A real-time monitoring method for multivariate curve resolution in the synthesis process of energetic compounds

Abstract

The present invention provides a multivariate curve resolution real-time monitoring method for the synthesis process of an energetic compound, which belongs to the field of spectral analysis technology, and includes: obtaining infrared spectrum data during the synthesis reaction of an energetic compound, and performing empirical mode decomposition denoising processing on the data; then performing baseline correction on the denoised data, and determining the optimal chemical level through KMeans clustering algorithm and differential analysis; using the determined optimal chemical level, performing interactive self-modeling hybrid analysis on the processed data to obtain a spectral matrix and a concentration matrix; then converting the two matrices into spectral graphs and chemical component concentration change graphs of different substances, respectively, and displaying the analysis results in real time through a graphical interface, while continuously monitoring and updating the results. The present invention realizes real-time data processing and display through spectral analysis technology and chemometric methods, can quickly obtain key indicators and change information in the reaction process, and improves the response capability to the reaction dynamics.

Description

Technical Field

[0001] The invention relates to the technical field of spectral analysis, and in particular to a multivariate curve resolution real-time monitoring method for an energetic compound synthesis process. Background Art

[0002] Energetic compounds refer to compounds that contain high-energy bonds or high-energy functional groups in their chemical structures. They usually have high energy release potential and can decompose or burn rapidly under certain conditions to release a large amount of energy. In the synthesis process of energetic compounds, real-time monitoring of the concentration changes of different substances in the system during the reaction is crucial to process improvement and product quality. However, there is still a lack of effective real-time monitoring methods in the existing technology.

[0003] Therefore, there is an urgent need to provide a real-time monitoring method to obtain accurate data support during the synthesis of energetic compounds, thereby optimizing the synthesis process and improving the quality of the final product. Summary of the invention

[0004] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide a multivariate curve resolution real-time monitoring method for the synthesis process of energetic compounds, which realizes real-time data processing and display through spectral analysis technology and chemometric methods, can quickly obtain key indicators and change information in the reaction process, and improve the response capability to reaction dynamics.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A multivariate curve resolution real-time monitoring method for an energetic compound synthesis process comprises the following steps:

[0007] S1, collecting infrared spectrum data of energetic compound synthesis reaction in real time, and performing empirical mode decomposition denoising processing on the infrared spectrum data;

[0008] S2, based on the iterative adaptive weighted penalty least squares algorithm, baseline correction is performed on the infrared spectrum data after denoising to obtain preprocessed data;

[0009] S3, according to the pre-processed data, determine the optimal chemical grade through KMeans clustering algorithm and differential analysis;

[0010] S4, using the determined optimal chemical level, performing interactive self-modeling hybrid analysis on the pre-processed data to obtain a spectral matrix and a concentration matrix;

[0011] S5. Convert the spectrum matrix and concentration matrix into spectrum diagrams and chemical composition concentration change diagrams of different substances, display the analysis results in real time through a graphical interface, and continuously monitor and update the results, thereby completing real-time monitoring of the energetic compound synthesis process.

[0012] Preferably, in step S1, real-time acquisition of infrared spectrum data during the energetic compound synthesis reaction includes: using an online infrared spectrometer to collect infrared spectrum data at 3000-650 cm within the first 20 minutes of the energetic compound synthesis reaction. -1 The spectral scanning range is 128 times, the number of scans is 128 times and the scanning time interval is 2 minutes, and the infrared spectrum data is collected in real time.

[0013] Preferably, the collected infrared spectrum data is subjected to empirical mode decomposition denoising processing, including:

[0014] S11, the collected infrared spectrum data is X (m×n), where m is the number of samples and n is the number of data points per spectrum;

[0015] S12. For any infrared spectrum data X j ={ x 1, x 2,…, x n} ( j =1,2,…,m), determine the local maximum and local minimum; if there is x p ( p =2,3,…,n-1) so that x p > x p-1 and x p > x p+1 ,but x p is a local maximum, if there is x p ( p =2,3,…,n-1) so that x p < x p-1 and x p < x p+1 ,but x p is a local minimum;

[0016] The upper envelope and the lower envelope are constructed by spline interpolation, where the upper envelope connects all local maximum points and the lower envelope connects all local minimum points, and the mean is calculated. :

[0017] ;

[0018] in, U represents the upper envelope, L represents the lower envelope;

[0019] S13. Using infrared spectroscopy data X j Subtract the mean Get the first intrinsic mode function:

[0020] ;

[0021] in, represents the intrinsic mode function;

[0022] S14, will get As new input data, repeat steps S11 to S13 to extract subsequent Components, until the set stop condition is met;

[0023] S15. All the extracted The components are reconstructed and combined to obtain the infrared spectrum data after empirical mode decomposition and denoising.

[0024] Preferably, in step S15, all the extracted The components are reconstructed and combined to obtain infrared spectrum data after empirical mode decomposition and denoising, including:

[0025] Before selection k indivual The components are used for reconstruction to obtain reconstructed data, and the obtained reconstructed data are combined to obtain infrared spectrum data after empirical mode decomposition and denoising. The formula is as follows:

[0026] ;

[0027] ;

[0028] in, Represents the reconstructed data, Represents the infrared spectrum data after denoising.

[0029] Preferably, in step S2, it specifically includes:

[0030] S21, initializing the parameters of the iterative adaptive weighted penalty least squares algorithm; including the penalty parameter λ , Iteration stop threshold ε and the maximum number of iterations T max ;

[0031] S22. Construct a second-order difference matrix, a diagonal weight matrix, and solve the least squares function with penalty terms:

[0032] ;

[0033] ;

[0034] ;

[0035] In the formula, represents the second-order difference matrix; represents the diagonal weight matrix; Indicates j The infrared spectrum data after empirical mode decomposition and denoising, j =1,2,…,m; Z =( z 1, z 2, z 3,…, z n ) represents the baseline of the fit; is the weight vector, if the current number of iterations t = 1, the weight vector are all 1. In each subsequent iteration, the weight vector Dynamically adjust according to the number of iterations;

[0036] S23, by finding the partial derivative vector and setting its value to 0, that is , calculate the baseline of the fit Z :

[0037] ;

[0038] ;

[0039] S24. Calculate residuals d , and update the weight vector :

[0040] ;

[0041] ;

[0042] In the formula, By residual d The negative element composition,t is the current iteration number;

[0043] S25, if the current number of iterations t Reached the maximum number of iterations T max Or when the termination condition provided by the following formula is met, the iteration stops and the fitted baseline is output Z Otherwise, return to step S22 to continue iteration;

[0044] The termination condition is:

[0045] ;

[0046] S26, No. j Infrared spectrum data after empirical mode decomposition denoising Subtract the fitted baseline Z , get the data after removing the baseline, and combine them to get the preprocessed data.

[0047] Preferably, in step S26, the formula used to obtain the data after baseline removal and the data after preprocessing is as follows:

[0048] ;

[0049] ;

[0050] In the formula, represents the data after removing the baseline. Represents the preprocessed data.

[0051] Preferably, in step S3, by using KMeans clustering algorithm and differential analysis, specifically including:

[0052] S31, using KMeans clustering algorithm, for different cluster numbers K ( K =1,2,3,…,m) to calculate cluster inertia I ( K ) to evaluate each K Clustering effect under value:

[0053] ;

[0054] In the formula, Representative o clusters, represents the cluster center;

[0055] S32. Calculate cluster inertia I ( K ) and find clustering relationships I (K )The acceleration and deceleration points of the changes:

[0056] ;

[0057] ;

[0058] In the formula, represents the first-order difference, represents the second-order difference;

[0059] S33, the second-order difference obtained by analysis , determine the cluster number where the minimum value is located as the optimal chemical level:

[0060] ;

[0061] In the formula, Indicates the cluster number where the minimum value is located.

[0062] Preferably, the optimal chemical level determined in step S33 is used to perform interactive self-modeling hybrid analysis on the preprocessed data to decompose a spectral matrix and a concentration matrix, wherein the spectral matrix includes the characteristic spectrum of each component in the synthesis of the energetic compound, and the concentration matrix includes the concentration of each component in the synthesis of the energetic compound at different time points.

[0063] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0064] (1) On the one hand, the present invention does not need to introduce additional intermediate capture agents during the synthesis of energetic compounds, which effectively reduces reagent consumption and saves experimental costs. On the other hand, by adopting real-time monitoring and analysis technology, key indicators and change information in the reaction process can be quickly captured, which significantly improves the response capability to the reaction dynamics, thereby facilitating timely discovery and adjustment of reaction conditions to ensure the smooth progress of the synthesis process.

[0065] (2) The present invention uses KMeans clustering and differential analysis methods to achieve automatic identification of chemical grades. This function reduces the reliance on manual intervention, not only improves the accuracy and efficiency of analysis, but also reduces the possibility of human error; at the same time, the method of the present invention is integrated into the real-time monitoring software, so that the real-time monitoring software can read, analyze and display experimental data in real time, so that users can easily obtain the required information, thereby enhancing the convenience of operation and user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be 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 labor.

[0067] Figure 1 It is a flow chart of a multivariate curve resolution real-time monitoring method for an energetic compound synthesis process of the present invention;

[0068] Figure 2 This is a spectrum data diagram provided in Example 1 of the present invention; wherein, Figure 2 (a) is the original spectrum data. Figure 2 (b) is the extracted baseline spectrum data. Figure 2 (c) is the spectral data after the empirical mode decomposition denoising process and the baseline removal by iterative adaptive weighted penalty least squares algorithm;

[0069] Figure 3 The best chemical grade determination diagram combining KMeans clustering algorithm and differential analysis provided in Example 1 of the present invention;

[0070] Figure 4 This is a real-time monitoring diagram of the first twenty minutes of the energetic compound synthesis reaction provided in Example 1 of the present invention; wherein, Figure 4 (a) is a schematic diagram of the three-dimensional results of spectral data. Figure 4 (b) is the relationship between spectral absorbance and wave number. Figure 4 (c) is the chemical composition concentration change diagram of different concentration matrices. Figure 4 (d) in the figure is the spectrum diagram of different substances converted from different spectral matrices. Component 1, component 2 and component 3 represent three different substances respectively. DETAILED DESCRIPTION

[0071] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0072] In order to make the purpose, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0073] Embodiment 1

[0074] like Figure 1 As shown, the present invention provides a multivariate curve resolution real-time monitoring method for the synthesis process of an energetic compound, comprising the following steps:

[0075] S1, collecting infrared spectrum data of energetic compound synthesis reaction in real time, and performing empirical mode decomposition denoising processing on the infrared spectrum data;

[0076] S2, based on the iterative adaptive weighted penalty least squares algorithm, baseline correction is performed on the infrared spectrum data after denoising to obtain preprocessed data;

[0077] S3, according to the pre-processed data, determine the optimal chemical grade through KMeans clustering algorithm and differential analysis;

[0078] S4, using the determined optimal chemical level, performing interactive self-modeling hybrid analysis on the pre-processed data to obtain a spectral matrix and a concentration matrix;

[0079] S5. Convert the spectrum matrix and concentration matrix into spectrum diagrams and chemical composition concentration change diagrams of different substances, display the analysis results in real time through a graphical interface, and continuously monitor and update the results, thereby completing real-time monitoring of the energetic compound synthesis process.

[0080] In step S1, real-time infrared spectrum data of energetic compound synthesis reaction is collected, including: using an online infrared spectrometer to collect infrared spectrum data of energetic compound synthesis reaction in the first 20 minutes at 3000-650 cm -1 The infrared spectrum data is collected in real time with a spectral scanning range of 128 scans and a scanning interval of 2 minutes. In this embodiment, the original spectrum data obtained is shown in FIG. Figure 2 As shown in (a) in .

[0081] Secondly, the collected infrared spectrum data is subjected to empirical mode decomposition (EMD) denoising, including:

[0082] S11, the collected infrared spectrum data is X (m×n), where m is the number of samples and n is the number of data points per spectrum;

[0083] S12. For any infrared spectrum data X j ={ x 1, x 2,…, x n} ( j =1,2,…,m), determine the local maximum and local minimum; if there is x p (p =2,3,…,n-1) so that x p > x p-1 and x p > x p+1 ,but x p is a local maximum, if there is x p ( p =2,3,…,n-1) so that x p < x p-1 and x p < x p+1 ,but x p is a local minimum;

[0084] The upper envelope and the lower envelope are constructed by spline interpolation, where the upper envelope connects all local maximum points and the lower envelope connects all local minimum points, and the mean is calculated. :

[0085] ;

[0086] in, U represents the upper envelope, L represents the lower envelope;

[0087] S13. Using infrared spectroscopy data X j Subtract the mean Get the first intrinsic mode function:

[0088] ;

[0089] in, represents the intrinsic mode function;

[0090] S14, will get As new input data, repeat steps S11 to S13 to extract subsequent Components, until the set stop condition is met;

[0091] S15. All the extracted The components are reconstructed and combined to obtain infrared spectrum data after empirical mode decomposition and denoising, including:

[0092] Before selection k indivual The components are used for reconstruction to obtain reconstructed data, and the obtained reconstructed data are combined to obtain infrared spectrum data after empirical mode decomposition and denoising. The formula is as follows:

[0093] ;

[0094] ;

[0095] in, Represents the reconstructed data, Represents the infrared spectrum data after denoising.

[0096] In this embodiment, the number of samples m is 10, and the number of data points of each spectrum is 589. There are 5 portions in total, select the first 4 The components are used for reconstruction, and the data after EMD denoising is X EMD (10×589).

[0097] Specifically, in step S2, the data is baseline corrected based on the iterative adaptive weighted penalized least squares (airPLS) algorithm, including:

[0098] S21, initializing the parameters of the iterative adaptive weighted penalty least squares algorithm; including the penalty parameter λ , Iteration stop threshold ε and the maximum number of iterations T max ;

[0099] S22. Construct a second-order difference matrix, a diagonal weight matrix, and solve the least squares function with penalty terms:

[0100] ;

[0101] ;

[0102] ;

[0103] In the formula, represents the second-order difference matrix; represents the diagonal weight matrix; Indicates j The infrared spectrum data after empirical mode decomposition and denoising, j =1,2,…,m; Z =( z 1, z 2, z 3,…, z n ) represents the baseline of the fit; is the weight vector, if the current number of iterations t = 1, the weight vector are all 1. In each subsequent iteration, the weight vector Dynamically adjust according to the number of iterations;

[0104] S23, by finding the partial derivative vector and setting its value to 0, that is , calculate the baseline of the fit Z :

[0105] ;

[0106] ;

[0107] S24. Calculate residuals d , and update the weight vector :

[0108] ;

[0109] ;

[0110] In the formula, By residual d The negative element composition, t is the current iteration number;

[0111] S25, if the current number of iterations t Reached the maximum number of iterations T max Or when the termination condition provided by the following formula is met, the iteration stops and the fitted baseline is output Z Otherwise, return to step S22 to continue iteration;

[0112] The termination condition is:

[0113] ;

[0114] S26, No. j Infrared spectrum data after empirical mode decomposition denoising Subtract the fitted baseline Z , get the data after removing the baseline, and combine to get the preprocessed data. The formula is as follows:

[0115] ;

[0116] ;

[0117] In the formula, represents the data after removing the baseline. Represents the preprocessed data.

[0118] In this embodiment, the penalty parameter λ Iteration stop threshold is 0.5 ε is 0.001, the maximum number of iterations T max is 20, and the fitted baseline is Figure 2 As shown in (b), the spectral data after denoising and baseline correction is as follows Figure 2 As shown in (c) in .

[0119] Specifically, in step S3, the KMeans clustering algorithm and differential analysis are used, which specifically include:

[0120] S31, using KMeans clustering algorithm, for different cluster numbers K ( K =1,2,3,…,m) to calculate cluster inertia I ( K ) to evaluate each K Clustering effect under value:

[0121] ;

[0122] In the formula, Representative o clusters, represents the cluster center;

[0123] S32. Calculate cluster inertia I ( K ) and find clustering relationships I ( K )The acceleration and deceleration points of the changes:

[0124] ;

[0125] ;

[0126] In the formula, represents the first-order difference, represents the second-order difference;

[0127] S33, the second-order difference obtained by analysis , determine the cluster number where the minimum value is located as the optimal chemical level:

[0128] ;

[0129] In the formula, Indicates the cluster number where the minimum value is located.

[0130] In this embodiment, the number of clusters where the minimum value is located is determined is 3, that is, the best chemical level is 3, and the results are as follows Figure 3 shown.

[0131] Finally, the optimal chemical level determined in step S33 is used to perform interactive self-modeling hybrid analysis (SIMPLISMA) on the pre-processed data to decompose the spectral matrix and the concentration matrix. The results are as follows: Figure 4 The spectrum matrix includes the characteristic spectrum of each component in the synthesis of the energetic compound, and the concentration matrix includes the concentration of each component at different time points in the synthesis of the energetic compound. The obtained spectrum matrix is ​​converted into spectrum diagrams of different substances, such as Figure 4 As shown in (d) in the figure, the obtained concentration matrix is ​​converted into different chemical component concentration change diagrams, such as Figure 4 The analysis results are displayed in real time through a graphical interface. Figure 4 (a), (b), (c) and (d) are continuously monitored and the analysis results are updated to complete the real-time monitoring of the energetic compound synthesis reaction process.

[0132] Therefore, the above-mentioned multivariate curve resolution real-time monitoring method for the synthesis process of an energetic compound is adopted, and real-time data processing and display are realized through spectral analysis technology and chemometric methods, which can quickly obtain key indicators and change information in the reaction process and improve the response capability to the reaction dynamics.

[0133] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A multivariate curve resolution real-time monitoring method for energetic compound synthesis process, characterized in that: The following steps are involved: S1, collecting infrared spectrum data of energetic compound synthesis reaction in real time, and performing empirical mode decomposition denoising processing on the infrared spectrum data; S2, based on the iterative adaptive weighted penalty least squares algorithm, baseline correction is performed on the infrared spectrum data after denoising to obtain preprocessed data; S3, according to the pre-processed data, determine the optimal chemical grade through KMeans clustering algorithm and differential analysis; S4, using the determined optimal chemical level, performing interactive self-modeling hybrid analysis on the pre-processed data to obtain a spectral matrix and a concentration matrix; S5. Convert the spectrum matrix and concentration matrix into spectrum diagrams and chemical composition concentration change diagrams of different substances, display the analysis results in real time through a graphical interface, and continuously monitor and update the results, thereby completing real-time monitoring of the energetic compound synthesis process.

2. The multivariate curve resolution real-time monitoring method for the synthesis process of an energetic compound according to claim 1, characterized in that: In step S1, real-time infrared spectrum data of energetic compound synthesis reaction is collected, including: using an online infrared spectrometer to collect infrared spectrum data of energetic compound synthesis reaction in the first 20 minutes at 3000-650 cm -1 The spectral scanning range is 128 times, the number of scans is 128 times and the scanning time interval is 2 minutes, and the infrared spectrum data is collected in real time.

3. The multivariate curve resolution real-time monitoring method for the synthesis process of an energetic compound according to claim 2, characterized in that: The collected infrared spectrum data is subjected to empirical mode decomposition and denoising, including: S11, the collected infrared spectrum data is X (m×n), where m is the number of samples and n is the number of data points per spectrum; S12. For any infrared spectrum data X j ={ x 1, x 2,…, x n } ( j =1,2,…,m), determine the local maximum and local minimum; if there is x p ( p =2,3,…,n-1) so that x p > x p-1 and x p > x p+1 ,but x p is a local maximum, if there is x p ( p =2,3,…,n-1) so that x p < x p-1 and x p < x p+1 ,but x p is a local minimum; The upper envelope and the lower envelope are constructed by spline interpolation, where the upper envelope connects all local maximum points and the lower envelope connects all local minimum points, and the mean is calculated. : ； in, U represents the upper envelope, L represents the lower envelope; S13. Using infrared spectroscopy data X j Subtract the mean Get the first intrinsic mode function: ； in, represents the intrinsic mode function; S14, will get As new input data, repeat steps S11 to S13 to extract subsequent Components, until the set stop condition is met; S15. All the extracted The components are reconstructed and combined to obtain the infrared spectrum data after empirical mode decomposition and denoising.

4. The multivariate curve resolution real-time monitoring method for the synthesis process of an energetic compound according to claim 3, characterized in that: In step S15, all the extracted The components are reconstructed and combined to obtain infrared spectrum data after empirical mode decomposition and denoising, including: Before selection k indivual The components are used for reconstruction to obtain reconstructed data, and the obtained reconstructed data are combined to obtain infrared spectrum data after empirical mode decomposition and denoising. The formula is as follows: ； ； in, Represents the reconstructed data, Represents the infrared spectrum data after denoising.

5. The multivariate curve resolution real-time monitoring method for the synthesis process of an energetic compound according to claim 4, characterized in that: In step S2, it specifically includes: S21, initializing the parameters of the iterative adaptive weighted penalty least squares algorithm; including the penalty parameter λ , Iteration stop threshold ε and the maximum number of iterations T max ; S22. Construct a second-order difference matrix, a diagonal weight matrix, and solve the least squares function with penalty terms: ； ； ； In the formula, represents the second-order difference matrix; represents the diagonal weight matrix; Indicates j The infrared spectrum data after empirical mode decomposition and denoising, j =1,2,…,m; Z =( z 1, z 2, z 3,…, z n ) represents the baseline of the fit; is the weight vector, if the current number of iterations t = 1, the weight vector are all 1. In each subsequent iteration, the weight vector Dynamically adjust according to the number of iterations; S23, by finding the partial derivative vector and setting its value to 0, that is , calculate the baseline of the fit Z : ； ； S24. Calculate residuals d , and update the weight vector : ； ； In the formula, By residual d The negative element composition, t is the current iteration number; S25, if the current number of iterations t Reached the maximum number of iterations T max Or when the termination condition provided by the following formula is met, the iteration stops and the fitted baseline is output Z Otherwise, return to step S22 to continue iteration; The termination condition is: ； S26, No. j Infrared spectrum data after empirical mode decomposition denoising Subtract the fitted baseline Z , get the data after removing the baseline, and combine them to get the preprocessed data.

6. The multivariate curve resolution real-time monitoring method for energetic compound synthesis process according to claim 5, characterized in that: In step S26, the formula used to obtain the data after baseline removal and the data after preprocessing is as follows: ； ； In the formula, represents the data after removing the baseline. Represents the preprocessed data.

7. The method for real-time monitoring of a synthesis process of an energetic compound by multivariate curve resolution according to claim 6, characterized in that: In step S3, the KMeans clustering algorithm and differential analysis are used, specifically including: S31, using KMeans clustering algorithm, for different cluster numbers K ( K =1,2,3…,m) to calculate cluster inertia I ( K ) to evaluate each K Clustering effect under value: ； In the formula, Representative o clusters, represents the cluster center; S32. Calculate cluster inertia I ( K ) and find clustering relationships I ( K )Changed acceleration and deceleration points: ； ； In the formula, represents the first-order difference, represents the second-order difference; S33, the second-order difference obtained by analysis , determine the cluster number where the minimum value is located as the optimal chemical level: ； In the formula, Indicates the cluster number where the minimum value is located.

8. The multivariate curve resolution real-time monitoring method for the synthesis process of an energetic compound according to claim 7, characterized in that: The optimal chemical level determined in step S33 is used to perform interactive self-modeling hybrid analysis on the preprocessed data to decompose a spectral matrix and a concentration matrix, wherein the spectral matrix includes the characteristic spectrum of each component in the synthesis of the energetic compound, and the concentration matrix includes the concentration of each component in the synthesis of the energetic compound at different time points.

Images