Interactive catalysis methods and platforms based on molecular simulations and artificial intelligence

Abstract

The application discloses an interactive catalytic method and platform based on molecular simulation and artificial intelligence, and realizes intelligent design of a catalyst through data classification, condition optimization, structure modeling and mechanism analysis. After being standardized, the catalytic data uploaded by a user is divided into two categories: data set A is used for recommending optimal reaction conditions by a multi-objective optimization algorithm; and the catalyst structure of data set B is optimized by quantum chemistry software to construct a three-dimensional model, and multiple intermediate, transition state structures and geometric / electronic descriptors are optimized. Through feature screening and dimension reduction technology, a double-model collaborative prediction system of model E and model F is constructed, and finally, model F is used to drive a molecular generation algorithm to screen high-performance catalysts. The application can realize flexible adaptation to specific application scenarios, meet the urgent needs of the industry for efficient and accurate catalyst performance prediction and intelligent design, and effectively help the industry to reduce costs and increase efficiency, and accelerate the catalyst research and development process.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of molecular catalysis, in particular to an interactive catalysis method and platform based on molecular simulation and artificial intelligence. BACKGROUND

[0002] Improvement and design of catalytic system plays a key role in chemical research and development process. However, systematic study of catalytic system faces many challenges, including complex and changeable influencing factors and wide catalytic structure space. With the continuous development of artificial intelligence technology, combination of molecular simulation method and AI algorithm has become a new trend in exploring catalytic system in academic and industrial circles. In order to reduce the technical threshold and improve the research and development efficiency, it is particularly important to develop a simple and easy-to-use integrated tool.

[0003] At present, there are many catalyst databases and platforms available, such as Catalysis-Hub, Open Catalyst Project, etc. (see review Nat. Catal. 2024, 7, 624-635), but these platforms still lack customization support, especially the flexible adaptation to specific application scenarios. In addition, existing tools often fail to effectively integrate molecular simulation and artificial intelligence methods, making it difficult to meet the urgent needs of the industry for efficient and accurate catalyst performance prediction and intelligent design.

[0004] Therefore, how to invent an interactive catalysis method based on molecular simulation and artificial intelligence to meet the urgent needs of the industry for efficient and accurate catalyst performance prediction and intelligent design, and effectively help the industry to reduce cost and increase efficiency and accelerate catalyst research and development process, has become a problem to be solved. SUMMARY

[0005] Therefore, the present application provides an interactive catalysis method and platform based on molecular simulation and artificial intelligence, which can realize flexible adaptation to specific application scenarios, meet the urgent needs of the industry for efficient and accurate catalyst performance prediction and intelligent design, and effectively help the industry to reduce cost and increase efficiency and accelerate catalyst research and development process.

[0006] In order to achieve the above purpose, the present application provides the following technical scheme: an interactive catalysis method based on molecular simulation and artificial intelligence, comprising:

[0007] standardizing the user uploaded catalytic research and development data to obtain standardized data; classifying the data with the same catalyst parameters but different experimental conditions in the standardized data into data set A; classifying the data with the same experimental conditions but different catalyst parameters in the standardized data into data set B;

[0008] determining the optimal catalytic conditions by a multi-objective optimization algorithm based on the data set A;

[0009] constructing a catalyst three-dimensional structure and an intermediate three-dimensional structure based on the catalyst structure information of the data set B; performing geometric structure optimization on the catalyst three-dimensional structure and the intermediate three-dimensional structure respectively by quantum chemistry software to obtain a catalyst structure molecular set and an intermediate structure molecular set;

[0010] performing transition state search and reaction path calculation on the intermediate molecules in the intermediate structure molecular set by the quantum chemistry software to generate a transition state structure molecular set and a candidate feature set C;

[0011] calculating geometric structure descriptors and electronic structure descriptors for the molecular structures in the catalyst structure molecular set, the intermediate structure molecular set and the transition state structure molecular set respectively to generate a candidate feature combination D;

[0012] performing integrated analysis on the data set B, the candidate feature set C and the candidate feature combination D by correlation analysis and dimensionality reduction technology to screen out a feature subset related to performance optimization parameters;

[0013] constructing a performance prediction model E and a performance prediction model F based on the feature subset; training the performance prediction model E and the performance prediction model F by a machine learning algorithm to obtain the trained performance prediction model E and the performance prediction model F;

[0014] generating a candidate catalyst structure by a molecular generation algorithm; constructing a catalyst candidate library based on the candidate catalyst structure; performing performance prediction on the catalyst structures in the catalyst candidate library by the trained performance prediction model F to screen out an optimal catalyst molecular structure.

[0015] As a preferred scheme of the interactive catalysis method based on molecular simulation and artificial intelligence, the catalytic research and development data includes experimental input parameters, catalyst and additive parameters, experimental output parameters and performance optimization parameters.

[0016] As a preferred scheme of the interactive catalysis method based on molecular simulation and artificial intelligence, in the process of determining the optimal catalytic conditions by the multi-objective optimization algorithm, the multi-objective optimization algorithm includes a genetic algorithm and a Bayesian optimization algorithm.

[0017] As a preferred scheme of the interactive catalysis method based on molecular simulation and artificial intelligence, the quantum chemistry software includes Gaussian and ORCA.

[0018] As a preferred solution of the interactive catalysis method based on molecular simulation and artificial intelligence, in the process of training the performance prediction model E and the performance prediction model F by the machine learning algorithm, the machine learning algorithm includes: random forest, XGBoost and neural network algorithm.

[0019] The application also provides an interactive catalysis platform based on molecular simulation and artificial intelligence, based on the above interactive catalysis method based on molecular simulation and artificial intelligence, comprising:

[0020] A catalyst data collection and standardization processing module is configured to standardize user-uploaded catalytic research and development data to obtain standardized data, classify data with the same catalyst parameters but different experimental conditions in the standardized data into a data set A, and classify data with the same experimental conditions but different catalyst parameters in the standardized data into a data set B.

[0021] An optimal catalytic condition determination module is configured to determine optimal catalytic conditions based on the data set A by a multi-objective optimization algorithm.

[0022] A molecular structure construction and optimization module is configured to construct catalyst three-dimensional structures and intermediate three-dimensional structures based on catalyst structure information of the data set B, and perform geometric structure optimization on the catalyst three-dimensional structures and the intermediate three-dimensional structures by quantum chemistry software to obtain a catalyst structure molecule set and an intermediate structure molecule set.

[0023] A reaction mechanism calculation module is configured to perform transition state search and reaction path calculation on intermediate molecules in the intermediate structure molecule set by the quantum chemistry software to generate a transition state structure molecule set and a candidate feature set C.

[0024] A molecular descriptor extraction module is configured to calculate geometric structure descriptors and electronic structure descriptors for molecular structures in the catalyst structure molecule set, the intermediate structure molecule set and the transition state structure molecule set, respectively, and generate a candidate feature combination D.

[0025] A data analysis and feature screening module is configured to perform integrated analysis on the data set B, the candidate feature set C and the candidate feature combination D by correlation analysis and dimensionality reduction techniques, and screen out a feature subset related to performance optimization parameters.

[0026] A performance prediction model construction and training module is configured to construct a performance prediction model E and a performance prediction model F based on the feature subset, and train the performance prediction model E and the performance prediction model F by a machine learning algorithm to obtain trained performance prediction model E and performance prediction model F.

[0027] The catalyst structure generation and optimization module is used to generate a candidate catalyst structure by a molecular generation algorithm; a catalyst candidate library is constructed based on the candidate catalyst structure; and a performance prediction model F is trained to predict the performance of the catalyst structure in the catalyst candidate library, and the optimal catalyst molecular structure is screened out.

[0028] As a preferred solution of the interactive catalysis platform based on molecular simulation and artificial intelligence, the catalyst data collection and standardization processing module includes experimental input parameters, catalyst and additive parameters, experimental output parameters, and performance optimization parameters.

[0029] As a preferred solution of the interactive catalysis platform based on molecular simulation and artificial intelligence, the optimal catalyst condition determination module includes a genetic algorithm and a Bayesian optimization algorithm in the process of determining the optimal catalyst condition by the multi-objective optimization algorithm.

[0030] As a preferred solution of the interactive catalysis platform based on molecular simulation and artificial intelligence, the molecular structure construction and optimization module includes Gaussian and ORCA.

[0031] As a preferred solution of the interactive catalysis platform based on molecular simulation and artificial intelligence, the performance prediction model construction and training module includes a random forest, XGBoost, and neural network algorithm in the process of training the performance prediction model E and the performance prediction model F by the machine learning algorithm.

[0032] The present application has the following advantages: the present application standardizes the user-uploaded catalytic research and development data to obtain standardized data; data with the same catalyst parameters and different experimental conditions in the standardized data are classified as data set A; data with different catalyst parameters and the same experimental conditions in the standardized data are classified as data set B; based on the data set A, the optimal catalytic conditions are determined by a multi-objective optimization algorithm; based on the catalyst structure information of the data set B, a catalyst three-dimensional structure and an intermediate three-dimensional structure are constructed; the catalyst three-dimensional structure and the intermediate three-dimensional structure are respectively subjected to geometric structure optimization by quantum chemistry software to obtain a catalyst structure molecular set and an intermediate structure molecular set; the intermediate molecules in the intermediate structure molecular set are subjected to transition state search and reaction path calculation by the quantum chemistry software to generate a transition state structure molecular set and a candidate feature set C; the molecular structures in the catalyst structure molecular set, the intermediate structure molecular set and the transition state structure molecular set are respectively calculated to generate a candidate feature combination D; the data set B, the candidate feature set C and the candidate feature combination D are subjected to integrated analysis by correlation analysis and dimensionality reduction technology to screen out a feature subset related to performance optimization parameters; based on the feature subset, a performance prediction model E and a performance prediction model F are constructed; the performance prediction model E and the performance prediction model F are trained by a machine learning algorithm to obtain the trained performance prediction model E and the performance prediction model F; a candidate catalyst structure is generated by a molecular generation algorithm; based on the candidate catalyst structure, a catalyst candidate library is constructed; the performance of the catalyst structure in the catalyst candidate library is predicted by the trained performance prediction model F to screen out an optimal catalyst molecular structure. The present application initiatively realizes the integration design of high customization and intelligence, effectively makes up for the shortcomings of the existing catalytic platform. It can realize flexible adaptation to specific application scenarios and meet the urgent needs of the industry for efficient and accurate catalyst performance prediction and intelligent design. By seamlessly integrating molecular simulation tools (such as quantum chemistry calculation) and advanced artificial intelligence algorithms, the platform allows users to upload and flexibly adapt their own research and development data, automatically complete the whole process from data standardization to catalyst structure optimization, and significantly reduce the technical threshold and the need for manual intervention. This design not only improves the accuracy of catalyst performance prediction, but also speeds up the intelligent recommendation of optimal reaction conditions and new structures, thereby significantly shortening the research and development cycle, reducing experimental trial and error costs, and providing strong support for the rapid iteration and cost reduction and efficiency improvement of catalysts in the industry. BRIEF DESCRIPTION OF DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only exemplary, and for those skilled in the art, other drawings can be obtained from the provided drawings without creative labor.

[0034] The structures, proportions, sizes, etc. shown in the specification are only used to cooperate with the content disclosed in the specification, so that those skilled in the art can understand and read, and are not used to limit the conditions that the present application can be implemented, so they do not have technical significance. Any modification of structure, change of proportion relationship or adjustment of size, without affecting the effect and purpose that the present application can produce, should still fall within the scope of the technical content disclosed by the present application.

[0035] Figure 1 A flowchart of the interactive catalysis method based on molecular simulation and artificial intelligence provided in embodiment 1 of the present application is shown.

[0036] Figure 2 A schematic diagram of the catalyst structure in a possible embodiment provided in embodiment 1 of the present application is shown.

[0037] Figure 3 A schematic diagram of the mechanism calculation in a possible embodiment provided in embodiment 1 of the present application is shown.

[0038] Figure 4 A schematic diagram of the relationship between the electronic descriptor and the transition state reaction enthalpy barrier in a possible embodiment provided in embodiment 1 of the present application is shown.

[0039] Figure 5 A schematic diagram of the new catalyst structure in a possible embodiment provided in embodiment 1 of the present application is shown.

[0040] Figure 6 A schematic diagram of the architecture of the interactive catalysis platform based on molecular simulation and artificial intelligence provided in embodiment 2 of the present application is shown. DETAILED DESCRIPTION

[0041] The embodiments of the present application will be described by specific specific embodiments, and those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in the specification. Obviously, the described embodiments are part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of the present application.

[0042] Embodiment 1

[0043] Referring to Figure 1 Embodiment 1 of the present application provides an interactive catalysis method based on molecular simulation and artificial intelligence, comprising the following steps:

[0044] S1, standardizing the user-uploaded catalysis research and development data to obtain standardized data; classifying data with the same catalyst parameters but different experimental conditions in the standardized data into a data set A; and classifying data with the same experimental conditions but different catalyst parameters in the standardized data into a data set B;

[0045] S2, determining optimal catalysis conditions by a multi-objective optimization algorithm based on the data set A;

[0046] S3, constructing a catalyst three-dimensional structure and an intermediate three-dimensional structure based on catalyst structure information of the data set B; and performing geometric structure optimization on the catalyst three-dimensional structure and the intermediate three-dimensional structure respectively by quantum chemistry software to obtain a catalyst structure molecular set and an intermediate structure molecular set;

[0047] S4, performing transition state search and reaction path calculation on intermediate molecules in the intermediate structure molecular set by the quantum chemistry software to generate a transition state structure molecular set and a candidate feature set C;

[0048] S5, calculating geometric structure descriptors and electronic structure descriptors for molecular structures in the catalyst structure molecular set, the intermediate structure molecular set and the transition state structure molecular set respectively to generate a candidate feature combination D;

[0049] S6, performing integrated analysis on the data set B, the candidate feature set C and the candidate feature combination D by correlation analysis and dimensionality reduction technology, and screening out a feature subset related to performance optimization parameters;

[0050] S7, constructing a performance prediction model E and a performance prediction model F based on the feature subset; and training the performance prediction model E and the performance prediction model F by a machine learning algorithm to obtain trained performance prediction model E and performance prediction model F;

[0051] S8, generating a candidate catalyst structure by a molecular generation algorithm; constructing a catalyst candidate library based on the candidate catalyst structure; and performing performance prediction on catalyst structures in the catalyst candidate library by the trained performance prediction model F to screen out an optimal catalyst molecular structure.

[0052] In this embodiment, in step S1, the user uploaded catalytic research and development data is standardized to obtain standardized data; the data with the same catalyst parameters and different experimental conditions in the standardized data is classified as data set A; and the data with different catalyst parameters and the same experimental conditions in the standardized data is classified as data set B.

[0053] Specifically, the user uploaded catalytic research and development data is received, and the heterogeneous data is uniformly formatted (such as SMILES standardization and unit conversion). The data with the same catalyst and additive parameters (or similar structure) but different experimental input parameters is classified as data set A; and the data with different catalyst and additive parameters (different structure or significant difference) but the same (or similar) experimental input parameters is classified as data set B.

[0054] The catalytic research and development data includes:

[0055] Experimental input parameters: such as reaction temperature, reaction time, reaction pressure, etc.

[0056] Catalyst and additive parameters: such as molecular structure (SMILES), XRD crystal structure, etc.

[0057] Experimental output parameters: such as catalytic activity, product conversion rate, selectivity, etc.

[0058] Performance optimization parameters: such as target performance indicators of catalytic efficiency, etc.

[0059] In this embodiment, in step S2, based on the data set A, the optimal catalytic condition is determined by a multi-objective optimization algorithm.

[0060] Specifically, based on the data set A, the function relationship between the experimental input parameters and the performance optimization indicators is constructed, and a multi-objective optimization algorithm (such as genetic algorithm, Bayesian optimization, etc.) is combined to realize the automatic search and recommendation of the optimal experimental conditions (such as the best temperature, time, and pressure combination), and the optimal catalytic condition is obtained.

[0061] In this embodiment, in step S3, based on the catalyst structure information of the data set B, the catalyst three-dimensional structure and the intermediate three-dimensional structure are constructed; the quantum chemistry software is used to respectively optimize the geometric structure of the catalyst three-dimensional structure and the intermediate three-dimensional structure, and the catalyst structure molecular set and the intermediate structure molecular set are obtained.

[0062] Specifically, the information of different catalysts and auxiliary agents in the data set B is encoded and expressed, and important reaction intermediate molecules are derived. The three-dimensional structure of the catalyst and the three-dimensional structure of the intermediate are constructed using the crystal structure of the catalyst (such as XRD) or the molecular structure representation (such as SMILES). The three-dimensional structure of the catalyst and the three-dimensional structure of the intermediate are geometrically optimized using quantum chemistry software (such as Gaussian, ORCA). The optimized catalyst structure is classified into a catalyst structure molecular set; the optimized intermediate structure is classified into an intermediate structure molecular set.

[0063] In this embodiment, in step S4, the intermediate molecules in the intermediate structure molecular set are subjected to transition state search and reaction path calculation by the quantum chemistry software, and a transition state structure molecular set and a candidate feature set C are generated;

[0064] Specifically, the key intermediate molecules are selected from the intermediate structure molecular set, combined with the known reaction sites, and the quantum chemistry software is called to perform transition state search and reaction path calculation to obtain the activation energy, reaction heat and other thermodynamic parameters of each elementary reaction, and form the candidate feature set C. This process simultaneously generates transition state structures to form a transition state structure molecular set.

[0065] In this embodiment, in step S5, the molecular structures in the catalyst structure molecular set, the intermediate structure molecular set and the transition state structure molecular set are calculated to generate a candidate feature combination D;

[0066] Specifically, the molecular structures in the catalyst structure molecular set, the intermediate structure molecular set and the transition state structure molecular set are calculated to generate a candidate feature combination D.

[0067] In this embodiment, in step S6, the data set B, the candidate feature set C and the candidate feature combination D are analyzed by correlation analysis and dimensionality reduction techniques to screen out a feature subset related to the performance optimization parameters;

[0068] Specifically, all the catalyst samples in the data set B are subjected to integrated analysis, and the samples include experimental input parameters, experimental output parameters, candidate feature set C and candidate feature combination D. Through correlation analysis and dimensionality reduction techniques (such as principal component analysis, mutual information evaluation, etc.), a feature subset highly related to the performance optimization parameters is screened out for subsequent model training.

[0069] In this embodiment, in step S7, a performance prediction model E and a performance prediction model F are constructed based on the feature subset; the performance prediction model E and the performance prediction model F are trained through a machine learning algorithm, and the trained performance prediction model E and the performance prediction model F are obtained;

[0070] Specifically, according to the screening result, two types of performance prediction models are constructed:

[0071] The first candidate feature group: containing experimental input parameters + candidate feature set C, D, used for training the performance prediction model E, which simultaneously considers the comprehensive influence of experimental conditions and catalyst structure on performance;

[0072] The second candidate feature group: containing only candidate feature set C, D, used for training the performance prediction model F, which focuses on the influence of structure on catalytic performance.

[0073] The machine learning algorithm (such as random forest, XGBoost, neural network, etc.) is used to train the above feature groups, and the trained performance prediction model E and performance prediction model F are obtained.

[0074] In this embodiment, in step S8, a candidate catalyst structure is generated through a molecular generation algorithm; a catalyst candidate library is constructed based on the candidate catalyst structure; the performance of the catalyst structure in the catalyst candidate library is predicted through the trained performance prediction model F, and the optimal catalyst molecular structure is screened out.

[0075] Specifically, based on the prediction ability of the performance prediction model F, combined with expert knowledge or through a molecular generation algorithm (such as a molecular generation model based on a graph neural network), a batch of candidate catalyst structures are generated, and a catalyst candidate library is constructed. Then, the performance prediction model F is used to perform high-throughput prediction and screening on the performance, and the catalyst molecular structure with the best performance is selected as the final recommended design scheme.

[0076] In one possible embodiment, an example of a catalyst with a defined geometry catalyzing the polymerization of ethylene into polyethylene is as follows:

[0077] T1, part of the collected catalyst structure is as shown in Figure 2 The researchers upload the research data to the system, and the data distribution is shown in Table 1 and Table 2:

[0078]

[0079] Table 1 Experimental input parameters

[0080] Catalyst number Yield Activity Weight average molecular weight Cat_1 30.2g 0.4 kg PE ·g Cat -1 ·h -1 ]]> 107 x 10 3 ]]

[0081] Table 2 Experimental output parameters

[0082] Table 1 and Table 2 are data distribution of catalyst experiment input parameters and experiment output parameters, performance optimization parameters are yield, and the uploaded data is grouped: the same or similar catalyst and its adjuvant parameters, and different experiment input parameters are recorded as CGC catalyst data set A; the different catalyst and its adjuvant parameters, and the same or similar experiment input parameters are recorded as CGC catalyst data set B.

[0083] T2, analyze CGC catalyst data set A, call the optimization algorithm module, and a set of optimal reaction conditions can be obtained: reaction time: 1 h; reaction temperature: 60 degrees Celsius; system pressure: 5 bar; catalyst amount: 10 mg; post-processing temperature: 20 degrees Celsius; adjuvant / catalyst ratio: 100.

[0084] T3, call CGC catalyst data set B, construct 3D structures of catalyst molecules and intermediate IM1 molecules of CGC catalysts inserted into ethylene, and here ORCA quantum chemistry software is used to expand the molecular optimization task: the calculation level is specifically: the density functional method of B3LYP, the basis set of def2svp, the solvent model based on density (SMD), the specific solvent is heptane, and the dispersion correction method of D3(BJ) is adopted. A batch of optimized CGC catalysts is recorded as CGC catalyst molecule set, and the intermediate IM1 molecules are recorded as intermediate IM1 molecule set.

[0085] T4, transition state calculation is performed on the IM1 molecules under different catalyst states, and the bonding site is the bonding between the 1st carbon atom and the 2nd carbon atom, as shown in Figure 3 , calculate the transition states of these catalysts, and these transition state structures are recorded as catalyst CGC transition state structure molecule set. The electronic energy barrier, free energy barrier and enthalpy barrier of these transition states are obtained and recorded as CGC catalyst candidate feature set C.

[0086] T5, based on the CGC catalyst molecule set, intermediate IM1 molecule set and catalyst CGC transition state structure molecule set obtained in steps T3 and T4, a batch of descriptors of these catalysts are calculated, including geometric descriptors and electronic descriptors, and a batch of candidate features are obtained and recorded as CGC catalyst candidate feature set D.

[0087] T6, sort the above data, select the catalyst samples of CGC catalyst data set B, all samples contain experiment input parameters, experiment output parameters, CGC catalyst candidate feature set C and CGC catalyst candidate feature set D. Analyze the respective correlations of these data, and retain the parts with higher correlation. Here, one of the correlation graphs is extracted, as shown in Figure 4 .

[0088] T7, This embodiment pays more attention to the relationship between catalytic structure and catalytic performance, so the CGC catalyst candidate feature set C and the CGC catalyst candidate feature set D are combined to establish a prediction model with the reaction yield. Through multiple rounds of feature engineering and machine learning model training, a prediction model F can be obtained, and the leave-one-out cross-validation R2 can reach 0.92.

[0089] T8, Based on the prediction model F of step T7, a new type of catalyst can be obtained through a molecular generation algorithm, and the structure is as shown in Figure 5 .

[0090] In summary, the present application obtains standardized data by standardizing the user-uploaded catalytic research and development data; classifies data with the same catalyst parameters but different experimental conditions in the standardized data as data set A; classifies data with different catalyst parameters but the same experimental conditions in the standardized data as data set B; determines optimal catalytic conditions based on the data set A through a multi-objective optimization algorithm; constructs catalyst three-dimensional structures and intermediate three-dimensional structures based on catalyst structure information of the data set B; performs geometric structure optimization on the catalyst three-dimensional structures and the intermediate three-dimensional structures through quantum chemistry software respectively to obtain a catalyst structure molecular set and an intermediate structure molecular set; performs transition state search and reaction path calculation on intermediate molecules in the intermediate structure molecular set through the quantum chemistry software to generate a transition state structure molecular set and a candidate feature set C; calculates geometric structure descriptors and electronic structure descriptors for molecular structures in the catalyst structure molecular set, the intermediate structure molecular set and the transition state structure molecular set respectively to generate a candidate feature combination D; integrates and analyzes the data set B, the candidate feature set C and the candidate feature combination D through correlation analysis and dimensionality reduction technology to screen out a feature subset related to performance optimization parameters; constructs performance prediction model E and performance prediction model F based on the feature subset; trains the performance prediction model E and the performance prediction model F through a machine learning algorithm to obtain trained performance prediction model E and performance prediction model F; generates a candidate catalyst structure through a molecular generation algorithm; constructs a catalyst candidate library based on the candidate catalyst structure; predicts the performance of catalyst structures in the catalyst candidate library through the trained performance prediction model F to screen out an optimal catalyst molecular structure. The present application initiates the integration design of high customization and intelligence, effectively makes up for the shortcomings of existing catalytic platforms. It can realize flexible adaptation to specific application scenarios and meet the urgent needs of the industry for efficient and accurate catalyst performance prediction and intelligent design. By seamlessly integrating molecular simulation tools (such as quantum chemistry calculation) and advanced artificial intelligence algorithms, the platform allows users to upload and flexibly adapt their own research and development data, automatically complete the whole process from data standardization to catalyst structure optimization, and significantly reduce the technical threshold and the need for manual intervention. This design not only improves the accuracy of catalyst performance prediction, but also speeds up the intelligent recommendation of optimal reaction conditions and new structures, thereby significantly shortening the research and development cycle, reducing experimental trial and error costs, and providing strong support for the rapid iteration and cost reduction and efficiency improvement of catalysts for the industry.

[0091] It should be noted that the method of the embodiments of the present disclosure can be executed by a single device, such as a computer or a server, etc. The method of the embodiments can also be applied to a distributed scenario, and be completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only execute one or more steps in the method of the embodiments of the present disclosure, and the multiple devices can interact with each other to complete the method.

[0092] It should be noted that some embodiments of the present disclosure are described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in a different order and still achieve desirable results. Additionally, the processes depicted in the figures do not necessarily require the particular order shown, or sequential order, to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous.

[0093] Embodiment 2

[0094] Referring to Figure 6 Embodiment 2 of the present disclosure also provides an interactive catalysis platform based on molecular simulation and artificial intelligence, comprising:

[0095] A catalyst data collection and standardization processing module 001 is configured to standardize user-uploaded catalysis research and development data to obtain standardized data, classify data with the same catalyst parameters but different experimental conditions in the standardized data into a data set A, and classify data with the same experimental conditions but different catalyst parameters in the standardized data into a data set B.

[0096] An optimal catalysis condition determination module 002 is configured to determine optimal catalysis conditions based on the data set A by using a multi-objective optimization algorithm.

[0097] A molecular structure construction and optimization module 003 is configured to construct a catalyst three-dimensional structure and an intermediate three-dimensional structure based on catalyst structure information of the data set B, and perform geometric structure optimization on the catalyst three-dimensional structure and the intermediate three-dimensional structure by using quantum chemistry software to obtain a catalyst structure molecule set and an intermediate structure molecule set.

[0098] A reaction mechanism calculation module 004 is configured to perform transition state searching and reaction path calculation on intermediate molecules in the intermediate structure molecule set by using the quantum chemistry software to generate a transition state structure molecule set and a candidate feature set C.

[0099] a molecule descriptor extraction module 005, configured to calculate geometric structure descriptors and electronic structure descriptors for the molecular structures in the catalyst structure molecule set, the intermediate structure molecule set and the transition state structure molecule set respectively, to generate candidate feature combinations D;

[0100] a data analysis and feature screening module 006, configured to perform integrated analysis on the data set B, the candidate feature set C and the candidate feature combinations D by correlation analysis and dimension reduction techniques, to screen a feature subset related to performance optimization parameters;

[0101] a performance prediction model construction and training module 007, configured to construct performance prediction model E and performance prediction model F based on the feature subset, and to train the performance prediction model E and the performance prediction model F by machine learning algorithms to obtain trained performance prediction model E and performance prediction model F;

[0102] a catalyst structure generation and optimization module 008, configured to generate candidate catalyst structures by a molecular generation algorithm, to construct a catalyst candidate library based on the candidate catalyst structures, and to perform performance prediction on the catalyst structures in the catalyst candidate library by the trained performance prediction model F to screen an optimal catalyst molecular structure.

[0103] In this embodiment, the catalyst data collection and standardization processing module 001, the catalyst research and development data includes: experimental input parameters, catalyst and additive parameters, experimental output parameters and performance optimization parameters.

[0104] In this embodiment, in the process of determining the optimal catalyst conditions by the multi-objective optimization algorithm in the optimal catalyst condition determination module 002, the multi-objective optimization algorithm includes genetic algorithm and Bayesian optimization algorithm.

[0105] In this embodiment, in the molecule structure construction and optimization module 003, the quantum chemistry software includes Gaussian and ORCA.

[0106] In this embodiment, in the process of training the performance prediction model E and the performance prediction model F by the machine learning algorithm in the performance prediction model construction and training module 007, the machine learning algorithm includes random forest, XGBoost and neural network algorithm.

[0107] It is to be noted that the information interaction and execution process between the modules of the above system, etc. are based on the same concept as the method embodiment in the first embodiment of the application, and the technical effects brought by them are the same as those of the method embodiment of the application. For specific content, please refer to the description in the method embodiment of the application as described above, which will not be repeated here.

[0108] Embodiment 3

[0109] Embodiment 3 provides a kind of non-transient computer readable storage medium, the program code of interactive catalytic method based on molecular simulation and artificial intelligence is stored in the computer readable storage medium, the program code includes the instruction for executing the interactive catalytic method based on molecular simulation and artificial intelligence of embodiment 1 or any possible implementation mode thereof.

[0110] Computer readable storage medium can be any available medium that computer can access or data storage device such as server, data center integrated with one or more available medium set, etc.Data storage device.The available medium can be magnetic medium, (for example, floppy disk, hard disk, magnetic tape), optical medium (for example, DVD), or semiconductor medium (for example, Solid State Disk (SSD)) and the like.

[0111] Embodiment 4

[0112] Embodiment 4 provides an electronic device, comprising: memory and processor;

[0113] The processor and the memory complete mutual communication by bus;The memory stores program instruction that can be executed by the processor, and the processor can execute the interactive catalytic method based on molecular simulation and artificial intelligence of embodiment 1 or any possible implementation mode thereof by calling the program instruction.

[0114] Specifically, the processor can be realized by hardware or software, when realized by hardware, the processor can be logic circuit, integrated circuit, etc.;When realized by software, the processor can be a general-purpose processor, realized by reading the software code stored in the memory, and the memory can be integrated in the processor, can be located outside the processor, and exist independently.

[0115] In the embodiments described above, all or some of the modules / units can be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or some of the modules / units can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded into and executed by a computer, all or some of the procedures or functions as described in the embodiments of the present application are generated. The computer can be a general purpose computer, a special purpose computer, a computer network, or other programmable system. The computer instructions can be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through a wired (for example, coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (for example, infrared, wireless, microwave, etc.) manner.

[0116] It is obvious that those skilled in the art should understand that the modules or steps of the present application described above can be implemented by a general computing system, which can be concentrated on a single computing system or distributed on a network composed of multiple computing systems, and optionally, they can be implemented by program codes executable by a computing system, so that they can be stored in a storage system and executed by a computing system, and in some cases, the steps shown or described can be executed in different order, or they can be made into individual integrated circuit modules, or multiple modules or steps can be made into a single integrated circuit module. Thus, the present application is not limited to any specific combination of hardware and software.

[0117] Although the present application has been described in detail by the above general description and specific embodiments, some modifications or improvements can be made on the basis of the present application, which is obvious to those skilled in the art. Therefore, these modifications or improvements made on the basis of not deviating from the spirit of the present application are within the scope of the present application.

Claims

1. An interactive catalysis method based on the interaction of molecular simulations and artificial intelligence, characterized in that, The method comprises the following steps: standardizing user-uploaded catalytic research and development data to obtain standardized data; classifying data in the standardized data with the same catalyst parameters and different experimental conditions as data set A, and classifying data in the standardized data with the same experimental conditions and different catalyst parameters as data set B; determining optimal catalytic conditions based on the data set A by a multi-objective optimization algorithm; constructing a catalyst three-dimensional structure and an intermediate three-dimensional structure based on catalyst structure information of the data set B; performing geometric structure optimization on the catalyst three-dimensional structure and the intermediate three-dimensional structure by quantum chemistry software to obtain a catalyst structure molecular set and an intermediate structure molecular set; performing transition state search and reaction path calculation on intermediate molecules in the intermediate structure molecular set by the quantum chemistry software to generate a transition state structure molecular set and a candidate feature set C; calculating geometric structure descriptors and electronic structure descriptors for molecular structures in the catalyst structure molecular set, the intermediate structure molecular set, and the transition state structure molecular set to generate a candidate feature combination D; integrating and analyzing the data set B, the candidate feature set C, and the candidate feature combination D by correlation analysis and dimension reduction technology to screen out a feature subset related to performance optimization parameters; constructing a performance prediction model E and a performance prediction model F based on the feature subset, and training the performance prediction model E and the performance prediction model F by a machine learning algorithm to obtain trained performance prediction model E and performance prediction model F; generating a candidate catalyst structure by a molecular generation algorithm, constructing a catalyst candidate library based on the candidate catalyst structure, and predicting the performance of catalyst structures in the catalyst candidate library by the trained performance prediction model F to screen out an optimal catalyst molecular structure.

2. The interactive catalysis method based on the interaction of molecular simulation and artificial intelligence according to claim 1, characterized in that, The catalytic research and development data includes experimental input parameters, catalyst and additive parameters, experimental output parameters, and performance optimization parameters.

3. The interactive catalysis method based on the interaction of molecular simulation and artificial intelligence according to claim 2, characterized in that, In the process of determining the optimal catalytic conditions by the multi-objective optimization algorithm, the multi-objective optimization algorithm includes a genetic algorithm and a Bayesian optimization algorithm.

4. The interactive catalysis method based on the interaction of molecular simulation and artificial intelligence according to claim 3, characterized in that, The quantum chemistry software includes Gaussian and ORCA.

5. The interactive catalysis method based on the interaction of molecular simulation and artificial intelligence according to claim 4, characterized in that, In the process of training the performance prediction model E and the performance prediction model F by the machine learning algorithm, the machine learning algorithm includes a random forest, XGBoost, and a neural network algorithm.

6. An interactive catalysis platform based on molecular simulation and artificial intelligence, using the interactive catalysis method based on molecular simulation and artificial intelligence according to any one of claims 1-5, characterized in that, The method comprises the following steps: a catalyst data acquisition and standardization processing module for standardizing user-uploaded catalytic research and development data to obtain standardized data; classifying data in the standardized data with the same catalyst parameters and different experimental conditions as data set A, and classifying data in the standardized data with the same experimental conditions and different catalyst parameters as data set B; an optimal catalytic condition determination module for determining optimal catalytic conditions based on the data set A by a multi-objective optimization algorithm; a molecular structure construction and optimization module for constructing a catalyst three-dimensional structure and an intermediate three-dimensional structure based on catalyst structure information of the data set B; The three-dimensional structure of the catalyst and the three-dimensional structure of the intermediate are optimized by quantum chemistry software to obtain a catalyst structure molecule set and an intermediate structure molecule set; A reaction mechanism calculation module is configured to perform transition state search and reaction path calculation on the intermediate molecules in the intermediate structure molecule set by the quantum chemistry software to generate a transition state structure molecule set and a candidate feature set C; A molecular descriptor extraction module is configured to calculate geometric structure descriptors and electronic structure descriptors for the molecular structures in the catalyst structure molecule set, the intermediate structure molecule set, and the transition state structure molecule set to generate a candidate feature combination D; A data analysis and feature screening module is configured to perform integrated analysis on the data set B, the candidate feature set C, and the candidate feature combination D by correlation analysis and dimensionality reduction techniques to screen a feature subset related to performance optimization parameters; A performance prediction model construction and training module is configured to construct performance prediction models E and F based on the feature subset, and train the performance prediction models E and F by machine learning algorithms to obtain trained performance prediction models E and F; A catalyst structure generation and optimization module is configured to generate candidate catalyst structures by a molecular generation algorithm, construct a catalyst candidate library based on the candidate catalyst structures, and perform performance prediction on the catalyst structures in the catalyst candidate library by the trained performance prediction model F to screen an optimal catalyst molecular structure.

7. The interactive catalysis platform based on the interaction of molecular simulation and artificial intelligence according to claim 6, characterized in that, In the catalyst data acquisition and standardization processing module, the catalyst research and development data includes experimental input parameters, catalyst and additive parameters, experimental output parameters, and performance optimization parameters.

8. The interactive catalysis platform based on the interaction of molecular simulation and artificial intelligence according to claim 7, characterized in that, In the optimal catalyst condition determination module, the multi-objective optimization algorithm includes genetic algorithms and Bayesian optimization algorithms in the process of determining the optimal catalyst conditions.

9. The interactive catalysis platform based on the interaction of molecular simulation and artificial intelligence according to claim 8, characterized in that, In the molecular structure construction and optimization module, the quantum chemistry software includes Gaussian and ORCA.

10. The interactive catalysis platform based on the interaction of molecular simulation and artificial intelligence according to claim 9, characterized in that, In the performance prediction model construction and training module, the machine learning algorithms include random forests, XGBoost, and neural network algorithms in the process of training the performance prediction models E and F by the machine learning algorithms.

Images