Method and system for enhancing prediction accuracy of degradation performance of a fenton-like system
By using a machine learning model with active sites as input features, the error problem in the prediction of the degradation performance of the biochar/PMS system was solved, the model accuracy was improved and the experimental cost was reduced.
Patent Information
- Application Number
- CN202411481563.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-23
- Publication Date
- 2025-10-10
- Estimated Expiration
- 2044-10-23
AI Technical Summary
The existing technology for predicting the degradation performance of organic matter in the biochar/PMS system still has an error of about 20% in model accuracy, and building a new model requires a deep theoretical foundation and complex algorithm development, which is quite difficult.
Active sites were used as input features instead of element ratios, and a machine learning model was used to construct a biochar/PMS system degradation performance prediction model. By obtaining data on biochar characteristics and physical descriptors of organic matter, dividing the training set and test set, determining the model parameters, and performing performance evaluation to obtain the optimal model.
The accuracy of the model in predicting the performance of the biochar/PMS system is improved, the trial-and-error probability of traditional experiments is reduced, and manpower, material resources and time are saved.
Smart Images
Figure CN119361010B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of environmental protection technology, and more particularly to a method and system for enhancing the prediction accuracy of degradation performance of a Fenton-like system. Background Art
[0002] Machine learning, as an artificial intelligence technology, has attracted widespread attention in the environmental field because of its ability to reveal complex and hidden relationships that are often overlooked by traditional analytical methods. Recently, researchers have applied machine learning technology to biochar / PMS systems to enhance the degradation of organic matter. However, after a series of model optimization (data optimization, model screening, feature engineering, cross-validation, and hyperparameter adjustment), the R of the model test set was 2 When the value exceeds 0.8, it is considered as a better prediction model, which means that there is still a prediction error of about 20%. Therefore, it is crucial to improve the accuracy of the prediction model for the degradation performance of the biochar / PMS system.
[0003] In addition to the aforementioned methods for improving accuracy, which have already been used in building biochar / PMS system performance prediction models, accuracy can be further improved by designing new models. However, building new models may require algorithm development, coding, debugging, and optimization from scratch, requiring developers to have a solid theoretical foundation and the ability to handle complex mathematical problems and algorithms in model design. This is a highly challenging task.
[0004] Research has shown that introducing unexplored features that may be relevant to outcome predictions can improve model performance. The active sites on the biochar surface directly influence its catalytic activity. However, in machine learning studies of biochar / PMS system performance enhancement, researchers have used C, O, and N elemental content or the O / C ratio to represent active sites on the biochar surface, which may be a cause of model errors.
[0005] Therefore, it is necessary to re-examine the model for predicting the performance of the biochar / PMS system constructed with elements as input features, and it is urgent to further explore the accuracy of the model in predicting the degradation performance of the biochar / PMS system when active sites are used as input features. Summary of the Invention
[0006] In view of this, the present invention provides a method and system for enhancing the prediction accuracy of degradation performance of a Fenton-like system, which solves the problems existing in the background technology.
[0007] In order to achieve the above object, the present invention provides the following technical solutions:
[0008] In one aspect, the present invention provides a method for enhancing the prediction accuracy of degradation performance of a Fenton-like system, comprising the following steps:
[0009] Obtain biochar characteristics, physical descriptors of organic matter and degradation performance of Fenton-like system lnk SA , construct parameter data set;
[0010] Divide the parameter data set into training set and test set according to the preset ratio;
[0011] Use the training set to train different machine learning models and determine the model parameters;
[0012] The prediction performance of the two machine learning models was evaluated using the test set, and finally the best model for predicting the degradation performance of the Fenton-like system was obtained.
[0013] Preferably, biochar characteristics, organic matter physical descriptors and Fenton-like degradation performance lnk are obtained SA The data is as follows:
[0014] Prepare various biochars by pyrolysis modification and obtain the properties of biochars through characterization methods;
[0015] Biochar was used to activate peroxymonosulfate to degrade different organic matter at different catalyst concentrations to obtain the reaction rate constant k. K was normalized using the specific surface area to obtain k SA , evaluate the degradation performance of the Fenton-like reaction system.
[0016] Preferably, the parameter data set includes a first data set and a second data set; wherein the first data set is C=C content, CN content, OC=O content, C=O content, CO content, pyridine nitrogen content, pyrrole nitrogen content, graphite nitrogen content, defect value I D / I G , the energy of the highest occupied molecular orbital E HOMO is the input feature, lnk SA As the target variable; the second data set is C content, O content, N content, C / N, C / O, defect value I D / I G , the energy of the highest occupied molecular orbital E HOMO is the input feature, lnk SA is the target variable.
[0017] Preferably, the ratio of the training set to the test set is 8:2.
[0018] Preferably, different machine learning models are trained using the training set, specifically:
[0019] The SVR algorithm was used to construct a model for predicting the performance of the biochar / PMS system using the first and second datasets, and the input variables were determined based on the evaluation indicators;
[0020] The hyperparameters are determined by using the training set, and the relationship between the characteristics of the biochar, the structure of the organic matter and the degradation performance of the biochar / PMS system is established.
[0021] Preferably, the hyperparameters include random_state, kernel, C, epsilon and tol.
[0022] Preferably, when the prediction performance of the two machine learning models is evaluated by using the test set, the performance of the model is tested by using the test set, and the accuracy of the model is evaluated by using the evaluation index.
[0023] Preferably, the evaluation index includes mean absolute error, root mean square error and determination coefficient.
[0024] In another aspect, the application also provides a system for enhancing the prediction accuracy of the degradation performance of the Fenton-like system, which is used to implement the method for enhancing the prediction accuracy of the degradation performance of the Fenton-like system according to any one of the above aspects, and includes a data set construction module, a division module, a training module and a test module connected in sequence.
[0025] The data set construction module is used to obtain the data of the characteristics of the biochar, the physical descriptors of the organic matter and the degradation performance lnk of the Fenton-like system, and construct a parameter data set. SA
[0026] The division module is used to divide the parameter data set into a training set and a test set according to a preset proportion.
[0027] The training module is used to train different machine learning models by using the training set, and determine the model parameters.
[0028] The test module is used to evaluate the prediction performance of the two machine learning models by using the test set, and finally obtain the best model for predicting the degradation performance of the Fenton-like system.
[0029] According to the above technical solution, compared with the prior art, the application provides a method and system for enhancing the prediction accuracy of the degradation performance of the Fenton-like system, which has the following beneficial effects:
[0030] By using the active site replacement element ratio as the input feature, the accuracy of the model for predicting the performance of the biochar / PMS system is improved, which helps to further clarify the structure-activity relationship of the biochar, greatly reduces the probability of traditional experimental trial and error, and saves the consumption of manpower, material resources and time. BRIEF DESCRIPTION OF DRAWINGS
[0031] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings described below are only a part of the embodiments of the present application, and all other drawings obtained by those of ordinary skill in the art without creative effort based on the provided drawings also belong to the protection scope of the present application.
[0032] Figure 1 A flowchart of a method for enhancing the prediction accuracy of the degradation performance of a Fenton-like system provided by the present application;
[0033] Figure 2 A prediction diagram of the performance of a biochar / PMS system by a model constructed with C content, O content, N content, C / N, C / O, I D / I G , E HOMO content as input parameters;
[0034] Figure 3 A prediction diagram of the performance of a biochar / PMS system by a model constructed with C=C content, C-N content, O-C=O content, C=O content, C-O content, pyridine nitrogen content, pyrrole nitrogen content, graphite nitrogen content, I D / I G , E HOMO content as input parameters;
[0035] Figure 4 An importance diagram of input parameters with a data set containing active sites as input provided by the present application;
[0036] Figure 5 A structural block diagram of a system for enhancing the prediction accuracy of the degradation performance of a Fenton-like system provided by the present application. DETAILED DESCRIPTION
[0037] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort belong to the protection scope of the present application.
[0038] The embodiments of the present application disclose a method for enhancing the prediction accuracy of the degradation performance of a Fenton-like system, as shown in the following steps: Figure 1
[0039] Obtain data of biochar characteristics, organic matter physical descriptors and Fenton-like system degradation performance lnk SA , and construct a parameter data set;
[0040] Divide the parameter data set into training set and test set according to the preset ratio;
[0041] Use the training set to train different machine learning models and determine the model parameters;
[0042] The prediction performance of the two machine learning models was evaluated using the test set, and finally the best model for predicting the degradation performance of the Fenton-like system was obtained.
[0043] Furthermore, the biochar characteristics, physical descriptors of organic matter and degradation performance of Fenton-like system lnk were obtained. SA The data, specifically:
[0044] Prepare various biochars by pyrolysis modification and obtain the properties of biochars through characterization methods;
[0045] Biochar was used to activate peroxymonosulfate to degrade different organic matter at different catalyst concentrations to obtain the reaction rate constant k. K was normalized using the specific surface area to obtain k SA , evaluate the degradation performance of the Fenton-like reaction system.
[0046] Furthermore, the parameter data set includes a first data set and a second data set; wherein the first data set is C=C content, CN content, OC=O content, C=O content, CO content, pyridine nitrogen content, pyrrole nitrogen content, graphite nitrogen content, defect value I D / I G , the energy of the highest occupied molecular orbital E HOMO is the input feature, lnk SA As the target variable; the second data set is C content, O content, N content, C / N, C / O, defect value I D / I G , the energy of the highest occupied molecular orbital E HOMO is the input feature, lnk SA is the target variable.
[0047] Furthermore, the division ratio of the training set and the test set is 8:2.
[0048] Furthermore, the training set is used to train different machine learning models, specifically:
[0049] The SVR algorithm was used to construct a model for predicting the performance of the biochar / PMS system using the first and second datasets, and the input variables were determined based on the evaluation indicators;
[0050] The training set was used to determine hyperparameters and establish the relationship between biochar characteristics, organic matter structure and the degradation performance of the biochar / PMS system.
[0051] Furthermore, the hyperparameters include: random_state, kernel, C, epsilon, and tol.
[0052] Furthermore, when using the test set to evaluate the prediction performance of the two machine learning models respectively, the performance of the model is tested using the test set, and the accuracy of the model is evaluated using the evaluation index.
[0053] Furthermore, the evaluation metrics include mean absolute error, root mean square error, and coefficient of determination.
[0054] Next, the technical solution of the present invention will be further described by referring to specific embodiments in conjunction with the accompanying drawings. It should be noted that the following specific embodiments are only for illustration and the protection scope of the present invention is not limited thereto.
[0055] Specific embodiment 1: A method for enhancing the prediction accuracy of degradation performance of a Fenton-like system in this embodiment is carried out by the following steps:
[0056] (1) Eight different biochars were prepared by pyrolysis modification, and the biochar properties were obtained by characterization. The prepared biochar catalysts were used to activate PMS to degrade pollutants with different structures, and the first-order rate constant k was obtained. K was normalized by specific surface area to obtain k. SA , set it as the target variable;
[0057] (2) According to the characteristics of biochar, it can be divided into element-related, active site-related and I D / I G . C content, O content, N content, C / N, C / O, I D / I G 、E HOMO As an input parameter, use lnk SA Construct a data set (data set A) with input variables;
[0058] (3) Using the SVR algorithm, the prediction model is constructed using the A dataset and the input variables are determined based on the evaluation indicators;
[0059] (4) Dataset A is randomly divided into training set and test set in a ratio of 8:2, with the training set accounting for 80% and the test set accounting for 20%;
[0060] (5) Using the SVR model, the hyperparameters were determined using the training set, and the biochar / PMS system lnk SA For output, the relationship between biochar characteristics, organic matter structure and degradation performance of the biochar / PMS system was established;
[0061] (6) Use the test set to test the performance of the model and use evaluation indicators to evaluate the accuracy of the model.
[0062] Figure 2 In this embodiment, the C content, O content, N content, C / N, C / O, I D / I G 、E HOMO As input parameters, the predicted biochar / PMS system lnk was constructed SA Model, RMSE, MAE and R of the training set 2 The RMSE, MAE and R of the test set are 0.64, 0.30 and 0.74 respectively. 2 They are 0.79, 0.44, and 0.66 respectively, so the model has low accuracy.
[0063] Specific embodiment 2: A method for enhancing the prediction accuracy of degradation performance of a Fenton-like system in this embodiment is carried out by the following steps:
[0064] (1) Eight different biochars were prepared by pyrolysis modification, and the biochar properties were obtained by characterization. The prepared biochar catalysts were used to activate PMS to degrade pollutants with different structures, and the first-order rate constant k was obtained. K was normalized by specific surface area to obtain k. SA , set it as the target variable;
[0065] (2) According to the characteristics of biochar, it can be divided into element-related, active site-related and I D / I G . C=C content, CN content, OC=O content, C=O content, CO content, pyridinic nitrogen content, pyrrolic nitrogen content, graphite nitrogen content, I D / I G 、E HOMO As an input parameter, use lnk SA Construct a data set (E data set) with input variables;
[0066] (3) Using the SVR algorithm, the prediction model is constructed using the E dataset and the input variables are determined based on the evaluation indicators;
[0067] (4) The E dataset is randomly divided into a training set and a test set in a ratio of 8:2, with the training set accounting for 80% and the test set accounting for 20%;
[0068] (5) Using the SVR model, the hyperparameters were determined using the training set, and the biochar / PMS system lnk SA For output, the relationship between biochar characteristics, organic matter structure and degradation performance of the biochar / PMS system was established;
[0069] (6) Use the test set to test the performance of the model and use evaluation indicators to evaluate the accuracy of the model.
[0070] Figure 3In this embodiment, the C=C content, CN content, OC=O content, C=O content, CO content, pyridinic nitrogen content, pyrrolic nitrogen content, graphite nitrogen content, I D / I G 、E HOMO As input parameters, the predicted biochar / PMS system lnk was constructed SA Model, RMSE, MAE and R of the training set 2 The RMSE, MAE and R of the test set are 0.19, 0.13 and 0.98 respectively. 2 They are 0.35, 0.23 and 0.94 respectively, so the model has high accuracy.
[0071] Figure 4 The effects of different input parameters on the biochar / PMS system lnk in the E dataset SA The feature importance diagram shows that C=C has a significant effect on lnk SA The most important, followed by CO and CN.
[0072] Reference Figure 5 As shown, the present invention also provides a system for enhancing the prediction accuracy of degradation performance of a Fenton-like system, the system is used to implement any of the above-mentioned methods for enhancing the prediction accuracy of degradation performance of a Fenton-like system, comprising a data set construction module, a partitioning module, a training module and a testing module connected in sequence;
[0073] Dataset building blocks for obtaining biochar characteristics, physical descriptors of organic matter, and Fenton-like degradation performance lnk SA , construct parameter data set;
[0074] A partitioning module is used to divide the parameter data set into a training set and a test set according to a preset ratio;
[0075] The training module is used to train different machine learning models using the training set and determine the model parameters;
[0076] The testing module is used to evaluate the prediction performance of the two machine learning models using the test set, and finally obtain the best model for predicting the degradation performance of the Fenton-like system.
[0077] Currently, machine learning methods for predicting the pollutant degradation performance of biochar / PMS systems use element content and element ratios as input features, which reduces the model's accuracy. In contrast, this example uses active sites instead of elements as input features, improving the model's accuracy in predicting biochar / PMS system performance. This helps further elucidate the structure-activity relationship of biochar, significantly reduces the trial-and-error nature of traditional experiments, and saves manpower, material resources, and time.
[0078] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.
[0079] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for enhancing the prediction accuracy of degradation performance of a Fenton-like system, characterized in that: The following steps are involved: Obtain biochar characteristics and Fenton-like degradation performance lnk through experiments SA By consulting the literature to obtain the physical descriptors of organic matter, a parameter data set is constructed; the parameter data set includes a first data set and a second data set; wherein, the first data set is C=C content, CN content, OC=O content, C=O content, CO content, pyridinic nitrogen content, pyrrolic nitrogen content, graphitic nitrogen content, defect value I D / I G , the energy of the highest occupied molecular orbital E HOMO is the input feature, lnk SA As the target variable; the second data set is C content, O content, N content, C / N, C / O, defect value I D / I G , the energy of the highest occupied molecular orbital E HOMO is the input feature, lnk SA is the target variable; Divide the parameter data set into training set and test set according to the preset ratio; Use the training set to train different machine learning models and determine the model parameters; The prediction performance of the two machine learning models was evaluated using the test set, and the optimal model for predicting the degradation performance of the Fenton-like system was finally obtained; Among them, obtaining biochar characteristics, physical descriptors of organic matter and degradation performance of Fenton-like system lnk SA The data is as follows: Prepare various biochars by pyrolysis modification and obtain the properties of biochars through characterization methods; Biochar was used to activate peroxymonosulfate to degrade different organic matter at different catalyst concentrations to obtain the reaction rate constant k. K was normalized using the specific surface area to obtain k SA , evaluate the degradation performance of the Fenton-like reaction system.
2. The method for enhancing the prediction accuracy of degradation performance of a Fenton-like system according to claim 1, characterized in that: The ratio of the training set to the test set is 8:
2.
3. The method for enhancing the prediction accuracy of degradation performance of a Fenton-like system according to claim 1, characterized in that: Use the training set to train different machine learning models, specifically: The SVR algorithm was used to construct a model for predicting the performance of the biochar / PMS system using the first and second datasets, and the input variables were determined based on the evaluation indicators; The training set was used to determine hyperparameters and establish the relationship between biochar characteristics, organic matter structure and the degradation performance of the biochar / PMS system.
4. The method for enhancing the prediction accuracy of degradation performance of a Fenton-like system according to claim 3, characterized in that: Hyperparameters include: random_state, kernel, C, epsilon, and tol.
5. The method for enhancing the prediction accuracy of degradation performance of a Fenton-like system according to claim 1, characterized in that: When using the test set to evaluate the prediction performance of two machine learning models respectively, the test set is used to test the performance of the model, and the evaluation index is used to evaluate the accuracy of the model.
6. The method for enhancing the prediction accuracy of degradation performance of a Fenton-like system according to claim 5, characterized in that: The evaluation indicators include mean absolute error, root mean square error and coefficient of determination.
7. A system for enhancing the prediction accuracy of degradation performance of Fenton-like systems, characterized in that: A method for enhancing the prediction accuracy of degradation performance of a Fenton-like system according to any one of claims 1 to 6, comprising a data set construction module, a partitioning module, a training module, and a testing module connected in sequence; Dataset building blocks for obtaining biochar characteristics, physical descriptors of organic matter, and Fenton-like degradation performance lnk SA , construct parameter data set; A partitioning module is used to divide the parameter data set into a training set and a test set according to a preset ratio; The training module is used to train different machine learning models using the training set and determine the model parameters; The testing module is used to evaluate the prediction performance of the two machine learning models using the test set, and finally obtain the best model for predicting the degradation performance of the Fenton-like system.
Citation Information
Patent Citations
Charcoal specific surface area prediction method, electronic equipment and storage medium
CN118016188A