Photosensitizer and pollutant quenching rate constant prediction method, system, equipment and medium

Abstract

The invention discloses a photosensitizer and pollutant quenching rate constant prediction method, system, equipment and medium, and the method comprises the steps: obtaining a data set of a photosensitizer and a pollutant, the data set comprising a characteristic variable and a target variable; simplifying the feature variables to obtain simplified feature variables, and constructing a simplified data set by the simplified feature variables and the target variables; constructing a quenching rate constant prediction model of the photosensitizer and the pollutants by utilizing an integrated machine learning model and the simplified data set; and predicting the quenching rate constant range of the existing pollutants and different photosensitizers according to the quenching rate constant prediction model so as to predict the interval of the reaction rate constants of the soluble organic matters in the natural water and the existing pollutants. According to the method, the quenching rate constant range of the dissolved organic matters and the pollutants in the natural water is rapidly predicted, and complexity and high cost of a large amount of experimental work are avoided.

Description

Technical Field

[0001] The invention relates to a method, system, electronic equipment and storage medium for predicting quenching rate constants of photosensitizers and pollutants, and belongs to the technical field of environmental risk monitoring. Background Art

[0002] With the development of new technologies, the number of new chemicals (such as pesticides, drugs, and cosmetic ingredients) has increased dramatically. Currently, there are about 350,000 chemicals used in production and use, and this number continues to grow. Chemicals that are confirmed to be harmful to the environment or human body are called priority controlled chemicals. Once these chemicals enter the environment, they become new pollutants. Due to the unknown nature of these chemicals, their risk assessment is particularly important for new chemicals that are about to be launched on the market to ensure that potential hazards to the environment and human health are prevented in a timely manner.

[0003] The presence of a large amount of dissolved organic matter in natural water is an important factor in initiating the photodegradation of pollutants. The photosensitized degradation rate constant between the excited triplet dissolved organic matter and the pollutant is crucial for determining the half-life of the chemical, which is of key significance for ecological risk assessment. However, the photosensitized degradation rate constant between the excited triplet dissolved organic matter and the pollutant is related to the structure of the pollutant as well as the structure of the dissolved organic matter. Studies have found that the photosensitized degradation rate constant of sulfadiazine in seawater is an order of magnitude larger than that in fresh water. Literature shows that the photosensitization properties of dissolved organic matter are closely related to its structure. Dissolved organic matter with a large molecular weight and strong aromaticity usually has a high oxidizability, but its excited triplet energy is low. This characteristic causes the photosensitized reaction mechanism of dissolved organic matter and pollutants to be divided into energy transfer, redox reaction, or both reactions. Therefore, the degradation rate of pollutants in natural water bodies is not a fixed value, but is within a range.

[0004] The structural differences of dissolved organic matter make it impossible to use the existing international standard humic acid to replace dissolved organic matter in natural water bodies for research. However, due to the complexity of the dissolved organic matter extraction process and the complex reaction mechanism between various dissolved organic matter and chemicals, the relevant measurement data are relatively scarce. Studies have shown that quenching rate constants can be used to replace reaction rate constants for risk assessment. Therefore, in ecological risk assessment, more comprehensive and flexible models are urgently needed to accurately predict the range of quenching rate constants of pollutants in natural water. Summary of the invention

[0005] In view of this, the present invention provides a method, device, electronic device and storage medium for predicting the quenching rate constants of photosensitizers and pollutants. The method adopts photosensitizers containing chromophores similar to dissolved organic matter, and represents the quenching rate constant range of pollutants in natural water by predicting the quenching rate constant range between these photosensitizers and pollutants. This can save experimental costs and time, and solve the problems in the prior art such as unclear structure of dissolved organic matter, difficulty in extraction, and challenge in calculating property parameters of dissolved organic matter.

[0006] The first object of the present invention is to provide a method for predicting the quenching rate constant of a photosensitizer and a pollutant

[0007] The second object of the present invention is to provide a system for predicting the quenching rate constants of photosensitizers and pollutants.

[0008] A third object of the present invention is to provide an electronic device.

[0009] A fourth object of the present invention is to provide a storage medium.

[0010] The first object of the present invention can be achieved by adopting the following technical solutions:

[0011] A method for predicting quenching rate constants of a photosensitizer and a pollutant, the method comprising:

[0012] Acquire a data set of photosensitizers and pollutants, wherein the data set includes characteristic variables and target variables, wherein the characteristic variables include structural and chemical characteristic data, sensitized photolysis mechanism characteristic data, and experimental data, and the target variable includes a quenching rate constant;

[0013] Simplify the feature variables to obtain simplified feature variables, and construct a simplified data set together with the simplified feature variables and the target variables;

[0014] Using an integrated machine learning model and a simplified data set, a prediction model for the quenching rate constants of photosensitizers and pollutants was constructed;

[0015] The quenching rate constant prediction model is used to predict the range of quenching rate constants between existing pollutants and different photosensitizers, thereby predicting the range of reaction rate constants between dissolved organic matter in natural water and existing pollutants.

[0016] Furthermore, simplifying the feature variables to obtain simplified feature variables, and constructing a simplified data set with the simplified feature variables and the target variables includes:

[0017] Calculate the feature importance value of each feature data;

[0018] Select multiple ensemble machine learning models and use the feature recursive elimination method to continuously remove unimportant feature data until the average coefficient of determination performance index of each ensemble machine learning model no longer improves;

[0019] A set of feature data that most significantly improves the effect of each integrated machine learning model is selected as the simplified feature variables, and the simplified feature variables are used together with the target variables to construct a simplified data set.

[0020] Furthermore, the feature recursive elimination method is used to continuously remove unimportant feature data until the average coefficient of determination performance index of each integrated machine learning model is no longer improved, including:

[0021] In each iteration, the least important feature data, that is, the feature data with the lowest weight or importance, is removed. By gradually reducing the feature data, each integrated machine learning model is repeatedly trained and updates the feature data. Through multiple rounds of iterations, unimportant feature data is continuously removed until the average determination coefficient performance index of various integrated machine learning models no longer improves.

[0022] Further, the ensemble machine learning model is constructed based on gradient boosted regression trees, extreme gradient boosting, lightweight gradient, and random forests;

[0023] The method uses an integrated machine learning model and a simplified data set to construct a quenching rate constant prediction model for photosensitizers and pollutants, including:

[0024] An integrated machine learning model was trained based on a simplified data set. After training, the average determination coefficient, average correlation coefficient, and average root mean square error of gradient boosting regression tree, extreme gradient enhancement, lightweight gradient, and random forest were compared. The model with the largest average determination coefficient and average correlation coefficient and the smallest average root mean square error was used as the prediction model for the quenching rate constant of photosensitizers and pollutants.

[0025] Furthermore, the quenching rate constant range of existing pollutants and different photosensitizers is predicted according to the quenching rate constant prediction model, including:

[0026] The quenching rate constants of existing pollutants and different photosensitizers are predicted according to the quenching rate constant prediction model. The interface components are called to present the importance of feature data and the maximum and minimum values ​​of the quenching rate constants of pollutants and photosensitizers, and the predicted value range of the quenching rate constants of pollutants and photosensitizers is obtained.

[0027] Furthermore, the structural and chemical characteristic data include the number of aromatic rings, the number of aliphatic rings, the number of saturated rings, the number of heavy atoms, the number of heteroatoms, the dissociation constant, the octanol-water partition coefficient, the molecular weight, and the dipole moment;

[0028] The characteristic data of the sensitized photolysis mechanism include the number of hydrogen bond donors, the number of hydrogen bond acceptors, the vertical ionization energy, the vertical affinity energy, the H bond dissociation enthalpy, the proton dissociation enthalpy, the oxidation potential, the reduction potential, the energy of the highest occupied molecular orbital, the energy of the lowest occupied molecular orbital, and the excited triplet energy;

[0029] The experimental data include pH.

[0030] Furthermore, the structural and chemical characteristic data are obtained by density functional theory calculations.

[0031] The second object of the present invention can be achieved by adopting the following technical solutions:

[0032] A photosensitizer and pollutant quenching rate constant prediction system, the system comprising:

[0033] An acquisition module, used to acquire a data set of photosensitizers and pollutants, wherein the data set includes characteristic variables and target variables, wherein the characteristic variables include structural and chemical characteristic data, sensitized photolysis mechanism characteristic data, and experimental data, and the target variable includes a quenching rate constant;

[0034] The first construction module is used to simplify the feature variables to obtain simplified feature variables, and construct a simplified data set together with the simplified feature variables and the target variables;

[0035] The second building module is used to construct a quenching rate constant prediction model for photosensitizers and pollutants using an integrated machine learning model and a simplified data set;

[0036] The prediction module is used to predict the range of quenching rate constants between existing pollutants and different photosensitizers based on the quenching rate constant prediction model, so as to predict the interval of reaction rate constants between dissolved organic matter and existing pollutants in natural water.

[0037] The third object of the present invention can be achieved by adopting the following technical solutions:

[0038] An electronic device comprises a processor and a memory for storing a program executable by the processor. When the processor executes the program stored in the memory, the above-mentioned method for predicting the quenching rate constant of a photosensitizer and a pollutant is implemented.

[0039] The fourth object of the present invention can be achieved by adopting the following technical solutions:

[0040] A storage medium stores a program, and when the program is executed by a processor, the above-mentioned method for predicting the quenching rate constant of a photosensitizer and a pollutant is implemented.

[0041] The present invention has the following beneficial effects compared with the prior art:

[0042] The method of the present invention obtains structural and chemical characteristic data of photosensitizers and pollutants, sensitized photolysis mechanism characteristic data and experimental data as characteristic variables, and obtains quenching rate constants of pollutants and photosensitizers as target variables, simplifies the characteristic variables, and constructs a simplified data set together with the target variables. A quenching rate constant prediction model for photosensitizers and pollutants is constructed by integrating machine learning models and simplified data sets, and successfully predicts the quenching rate constant value range of dissolved organic matter and pollutants in natural water. In addition, since the acquisition of input data does not rely on complex detection technology, experimental costs and experimental time are saved. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying creative work.

[0044] Figure 1 This is a scene diagram of the method for predicting the quenching rate constants of photosensitizers and pollutants according to Example 1 of the present invention.

[0045] Figure 2 This is a flow chart of the method for predicting the quenching rate constants of photosensitizers and pollutants according to Example 1 of the present invention.

[0046] Figure 3 Schematic diagram of feature importance ranking and weighting in Example 1 of the present invention.

[0047] Figure 4a to Figure 4d This is a graph showing the results of using the feature recursive elimination method in Example 1 of the present invention to screen feature data.

[0048] Figure 5 This is a prediction evaluation diagram of the integrated machine learning model of Example 1 of the present invention.

[0049] Figure 6 This is a graph of prediction accuracy based on the random forest model of Example 1 of the present invention.

[0050] Figure 7 This is a feature importance graph based on the random forest model of Example 1 of the present invention.

[0051] Figure 8 This is a structural block diagram of the photosensitizer and pollutant quenching rate constant prediction system of Example 2 of the present invention.

[0052] Fig. 9 This is a structural block diagram of an electronic device according to Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0054] It should be noted that, in the embodiments of the present invention, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0055] Embodiment 1:

[0056] like Figure 1 As shown, this is a scene diagram of the method for predicting the quenching rate constant of the photosensitizer and the pollutant in this embodiment. The scene may include an electronic device 102 and an experimental device 103. The electronic device 102 and the experimental device 103 are connected. The electronic device 102 may be a laptop computer or a desktop computer. The connection method may be a wired connection or a wireless connection, which is not limited here.

[0057] First, the electronic device 102 obtains structural and chemical characteristic data, sensitized photolysis mechanism characteristic data and experimental data as characteristic variables, obtains the quenching rate constant of the pollutant and the photosensitizer as the target variable, and constructs a data set according to the characteristic variables and the target variables.

[0058] Secondly, the electronic device 102 simplifies the characteristic variables to obtain simplified characteristic variables, and constructs a simplified data set together with the simplified characteristic variables and the target variables, and uses the integrated machine learning model and the simplified data set to construct a quenching rate constant prediction model for the photosensitizer and the pollutant.

[0059] Finally, the electronic device 102 predicts the numerical range of the quenching rate constants of pollutants and dissolved organic matter according to the quenching rate constant prediction model and the data to be predicted input by the first user 101.

[0060] Among them, the structural and chemical characteristic data include the number of aromatic rings, the number of aliphatic rings, the number of saturated rings, the number of heavy atoms, the number of heteroatoms, the dissociation constant, the octanol-water partition coefficient, the molecular weight, and the dipole moment. The structural and chemical characteristic data are calculated using density functional theory; the sensitized photolysis mechanism characteristic data include the number of hydrogen bond donors, the number of hydrogen bond acceptors, the vertical ionization energy, the vertical affinity energy, the H bond dissociation enthalpy, the proton dissociation enthalpy, the oxidation potential, the reduction potential, the energy of the highest occupied molecular orbital, the energy of the lowest occupied molecular orbital, and the excited triplet energy; the experimental data include pH.

[0061] Among them, the integrated machine learning model is built based on Gradient Boosting Regression (GBR), extreme gradient boosting (XGB), lightweight gradient (Light GradientBoosting Machine, LGBM) and random forest (Random Forest, RF), that is, there are four integrated machine learning models.

[0062] The second user 104 operates the experimental equipment 103 and the electronic device 102 to obtain the data set of the above embodiment, and the data set includes: 1. Experimental conditions for quenching rate constant: pH; 2. Structural and chemical characteristic data of pollutants and photosensitizers including aromatic rings, aliphatic rings, pollutant electrotopological state descriptors (Hall Kier Alpha), sp 3 Carbon fraction (FractionCSP3), valence electrons, saturated rings, heavy atom count, heteroatom count, dissociation constant (pKa), octanol-water partition coefficient (Logkow), molecular weight (MW) and dipole moment (μ); 3. Characteristic data of sensitized photolysis mechanism include hydrogen bond donor number (HBDC), hydrogen bond acceptor number (HBAC), vertical ionization energy (VIE), vertical affinity energy (VEA), H bond dissociation enthalpy (BDE), proton dissociation enthalpy (PDE), redox potential (ROX), energy of highest occupied molecular orbital (E HOMO ), the energy of the lowest occupied molecular orbital (E LUMO ), excited triplet energy (E T1);4. Quenching rate constant of pollutants and photosensitizers;Illustratively, the experimental equipment 103 includes a laser flash photolysis instrument and a high performance liquid chromatography, the quenching rate constant of pollutants and photosensitizers is obtained by exciting the flash photolysis instrument, and the structural and chemical characteristic data and the sensitized photolysis mechanism characteristic data are all calculated by the electronic device 102.

[0063] The structural and chemical characteristic data were calculated as follows: the RDKit library was installed and imported in the Python environment, the molecular structure of the pollutant was converted into the SMILES format, the string was read and the relevant molecular structure parameters were calculated through the Descriptors module of RDKit.

[0064] The characteristic data of the sensitized photolysis mechanism were calculated as follows: the initial rough three-dimensional structure of the photosensitizer and the pollutant was drawn by ChemDraw 19, and then the molecular conformation was adjusted by MM2 to minimize the energy. The molecular configuration of each compound was optimized using density functional theory (DFT) calculations at the B3LYP / 6-31+G(d,p) level, and IEFPCM was selected as the solvation model. The optimized molecular structure was verified to have the minimum energy when there was no imaginary frequency by frequency analysis. The dipole moment (μ), the energy of the highest occupied molecular orbital (E HOMO ), the energy of the lowest unoccupied molecular orbital (E LUMO ) and the HOMO-LUMO separation energy (E HOMO -E LUMO )5 quantum chemical descriptors.

[0065] The above is only an example, and the selection and use of equipment in actual applications can be more flexible.

[0066] like Figure 2 FIG. 1 is a flow chart of a method for predicting the quenching rate constant of a photosensitizer and a pollutant. The method is executed by the above-mentioned electronic device and includes the following steps:

[0067] S201. Obtain data sets of photosensitizers and pollutants.

[0068] As described above, the data set of this embodiment includes characteristic variables and target variables, the characteristic variables include structural and chemical characteristic data, sensitized photolysis mechanism characteristic data and experimental data, and the target variable includes the quenching rate constant.

[0069] S202: Simplify the feature variables to obtain simplified feature variables, and construct a simplified data set together with the simplified feature variables and the target variables.

[0070] In one embodiment, the feature importance value of each feature data is calculated; multiple integrated machine learning models are selected, and unimportant feature data are continuously removed according to the feature recursive elimination method until the average determination coefficient performance index of each integrated machine learning model is no longer improved; a set of feature data that most significantly improves the effect of each integrated machine learning model is selected as a simplified feature variable, and the simplified feature variables are used together with the target variable to construct a simplified data set. The specific steps are as follows:

[0071] S2021. The feature importance value of each feature data is calculated internally in the integrated machine learning model.

[0072] S2022. Remove the least important feature data in the integrated machine learning model, that is, the feature data with the lowest weight or importance. The calculation results refer to Figure 3 shown.

[0073] S2023. Retrain the integrated machine learning model using the remaining feature data and evaluate the feature importance again.

[0074] S2024. Loop through steps S2022 to S2023 step by step, eliminating the least important feature data in turn until the average determination coefficient performance index of each integrated machine learning model no longer improves; finally, select a set of feature data that most significantly improves the effect of the integrated machine learning model through the average determination coefficient of the integrated machine learning model as simplified feature variables, and construct a simplified data set together with the target variable.

[0075] After the above feature engineering, the calculation results refer to Figure 4a-4d As shown, RF retains 7 eigenvalues, LGBM retains 10 eigenvalues, GBR retains 6 eigenvalues, and XGB retains 9 eigenvalues, completing the simplification of feature variables.

[0076] S203. Using an integrated machine learning model and a simplified data set, a prediction model for the quenching rate constants of photosensitizers and pollutants is constructed;

[0077] The ensemble algorithm provides more accurate model output by integrating a large number of weak learners into a strong learner. The gradient boosting regression tree is a model that builds decision trees in a progressive manner. The residual of the prediction result of the previous tree is used as the true value of the next tree and predicted. Finally, the prediction results of all trees are added to get the final prediction result. Extreme gradient enhancement uses a large number of decision trees, and helps parallel and distributed computing through weighted quantile search, and adds regular terms in the loss function to reduce the complexity of the tree model, thereby improving the generalization ability of the model. Lightweight gradient builds multiple decision trees, gradually fits the error and accumulates the prediction results of each tree to achieve accurate prediction of the model. It uses the histogram algorithm and the leaf-first splitting strategy to speed up the calculation and improve the convergence efficiency of the model. Random forest applies the bagging theory in the modeling process. By averaging the performance of decision trees on each sub-sample of the data set, it can improve the prediction accuracy and control overfitting.

[0078] In one embodiment, an integrated machine learning model is trained based on a simplified data set, and 100 random seeds are used to randomly divide the entire simplified data set into a training set (80%) and a test set (20%). After training, the average determination coefficient, average correlation coefficient, and average root mean square error of the gradient boosting regression tree, extreme gradient enhancement, lightweight gradient, and random forest are compared, and the model with the largest average determination coefficient and average correlation coefficient and the smallest average root mean square error is used as the prediction model for the quenching rate constant of the photosensitizer and the pollutant.

[0079] The calculation results are as follows Figure 5 and Figure 6 As shown, among the four integrated machine learning models, the average determination coefficient of random forest is 0.619, which has the best prediction accuracy. In the Taylor diagram, the random forest model has the best prediction effect, characterized by the maximum average correlation coefficient of 0.805 and the minimum average root mean square error of 0.533. Its Taylor diagram results show that the points are closest to the standard value, so it is used as the quenching rate constant prediction model.

[0080] In one embodiment, the contribution of each feature to the prediction of the quenching rate constants of pollutants and dissolved organic matter in natural water is explained by using SHapley Additive exPlanations (SHAP) feature importance analysis, such as Figure 7 As shown, the importance of input features can be divided into three categories, and the priority order is: sensitized photolysis mechanism characteristic data > structural and chemical characteristic data > experimental data, which indicates that the reaction mechanism of pollutants and dissolved organic matter makes the greatest contribution to the prediction of the quenching rate constants of pollutants and dissolved organic matter in natural water, further emphasizing the key role of mechanistic factors in prediction.

[0081] S204. Predict the range of quenching rate constants between existing pollutants and different photosensitizers based on the quenching rate constant prediction model, so as to predict the interval of reaction rate constants between dissolved organic matter in natural water and existing pollutants.

[0082] In one embodiment, the quenching rate constants of existing pollutants and different photosensitizers are predicted based on the quenching rate constant prediction model, and the interface component is called to present the importance of feature data and the maximum and minimum values ​​of the quenching rate constants of pollutants and photosensitizers to obtain the predicted value range of the quenching rate constants of pollutants and photosensitizers.

[0083] Those skilled in the art will appreciate that all or part of the steps in the method for implementing the above embodiments may be completed by instructing related hardware through a program, and the corresponding program may be stored in a computer-readable storage medium.

[0084] It should be noted that although the method operations of the above embodiments are described in a specific order, this does not require or imply that these operations must be performed in this specific order, or that all the operations shown must be performed to achieve the desired results. On the contrary, the steps depicted can be performed in a different order. Additionally or alternatively, some steps can be omitted, multiple steps can be combined into one step, and / or one step can be decomposed into multiple steps.

[0085] Embodiment 2:

[0086] like Figure 8 As shown, this embodiment provides a system for predicting the quenching rate constant of a photosensitizer and a pollutant, the system comprising an acquisition module 801, a first construction module 802, a second construction module 803 and a prediction module 804, and the specific description of each module is as follows:

[0087] The acquisition module 801 is used to acquire a data set of photosensitizers and pollutants, wherein the data set includes characteristic variables and target variables, wherein the characteristic variables include structural and chemical characteristic data, sensitized photolysis mechanism characteristic data and experimental data, and the target variable includes a quenching rate constant.

[0088] The first construction module 802 is used to simplify the feature variables to obtain simplified feature variables, and construct a simplified data set together with the simplified feature variables and the target variables.

[0089] The second construction module 803 is used to construct a quenching rate constant prediction model for photosensitizers and pollutants by using an integrated machine learning model and a simplified data set.

[0090] The prediction module 804 is used to predict the range of quenching rate constants between existing pollutants and different photosensitizers according to the quenching rate constant prediction model, so as to predict the interval of the reaction rate constants between dissolved organic matter in natural water and existing pollutants.

[0091] It should be noted that the system provided in this embodiment is only illustrated by the division of the above-mentioned functional modules. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure can be divided into different functional modules to complete all or part of the functions described above.

[0092] Embodiment 3:

[0093] This embodiment provides an electronic device, such as Fig. 9 As shown, it includes a processor 902, a memory, an input device 903, a display device 904 and a network interface 905 connected through a device bus 901. The processor is used to provide computing and control capabilities. The memory includes a non-volatile storage medium 906 and an internal memory 907. The non-volatile storage medium 906 stores an operating device, a computer program and a database. The internal memory 907 provides an environment for the operation of the operating device and the computer program in the non-volatile storage medium. When the processor 902 executes the computer program stored in the memory, the method for predicting the quenching rate constant of the photosensitizer and the pollutant in the above-mentioned embodiment 1 is implemented as follows:

[0094] A data set of photosensitizers and pollutants is obtained, wherein the data set includes characteristic variables and target variables, wherein the characteristic variables include structural and chemical characteristic data, sensitized photolysis mechanism characteristic data and experimental data, and the target variable includes a quenching rate constant; the characteristic variables are simplified to obtain simplified characteristic variables, and a simplified data set is constructed together with the simplified characteristic variables and the target variables; a quenching rate constant prediction model of photosensitizers and pollutants is constructed; and the quenching rate constant range of existing pollutants and different photosensitizers is predicted according to the quenching rate constant prediction model, so as to predict the interval of the reaction rate constant of dissolved organic matter in natural water and existing pollutants.

[0095] Embodiment 4:

[0096] This embodiment provides a storage medium, which is a computer-readable storage medium, storing a computer program. When the computer program is executed by a processor, the method for predicting the quenching rate constant of the photosensitizer and the pollutant in the above embodiment 1 is implemented as follows:

[0097] A data set of photosensitizers and pollutants is obtained, wherein the data set includes characteristic variables and target variables, wherein the characteristic variables include structural and chemical characteristic data, sensitized photolysis mechanism characteristic data and experimental data, and the target variable includes a quenching rate constant; the characteristic variables are simplified to obtain simplified characteristic variables, and a simplified data set is constructed together with the simplified characteristic variables and the target variables; a quenching rate constant prediction model of photosensitizers and pollutants is constructed; and the quenching rate constant range of existing pollutants and different photosensitizers is predicted according to the quenching rate constant prediction model, so as to predict the interval of the reaction rate constant of dissolved organic matter in natural water and existing pollutants.

[0098] It should be noted that the computer-readable storage medium of the present embodiment may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0099] In this embodiment, the computer-readable storage medium may be any tangible medium that includes or stores a program that can be used by or in combination with an instruction execution system, device, or device. In this embodiment, the computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium may also be any computer-readable storage medium other than a computer-readable storage medium, which may send, propagate, or transmit a program used by or in combination with an instruction execution system, device, or device. The computer program included on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0100] The computer readable storage medium can be written in one or more programming languages ​​or a combination thereof to execute the computer program of the present embodiment, and the programming language includes an object-oriented programming language, such as Java, Python, C++, and also includes a conventional procedural programming language, such as C language or a similar programming language. The program can be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., using an Internet service provider to connect through the Internet).

[0101] In summary, the method of the present invention established a machine learning model based on small sample experimental data, successfully predicted the range of quenching rate constants of dissolved organic matter and pollutants in natural water, and did not rely on complex detection technology. It can quickly and accurately obtain the range of quenching rate constants through calculations of the physicochemical properties of pollutants and even new chemicals, thereby greatly reducing experimental costs and time. This method is not only applicable to existing pollutants, but can also be applied to the prediction of quenching rate constants of new chemicals; based on this quenching rate constant, the half-life of chemicals in natural water bodies can be further calculated, providing a scientific basis for the ecological risk assessment of these chemicals.

[0102] The above is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical solution and invention concept of the present invention within the scope disclosed by the present invention, which shall fall within the protection scope of the present invention.

Claims

1. A method for predicting the quenching rate constant of a photosensitizer and a pollutant, characterized in that: The method comprises: Acquire a data set of photosensitizers and pollutants, wherein the data set includes characteristic variables and target variables, wherein the characteristic variables include structural and chemical characteristic data, sensitized photolysis mechanism characteristic data, and experimental data, and the target variable includes a quenching rate constant; Simplify the feature variables to obtain simplified feature variables, and construct a simplified data set together with the simplified feature variables and the target variables; Using an integrated machine learning model and a simplified data set, a prediction model for the quenching rate constants of photosensitizers and pollutants was constructed; The quenching rate constant prediction model is used to predict the range of quenching rate constants between existing pollutants and different photosensitizers, thereby predicting the range of reaction rate constants between dissolved organic matter in natural water and existing pollutants.

2. The method for predicting the quenching rate constant of a photosensitizer and a pollutant according to claim 1, characterized in that: The step of simplifying the feature variables to obtain simplified feature variables, and constructing a simplified data set using the simplified feature variables and the target variables includes: Calculate the feature importance value of each feature data; Select multiple ensemble machine learning models and use the feature recursive elimination method to continuously remove unimportant feature data until the average coefficient of determination performance index of each ensemble machine learning model no longer improves; A set of feature data that most significantly improves the effect of each integrated machine learning model is selected as the simplified feature variables, and the simplified feature variables are used together with the target variables to construct a simplified data set.

3. The method for predicting the quenching rate constant of a photosensitizer and a pollutant according to claim 2, characterized in that: The recursive feature elimination method is used to continuously remove unimportant feature data until the average coefficient of determination performance index of each integrated machine learning model no longer improves, including: In each iteration, the least important feature data, that is, the feature data with the lowest weight or importance, is removed. By gradually reducing the feature data, each integrated machine learning model is repeatedly trained and updates the feature data. Through multiple rounds of iterations, unimportant feature data is continuously removed until the average determination coefficient performance index of various integrated machine learning models no longer improves.

4. The method for predicting the quenching rate constant of a photosensitizer and a pollutant according to claim 1, characterized in that: The ensemble machine learning model is built based on gradient boosted regression trees, extreme gradient boosting, lightweight gradient, and random forests; The method uses an integrated machine learning model and a simplified data set to construct a quenching rate constant prediction model for photosensitizers and pollutants, including: An integrated machine learning model was trained based on a simplified data set. After training, the average determination coefficient, average correlation coefficient, and average root mean square error of gradient boosting regression tree, extreme gradient enhancement, lightweight gradient, and random forest were compared. The model with the largest average determination coefficient and average correlation coefficient and the smallest average root mean square error was used as the prediction model for the quenching rate constant of photosensitizers and pollutants.

5. The method for predicting the quenching rate constant of a photosensitizer and a pollutant according to claim 1, characterized in that: The quenching rate constant ranges of existing pollutants and different photosensitizers are predicted according to the quenching rate constant prediction model, including: The quenching rate constants of existing pollutants and different photosensitizers are predicted according to the quenching rate constant prediction model. The interface components are called to present the importance of feature data and the maximum and minimum values ​​of the quenching rate constants of pollutants and photosensitizers, and the predicted value range of the quenching rate constants of pollutants and photosensitizers is obtained.

6. The method for predicting the quenching rate constant of a photosensitizer and a pollutant according to claim 1, characterized in that: The structural and chemical characteristic data include the number of aromatic rings, the number of aliphatic rings, the number of saturated rings, the number of heavy atoms, the number of heteroatoms, the dissociation constant, the octanol-water partition coefficient, the molecular weight, and the dipole moment; The characteristic data of the sensitized photolysis mechanism include the number of hydrogen bond donors, the number of hydrogen bond acceptors, the vertical ionization energy, the vertical affinity energy, the H bond dissociation enthalpy, the proton dissociation enthalpy, the oxidation potential, the reduction potential, the energy of the highest occupied molecular orbital, the energy of the lowest occupied molecular orbital, and the excited triplet energy; The experimental data include pH.

7. The method for predicting the quenching rate constant of a photosensitizer and a pollutant according to claim 6, characterized in that: The structural and chemical characteristic data were obtained by density functional theory calculations.

8. A photosensitizer and pollutant quenching rate constant prediction system, characterized in that: The system comprises: An acquisition module, used to acquire a data set of photosensitizers and pollutants, wherein the data set includes characteristic variables and target variables, wherein the characteristic variables include structural and chemical characteristic data, sensitized photolysis mechanism characteristic data, and experimental data, and the target variable includes a quenching rate constant; The first construction module is used to simplify the feature variables to obtain simplified feature variables, and construct a simplified data set together with the simplified feature variables and the target variables; The second building module is used to construct a quenching rate constant prediction model for photosensitizers and pollutants using an integrated machine learning model and a simplified data set; The prediction module is used to predict the range of quenching rate constants between existing pollutants and different photosensitizers based on the quenching rate constant prediction model, so as to predict the interval of reaction rate constants between dissolved organic matter and existing pollutants in natural water.

9. An electronic device comprising a processor and a memory for storing a program executable by the processor, characterized in that: When the processor executes the program stored in the memory, the method for predicting the quenching rate constant of a photosensitizer and a pollutant according to any one of claims 1 to 7 is implemented.

10. A storage medium storing a program, characterized in that: When the program is executed by a processor, the method for predicting the quenching rate constant of a photosensitizer and a pollutant according to any one of claims 1 to 7 is implemented.

Images