Method for determining reaction parameters of reaction liquid and hydrogenation efficiency of hydrogenation liquid in hydrogen peroxide production process

Abstract

The invention relates to the technical field of spectrum, and discloses a method for determining reaction parameters of a reaction solution and hydrogenation efficiency of a hydrogenation solution in a hydrogen peroxide production process. The method comprises the following steps: (1) providing a plurality of standard reaction solutions, and respectively collecting near infrared spectrograms; selecting the characteristic spectrum wave band; (2) performing three-dimensional processing on the characteristic spectral band to obtain a three-dimensional spectrogram; carrying out regression analysis on the three-dimensional spectrogram and reaction parameters of a standard reaction solution; and (3) taking a reaction solution to be detected, collecting a near infrared spectrogram, selecting a corresponding characteristic spectral band, processing according to the step (2) to obtain a three-dimensional spectrogram, and substituting the three-dimensional spectrogram into the model obtained after regression analysis in the step (2) to obtain reaction parameters to be detected. The method is simple to operate, high in analysis speed, accurate in prediction and high in model stability.

Description

Technical Field

[0001] The invention relates to the field of spectral technology, and in particular to a method for determining reaction parameters of a reaction liquid and hydrogenation efficiency of a hydrogenated liquid in a hydrogen peroxide production process. Background Art

[0002] The domestic preparation method of hydrogen peroxide mainly adopts the anthraquinone method, which has the advantages of advanced technology, high degree of automation control, low product cost and energy consumption, and is suitable for large-scale production. However, since the raw materials, intermediate products and final products are almost all flammable and explosive or combustion-supporting substances, the production process is also relatively complicated, so it is particularly important to accurately and timely monitor and adjust the quality indicators of the hydrogen peroxide production process.

[0003] Hydrogenation efficiency, oxidation efficiency and raffinate concentration are three important analysis items in the anthraquinone method hydrogen peroxide production process, which directly affect the output, quality and production safety of hydrogen peroxide. The analysis frequency is very high, and an average of once every 2 hours is required.

[0004] The hydrogenated liquid is a liquid product formed after the hydrogenation reaction in the hydrogen peroxide production process. Its main component is anthrahydroquinone, and it also contains a small amount of unreacted anthraquinone and impurities such as catalysts. The general measurement method of hydrogenation efficiency is: take a sample of the hydrogenated liquid, pass oxygen into it for oxidation (the hydrogenated liquid itself does not contain hydrogen peroxide, so oxygen needs to be added to oxidize it into hydrogen peroxide. The hydrogenation efficiency is the content of hydrogen peroxide produced after passing oxygen, that is, the potential content), and then add potassium permanganate titration solution until the hydrogen peroxide reaction is complete. The hydrogenation efficiency of the hydrogenated liquid (that is, the potential hydrogen peroxide content) is calculated based on the content of the titration solution consumed at the titration endpoint, and the unit is g / L.

[0005] The oxidizing liquid is a liquid product formed after the oxidation reaction during the production of hydrogen peroxide. Its main component is hydrogen peroxide, and it also contains a small amount of impurities such as unreacted anthraquinone, decomposition products of hydrogen peroxide and catalysts.

[0006] The raffinate is the liquid remaining after hydrogen peroxide is extracted from the material obtained from the oxidation reaction during the hydrogen peroxide production process. The main components are water and some organic impurities (such as organic acids) and catalysts, etc., and it also contains a small amount of hydrogen peroxide.

[0007] Among them, both the oxidizing liquid and the raffinate actually contain hydrogen peroxide, and the two can be calculated using the same method. The method for determining the oxidation efficiency and the raffinate concentration is generally: take a part of the sample, add potassium permanganate titration solution, and calculate the corresponding oxidation efficiency and raffinate concentration (that is, the actual hydrogen peroxide content in the sample) according to the content of the titration solution consumed at the titration endpoint, in units of g / L.

[0008] At present, the manual titration method is mainly used to determine the reaction parameters of the reaction liquid in the production process. However, the manual titration method is cumbersome, has high requirements on the operator, has a slow analysis speed and is not environmentally friendly. It does not have an advantage in the actual determination of the reaction parameters in the hydrogen peroxide production process.

[0009] Therefore, it is necessary to develop a method that can quickly, simply and accurately determine the reaction parameters (mainly including hydrogenation efficiency, oxidation efficiency and raffinate concentration) in each step of the process in hydrogen peroxide. Summary of the invention

[0010] The purpose of the present invention is to overcome the above problems existing in the prior art and provide a method for determining the reaction parameters of the reaction liquid and the hydrogenation efficiency of the hydrogenated liquid in the production process of hydrogen peroxide. The method is simple to operate, fast in analysis speed, accurate in prediction, and has strong model stability. Compared with the manual titration method, the method provided by the present invention can also effectively improve the environment of the analysis operation.

[0011] In order to achieve the above object, the first aspect of the present invention provides a method for determining reaction parameters of a reaction liquid in a hydrogen peroxide production process, the method comprising:

[0012] (1) providing a plurality of standard reaction solutions with known reaction parameters and different reaction parameters, collecting near-infrared spectra of the standard reaction solutions respectively; and selecting characteristic spectral bands of the near-infrared spectra;

[0013] (2) performing three-dimensional processing on the characteristic spectral band to obtain a three-dimensional spectrum;

[0014] Perform regression analysis on the three-dimensional spectrum and the reaction parameters of the standard reaction solution;

[0015] (3) Take the reaction liquid to be tested, collect a near-infrared spectrum and select the corresponding characteristic spectral band, and process it according to step (2) to obtain a three-dimensional spectrum, and substitute the three-dimensional spectrum into the model obtained after the regression analysis in step (2) to obtain the reaction parameters to be tested.

[0016] A second aspect of the present invention provides a method for determining the hydrogenation efficiency of a hydrogenated liquid in a hydrogen peroxide production process, the method comprising:

[0017] (I) providing a plurality of standard hydrogenation liquids with known and different hydrogenation efficiencies, collecting near-infrared spectra of the standard hydrogenation liquids respectively, and selecting characteristic spectral bands of the near-infrared spectra;

[0018] The characteristic spectral bands of the near-infrared spectrum are subjected to second-order differentiation and Fourier transformation to obtain the real and imaginary parts of the characteristic spectral bands;

[0019] (II) The real and imaginary parts of the near-infrared spectrum are stacked to obtain a three-dimensional spectrum X dz; The three-dimensional spectrum X dz , and the hydrogenation efficiency of the hydrogenation liquid were analyzed by multidimensional partial least squares to obtain the prediction model Y dz ;

[0020] The real and imaginary parts corresponding to the near-infrared spectrum are processed by outer product to obtain a three-dimensional spectrum X wj , the three-dimensional spectrum X wj The multidimensional partial least squares analysis was performed on the hydrogenation efficiency of the hydrogenation liquid to obtain the prediction model Y wj ;

[0021] The real and imaginary parts corresponding to the near-infrared spectrum are processed externally to obtain a three-dimensional spectrum X wh , and X wh The multidimensional partial least squares analysis was performed on the hydrogenation efficiency of the hydrogenation liquid to obtain the prediction model Y wh ;

[0022] (III) X dz , substitute into Y dz The prediction result P is obtained dz ,

[0023] X wj , substitute into Y wj The prediction result P is obtained wj ,

[0024] X wh , substitute into Y wh The prediction result P is obtained wh ,

[0025] The above prediction results are assigned weights a, b and c respectively. dz +bP wj +cP wh Adjust a, b and c to aP dz +bP wj +cP wh The calculation result is close to the real hydrogenation efficiency of the hydrogenated liquid, where a, b, and c are respectively greater than 0 and less than 1, and a+b+c=1;

[0026] The fusion model is obtained as M = aY dz +bY wj +cY wh ;

[0027] (IV) Collect a near infrared spectrum of the reaction solution to be tested, and process it according to steps (I)-(II) to obtain a three-dimensional spectrum X dz ', X wj ' and X wh ', and substitute the above three-dimensional spectrum into the fusion model in step (III) to obtain the hydrogenation efficiency to be measured.

[0028] When the fusion model obtained by the method provided by the present invention is used to predict the reaction parameters of the liquid to be tested, the speed is fast, the results are accurate, and the operation is simple. Compared with the traditional manual titration method, the method provided by the present invention has a fast analysis speed, high efficiency, and can greatly improve the analysis operation environment, which has obvious advantages. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a correlation diagram of the measured values ​​and calculated values ​​of the verification set in Example 1. DETAILED DESCRIPTION

[0030] The endpoints and any values ​​of the ranges disclosed in this article are not limited to the precise ranges or values, and these ranges or values ​​should be understood to include values ​​close to these ranges or values. For numerical ranges, the endpoint values ​​of each range, the endpoint values ​​of each range and the individual point values, and the individual point values ​​can be combined with each other to obtain one or more new numerical ranges, which should be considered as specifically disclosed in this article.

[0031] In a first aspect, the present invention provides a method for determining reaction parameters of a reaction solution in a hydrogen peroxide production process, the method comprising:

[0032] (1) providing a plurality of standard reaction solutions with known reaction parameters and different reaction parameters, collecting near-infrared spectra of the standard reaction solutions respectively; and selecting characteristic spectral bands of the near-infrared spectra;

[0033] (2) performing three-dimensional processing on the characteristic spectral band to obtain a three-dimensional spectrum;

[0034] Perform regression analysis on the three-dimensional spectrum and the reaction parameters of the standard reaction solution;

[0035] (3) Take the reaction liquid to be tested, collect a near-infrared spectrum and select the corresponding characteristic spectral band, and process it according to step (2) to obtain a three-dimensional spectrum, and substitute the three-dimensional spectrum into the model obtained after the regression analysis in step (2) to obtain the reaction parameters to be tested.

[0036] The inventors of the present invention have found in their research that when the above-mentioned method is adopted, the near-infrared spectrum of the standard reaction solution is first processed to obtain a three-dimensional spectrum, and the three-dimensional spectrum is used for regression analysis. Then, for samples with unknown reaction parameters, the reaction parameters can be accurately and quickly obtained by substituting the three-dimensional spectrum into the model obtained after regression analysis. In particular, by constructing a three-dimensional spectrum in the above-mentioned way, the influence of overlapping peaks can be effectively avoided, and more comprehensive information can be provided to ensure the accuracy of the results. In general, in the three-dimensional processing of the spectrum, the three dimensions of the obtained three-dimensional spectrum are wave number (which can be an X-axis, and the range can be 1000-10000cm -1 , such as 1000-2500nm or 3500-10000cm -1 etc.), absorbance (which can be the Y axis, and the range can be 0-100%) and band type (which can be the Z axis). Compared with other materials, the method of the present invention can obtain more accurate results for the reaction liquid in the hydrogen peroxide production process.

[0037] In step (1), after obtaining the standard reaction solution, the reaction parameters can be obtained by the currently available manual titration method. In addition, it can be understood that after selecting the characteristic spectral band, a near-infrared spectrum containing the characteristic spectral band can be obtained, and the characteristic spectral band is processed in three dimensions, that is, the near-infrared spectrum containing the characteristic spectral band is processed in three dimensions.

[0038] According to the present invention, preferably, the reaction liquid is a hydrogenation liquid in a hydrogen peroxide production process, the reaction parameter is a hydrogenation efficiency, and the hydrogenation efficiency is 0-15 g / L, preferably 0.3-12.5 g / L.

[0039] According to the present invention, preferably, the reaction liquid is an oxidizing liquid in a hydrogen peroxide production process, the reaction parameter is an oxidation efficiency, and the oxidation efficiency is 0.5-7 g / L, preferably 0.9-6.4 g / L.

[0040] According to the present invention, preferably, the reaction liquid is the raffinate in the hydrogen peroxide production process, the reaction parameter is the raffinate concentration, and the raffinate concentration is 0.01-1 g / L, preferably 0.03-0.9 g / L.

[0041] Among them, the currently commonly used method for determining the hydrogenation efficiency of the hydrogenated liquid, the oxidation efficiency of the oxidized liquid and the raffinate concentration of the raffinate can be a conventional titration method in the art, and the present invention has no particular limitation.

[0042] Hydrogen peroxide is mainly produced by the anthraquinone method, and the reaction liquid generally includes the above hydrogenation liquid, oxidation liquid and raffinate. The main process of the anthraquinone method includes: anthraquinone (generally 2-ethyl anthraquinone) and an organic solvent (such as a mixed solvent of C9-C10 heavy aromatics, trioctyl phosphate and tetrabutyl urea) are prepared into a working solution, and the total effective anthraquinone content is controlled to be about 160g / L. Hydrogenation stage: under the conditions of 0.3MPa or above, 40-80°C and a catalyst (such as Pd catalyst), the anthraquinone in the working solution is hydrogenated and reduced by H2 to obtain anthrahydroquinone or tetrahydroanthrahydroquinone; in the oxidation stage, the material after the hydrogenation stage is subjected to an oxidation reaction with O2 at 30-60°C and slight compression to oxidize anthrahydroquinone and tetrahydroanthrahydroquinone to generate H2O2 and anthraquinone; then the material after the oxidation stage is subjected to extraction (the raffinate is the solution remaining after the H2O2 is extracted), regeneration, refining and concentration to obtain a 20wt%-50wt% H2O2 aqueous solution.

[0043] It can be understood that when collecting samples, it is generally required that the conditions of the samples are distributed relatively evenly within a possible range, and therefore the reaction parameters in the standard reaction solution are distributed relatively evenly within the above range.

[0044] According to the present invention, preferably, in step (1), the number of the standard reaction solutions is ≥30, preferably 40-200 (for example, it can be 40, 42, 45, 47, 50, 60, 80, 100, 120, 140, 160, 180, 200, and the range formed by any two of the above values ​​and the value within the range). When it is within the above range, the accuracy of the result can be further guaranteed.

[0045] When collecting near-infrared spectra, the operating conditions are not particularly limited and can be carried out under conventional conditions in the art. Preferably, the conditions for collecting near-infrared spectra include: a collection temperature of 15-50°C, more preferably 22-27°C; a wave number range of 3500-10000cm -1 .

[0046] Preferably, the resolution of the acquisition is 2-16cm -1 , more preferably 6-10cm -1 . Can scan 100-140 times.

[0047] When collecting near infrared spectra, the sample can be injected into about two-thirds of the volume of the cuvette, and the cuvette containing the sample is sealed, and the sealed cuvette can be placed in a temperature-controlled sample cell rack for spectrum collection. The optical path of the cuvette can be 1-5 mm. The instrument when collecting near infrared spectra is not particularly limited and can be a conventional choice in the art, for example, Antaris II near infrared spectrometer (Thermo Fisher Company) can be used.

[0048] According to the present invention, preferably, the reaction liquid is a hydrogenation liquid in the production process of hydrogen peroxide, the reaction parameter is the hydrogenation efficiency, and the wave number interval of the characteristic spectral band is 3880-4204cm -1 and 4458-6194cm -1 .

[0049] According to the present invention, preferably, the reaction liquid is an oxidizing liquid in the production process of hydrogen peroxide, the reaction parameter is the oxidation efficiency, and the wave number interval of the characteristic spectral band is 3880-4204cm -1 and 4458-6194cm -1 .

[0050] According to the present invention, preferably, the reaction liquid is the raffinate in the hydrogen peroxide production process, the reaction parameter is the raffinate concentration, and the wave number interval of the characteristic spectral band is 3880-4204cm -1 and 4458-6194cm -1 .

[0051] In the hydrogenated liquid, the oxidized liquid and the raffinate, special attention is paid to the presence of hydrogen peroxide, and therefore the same characteristic spectral bands as described above may be selected.

[0052] According to the present invention, preferably, before the three-dimensional processing is performed, the method further includes: preprocessing the characteristic spectral band to eliminate redundant information and noise in the spectrum. Generally, through preprocessing, baseline drift, spectral overlap, noise, etc. in the spectrum can be eliminated to further improve the accuracy of subsequent model construction.

[0053] Preferably, the preprocessing method includes first-order differential and / or second-order differential, more preferably second-order differential.

[0054] Preferably, the window width of the second-order differential processing is 18-23 (for example, it can be 18, 19, 20, 21, 22, 23).

[0055] The specific method of regression analysis is not particularly limited and can be a conventional method. However, preferably, the regression analysis method is multidimensional partial least squares analysis. The inventors of the present invention further found that, compared with other regression analysis methods, the use of multidimensional partial least squares analysis can further ensure the accuracy of the results.

[0056] According to the present invention, preferably, in step (2), N different three-dimensional processing methods are used to perform N three-dimensional processing, and N three-dimensional spectra are obtained accordingly; each three-dimensional spectra is respectively subjected to regression analysis with the reaction parameters of the standard reaction solution to obtain N prediction models, and the N prediction models are weightedly fused to obtain a fusion model;

[0057] The weighted fusion method includes: substituting the three-dimensional spectrograms obtained by the three-dimensional processing into the corresponding prediction models respectively, and obtaining the prediction results of each prediction model respectively;

[0058] The multiple prediction results obtained above are assigned weights respectively, and the weight values ​​are adjusted until the sum of the products of each weight and the prediction result (that is, the weight value is multiplied by the corresponding prediction result respectively, so that multiple products are obtained, and then the products are added and summed) is close to the true reaction parameter, wherein each weight value satisfies that each is greater than 0 and less than 1, and the sum of each weight value is 1;

[0059] The fusion model is the sum of the products of each weight value and the corresponding prediction model (each weight value is multiplied by the corresponding prediction model, and then multiple products are added together);

[0060] In step (3), the three-dimensional spectrum corresponding to the reaction solution to be tested is substituted into the fusion model to obtain the reaction parameters to be tested.

[0061] The inventors of the present invention also found that if different three-dimensional processing methods are used in the above manner to obtain N prediction models, and then the N prediction models are weighted and fused to obtain a fusion model, and the three-dimensional spectrum corresponding to the reaction liquid to be tested is substituted into the fusion model, compared with only one three-dimensional processing to obtain a prediction model, the three-dimensional spectrum corresponding to the reaction liquid to be tested is substituted into the prediction model. A fusion model with higher effective accuracy can be further obtained, thereby improving the accuracy of the obtained results, and the stability of the fusion model obtained in this way is higher. The three-dimensional processing method can be selected from at least one of stacked arrays, outer products, and outer sums.

[0062] According to the present invention, preferably, N is 2-3 (for example, it can be 2 or 3).

[0063] Preferably, before the three-dimensional processing, the method further comprises: preprocessing the selected characteristic spectral band to eliminate redundant information and / or noise in the spectrum; and performing Fourier transform on the preprocessed characteristic spectral band to obtain a real part and an imaginary part.

[0064] The specific meaning of the band types in the three dimensions obtained by the three-dimensional processing is related to the specific method performed in step (2). For example, if Fourier transform is performed in the process, the band type may refer to the real part and imaginary part obtained by Fourier transform; if wavelet transform is performed in the process, the band type may refer to the wavelet coefficients obtained by the wavelet transform that can reflect different information.

[0065] Preferably, the three-dimensional processing method includes: stacking the real part and the imaginary part.

[0066] Preferably, the three-dimensional processing method includes: performing outer product processing on the real part and the imaginary part.

[0067] Preferably, the three-dimensional processing method includes: performing external sum processing on the real part and the imaginary part.

[0068] According to a particularly preferred embodiment of the present invention, according to the above three three-dimensional processing, a three-dimensional spectrum after stacking processing, a three-dimensional spectrum after outer product processing, and a three-dimensional spectrum after outer sum processing are obtained respectively. The three three-dimensional spectra as mentioned above are respectively subjected to regression analysis with the reaction parameters of the standard reaction solution to obtain three prediction models in total.

[0069] In a second aspect, the present invention provides a method for determining the hydrogenation efficiency of a hydrogenated liquid in a hydrogen peroxide production process, the method comprising:

[0070] (I) providing a plurality of standard hydrogenation liquids with known and different hydrogenation efficiencies, collecting near-infrared spectra of the standard hydrogenation liquids respectively, and selecting characteristic spectral bands of the near-infrared spectra;

[0071] The characteristic spectral bands of the near-infrared spectrum are subjected to second-order differentiation and Fourier transformation to obtain the real and imaginary parts of the characteristic spectral bands;

[0072] (II) The real and imaginary parts of the near-infrared spectrum are stacked to obtain a three-dimensional spectrum X dz ; The three-dimensional spectrum X dz , and the hydrogenation efficiency of the hydrogenation liquid were analyzed by multidimensional partial least squares to obtain the prediction model Y dz ;

[0073] The real and imaginary parts corresponding to the near-infrared spectrum are processed by outer product to obtain a three-dimensional spectrum X wj , the three-dimensional spectrum X wj The multidimensional partial least squares analysis was performed on the hydrogenation efficiency of the hydrogenation liquid to obtain the prediction model Y wj ;

[0074] The real and imaginary parts corresponding to the near-infrared spectrum are processed externally to obtain a three-dimensional spectrum X wh , and X wh The multidimensional partial least squares analysis was performed on the hydrogenation efficiency of the hydrogenation liquid to obtain the prediction model Y wh ;

[0075] (III) X dz , substitute into Y dz The prediction result P is obtained dz ,

[0076] X wj , substitute into Y wjThe prediction result P is obtained wj ,

[0077] X wh , substitute into Y wh The prediction result P is obtained wh ,

[0078] The above prediction results are assigned weights a, b and c respectively. dz +bP wj +cP wh Adjust a, b and c to aP dz +bP wj +cP wh The calculation result is close to the real hydrogenation efficiency of the hydrogenated liquid, where a, b, and c are respectively greater than 0 and less than 1, and a+b+c=1;

[0079] The fusion model is obtained as M = aY dz +bY wj +cY wh ;

[0080] (IV) Collect a near infrared spectrum of the reaction solution to be tested, and process it according to steps (I)-(II) to obtain a three-dimensional spectrum X dz ', X wj ' and X wh ', and substitute the above three-dimensional spectrum into the fusion model in step (III) to obtain the hydrogenation efficiency to be measured.

[0081] It can be understood that the real hydrogenation efficiency refers to the hydrogenation efficiency obtained by using the currently available manual titration method after obtaining the standard solution.

[0082] In the process of adjusting a, b and c, there are many possible combinations that can be adopted, and the values ​​of a, b and c that can make the calculation results closer to the actual hydrogenation efficiency can be selected from these combinations.

[0083] It is understandable that the main component of the hydrogenated liquid is anthrahydroquinone, and it also contains a small amount of impurities such as unreacted anthraquinone and catalyst. Since the hydrogenated liquid itself does not contain hydrogen peroxide, oxygen needs to be added to oxidize it into hydrogen peroxide. The hydrogenation efficiency is the content of hydrogen peroxide produced after oxygenation, which characterizes the potential hydrogen peroxide content in the original hydrogenated liquid.

[0084] The method provided by the present invention is particularly suitable for the above standard hydrogenation liquid.

[0085] Among them, the range of hydrogenation efficiency in the standard hydrogenation liquid, the number of standard reaction liquids, the conditions for collecting near-infrared spectra, the selected characteristic spectral bands, the preprocessing method of the spectral bands, and the regression analysis method can be referred to the first aspect and will not be repeated here.

[0086] For the hydrogenated liquid in the process of preparing hydrogen peroxide by the anthraquinone method, by selecting the specific characteristic spectral band defined by the present invention, and coordinating the above second-order differential, Fourier transform, and spectral processing methods of stacked arrays, outer products and outer sums, especially coordinating the multidimensional partial least squares analysis applicable to three-dimensional spectra, and fusing the obtained prediction models in the manner defined by the present invention, the obtained fusion model has a high accuracy. Compared with other methods, such as using other processing methods for the spectrum, or not converting the spectrum into three dimensions and using a method applicable to constructing a prediction model corresponding to a two-dimensional spectrum (such as partial least squares analysis), the specific method provided by the present invention has a higher accuracy.

[0087] The present invention will be described in detail below through examples.

[0088] In the following examples, the hydrogenated liquid was obtained during the process of preparing hydrogen peroxide by the anthraquinone process.

[0089] In step (I) of the following examples, the hydrogenation efficiency in the standard hydrogenation solution is determined by potassium permanganate titration, and the hydrogenation efficiency value obtained by potassium permanganate titration is recorded as the measured value.

[0090] The instrument used to collect near-infrared spectra was Antaris II near-infrared spectrometer (Thermo Fisher).

[0091] The operations after spectrum acquisition were performed in MATLAB.

[0092] Example 1

[0093] The method for determining the hydrogenation efficiency of the hydrogenation liquid in the hydrogen peroxide production process provided by the present invention is used to illustrate the method

[0094] (I) 64 hydrogenated liquids (oxidized by oxygen to characterize the potential content of hydrogen peroxide) were obtained as standard hydrogenated liquids, and the hydrogenation efficiency thereof was determined (the hydrogenation efficiency values ​​were distributed in the range of 0.39-12.15 g / L, and the difference between every two adjacent hydrogenation efficiency values ​​was in the range of 0.3-1 g / L).

[0095] The sample was injected into the cuvette (optical path length 2 mm) until it reached two-thirds of the cuvette volume, the cuvette containing the sample was sealed, and the sealed cuvette was placed in a temperature-controlled sample cell rack for spectrum collection. The conditions for collecting the near-infrared spectrum of the standard hydrogenated liquid include: temperature 15°C, and the collected wavenumber range 3500-10000cm -1 , resolution 8cm -1 , scanned 128 times. For the obtained near-infrared spectrum, the characteristic spectrum band 3880-4204cm was selected -1 and 4458-6194cm -1.

[0096] For the above characteristic spectral bands, a second-order differential with a window width of 21 is first performed, and then a Fourier transform is performed to obtain the real part and the imaginary part.

[0097] (II) The real part and the imaginary part are stacked to obtain a three-dimensional spectrum X dz , the three-dimensional spectrum X dz The multidimensional partial least squares analysis was performed on the hydrogenation efficiency of the hydrogenation liquid to obtain the prediction model Y dz ;

[0098] Perform outer product processing on the real and imaginary parts to obtain the three-dimensional spectrum X wj , the three-dimensional spectrum X wj The multidimensional partial least squares analysis was performed on the hydrogenation efficiency of the hydrogenation liquid to obtain the prediction model Y wj ;

[0099] The real and imaginary parts are processed externally to obtain a three-dimensional spectrum X wh , and the three-dimensional spectrum X wh The multidimensional partial least squares analysis was performed on the hydrogenation efficiency of the hydrogenation liquid to obtain the prediction model Y wh ;

[0100] (III) X dz , substitute into Y dz The prediction result P is obtained dz ,

[0101] X wj , substitute into Y wj The prediction result P is obtained wj ,

[0102] X wh , substitute into Y wh The prediction result P is obtained wh ,

[0103] P dz , P wj and P wh , respectively assign weights a, b and c, according to aP dz +bP wj +cP wh , adjust a, b and c to aP dz +bP wj +cP wh The calculated results are close to the actual hydrogenation efficiency of the hydrogenated liquid, and a is 0.924, b is 0.001, and c is 0.075.

[0104] Then we get the fusion model M = aY dz +bY wj +cY wh.

[0105] (IV) Fifteen hydrogenated liquids from the production of hydrogen peroxide by the anthraquinone process with unknown hydrogenation efficiency were selected as the validation set and processed as follows:

[0106] The method in step (I) is adopted, the hydrogenation efficiency is first determined by potassium permanganate titration, and a near-infrared spectrum is collected, a characteristic spectral band is selected, and the characteristic spectral band is subjected to the same second-order differential and Fourier transform as in step (I) to obtain the real part and the imaginary part.

[0107] The real part and the imaginary part are stacked to obtain the three-dimensional spectrum X dz ';

[0108] Perform outer product processing on the real and imaginary parts to obtain the three-dimensional spectrum X wj ';

[0109] The real part and the imaginary part are processed externally to obtain the three-dimensional spectrum X wh ';

[0110] Substitute the three-dimensional spectrum obtained above into the fusion model in step (III) to obtain the hydrogenation efficiency (i.e., the calculated value in Table 1).

[0111] The measured and calculated values ​​of the 15 samples in the validation set, as well as the deviation between the two (measured value minus calculated value) are shown in Table 1. In addition, the root mean square error of prediction (RMSEP) and R (correlation coefficient) are used to evaluate the performance of the fusion model, where

[0112]

[0113] Among them, n is the total number of samples in the validation set, y i is the measured value of the i-th sample, is the calculated value of the ith sample.

[0114]

[0115] Among them, n is the total number of samples in the validation set, y i is the measured value of the i-th sample, y i ^ is the calculated value of the i-th sample, y — is the average value of the actual values ​​of the validation set samples. By calculation, we can get R = 0.992.

[0116] Table 1

[0117]

[0118] In addition, a correlation diagram is drawn with the measured values ​​(x-axis) and calculated values ​​(y-axis) of the 15 samples in the validation set, see Figure 1 (where R is the correlation coefficient).

[0119] It can be seen that the method provided by the present invention can quickly, simply and accurately determine the parameters in the reaction liquid in the hydrogen peroxide production process, and the prediction results at various concentrations are relatively accurate and have high stability.

[0120] Comparative Example 1

[0121] According to the method in step (1) of Example 1, proceed to "first perform a second-order differential with a window width of 21 on the above characteristic spectral band", and then perform a partial least squares analysis on the band after the second-order differential and the hydrogenation efficiency of the hydrogenated liquid to obtain the prediction model Y D .

[0122] Then, 15 hydrogenated liquids from the production of hydrogen peroxide by the anthraquinone process with unknown hydrogenation efficiency were taken as the validation set. According to step (IV) of Example 1, the process proceeded to "and the characteristic spectral band was subjected to the same second-order differential as in step (I)". The band after the second-order differential at this time was substituted into the prediction model Y D Get the prediction model Y D The calculated values ​​are shown in Table 2.

[0123] Calculate the prediction model Y in the manner described in Example 1 D The root mean square error of prediction (RMSEP) and the bias of the obtained results are shown in Table 2.

[0124] Table 2

[0125]

[0126] It can be understood that RMSEP can characterize the accuracy of the method. It can be seen that the method provided by the present invention has a smaller deviation from the actual measured value of the standard method, and the RMSEP is also smaller. The method provided by the present invention has better accuracy.

[0127] The preferred embodiments of the present invention are described in detail above, but the present invention is not limited thereto. Within the technical concept of the present invention, the technical solution of the present invention can be subjected to a variety of simple modifications, including the combination of various technical features in any other suitable manner, and these simple modifications and combinations should also be regarded as the contents disclosed by the present invention and belong to the protection scope of the present invention.

Claims

1. A method for determining reaction parameters of a reaction solution in a hydrogen peroxide production process, characterized in that: The method includes: (1) providing a plurality of standard reaction solutions with known reaction parameters and different reaction parameters, collecting near-infrared spectra of the standard reaction solutions respectively; and selecting characteristic spectral bands of the near-infrared spectra; (2) performing three-dimensional processing on the characteristic spectral band to obtain a three-dimensional spectrum; Perform regression analysis on the three-dimensional spectrum and the reaction parameters of the standard reaction solution; (3) Take the reaction liquid to be tested, collect a near-infrared spectrum and select the corresponding characteristic spectral band, and process it according to step (2) to obtain a three-dimensional spectrum, and substitute the three-dimensional spectrum into the model obtained after the regression analysis in step (2) to obtain the reaction parameters to be tested.

2. The method according to claim 1, wherein: The reaction liquid is a hydrogenation liquid in a hydrogen peroxide production process, the reaction parameter is a hydrogenation efficiency, and the hydrogenation efficiency is 0-15 g / L, preferably 0.3-12.5 g / L.

3. The method according to claim 1, wherein: The reaction liquid is an oxidizing liquid in a hydrogen peroxide production process, the reaction parameter is an oxidation efficiency, and the oxidation efficiency is 0.5-7 g / L, preferably 0.9-6.4 g / L.

4. The method according to claim 1, wherein: The reaction liquid is the raffinate in the hydrogen peroxide production process, the reaction parameter is the raffinate concentration, and the raffinate concentration is 0.01-1 g / L, preferably 0.03-0.9 g / L.

5. The method according to claim 1 or 2, wherein: In step (1), the number of the standard reaction solutions is ≥30, preferably 40-200.

6. The method according to claim 1 or 2, wherein: The conditions for collecting near-infrared spectra include: the collection temperature is 15-50°C, preferably 22-27°C; the wave number range is 3500-10000cm -1 ; and / or, the acquisition resolution is 2-16cm -1 , preferably 6-10cm -1 .

7. The method according to claim 1 or 2, wherein: The reaction liquid is a hydrogenation liquid in the production process of hydrogen peroxide, the reaction parameter is the hydrogenation efficiency, and the wave number interval of the characteristic spectral band is 3880-4204cm -1 and 4458-6194cm -1 ; And / or, the reaction liquid is an oxidizing liquid in a hydrogen peroxide production process, the reaction parameter is an oxidation efficiency, and the wave number interval of the characteristic spectral band is 3880-4204cm -1 and 4458-6194cm -1 ; And / or, the reaction liquid is the raffinate in the hydrogen peroxide production process, the reaction parameter is the raffinate concentration, and the wave number interval of the characteristic spectral band is 3880-4204cm -1 and 4458-6194cm -1 .

8. The method according to claim 1, wherein: Before performing the three-dimensional processing, the method further includes: pre-processing the characteristic spectral band to eliminate redundant information and / or noise in the spectrum; Preferably, the preprocessing method includes first-order differential and / or second-order differential, more preferably second-order differential; Preferably, the window width of the second-order differential processing is 18-23.

9. The method according to claim 1 or 8, wherein: The regression analysis method is multidimensional partial least squares analysis.

10. The method according to claim 1 or 8, wherein: In step (2), N three-dimensional processing methods are used to perform N three-dimensional processing, and N three-dimensional spectra are obtained accordingly; each three-dimensional spectra is respectively subjected to regression analysis with the reaction parameters of the standard reaction solution to obtain N prediction models, and the N prediction models are weightedly fused to obtain a fusion model; The weighted fusion method includes: substituting the three-dimensional spectrograms obtained by the three-dimensional processing into the corresponding prediction models respectively, and obtaining the prediction results of each prediction model respectively; The multiple prediction results obtained above are assigned weights respectively, and the weight values ​​are adjusted so that the sum of the products of each weight and the prediction result is close to the true reaction parameter, wherein each weight value satisfies that each weight value is greater than 0 and less than 1, and the sum of each weight value is 1; The fusion model is the sum of the products of each weight value and the corresponding prediction model; In step (3), the three-dimensional spectrum corresponding to the reaction solution to be tested is substituted into the fusion model to obtain the reaction parameters to be tested.

11. The method according to claim 10, wherein: N is 2-3.

12. The method according to claim 1, wherein: Before the three-dimensional processing, the method further includes: preprocessing the selected characteristic spectral band to eliminate redundant information and / or noise in the spectrum; performing Fourier transformation on the preprocessed characteristic spectral band to obtain a real part and an imaginary part; Preferably, the three-dimensional processing method includes: stacking the real part and the imaginary part; Preferably, the three-dimensional processing method includes: performing outer product processing on the real part and the imaginary part; Preferably, the three-dimensional processing method includes: performing external sum processing on the real part and the imaginary part.

13. A method for determining the hydrogenation efficiency of a hydrogenation liquid in a hydrogen peroxide production process, characterized in that The method includes: (I) providing a plurality of standard hydrogenation liquids with known and different hydrogenation efficiencies, collecting near-infrared spectra of the standard hydrogenation liquids respectively, and selecting characteristic spectral bands of the near-infrared spectra; The characteristic spectral bands of the near-infrared spectrum are subjected to second-order differentiation and Fourier transformation to obtain the real and imaginary parts of the characteristic spectral bands; (II) The real and imaginary parts of the near-infrared spectrum are stacked to obtain a three-dimensional spectrum X dz ; The three-dimensional spectrum X dz , and the hydrogenation efficiency of the hydrogenation liquid were analyzed by multidimensional partial least squares to obtain the prediction model Y dz ; The real and imaginary parts corresponding to the near-infrared spectrum are processed by outer product to obtain a three-dimensional spectrum X wj , the three-dimensional spectrum X wj The multidimensional partial least squares analysis was performed on the hydrogenation efficiency of the hydrogenation liquid to obtain the prediction model Y wj ; The real and imaginary parts corresponding to the near-infrared spectrum are processed externally to obtain a three-dimensional spectrum X wh , and X wh The multidimensional partial least squares analysis was performed on the hydrogenation efficiency of the hydrogenation liquid to obtain the prediction model Y wh ; (III) X dz , substitute into Y dz The prediction result P is obtained dz , X wj , substitute into Y wj The prediction result P is obtained wj , X wh , substitute into Y wh The prediction result P is obtained wh , The above prediction results are assigned weights a, b and c respectively. dz +bP wj +cP wh Adjust a, b and c to aP dz +bP wj +cP wh The calculation result is close to the real hydrogenation efficiency of the hydrogenated liquid, where a, b, and c are respectively greater than 0 and less than 1, and a+b+c=1; The fusion model is obtained as M = aY dz +bY wj +cY wh ; (IV) Collect a near infrared spectrum of the reaction solution to be tested, and process it according to steps (I)-(II) to obtain a three-dimensional spectrum X dz ', X wj ' and X wh ', and substitute the above three-dimensional spectrum into the fusion model in step (III) to obtain the hydrogenation efficiency to be measured.

Images