Optimization Method and System for C4 Olefin Preparation Conditions Based on Partial Least Squares Analysis

Abstract

This invention belongs to the field of ethanol coupling for the preparation of C4 olefins, and particularly relates to a method and system for optimizing C4 olefin preparation conditions based on partial least squares analysis. The method includes: acquiring experimental data on the ethanol coupling for C4 olefin preparation; performing correlation analysis on the experimental data and selecting strongly correlated reaction influencing factors as independent variables; classifying the by-products of the ethanol coupling for C4 olefin preparation and selecting dependent variables based on the main reaction products and by-products; performing regression modeling on the independent and dependent variables using partial least squares to obtain a multiple linear regression equation; establishing a single-objective optimization model to maximize the C4 olefin yield based on the multiple linear regression equation; and solving the single-objective optimization model to obtain the optimal preparation conditions that maximize the C4 olefin yield. This invention, based on partial least squares regression analysis, is beneficial for studying the relationship between multiple dependent variables and provides a deeper understanding of the potential correlations between them.

Description

Technical Field

[0001] This invention belongs to the field of ethanol coupling for the preparation of C4 olefins, and particularly relates to an optimization method and system for C4 olefin preparation conditions based on partial least squares analysis. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] C4 olefins are important basic chemical raw materials that can be used in chemical industrial production and pharmaceutical manufacturing, such as in the production of polypropylene and other chemical products. C4 olefins are also the foundation of the petrochemical industry and are high-quality gasoline blending components.

[0004] In recent years, with the continuous progress of China's chemical industry, the comprehensive utilization of C4 olefins has received increasing attention. Selecting appropriate catalytic production processes to achieve stable and efficient production will help improve the economic benefits of relevant enterprises and promote the development of my country's chemical industry. Furthermore, due to the increasingly serious global environmental pollution problem, traditional olefin production methods using fossil fuels are no longer applicable, and people are actively seeking preparation methods using clean energy as raw materials.

[0005] Ethanol, as a biorenewable energy source, is cleaner and more environmentally friendly than traditional fossil fuels, effectively reducing emissions of gases such as carbon dioxide and mitigating the global greenhouse effect. Therefore, exploring the effects of catalyst combinations (i.e., the combination of Co loading, Co / SiO2 and HAP charge ratios, and ethanol concentration) and temperature on the coupling of ethanol to olefins is of significant research value. Summary of the Invention

[0006] To overcome the shortcomings of the prior art, the present invention provides an optimization method and system for C4 olefin preparation conditions based on partial least squares analysis.

[0007] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:

[0008] The first aspect of this invention provides a method for optimizing C4 olefin preparation conditions based on partial least squares analysis, comprising:

[0009] To obtain experimental data on the preparation of C4 olefins by ethanol coupling under different catalyst combinations and reaction temperatures;

[0010] Correlation analysis was performed on the obtained experimental data, and factors showing strong correlation were selected as independent variables.

[0011] Byproducts of the coupling of ethanol to prepare C4 olefins were classified and merged, and dependent variables were selected based on the main reaction products and byproducts.

[0012] Regression modeling of independent and dependent variables is performed based on partial least squares method to obtain multiple linear regression equation;

[0013] A single-objective optimization model for maximizing the yield of C4 olefins was established based on a multiple linear regression equation.

[0014] Solving the single-objective optimization model maximizes the optimal preparation conditions that maximize the yield of C4 olefins.

[0015] A second aspect of the present invention provides an optimization system for C4 olefin preparation conditions based on partial least squares analysis, comprising:

[0016] The experimental data acquisition module is configured to acquire experimental data on the preparation of C4 olefins by ethanol coupling under different catalyst combinations and reaction temperatures.

[0017] The correlation analysis module is configured to perform correlation analysis on the acquired experimental data and select response factors that show strong correlation as independent variables.

[0018] The classification module is configured to classify the by-products of the ethanol coupling to prepare C4 olefins, and select dependent variables based on the main reaction products and by-products.

[0019] The partial least squares analysis module is configured to: perform regression modeling on independent and dependent variables based on the partial least squares method to obtain a multiple linear regression equation;

[0020] The optimization model building module is configured to: build a single-objective optimization model for maximizing the yield of C4 olefins based on a multiple linear regression equation;

[0021] The model solver module is configured to solve the single-objective optimization model to obtain the optimal preparation conditions that maximize the yield of C4 olefins.

[0022] A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps in a method for optimizing C4 olefin preparation conditions based on partial least squares analysis as described in the first aspect of the present invention.

[0023] The fourth aspect of the present invention provides an electronic device including a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in the method for optimizing C4 olefin preparation conditions based on partial least squares analysis as described in the first aspect of the present invention.

[0024] The above one or more technical solutions have the following beneficial effects:

[0025] (1) In analyzing the effects of different catalyst combinations and temperatures on ethanol conversion and C4 olefin selectivity, this invention applies partial least squares regression analysis to optimize conventional analytical methods and explore the interactions and influences of multiple dependent variables during the experiment. From a horizontal perspective, the addition of partial least squares regression analysis is beneficial for studying the relationships between multiple dependent variables, allowing for a deeper focus on the potential correlations between them. This helps us to more comprehensively analyze the relationships between various products when studying chemical reactions, thereby extending and expanding the model to play a greater role in medical diagnosis and chemical experiment prediction, achieving "predictive foresight."

[0026] (2) When considering the numerical relationship between the selectivity of reaction byproducts, in order to avoid the problem of excessive redundancy of variables, this invention classifies and merges each reaction product, uses SPSS software to perform Q-type clustering, creates a data spectrum of all byproducts, and classifies the byproducts produced by the reaction, thereby reducing the dimensionality of the data and making it easy to understand and use the data; it is only necessary to consider the influence of two types of products on ethanol conversion and C4 olefin selectivity.

[0027] (3) The C4 olefin yield is obtained by multiplying the ethanol conversion rate and the C4 olefin selectivity. The C4 olefin yield is directly limited by the coefficient between the ethanol conversion rate and the C4 olefin conversion rate. Therefore, we introduce analysis of variance to study whether there is an interaction relationship between these variables, thereby optimizing the single-objective optimization model for C4 olefins, improving the fitting effect, facilitating the solution of the optimal catalyst combination and the most suitable temperature for the ethanol-coupled preparation of C4 olefins, and simple verification can be used to improve the process conditions for the ethanol-coupled preparation of C4 olefins.

[0028] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0029] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0030] Figure 1 This is a flowchart of the method in the first embodiment.

[0031] Figure 2 (a) and (b) in the figure are fitting curves of ethanol conversion and C4 olefin selectivity, respectively.

[0032] Figure 3 This is a clustering diagram of the reactants in the first embodiment.

[0033] Figure 4(a) and (b) in the figure are the prediction graphs of ethanol conversion rate and olefin conversion rate, respectively.

[0034] Figure 5 This is the partial least squares regression coefficient histogram for the first embodiment. Detailed Implementation

[0035] In order to construct a more accurate mathematical model, this invention makes the following reasonable assumptions:

[0036] 1. In the process of catalytic coupling of ethanol to prepare C4 olefins, apart from the given catalyst combination, temperature and reaction time, the influence of other conditions on the experimental results can be ignored.

[0037] 2. During the reaction, the ambient temperature and pressure in the laboratory will not affect the reaction.

[0038] 3. The experimental equipment is in good working order and there will be no leakage, damage or other issues during the experiment that could lead to experimental failure or data errors.

[0039] Symbol explanation:

[0040] Table 1

[0041]

[0042]

[0043] Note: Undeclared variables are subject to the specific description at the place where the symbol appears.

[0044] Let y1 represent ethanol conversion, y2 represent C4 olefin selectivity, y3 represent fatty alcohols with 4-12 carbon atoms, and y4 represent the combined selectivity of methylbenzaldehyde and methylbenzyl alcohol, acetaldehyde selectivity, ethylene selectivity, and selectivity of other products. Let x1 represent the Co / SiO2 mass, x2 represent the Co loading, x3 represent the HAP mass, x4 represent the ethanol concentration, and x5 represent the reaction temperature.

[0045] Example 1

[0046] To achieve the maximum yield of C4 olefins, the influence of the preparation environment on the C4 yield was analyzed and integrated, and a model was established to determine the optimal preparation environment. That is, the relationship between environmental factors such as catalyst combination and temperature and the C4 olefin yield was analyzed through experimental data, and then an optimization model was established based on the obtained relationship and the actual situation to obtain the preparation environment that maximizes the C4 olefin yield.

[0047] like Figure 1 As shown, this embodiment discloses an optimization method for C4 olefin preparation conditions based on partial least squares analysis, including:

[0048] Step 1: Obtain experimental data on the preparation of C4 olefins by ethanol coupling under different catalyst combinations and reaction temperatures;

[0049] Step 2: Perform correlation analysis on the obtained experimental data and select response factors that show strong correlation as independent variables;

[0050] Step 3: Classify the by-products of the coupling of ethanol to prepare C4 olefins, and select dependent variables based on the main reaction products and by-products;

[0051] Step 4: Perform regression modeling on the independent and dependent variables based on the partial least squares method to obtain the multiple linear regression equation;

[0052] Step 5: Establish a single-objective optimization model to maximize the yield of C4 olefins based on the multiple linear regression equation;

[0053] Step 6: Solve the single-objective optimization model to obtain the optimal preparation conditions that maximize the yield of C4 olefins.

[0054] In step 1, a series of experiments were conducted in the chemical laboratory with different catalysts and temperatures. The recorded data included the conversion rates of reaction products such as ethylene, C4 olefins, acetaldehyde, and fatty alcohols with 4-12 carbon atoms under different catalyst combinations and temperatures, as well as the test data of a given catalyst combination at 350 degrees Celsius over time.

[0055] Wherein, temperature is the reaction temperature; selectivity is the proportion of a certain product in all products; time is the reaction time of the catalyst in an ethanol atmosphere; Co loading is the weight ratio of Co to SiO2; HAP is a catalyst support, Chinese name hydroxyapatite; Co / SiO2 and HAP loading ratio refers to the mass ratio of Co / SiO2 and HAP; ethanol conversion rate is the single-pass conversion rate of ethanol per unit time, and its value is 100% (ethanol inlet gas rate - ethanol residue) / ethanol inlet gas rate; C4 olefin yield is its value as ethanol conversion rate × C4 olefin selectivity.

[0056] Step 2 includes: Step 201: Correlation analysis

[0057] By calculating the Pearson correlation coefficient, the correlation between ethanol conversion, C4 olefin selectivity and temperature was analyzed for each catalyst combination.

[0058]

[0059] In the formula, X represents temperature, and Y represents ethanol conversion and C4 olefin selectivity, respectively.

[0060] Of the 21 sets of data obtained in the laboratory, 19 sets had Pearson correlation coefficients greater than 0.9. Therefore, under different catalyst combinations, reaction temperature and ethanol conversion and C4 olefin selectivity showed a strong correlation.

[0061] Based on the components of each catalyst combination, the present invention classifies the catalyst combinations into four categories, as shown in Table 2:

[0062] Table 2

[0063]

[0064] Step 202: Fitting the nonlinear regression curve

[0065] The above analysis shows a strong correlation between reaction temperature, ethanol conversion rate, and C4 olefin selectivity, with the data exhibiting a positive correlation trend. Therefore, nonlinear regression was used for curve fitting, employing linear equations, S-shaped curves, and quadratic curves for fitting. Grouped by catalyst combination, 21 fitted images were obtained, and the goodness of fit R0 was satisfactory. 2 The differences are not significant, both exceeding 93%. Figure 2 As shown in (a) and (b) in the figure.

[0066] The formula for C4 olefin yield is directly limited by the coefficient between ethanol conversion rate and C4 olefin conversion rate. Therefore, analysis of variance is introduced to investigate whether there are any interactions between these variables, thereby optimizing the single-objective optimization model for C4 olefins.

[0067] Step 203: Goodness-of-fit test

[0068] This embodiment calculates the goodness-of-fit R of the two representative fitted images mentioned above. 2 As shown in Table 3, the goodness of fit Ri using the quadratic regression function is... 2 The goodness of fit is greater than that of linear and "S"-shaped curves. Therefore, the ethanol conversion rate and the selectivity of C4 olefins have a quadratic regression function form with temperature, and are positively correlated.

[0069] Table 3

[0070]

[0071] Step 204: Data Preprocessing

[0072] This embodiment selects all types of experiments and uses the pandas library in Python to decompose all catalyst combinations into groups. Since the charge ratio exists in the data as a ratio, which is not conducive to data processing, the charge ratio is decomposed into Co / SiO2 mass and HAP mass, resulting in two separate independent variables. Ultimately, five independent variables are obtained.

[0073] It should be noted that the quartz sand acted as an anti-boiling agent in the experiment, and the efficiency did not significantly increase after adding quartz sand, thus the ethanol conversion rate and C4 olefin selectivity did not exhibit catalytic activity. Therefore, this embodiment discards the quartz sand variable.

[0074] Step 3 includes: Considering the numerical relationships between the selectivity of reaction byproducts, but with excessively redundant variables, this embodiment classifies and merges the various reaction products, uses SPSS software for Q-type clustering, and generates a phylogenetic map of all byproduct data (e.g., Figure 3 (as shown), and classify the byproducts produced by the reaction.

[0075] Analysis of the pedigree chart reveals that the selectivity of ethylene, methylbenzaldehyde and methylbenzyl alcohol, acetaldehyde, and other products can be grouped into one category, called minor byproducts; the selectivity of fatty alcohols can be grouped into another category, called major byproducts. This achieves data dimensionality reduction, requiring only consideration of the impact of these two types of products on ethanol conversion and C4 olefin selectivity.

[0076] Through the above steps, the dependent and independent variables selected in this embodiment are shown in Table 4:

[0077] Table 4

[0078]

[0079]

[0080] In step 4, the data obtained after data preprocessing is used as independent variables to analyze the main factors affecting ethanol conversion rate and C4 olefin selectivity. For ethanol conversion rate and C4 olefin selectivity, we optimize the conventional analysis method by partial least squares method to explore the analysis of multiple dependent variables. Finally, the regression coefficients are used to analyze the main factors affecting ethanol conversion rate and C4 olefin conversion rate.

[0081] In a chemical context, when studying multi-dependent variable problems, the dependent variables are likely to influence each other. For the two dependent variables in this invention, partial least squares method is used to improve the original analysis method, explore the analysis of multi-dependent variables, and investigate whether there is an influence between the dependent variables on the original basis.

[0082] The advantages of partial least squares method in processing data are as follows:

[0083] [1] A regression modeling method for multiple dependent variables on multiple independent variables is provided;

[0084] [2] Effectively solves the problem of multicollinearity among variables;

[0085] [3] Suitable for regression modeling when the number of sample points is less than the number of independent variables;

[0086] [4] The final model contains all the original independent variables, and the regression coefficients are easy to interpret.

[0087] Following the method in step 2, the dependent and independent variables are categorized and stored in the following two matrices:

[0088] The observed data matrix A of the independent variable is (a ij ) 108×5 The observed data matrix of the dependent variable is B = (b ij ) 108×4 .

[0089] Step 401: Data Standardization

[0090] Let each indicator value a ij Convert into standardized indicator values have

[0091]

[0092] in, Right now For the j-th independent variable x j The sample mean and sample standard deviation.

[0093] Correspondingly, it is called:

[0094]

[0095] These are standardized variables.

[0096] Similarly, there are standardized index values:

[0097]

[0098] in, Right now For the j-th dependent variable y j The sample mean and sample standard deviation.

[0099] Correspondingly, it is called:

[0100]

[0101] These are standardized variables.

[0102] Step 402: Calculate the correlation coefficient matrix

[0103] Table 5 presents the simple correlation coefficient matrix of these nine variables:

[0104] Table 5

[0105]

[0106]

[0107] The correlation coefficient matrix shows that Co / SiO2 mass (x1) is positively correlated with Co loading (x2), Co loading (x2) is positively correlated with HAP mass (x3) and ethanol concentration (x4), HAP mass (x3) is negatively correlated with ethanol concentration (x4), ethanol conversion rate (y1) is positively correlated with Co / SiO2 mass (x1), Co loading (x2), HAP mass (x3) and temperature (x5), and negatively correlated with ethanol concentration (x4); C4 olefin selectivity (y2) is positively correlated with Co / SiO2 mass (x1), HAP mass (x3) and temperature (x5), and negatively correlated with Co loading (x2) and ethanol concentration (x4).

[0108] Step 403: Extract the principal components of the independent variable group and the dependent variable group respectively, and maximize their correlation.

[0109] The principal components [u1,v1], [u2,v2], [u3,v3], [u4,v4], and [u5,v5] can be obtained using the Matlab program.

[0110]

[0111] Simplified to:

[0112]

[0113] The first four components explain 98.69% of the independent variables, so selecting only the first four pairs of components is sufficient to meet the analytical requirements.

[0114] Specifically: Suppose that the first pair of components, u1 and v1, are extracted from two sets of variables, where u1 is the set of independent variables X = [x1, ..., x...]. m ] T Linear combination: u1=α 11 x1+…+α 1m x m =ρ (1)T X and v1 are the dependent variable set Y = [y1, ..., y1]. p ] T Linear combination: v1 = β 11y1+…+β 1p y p =γ (1)T Y.

[0115] For the purposes of regression analysis, the following is required:

[0116] i) Extract as much variation information as possible from the variable groups to which u1 and v1 belong;

[0117] ii) The correlation between u1 and v1 reaches its maximum.

[0118] From the standardized observation data matrices A and B of the two variable sets, the score vector of the first pair of components can be calculated, denoted as... and

[0119]

[0120]

[0121] The covariance Cov(u1,v1) of the first pair of components u1 and v1 can be expressed as the score vector of the first pair of components. and The inner product is used for calculation.

[0122] Therefore, the above two requirements can be transformed into a mathematical problem of conditional extrema:

[0123]

[0124]

[0125] Step 404: Find the regression equation between the principal component pairs and the standardized index variables.

[0126] Establish y1,…,y p Regression on u1 and x1,…,x m Regression on u1.

[0127] Assume the regression model is as follows:

[0128]

[0129] Where, σ (1) =[σ 11 ,…,σ 1m ] T , τ (1) =[τ 11 ,…,τ 1p ] T These are the parameter vectors in a many-to-one regression model, and A1 and B1 are the residual matrices.

[0130] Regression coefficient vector σ(1) ,τ (1) The least squares estimate is:

[0131]

[0132] σ (1) ,τ (1) This represents the model effect load.

[0133] Replace A and B with residual matrices A1 and B1 and repeat the above steps. Then the residual matrix If the absolute values ​​of the elements in residual matrix B1 are approximately 0, then the accuracy of the regression formula established using the first component is considered sufficient, and component extraction can be stopped. Otherwise, replace A and B with residual matrices A1 and B1 and repeat the above steps to obtain ρ. (2) =[α 21 ,…,α 2m ] T γ (2) =[β 21 ,…,β 2p ] T ,

[0134] and This is the score vector for the second pair of components. Let X and Y be the loading amounts of the second pair of components, respectively. Then we have...

[0135]

[0136] Therefore, since the absolute values ​​of the elements in residual matrix B5 are approximately 0, the accuracy of the regression equation established using the first four components u1, u2, u3, u4, u5 is considered sufficient, and component extraction can be stopped. Subsequently, the regression equations between the independent and dependent variable groups and u1, u2, u3, u4, u5 are obtained as follows:

[0137]

[0138] Step 405: Find the regression equation between the dependent variable group and the independent variable group.

[0139] Will Substitution The regression equation is obtained by regressing the standardized index variables;

[0140] Standardized variables Restored to the original variable y i ,x j The regression equation is obtained as follows:

[0141]

[0142] Step 406: Interpretation of the Model

[0143] When modeling using partial least squares, a discrete graph about the central axis is plotted based on the predicted and actual values, resulting in predicted ethanol conversion and olefin conversion graphs, shown in figures 4(a) and (b), respectively. Figure 5 The data distribution indicates a good fit.

[0144] The regression coefficient graphs reveal that temperature plays a crucial role in explaining the two regression equations for ethanol conversion and C4 olefin selectivity. When temperature is combined with other independent variables, it significantly impacts the dependent variable. Furthermore, the explanatory power of the Co / SiO2 and HAP loading ratios is similar in both regression equations, showing a positive correlation. Co loading offers virtually no explanation for ethanol conversion, but exhibits a strong negative correlation in C4 olefin selectivity. Ethanol concentration, on the other hand, has a negative explanatory value for ethanol conversion but offers little explanation for C4 olefin selectivity.

[0145] Step 5 includes:

[0146] Step 501: Establish a basic maximization single-objective optimization model

[0147] Based on the clustering results above, the products were divided into C4 olefins, major byproducts (fatty alcohols), and minor byproducts, and data analysis was performed.

[0148] The yield formula: C4 olefin yield = ethanol conversion × C4 olefin selectivity, allows us to calculate the C4 olefin yield for any catalyst combination. Therefore, the latter two factors directly influence the C4 olefin yield. Considering the need for optimal data fitting to achieve the maximum C4 olefin yield, and the existence of some poorly fitted data in the experiments, we ranked 109 C4 olefin yields and set a threshold of 1%. This threshold was used as a preliminary screening criterion. Simultaneously, some control experiments with poor results were also removed, ultimately yielding 61 valid data sets.

[0149] Let ρ represent the yield of C4 olefins, then the following relationship can be obtained:

[0150] ρ=y1×y2 (10)

[0151] Establish a single-objective optimization model to maximize the yield of C4 olefins:

[0152] 1. Determining the objective function

[0153] maxρ=y1×y2 (11)

[0154] 2. Determining Constraints

[0155] Constraint 1: If the sum of the selectivity of all products equals 100%, then we have

[0156]

[0157] Constraint 2: All independent variables must satisfy the range of values ​​specified in the existing experimental design. Based on the data splitting in Appendix 1 of Question 2, the actual range of values ​​for each independent variable is obtained. Therefore, one constraint is:

[0158] Xl i ≤x i ≤Xl r , i = 1, 2, ..., 5. (13)

[0159] Among them, Xl i ,Xl r These represent the left and right intervals of the independent variable, respectively.

[0160] In summary, the basic single-objective optimization model for maximizing C4 olefin yield is obtained as follows:

[0161]

[0162] Step 502: Analyze the interaction relationships between the independent variables, and optimize the basic maximization single-objective optimization model based on the analysis results to obtain an improved maximization single-objective optimization model.

[0163] Step 5021: Optimization Mechanism Analysis

[0164] The C4 olefin yield is obtained by multiplying the ethanol conversion rate and the C4 olefin selectivity. Therefore, the C4 olefin yield is directly limited by the coefficient of the ethanol conversion rate and the C4 olefin conversion rate. Thus, analysis of variance is introduced to study whether there is an interaction relationship between these variables, so as to optimize the single-objective optimization model for C4 olefins.

[0165] Step 5022: Study the yield of C4 olefins using analysis of variance.

[0166] Based on the ANOVA results in step two, interactions exist between x2*x5, x4*x5, and x3*x5. Furthermore, the F-values ​​and root mean squared errors of these interactions indicate a significant impact on the yield of C4 olefins. Therefore, these three interactive components are considered as interaction factors, and x6 = x2x5, x7 = x3x5, x8 = x4x5. Through data merging, eight independent variables are obtained. The following multiple linear regression model is then calculated using SPSS software:

[0167]

[0168] Step 5023: Obtain the optimization model for the maximum yield of C4 olefins improved by analysis of variance.

[0169] The model was improved through analysis of variance. The single-objective optimization model for the maximum yield of C4 olefins can be obtained from formula (15):

[0170]

[0171] Step 6: Under the same experimental conditions, determine the optimal ratio of catalyst combination to temperature. That is, under the same experimental conditions, the values ​​of all independent variables are subject to the same restrictions as the data values ​​in the laboratory. This indicates that there are no special restrictions on variables that mainly affect the yield of C4 olefins, such as temperature. Therefore, it is reasonable to choose an optimization model for the yield of C4 olefins based on analysis of variance.

[0172] The optimal catalyst combination and temperature were determined using the Lingo program as follows:

[0173]

[0174] Secondly, limiting the temperature to below 350 degrees Celsius, we need to find the optimal ratio of catalyst combination to temperature.

[0175] Since the reaction temperature was limited, the optimization model for the maximum yield of C4 olefins, which was improved by variance analysis and had an enhanced effect on the main influencing factors such as temperature, was used to solve the problem. That is, the upper limit of the temperature constraint was adjusted to 350°.

[0176] The optimal catalyst combination and temperature were determined using the Lingo program as follows:

[0177]

[0178] Example 2

[0179] This embodiment discloses an optimization system for C4 olefin preparation conditions based on partial least squares analysis, including:

[0180] The experimental data acquisition module is configured to acquire experimental data on the preparation of C4 olefins by ethanol coupling under different catalyst combinations and reaction temperatures.

[0181] The correlation analysis module is configured to perform correlation analysis on the acquired experimental data and select factors that show strong correlation as independent variables.

[0182] The classification processing module is configured to classify and merge the by-reaction products of ethanol coupling to prepare C4 olefins, and select dependent variables based on the main reaction products and by-reaction products.

[0183] The partial least squares analysis module is configured to: perform regression modeling on independent and dependent variables based on the partial least squares method to obtain a multiple linear regression equation;

[0184] The optimization model building module is configured to: build a single-objective optimization model for maximizing the yield of C4 olefins based on a multiple linear regression equation;

[0185] The model solver module is configured to solve the single-objective optimization model to obtain the optimal preparation conditions that maximize the yield of C4 olefins.

[0186] Example 3

[0187] The purpose of this embodiment is to provide a computer-readable storage medium.

[0188] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in a method for optimizing C4 olefin preparation conditions based on partial least squares analysis as described in Embodiment 1 of this disclosure.

[0189] Example 4

[0190] The purpose of this embodiment is to provide an electronic device.

[0191] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in a method for optimizing C4 olefin preparation conditions based on partial least squares analysis as described in Embodiment 1 of this disclosure.

[0192] The steps and methods involved in the apparatuses of Embodiments 2, 3, and 4 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0193] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0194] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for optimizing C4 olefin preparation conditions based on partial least squares analysis, characterized in that, include: To obtain experimental data on the preparation of C4 olefins by ethanol coupling under different catalyst combinations and reaction temperatures; Correlation analysis was performed on the obtained experimental data, and response factors that showed strong correlation were selected as independent variables. By-products of the coupling of ethanol to prepare C4 olefins were classified and processed, and dependent variables were selected based on the main reaction products and by-products. The classification of by-products in the coupling of ethanol to prepare C4 olefins includes: performing Q-type clustering on all by-products to obtain a data phylogenetic diagram of by-products, and classifying the by-products into major by-products and minor by-products based on the phylogenetic diagram. Regression modeling of independent and dependent variables is performed based on partial least squares method to obtain multiple linear regression equation; A single-objective optimization model for maximizing the yield of C4 olefins was established based on a multiple linear regression equation. The single-objective optimization model for maximizing C4 olefin yield based on a multiple linear regression equation includes: Establish a basic maximization single-objective optimization model; The interaction relationships between independent variables are analyzed, and the model is improved through analysis of variance. Based on the analysis results, the basic maximization single-objective optimization model is optimized to obtain the improved maximization single-objective optimization model. Solving the single-objective optimization model to maximize the yield of C4 olefins yields the optimal preparation conditions. The improved maximization single-objective optimization model is as follows: In the formula, Indicates the yield of C4 olefins. express quality, express Load capacity express quality, Indicates ethanol concentration. Indicates the reaction temperature. Indicates the interaction factor; Represent the independent variable The range of values ​​satisfies the set value.

2. The method for optimizing C4 olefin preparation conditions based on partial least squares analysis as described in claim 1, characterized in that, The dependent variables include ethanol conversion rate, C4 olefin selectivity, major by-product selectivity, and minor by-product selectivity.

3. The method for optimizing C4 olefin preparation conditions based on partial least squares analysis as described in claim 1, characterized in that, The regression modeling of independent and dependent variables based on the partial least squares method includes: Construct the observation data matrices for the independent and dependent variables respectively; Calculate the sample mean and sample standard deviation of the variables; The independent and dependent variables are standardized based on the sample mean and sample standard deviation; Calculate and analyze the correlation coefficient matrix of standardized independent and dependent variables to obtain principal component pairs; Establish regression equations between principal component pairs and standardized independent and dependent variables; By restoring the standardized independent and dependent variables in the regression equation to their original values, we obtain the multiple linear regression equation between the independent and dependent variables.

4. The method for optimizing C4 olefin preparation conditions based on partial least squares analysis as described in claim 1, characterized in that, The optimal preparation conditions include obtaining the optimal catalyst combination and reaction temperature.

5. An optimization system for C4 olefin preparation conditions based on partial least squares analysis, characterized in that, include: The experimental data acquisition module is configured to acquire experimental data on the preparation of C4 olefins by ethanol coupling under different catalyst combinations and reaction temperatures. The correlation analysis module is configured to perform correlation analysis on the acquired experimental data and select response factors that show strong correlation as independent variables. The classification module is configured to classify the by-products of the ethanol coupling to prepare C4 olefins, and select dependent variables based on the main reaction products and by-products. The classification of by-products in the coupling of ethanol to prepare C4 olefins includes: performing Q-type clustering on all by-products to obtain a data phylogenetic diagram of by-products, and classifying the by-products into major by-products and minor by-products based on the phylogenetic diagram. The partial least squares analysis module is configured to: perform regression modeling on independent and dependent variables based on the partial least squares method to obtain a multiple linear regression equation; The optimization model building module is configured to: build a single-objective optimization model for maximizing the yield of C4 olefins based on a multiple linear regression equation; The single-objective optimization model for maximizing C4 olefin yield based on a multiple linear regression equation includes: Establish a basic maximization single-objective optimization model; The interaction relationships between independent variables are analyzed, and the model is improved through analysis of variance. Based on the analysis results, the basic maximization single-objective optimization model is optimized to obtain the improved maximization single-objective optimization model. The model solving module is configured to solve the single-objective optimization model to obtain the optimal preparation conditions that maximize the yield of C4 olefins. The improved maximization single-objective optimization model is as follows: In the formula, Indicates the yield of C4 olefins. express quality, express Load capacity express quality, Indicates ethanol concentration. Indicates the reaction temperature. Indicates the interaction factor; Represent the independent variable The range of values ​​satisfies the set value.

6. A computer-readable storage medium having a program stored thereon, characterized in that, When executed by a processor, the program implements the steps in the method for optimizing C4 olefin preparation conditions based on partial least squares analysis as described in any one of claims 1-4.

7. An electronic device, comprising a memory, a processor, and a program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the method for optimizing C4 olefin preparation conditions based on partial least squares analysis as described in any one of claims 1-4.

Images