A method for predicting the specific surface area and the amount of adsorbed aromatic compounds of a biochar

Abstract

The application discloses a method for predicting specific surface area and aromatic compound adsorption capacity of biochar. First, the experimental data set of biochar is obtained, and complete biochar adsorption data and complete pyrolysis experimental data of biochar are obtained through data processing; the complete data is substituted into a machine learning algorithm to model, a prediction model of different research targets is obtained, and performance evaluation and model explainability analysis are performed on the model; a to-be-predicted database is established according to the analysis purpose, the to-be-predicted database is substituted into the prediction model of different research targets, and prediction data is obtained. The application proposes a data model of the specific surface area of biochar material and the maximum adsorption capacity of biochar to aromatic compounds, which greatly saves the trial and error cost of experiments.

Description

A method for predicting the specific surface area and adsorption capacity of aromatic compounds in biochar Technical Field

[0001] This invention belongs to the fields of machine learning and biochar adsorption technology, specifically relating to a method for predicting the specific surface area of ​​biochar and the adsorption capacity of aromatic compounds. Background Technology

[0002] Aromatic compounds are a very important class of pollutants in the environment. They are widely present in the atmosphere, water bodies and soil, and can have very adverse effects on human health, atmospheric optical properties and environmental climate.

[0003] Biochar, as an effective biomass material, has always had unique advantages in various fields. The maximum adsorption capacity of biochar materials for pollutants and the specific surface area of ​​biochar materials are important indicators for evaluating the pollutant removal performance and overall performance of biochar materials, and are key research topics for many researchers.

[0004] Compared with traditional methods, machine learning offers significant advantages in biochar preparation and adsorption due to its high efficiency, complex modeling capabilities, and cost savings. Therefore, establishing models to predict the performance and adsorption capacity of biomass materials is a pressing issue that needs to be addressed.

[0005] Existing technologies suffer from the following problems: Typically, researchers first select various biomass materials and prepare different types of biochar under varying conditions. Next, the biochar is characterized to obtain relevant parameters such as its specific surface area. Based on experience, suitable biochar is selected for batch adsorption experiments. Finally, adsorption isotherms are plotted using experimental data, and the maximum adsorption capacity is obtained by fitting the data using equations such as Langmuir. The key parameter of biochar's specific surface area requires expensive experimental equipment for measurement, and numerous factors influence this result. Furthermore, selecting suitable biochar for adsorption experiments and determining the maximum adsorption capacity requires a series of trial-and-error experiments, resulting in extremely high trial-and-error costs, numerous uncertainties, and the possibility of rework. Summary of the Invention

[0006] To address the problems mentioned in the background section, this invention provides a method for predicting the specific surface area of ​​biochar and the adsorption capacity of aromatic compounds, which greatly reduces experimental costs, provides accurate predictions, and offers a basis for finding the optimal preparation conditions for products with the best performance.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] In a first aspect, the present invention provides a method for predicting the specific surface area and adsorption capacity of aromatic compounds in biochar, comprising,

[0009] Step (1) Obtain the biochar experimental dataset, which includes a biochar adsorption of aromatic compounds dataset and a biochar pyrolysis experimental dataset;

[0010] Step (2) The biochar adsorption aromatic compound dataset and the biochar pyrolysis experimental dataset are processed to obtain complete biochar adsorption data and complete biochar pyrolysis experimental data.

[0011] Step (3) Substitute the complete biochar adsorption data and complete biochar pyrolysis experimental data into the machine learning algorithm to model, obtain the prediction model, and perform performance evaluation and model interpretability analysis on the model.

[0012] Step (4) Establish a database to be predicted according to the purpose of analysis, substitute the database to be predicted into the prediction model, and obtain the predicted data of biochar specific surface area and aromatic compound adsorption amount.

[0013] In some embodiments, in step (1), the biochar experimental dataset is obtained by downloading corresponding Chinese and English literature and master's and doctoral dissertations from publicly available databases by searching for keywords; the biochar adsorption of aromatic compounds dataset must include data on the maximum adsorption capacity of biochar for aromatic compounds fitted by the Langmuir equation, and the Langmuir equation fitting coefficient R. 2 It needs to be greater than 0.95.

[0014] In some embodiments, in step (1), the biochar adsorbed aromatic compound dataset includes a biochar material characterization descriptor, an aromatic compound solvation parameter descriptor, and a maximum adsorption capacity descriptor.

[0015] In some embodiments, in step (1), the biochar pyrolysis experimental dataset includes a biomass material characterization descriptor, a biochar preparation condition descriptor, and a biochar material specific surface area descriptor.

[0016] In some embodiments, step (2) involves data processing, which includes supplementing missing data using the KNN algorithm.

[0017] In some embodiments, in step (3), the multiple predictive performance models include: a biochar specific surface area prediction model and a biochar maximum adsorption capacity prediction model for aromatic compounds; the biochar specific surface area prediction model and the biochar maximum adsorption capacity prediction model for aromatic compounds are established using complete biochar pyrolysis experimental data and complete biochar adsorption data; the performance evaluation uses RMSE, R 2 The calculation formula is as follows:

[0018]

[0019] y、 and are the predicted value, actual value, and average value of the target descriptor, respectively; n is the number of data points for any given instance; and N is the total number of data points.

[0020] In some embodiments, in step (3), the model interpretability analysis includes analyzing the contribution of different features in the model and the impact of the Shapley values ​​of different features on the model.

[0021] In some embodiments, in step (4), the database to be predicted established according to the analytical purpose includes two analytical purposes: S1 to obtain the specific surface area prediction of biochar and the optimal preparation conditions; S2 to obtain the maximum adsorption capacity of biochar materials for aromatic compounds.

[0022] In some embodiments, the specific surface area prediction and optimal preparation conditions obtained by the biochar are to be included in the specific surface area prediction database, which includes: biomass material characterization descriptors and biochar preparation condition descriptors; the maximum adsorption capacity of the biochar material for aromatic compounds obtained by the biochar material is to be included in the adsorption capacity database, which includes: biochar characterization descriptors and pollution physicochemical parameter descriptors.

[0023] In a second aspect, the present invention provides a device for predicting the specific surface area of ​​biochar and the adsorption capacity of aromatic compounds, including a processor and a storage medium.

[0024] The storage medium is used to store instructions;

[0025] The processor is configured to operate according to the instructions to execute the method according to the first aspect.

[0026] Thirdly, the present invention provides an apparatus comprising,

[0027] Memory;

[0028] processor;

[0029] as well as

[0030] Computer programs;

[0031] The computer program is stored in the memory and configured to be executed by the processor to implement the method described in the first aspect above.

[0032] Fourthly, the present invention provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described in the first aspect.

[0033] Compared with the prior art, the beneficial effects of the present invention are:

[0034] (1) This invention establishes a predictive model for the maximum adsorption capacity of biochar for aromatic compounds. By inputting the prediction database into the established model, the maximum adsorption capacity of the biochar for this pollutant can be obtained, which greatly saves experimental costs.

[0035] (2) The present invention establishes a specific surface area prediction model for biochar materials, which can effectively predict the specific surface area of ​​biochar. At the same time, the data in the biochar preparation condition descriptor corresponding to the maximum specific surface area output by the model can be regarded as the optimal preparation conditions for this type of biochar. This method greatly reduces the experimental trial and error cost and provides a certain basis for the subsequent modification and activation of biochar.

[0036] (3) This invention uses model interpretability analysis to analyze the main factors affecting the prediction model of biochar specific surface area and the prediction model of maximum adsorption capacity of aromatic compounds, providing certain theoretical guidance for the preparation of biochar. Attached Figure Description

[0037] Figure 1 is a schematic diagram of the method flow according to an embodiment of the present invention;

[0038] Figure 2 shows the predictive performance of the model in this embodiment of the invention, represented by the goodness of fit between the actual and predicted values ​​of the model.

[0039] Figure 3 is a schematic diagram of the feature contribution of each feature of the model in the embodiment of the present invention;

[0040] Figure 4 is a schematic diagram of the Shapley values ​​of each feature in the model of the embodiment of the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Example 1

[0043] This embodiment uses the prediction of specific surface area of ​​rice husk biochar and phenol adsorption by bamboo biochar as examples.

[0044] Please refer to Figure 1. This invention provides the following technical solution: a method for predicting the specific surface area and adsorption capacity of aromatic compounds in biochar, comprising:

[0045] Step (1) Obtain the biochar experimental dataset, which includes a biochar adsorption of aromatic compounds dataset and a biochar pyrolysis experimental dataset.

[0046] Step (2) involves processing the biochar adsorption aromatic compound dataset and the biochar pyrolysis experimental dataset to obtain complete biochar adsorption data and complete biochar pyrolysis experimental data.

[0047] Step (3) Substitute the complete biochar adsorption data and complete biochar pyrolysis experimental data into the machine learning algorithm to model, obtain prediction models for different research objectives, and conduct performance evaluation and model interpretability analysis on the models.

[0048] Step (4) Establish a database to be predicted according to the purpose of analysis, and substitute the database to be predicted into the prediction model of different research objectives to obtain the prediction data.

[0049] Specifically, the biochar experimental dataset is obtained by searching for relevant Chinese and English literature and master's and doctoral dissertations from publicly available databases using keywords; the biochar adsorption of aromatic compounds dataset must include data on the maximum adsorption capacity of biochar for aromatic compounds fitted by the Langmuir equation, with the Langmuir equation fitting coefficient R0. 2 It needs to be greater than 0.95.

[0050] By adopting the above technical solution, the obtained dataset comes from a publicly available database, which ensures the authenticity and reliability of the data. The Langmuir equation fitting coefficient R in the collected dataset is [data missing]. 2 A value greater than 0.95 can effectively improve the model's predictive performance.

[0051] Specifically, the dataset for biochar adsorption of aromatic compounds includes biochar material characterization descriptors, aromatic compound solvation parameter descriptors, and maximum adsorption capacity descriptors;

[0052] In this implementation case, the biochar material characterization descriptors include: ash content (Ash), carbon content (C), hydrogen-to-carbon ratio (H / C), oxygen-to-carbon ratio (O / C), oxygen-to-nitrogen ratio (O+N) / C, specific surface area (SA), and total pore volume (V). total ), micropore volume (V) micro The solvation parameters for aromatic compounds include: molar refractivity (E), polarity / polarizability (S), total solute H-bond acidity (A), total solute H-bond basicity (B), and McGowan characteristic volume (V); the maximum adsorption capacity descriptor is Q. max .

[0053] Specifically, the biochar pyrolysis experimental dataset includes biomass material characterization descriptors, biochar preparation condition descriptors, and biochar material specific surface area descriptors.

[0054] In this implementation case, the biomass material characterization descriptors include: carbon content (C), hydrogen content (H), oxygen content (O), nitrogen content (N), volatile matter (VM), ash content (Ash), and fixed carbon content (FC); the biochar preparation condition descriptors include: heating rate (HR), pyrolysis temperature (T), and holding time (RT); and the specific surface area descriptor is SA.

[0055] Specifically, the missing data is supplemented using the KNN algorithm;

[0056] By adopting the above technical solution, using the KNN algorithm to supplement missing values ​​helps to improve the robustness of the model.

[0057] Specifically, the various predictive performance models include: a biochar specific surface area prediction model and a biochar maximum adsorption capacity prediction model for aromatic compounds; the biochar specific surface area prediction model and the biochar maximum adsorption capacity prediction model for aromatic compounds are established using complete biochar pyrolysis experimental data and complete biochar adsorption data; the performance evaluation uses RMSE, R... 2 The calculation formula is as follows:

[0058]

[0059] y、 and are the predicted value, actual value, and average value of the target descriptor, respectively; n is the number of data points for any given instance; and N is the total number of data points.

[0060] By adopting the above technical solution, the complete biochar pyrolysis experimental data and complete biochar adsorption data were modeled using the random forest algorithm. The model was built using Python and the PyCharm software platform, resulting in two prediction models: a biochar specific surface area prediction model and a biochar maximum adsorption capacity prediction model for aromatic compounds. The performance evaluation of the two models is shown in Table 1 below.

[0061]

[0062] Specifically, interpretability analysis was conducted on the two models, including analyzing the contribution of different features and the impact of Shapley values ​​of different features on the models.

[0063] By employing the above technical solutions, the key factors influencing the maximum adsorption capacity model and the biochar specific surface area model are explained. Feature contribution analysis and eigenvalue contribution analysis are performed on the two prediction models by calling functions from the Sklearn library. Figure 3 shows the feature contribution analysis; the left figure is the adsorption capacity prediction model, and the right figure is the biochar specific surface area prediction model. Figure 4 shows the Shapley value analysis of features in different models; the left figure is the adsorption capacity prediction model, and the right figure is the biochar specific surface area prediction model. Visualizing the established models helps to explain their prediction mechanisms.

[0064] Specifically, the database to be predicted, established according to the analytical purpose, includes two analytical purposes:

[0065] S1 yielded the predicted specific surface area of ​​biochar and the optimal preparation conditions;

[0066] S2 yielded the maximum adsorption capacity of the biochar material for aromatic compounds.

[0067] Specifically, in step S1, the predicted specific surface area and optimal preparation conditions of biochar are obtained, and the corresponding specific surface area prediction database should include: biomass material characterization descriptor and biochar preparation condition descriptor; the maximum adsorption capacity of biochar material for aromatic compounds is obtained, and the corresponding adsorption capacity database should include: biochar characterization descriptor and pollution physicochemical parameter descriptor.

[0068] By adopting the above technical solution and substituting the specific surface area database into the biochar specific surface area prediction model, the specific surface area prediction data of biochar can be obtained. Furthermore, the highest specific surface area prediction result corresponds to the biochar preparation conditions, which can provide a preparation method for this type of biochar. The specific output style of the biochar specific surface area prediction model is shown in Table 2 below:

[0069]

[0070] Based on the surface area prediction values ​​output by the biochar specific surface area prediction model in Table 2, the data in the biochar preparation condition descriptor corresponding to the maximum specific surface area in the output results is selected as the optimal preparation conditions for this type of biochar. This method greatly reduces the experimental trial and error costs and provides a certain basis for the subsequent modification and activation of biochar.

[0071] By substituting the database of adsorption capacities to be predicted into the maximum adsorption capacity prediction model, the specific surface area prediction data of biochar can be obtained. The output format of the maximum adsorption capacity prediction data of biochar for aromatic compounds is as follows.

[0072] Table 3:

[0073]

[0074] It can be seen that No. 2 bamboo biochar has the best adsorption effect on phenol. In practical applications, No. 2 bamboo biochar can be used to remove phenol from water. At the same time, this method also provides some reference value for further adsorption experiments.

[0075] Example 2

[0076] Secondly, based on Example 1, this embodiment provides a device for predicting the specific surface area of ​​biochar and the adsorption capacity of aromatic compounds, including a processor and a storage medium.

[0077] The storage medium is used to store instructions;

[0078] The processor is configured to operate according to the instructions to execute the method according to Embodiment 1.

[0079] Example 3

[0080] Thirdly, based on Embodiment 1, this embodiment provides a device, including,

[0081] Memory;

[0082] processor;

[0083] as well as

[0084] Computer programs;

[0085] The computer program is stored in the memory and configured to be executed by the processor to implement the method described in Embodiment 1.

[0086] Example 4

[0087] Fourthly, based on Embodiment 1, this embodiment provides a storage medium on which a computer program is stored, and when the computer program is executed by a processor, it implements the method described in Embodiment 1.

[0088] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0089] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in one or more flowchart illustrations and / or one or more block diagrams.

[0090] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0091] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0092] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for predicting the specific surface area and adsorption capacity of aromatic compounds in biochar, characterized in that, include: Step (1) Obtain the biochar experimental dataset, which includes a biochar adsorption of aromatic compounds dataset and a biochar pyrolysis experimental dataset; wherein, the biochar adsorption of aromatic compounds dataset includes biochar material characterization descriptors, aromatic compound solvation parameter descriptors, and maximum adsorption capacity descriptors; the biochar material characterization descriptors include ash content (Ash), carbon content (C), hydrogen-to-carbon ratio (H / C), oxygen-to-carbon ratio (O / C), oxygen-to-nitrogen ratio (O+N) / C, specific surface area (SA), and total pore volume (V). total Micropore volume V micro The solvation parameter descriptors for aromatic compounds include molar refractivity E, polarity / polarizability S, total solute H-bond acidity A, total solute H-bond basicity B, and McGowan characteristic volume V. The maximum adsorption capacity descriptor is Q. max The biochar pyrolysis experimental dataset includes biomass material characterization descriptors, biochar preparation condition descriptors, and biochar material specific surface area descriptors. The biomass material characterization descriptors include carbon content (C), hydrogen content (H), oxygen content (O), nitrogen content (N), volatile matter (VM), ash content (Ash), and fixed carbon content (FC). The biochar preparation condition descriptors include heating rate (HR), pyrolysis temperature (T), and holding time (RT). The biochar material specific surface area descriptor is SSA. Step (2) processes the biochar adsorption aromatic compound dataset and the biochar pyrolysis experimental dataset to obtain complete biochar adsorption data and complete... The complete biochar pyrolysis experimental data; Step (3) Substitute the complete biochar adsorption data and complete biochar pyrolysis experimental data into the machine learning algorithm to model, obtain the prediction model, and perform performance evaluation and model interpretability analysis on the model; The prediction model includes a biochar specific surface area prediction model and a biochar maximum adsorption amount prediction model for aromatic compounds, which are established using complete biochar pyrolysis experimental data and complete biochar adsorption data; Step (4) Establish a database to be predicted according to the analysis purpose, substitute the database to be predicted into the prediction model, and obtain the predicted data of biochar specific surface area and aromatic compound adsorption amount.

2. The method for predicting the specific surface area and aromatic compound adsorption capacity of biochar according to claim 1, characterized in that, In step (1), the biochar experimental dataset is obtained by downloading corresponding Chinese and English literature and master's and doctoral dissertations from publicly available databases by searching for keywords; the biochar adsorption of aromatic compounds dataset must include data on the maximum adsorption capacity of biochar for aromatic compounds fitted by the Langmuir equation, and the Langmuir equation fitting coefficient R. 2 It needs to be greater than 0.

95.

3. The method for predicting the specific surface area and aromatic compound adsorption capacity of biochar according to claim 1, characterized in that: In step (2), the data processing includes supplementing the missing data parts using the KNN algorithm.

4. The method for predicting the specific surface area and aromatic compound adsorption capacity of biochar according to claim 1, characterized in that: In step (3), the performance evaluation uses the root mean square error (RMSE) and the linear correlation coefficient (R²). 2 .

5. The method for predicting the specific surface area and aromatic compound adsorption capacity of biochar according to claim 1, characterized in that: In step (3), the model interpretability analysis includes: analyzing the contribution of different features in the model and the impact of the Shapley values ​​of different features on the model.

6. The method for predicting the specific surface area and aromatic compound adsorption capacity of biochar according to claim 1, characterized in that: In step (4), the database to be predicted established according to the analysis purpose includes two analysis purposes: S1 to obtain the prediction of biochar specific surface area and the optimal preparation conditions; S2 to obtain the maximum adsorption capacity of biochar material for aromatic compounds.

7. The method for predicting the specific surface area and aromatic compound adsorption capacity of biochar according to claim 6, characterized in that: The specific surface area prediction and optimal preparation conditions for obtaining biochar should be included in the specific surface area prediction database, which should include: biomass material characterization descriptors and biochar preparation condition descriptors; the maximum adsorption capacity of biochar material for aromatic compounds should be included in the maximum adsorption capacity database, which should include: biochar characterization descriptors and pollution physicochemical parameter descriptors.

8. A storage medium, characterized in that, It stores a computer program thereon, which, when executed by a processor, implements the method described in any one of claims 1 to 7.

Images