficient is used as the input of the regression model, and the output is the actual concentration of ammonia at the current moment.

4. An ammonia escape in-situ measurement system based on laser spectroscopy technology, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 3 are implemented.