Reaction identification method and related device

Abstract

The invention discloses a reaction recognition method and a related device, and the method comprises the steps: firstly obtaining N molecular maps of a multi-molecular system in a chemical simulation reaction corresponding to N reaction frames, and enabling the molecular maps to be in one-to-one correspondence with the reaction frames; the molecular graph comprises a first node for identifying atoms, a connecting line for identifying a keying relationship and a sub-graph for identifying molecules. Determining a reactant according to the sub-graph of the first molecular graph, determining a product according to the sub-graph of the i-th molecular graph, generating a reaction graph in the i-th reaction frame based on the reactant and the product, the reaction graph comprising a second node identifying the reactant and a third node identifying the product, a connecting line between the second node and the third node is used for identifying candidate reaction to be identified. For a target second node and a target third node having candidate reactions, a reaction event is determined by graph isomorphism between sub-graphs. Recognition of the reaction events generated in all the reaction frames of the chemical simulation reaction can be efficiently and accurately achieved.

Description

Technical Field

[0001] The present application relates to the field of data processing, and in particular to a reaction identification method and related devices. Background Art

[0002] In the field of chemistry, we often face situations where the chemical reaction process is complex, but we need to conduct a detailed analysis of the chemical reaction process. Therefore, we choose to analyze the actual chemical reaction based on chemical simulation reactions. Therefore, the technical demand for separating and identifying the process of chemical simulation reactions is becoming increasingly strong.

[0003] In related technologies, separation and identification of complex chemical simulation reactions require a complex process that requires high computational costs, while the accuracy of the chemical reaction identification results is low.

[0004] Therefore, due to the limitations of the relevant technologies themselves, it is impossible to efficiently and accurately separate and identify the chemical reactions involved in the chemical simulation reaction process. Summary of the Invention

[0005] In order to solve the above technical problems, the present application provides a reaction identification method and related devices, which can quickly and automatically identify the reaction events generated in each reaction frame of a completed chemical simulation reaction without the need for the complex computational costs in related technologies, thereby effectively improving the reaction event identification efficiency of the chemical simulation reaction.

[0006] The embodiments of this application disclose the following technical solutions:

[0007] In one aspect, an embodiment of the present application provides a reaction identification method, the method comprising:

[0008] Obtaining N molecular graphs of a multi-molecule system in a chemical simulation reaction, where the chemical simulation reaction is completed through N reaction frames, the molecular graphs corresponding one-to-one to the reaction frames, the molecular graphs including first nodes identifying atoms, a line between two first nodes used to identify a bond relationship between the atoms corresponding to the two first nodes, the i-th molecular graph used to identify the bond relationship between atoms in the multi-molecule system in the i-th reaction frame of the chemical simulation reaction, the molecular graphs including subgraphs used to identify molecules, and N>1;

[0009] Determine the reactants of the chemical simulation reaction according to the subgraph of the first molecular graph, determine the products of the chemical simulation reaction according to the subgraph of the i-th molecular graph, and generate a reaction graph of the chemical simulation reaction in the i-th reaction frame based on the reactants and the products, wherein the reaction graph includes a second node identifying the reactants and a third node identifying the products, and a line between the second node and the third node is used to identify that there is a candidate reaction to be identified between the molecules corresponding to the second node and the third node, respectively. <i≤N;

[0010] For the target second node and the target third node having the candidate reaction, the reaction event of obtaining the product corresponding to the target third node from the reactant corresponding to the target second node is determined through the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node.

[0011] On the other hand, an embodiment of the present application provides a reaction identification device, the device comprising: an acquisition module, a generation module, and a determination module;

[0012] The acquisition module is used to acquire N molecular graphs of a multi-molecule system in a chemical simulation reaction, where the chemical simulation reaction is completed through N reaction frames, the molecular graphs correspond to the reaction frames one-to-one, the molecular graphs include first nodes that identify atoms, a line between two first nodes is used to identify that the atoms corresponding to the two first nodes have a bond relationship, the i-th molecular graph is used to identify the bond relationship between atoms in the multi-molecule system in the i-th reaction frame of the chemical simulation reaction, the molecular graph includes subgraphs, and the subgraphs are used to identify molecules, where N>1;

[0013] The generation module is used to determine the reactants of the chemical simulation reaction according to the subgraph of the first molecular graph, determine the products of the chemical simulation reaction according to the subgraph of the i-th molecular graph, and generate a reaction graph of the chemical simulation reaction in the i-th reaction frame based on the reactants and the products, wherein the reaction graph includes a second node identifying the reactants and a third node identifying the products, and a line between the second node and the third node is used to identify that there is a candidate reaction to be identified between the molecules corresponding to the second node and the third node respectively. <i≤N;

[0014] The determination module is used to determine, for the target second node and the target third node having the candidate reaction, a reaction event in which the product corresponding to the target third node is obtained from the reactant corresponding to the target second node through the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node.

[0015] In another aspect, an embodiment of the present application provides a computer device, comprising a processor and a memory:

[0016] Memory is used to store computer programs;

[0017] The processor is configured to execute the above-described method according to the computer program.

[0018] In another aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium is used to store a computer program, and the computer program is used to execute the method described in the above aspects.

[0019] On the other hand, an embodiment of the present application provides a computer program product including a computer program, which, when executed on a computer device, enables the computer device to execute the method described in the above aspects.

[0020] It can be seen from the above technical solution that for the N reaction frames that complete the chemical simulation reaction, a molecular graph is used to express the molecules involved in the multi-molecule system in each reaction frame. The specific expression is: the first node is used to express the atoms in the molecular graph, and the line between the two first nodes is used to express the bond relationship between the two atoms. The molecular graph includes one or more subgraphs, and each individual subgraph is used to identify a molecule, that is, the independent expression of molecules can be achieved in the molecular graph. In order to accurately identify the reaction events generated in the chemical simulation reaction, for the i-th reaction frame including the product, the subgraph in the molecular graph of the reaction frame and the subgraph of the first molecular graph constitute the reaction graph of the reaction frame. The reaction graph can accurately express the molecular changes from the first reaction frame to the i-th reaction frame from the molecular level. In the reaction graph, the second node is used to identify the reactants, and the third node is used to identify the products. The line between the second node and the third node identifies the candidate reactions to be identified between the molecules corresponding to the second node and the third node. Graph isomorphism can accurately determine whether changes have occurred between the reactant corresponding to the second node and the product corresponding to the third node, thereby determining whether a reaction event has actually occurred between the reactants and products associated with the candidate reaction. By subdividing the molecular graph into subgraphs to identify independent molecules and using graph isomorphism to identify the occurrence of reaction events between the molecules corresponding to reactants and products in the reaction graph, the reaction events generated in each reaction frame of a completed chemical simulation reaction can be quickly and automatically identified without the complex computational costs of related technologies, effectively improving the efficiency of reaction event identification in chemical simulation reactions. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0022] Figure 1 A schematic diagram of a reaction recognition scenario provided in an embodiment of the present application;

[0023] Figure 2 A flow chart of a reaction identification method provided in an embodiment of the present application;

[0024] Figure 3 A schematic diagram of setting an initial index provided in an embodiment of the present application;

[0025] Figure 4 A schematic diagram of subgraph division provided in an embodiment of the present application;

[0026] Figure 5 A schematic diagram of a reaction diagram provided in an embodiment of the present application;

[0027] Figure 6 A flowchart of a reaction event determination provided in an embodiment of the present application;

[0028] Figure 7 A flowchart of another reaction event determination provided in an embodiment of the present application;

[0029] Figure 8 A schematic diagram of a reaction event provided in an embodiment of the present application;

[0030] Figure 9 A schematic diagram of another reaction event occurrence provided in an embodiment of the present application;

[0031] Figure 10 A schematic diagram of chemical reaction impact statistics provided in an embodiment of the present application;

[0032] Figure 11 A schematic diagram of a reaction identification device provided in an embodiment of the present application;

[0033] Figure 12 A structural diagram of a terminal device provided in an embodiment of the present application;

[0034] Figure 13 A structural diagram of a server provided in an embodiment of the present application. DETAILED DESCRIPTION

[0035] The embodiments of the present application are described below with reference to the accompanying drawings.

[0036] In the field of chemistry, due to research needs, it is often necessary to analyze and study real-world chemical reaction mechanisms and reaction events based on chemical simulation reactions. Therefore, chemical simulation reactions can simulate complex chemical reaction processes. By separating and identifying the chemical reactions involved in chemical simulation reactions to obtain reaction events, researchers can understand the chemical reaction mechanisms involved in the chemical reaction process based on the identified reaction events.

[0037] In related technologies, existing models are used to identify chemical simulation reactions, such as ChemTraYzer (a software package for processing and analyzing chemical processes). However, these technologies rely on information provided by other software. Furthermore, due to the limitations of these technologies themselves, the identification of chemical simulation reactions involves complex calculations and suffers from poor accuracy, making it impossible to efficiently and accurately separate and identify chemical reactions within a chemical reaction process.

[0038] To this end, an embodiment of the present application provides a reaction identification method and related devices, which presents a chemical simulation reaction in the form of a reaction frame and independently divides the molecules in the molecular graph corresponding to the reaction frame to realize the identification of reaction events between the molecules corresponding to the reactants and products in the reaction graph based on graph isomorphism. The reaction events generated in each reaction frame of the completed chemical simulation reaction can be quickly and automatically identified without the complex computational cost in the related technology, thereby effectively improving the reaction event identification efficiency of the chemical simulation reaction.

[0039] The reaction recognition method provided in the embodiments of the present application can be implemented by a computer device, which can be a terminal device or a server, wherein the server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services. Terminal devices include but are not limited to mobile phones, computers, intelligent voice interaction devices, smart home appliances, vehicle-mounted terminals, aircraft, etc. The terminal device and the server can be directly or indirectly connected via wired or wireless communication, and this application does not impose any restrictions on this.

[0040] First, several noun terms that may be involved in the embodiments below in this application are explained.

[0041] 3T (Tiered tensor transform): An efficient multiscale structural energy minimization algorithm that can be used to optimize trajectories. In the chemical simulation reactions mentioned in the examples of this application, when the goal is to optimize the trajectory, an energy minimization algorithm is needed to induce structural changes in the molecules in the multi-molecule system. In this case, the 3T algorithm can be used.

[0042] Energy minimization: Adjusting the three-dimensional coordinates of atoms so that the total energy of the entire multi-molecule system reaches a local minimum. Energy evaluation can be done using various methods, including classical force fields.

[0043] Classical force fields: commonly used to control the interactions between atoms in atomistic molecular dynamics simulations.

[0044] Undirected graph: In graph theory, an undirected graph is a type of graph in which the edges have no direction. It consists of a set of vertices (also called nodes) and edges connecting these vertices. In an undirected graph, edge (A, B) is equivalent to edge (B, A), indicating that there is a line between vertex A and vertex B. The molecular graphs and reaction graphs in the embodiments of the present application are both undirected graphs. In the molecular graphs and reaction graphs, vertices are nodes used to identify atoms (such as the first node, the second node, etc.), and edges are lines used to identify the bond relationships between atoms.

[0045] Connected components of a graph: In graph theory, connectivity is a fundamental property of a graph. If every pair of vertices in an undirected graph is connected by a path, the graph is said to be connected. A graph that is not fully connected can be decomposed into several connected components. In the embodiments of the present application, the aforementioned graph that is not fully connected refers to a molecular graph, and the connected components decomposed into it refer to subgraphs obtained by dividing the molecular graph, and the subgraphs are used to identify independent molecules.

[0046] Graph isomorphism: used to describe the structural similarity between two subgraphs in the embodiments of the present application. If two subgraphs are isomorphic, then in the two subgraphs, the vertices (i.e., nodes) and edges (i.e., lines) in one graph can completely correspond to the vertices and edges in the other graph, keeping the structure of the graph unchanged. The properties of graph isomorphism include: the same number of vertices (i.e., nodes), the same number of edges (i.e., lines), and the same degree of vertices (i.e., the same number of lines each node has).

[0047] Molecular graph: A molecular representation of a compound using graph theory. A molecular graph is a labeled graph. Generally, vertices (i.e., first nodes) correspond to atoms in the compound. They can be labeled with the corresponding atomic type or other numbers. Edges in a molecular graph correspond to the bonding relationships (i.e., chemical bonds) between atoms, and are represented by connecting lines in the examples of this application.

[0048] Bond order: In chemistry, the bond order (also known as bond sequence) of a chemical bond refers to the number of electron pairs shared between two atoms. Bond order is a measure of the strength of a chemical bond and is related to the length and energy of the bond. Common chemical bond orders are single bonds as primary, double bonds as secondary, and triple bonds as tertiary. These different bond orders indicate the number of electron pairs shared between atoms, reflecting the strength and length of the bond. Single bonds are longer and weaker than double bonds, while triple bonds are shorter and stronger.

[0049] SMILES (Simplified molecular-input line-entry system): A specification that uses ASCII strings to clearly describe molecular structures. In the examples of this application, SMILES can be used to determine equivalent reactions.

[0050] Molecular dynamics: A computational simulation method used to study the motion of atoms and molecules in time and space. Molecular dynamics simulations can be used to simulate and analyze the interactions between atoms and molecules.

[0051] Artificial Intelligence (AI) refers to the theories, methods, techniques, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, to perceive the environment, acquire knowledge, and use that knowledge to achieve optimal results. In other words, AI is a comprehensive technology within computer science that seeks to understand the essence of intelligence and produce new intelligent machines that can respond in a manner similar to human intelligence. AI also studies the design principles and implementation methods of various intelligent machines, enabling them to possess the capabilities of perception, reasoning, and decision-making.

[0052] Artificial intelligence (AI) technology is a comprehensive discipline encompassing a wide range of fields, encompassing both hardware and software technologies. Foundational AI technologies generally include sensors, specialized AI chips, cloud computing, distributed storage, big data processing, operating / interaction systems, and mechatronics. AI software technologies primarily encompass computer vision, speech processing, natural language processing, and machine learning / deep learning.

[0053] Machine learning (ML) is a multidisciplinary field that encompasses probability theory, statistics, approximation theory, convex analysis, and algorithmic complexity theory. It specifically studies how computers can simulate or implement human learning behaviors to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their performance. Machine learning is at the core of artificial intelligence and the fundamental way to make computers intelligent. Its applications span all areas of AI. Machine learning and deep learning typically include techniques such as artificial neural networks, belief networks, reinforcement learning, transfer learning, inductive learning, and self-learning.

[0054] With the research and advancement of artificial intelligence technology, artificial intelligence technology has been studied and applied in many fields, such as common smart homes, smart wearable devices, virtual assistants, smart speakers, smart marketing, unmanned driving, autonomous driving, drones, robots, smart medical care, smart customer service, etc. It is believed that with the development of technology, artificial intelligence technology will be applied in more fields and play an increasingly important role.

[0055] The solutions provided in the embodiments of this application involve cloud computing and big data processing technologies for artificial intelligence. For example, when performing reaction identification for a chemical simulation reaction, a molecular graph corresponding to a multi-molecule system is obtained through processing. The solutions provided in the embodiments of this application require identifying graph isomorphisms between subgraphs corresponding to reactants and products of candidate reactions. This is specifically illustrated by the following examples:

[0056] Figure 1 A schematic diagram of a reaction recognition scenario provided in an embodiment of the present application, wherein the aforementioned computer device is a server.

[0057] During the reaction recognition process, the server will obtain N molecular graphs in the chemical simulation reaction. Each molecular graph has a corresponding reaction frame. Figure 1 As shown, a molecular graph includes first nodes for identifying atoms, and a line between two first nodes is used to identify the bond relationship between two atoms. At the same time, a molecular graph may include one or more subgraphs, each of which is used to identify an independent molecule.

[0058] Below Figure 1 Molecules in Figure 1 Take the example to explain in detail, in the molecule Figure 1The graph includes two subgraphs, subgraph A and subgraph B. Subgraph A is used to identify molecule A, and subgraph B is used to identify molecule B. Molecule A includes six atoms (i.e., the first nodes): atom 1, atom 2, atom 3, atom 4, atom 5, and atom 6. Molecule B includes four atoms (i.e., the first nodes): atom 7, atom 8, atom 9, and atom 10. As can be seen from the figure, a single atom can establish one or more bond relationships with other atoms (in actual applications, there are cases where molecules are composed of a single atom, so a single atom can also be a molecule without establishing bond relationships with other atoms), and these are represented by lines.

[0059] The corresponding relationship between the molecular graph and the reaction frame is also shown in the figure, where the molecular Figure 1 Corresponding reaction frame 1, molecule Figure 2 Corresponding to reaction frame 2...Molecular graph i corresponds to reaction frame i. Generally, in the first molecular graph (i.e., molecular Figure 1 ) determines the reactants of the chemical simulation reaction, and determines the products of the chemical simulation reaction based on the other molecular graphs except the first molecular graph. Then, a reaction graph for the corresponding reaction frame is generated based on the reactants and products. The figure shows the reaction graph of reaction frame 2 (i.e., the second reaction frame). The reaction graph of reaction frame 2 is generated based on the first molecular graph and the second reaction graph.

[0060] The reaction graph includes a second node identifying the reactant and a third node identifying the product, and the line between the second and third nodes is used to identify the candidate reaction to be identified. Then, based on the graph isomorphism between the subgraphs with the target second node and the target third node of the candidate reaction, the reaction event that occurs from the reactant corresponding to the target second node to the product corresponding to the target third node is determined. For example, with molecule A as the target second node and molecule E as the target third node in the reaction graph, the reaction event that occurs from the reactant corresponding to molecule A to the product corresponding to molecule E can be determined based on the graph isomorphism between molecules A and E.

[0061] Independent expression of molecules can be achieved through subgraphs in the molecular graph. In order to identify reaction events generated in chemical simulation reactions, a reaction graph is constructed for the reaction frame that has a corresponding relationship with the molecular graph. Based on the reaction graph, the molecular changes from the first reaction frame to the reaction frame are accurately expressed at the molecular level. Finally, the graph isomorphism can be used to determine the reaction events that occur from the reactants of the first reaction frame to the products of the reaction frame. This process does not require the complex computational costs used in related technologies, effectively improving the efficiency of reaction event identification in chemical simulation reactions.

[0062] It should be noted that, for the sake of convenience, Figure 1 The molecular diagram and reaction diagram shown in are both in two-dimensional form, but in fact both the molecular diagram and reaction diagram are in three-dimensional form.

[0063] Figure 2 This is a flowchart of a reaction identification method provided in an embodiment of the present application. The method can be executed by a computer device. In this embodiment, the computer device is described as a server.

[0064] The method comprises:

[0065] S201: Obtain N molecular graphs of a multi-molecule system in a chemical simulation reaction, where N>1.

[0066] Chemical simulation reaction refers to the process of utilizing computer program or software to simulate chemical reaction, through the virtual reaction environment of construction to predict the product of reaction event or reaction conditions and other information. Chemical simulation reaction can facilitate understanding and predict the behavior of actual reaction event. In the embodiment of the present application, it is necessary to identify the reaction event involved in the process of chemical simulation reaction. In chemical simulation reaction, according to the complexity of the reaction, the occurrence of the reaction event involved is also different, and one or more reaction events can be involved in a chemical simulation reaction. Multi-molecule system can be understood as the system composed of all substances (including reactants and products) participating in chemical simulation reaction in the virtual reaction environment. In the present application, multi-molecule system is expressed by molecular graph, and molecular graph includes subgraph, and subgraph is used to identify molecule, including atom in molecule.

[0067] The chemical simulation reaction is completed via N reaction frames, and the molecular graph corresponds one-to-one to the reaction frames. Reaction frames are used to describe changes during a chemical simulation reaction. Changes here can refer to changes in the bonding relationships between atoms in a single molecule. When a new chemical bond forms between two atoms or an old chemical bond breaks, it means that the bonding relationship between the two atoms has changed. In other words, the bonding relationships between atoms involved in different reaction frames vary. Reaction frames are divided based on changes and are arranged in time sequence. Specific changes can include changes in bonding relationships, changes in molecular positions, and changes in atomic positions. A reaction frame can be understood as the coordinate structure of any independently stored multi-molecular system. The collection of reaction frames arranged in time sequence can be called a trajectory, which can reflect the changing process of reactants in a chemical simulation reaction. Trajectories can include molecular dynamics trajectories and optimization trajectories. A molecular dynamics trajectory refers to the trajectory that describes the time-dependent evolution of molecules in a multi-molecular system during a molecular dynamics simulation. An optimization trajectory refers to the trajectory that uses optimization methods to find the stable configuration or lowest energy state of a multi-molecular system. The optimization trajectory can be generated based on the 3T algorithm. As an efficient energy minimization algorithm, the 3T algorithm can realize rapid search for products in complex multi-molecular systems.

[0068] The molecular graph identifies atoms and the bonding relationships between atoms. Therefore, by means of the molecular graph of the chemical simulation reaction, the reactants and products involved in the chemical simulation reaction process can be grasped. A molecular graph may include one or more subgraphs. The molecular graph includes first nodes that identify atoms, and the connection lines between two first nodes are used to identify that there is a bonding relationship between the atoms corresponding to the two first nodes. The i-th molecular graph is used to identify the bonding relationships between atoms in the multi-molecule system at the i-th reaction frame of the chemical simulation reaction, where N > 1. The connection line identifies the bonding relationship between atoms and has no directionality itself.

[0069] S202: Determine the reactants of the chemical simulation reaction according to the subgraph of the first molecular graph, determine the products of the chemical simulation reaction according to the subgraph of the i-th molecular graph, and generate a reaction graph of the chemical simulation reaction at the i-th reaction frame based on the reactants and the products.

[0070] As mentioned above, the division of reaction frames is based on changes and is arranged in chronological order. There is a one-to-one correspondence between the molecular graph and the reaction frame, which means that there is a chronological relationship between different molecular graphs. Generally, the first molecular graph can be used as the starting molecular graph. Then, the reactants of the chemical simulation reaction can be determined from the subgraph of the first molecular graph, and the products of the chemical simulation reaction can be determined according to the subgraphs of other molecular graphs except the first molecular graph. A reaction graph of the chemical simulation reaction at the reaction frame corresponding to the other molecular graph is generated based on the determined reactants and products. In the embodiments of the present application, by comparing the changes in the products in each reaction frame relative to the reactants in the first reaction frame, the reaction events occurring in different reaction frames are determined.

[0071] The reaction graph includes second nodes that identify reactants and third nodes that identify products. The connection lines between the second nodes and the third nodes are used to identify that there is a candidate reaction to be recognized between the molecules corresponding to the second nodes and the third nodes, where 1 < i ≤ N. When the bonding relationship of the reactants changes, it means that new products are generated based on the reactants, that is, a chemical reaction (i.e., the candidate reaction to be recognized) occurs between the reactants and the products. The method for determining the products corresponding to the reactants can be: determining whether the subgraph of the i-th molecular graph contains atoms consistent with those in the subgraph of the first molecular graph.

[0072] To facilitate judgment, an initial index can be set for each atom in the reactants. The initial index corresponding to the atom does not change during the chemical simulation reaction. Based on this initial index, it can be determined whether the subgraph of the i-th molecular graph contains an atom with the same initial index as the subgraph of the first molecular graph. If so, the molecules with the same initial index correspond to a set of reactants and products. During the chemical simulation reaction, a product can be obtained from multiple reactants, and multiple products can be obtained from a single reactant. In other words, a set of reactants and products with the same initial index can include more than two molecules.

[0073] Figure 3 A schematic diagram of setting an initial index provided in an embodiment of the present application is provided. Figure 3 As shown in the figure, the subgraph corresponding to CH4 is shown, in which an initial index is set for each atom, the initial index corresponding to the carbon atom is 1, and the initial indexes corresponding to the four hydrogen atoms are 2, 3, 4 and 5 respectively. It can be seen that when setting the initial index in the embodiment of the present application, it is not based on the type of atom, but a corresponding initial index is set for each independent atom.

[0074] S203: For the target second node and the target third node having the candidate reaction, determine the reaction event of obtaining the product corresponding to the target third node from the reactant corresponding to the target second node through the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node.

[0075] A reaction graph can include multiple second and third nodes. A reactant can be identified as the target second node, and the product corresponding to that reactant can be identified as the target third node. By determining the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node, the reaction event that produces the product corresponding to the target third node from the reactant corresponding to the target second node can be determined.

[0076] Graph isomorphism is used to determine the structural similarity between two subgraphs. When two subgraphs are determined to be in an isomorphic relationship, it means that the number of nodes and the number of bond relationships between the molecules corresponding to the two subgraphs are exactly the same, that is, the two molecular structures are exactly the same. Graph isomorphism measures the degree of similarity between two subgraphs based on the aforementioned properties of graph isomorphism. In the measurement process, it is necessary to consider both the nodes and the connections between the nodes. By measuring the graph isomorphism between the subgraphs corresponding to the target second node and the target third node, that is, the nodes and the connections between the subgraphs corresponding to the target second node and the target third node, the reaction events from the target second node and the target third node can be determined. Specifically, reaction events can include the following three types: chemical reaction events, ionization reaction events, and no-reaction events.

[0077] It can be seen from the above technical solution that for the N reaction frames that complete the chemical simulation reaction, a molecular graph is used to express the molecules involved in the multi-molecule system in each reaction frame. The specific expression is: the first node is used to express the atoms in the molecular graph, and the line between the two first nodes is used to express the bond relationship between the two atoms. The molecular graph includes one or more subgraphs, and each individual subgraph is used to identify a molecule, that is, the independent expression of molecules can be achieved in the molecular graph. In order to accurately identify the reaction events generated in the chemical simulation reaction, for the i-th reaction frame including the product, the subgraph in the molecular graph of the reaction frame and the subgraph of the first molecular graph constitute the reaction graph of the reaction frame. The reaction graph can accurately express the molecular changes from the first reaction frame to the i-th reaction frame from the molecular level. In the reaction graph, the second node is used to identify the reactants, and the third node is used to identify the products. The line between the second node and the third node identifies the candidate reactions to be identified between the molecules corresponding to the second node and the third node. Graph isomorphism can accurately determine whether changes have occurred between the reactant corresponding to the second node and the product corresponding to the third node, thereby determining whether a reaction event has actually occurred between the reactants and products associated with the candidate reaction. By subdividing the molecular graph into subgraphs to identify independent molecules and using graph isomorphism to identify the occurrence of reaction events between the molecules corresponding to reactants and products in the reaction graph, the reaction events generated in each reaction frame of a completed chemical simulation reaction can be quickly and automatically identified without the complex computational costs of related technologies, effectively improving the efficiency of reaction event identification in chemical simulation reactions.

[0078] In the aforementioned S201, it is mentioned that “N molecular graphs of a multi-molecule system in a chemical simulation reaction are obtained”. Specifically, the methods of obtaining molecular graphs may include the following two: ① obtaining molecular graphs with the help of additional software costs; ② obtaining molecular graphs based on the simplest trajectory file. Compared with the first method, the method of obtaining molecular graphs based on the simplest trajectory file is more convenient and efficient. The second method of obtaining molecular graphs is described in detail below. In one possible implementation method, the method includes: first obtaining a trajectory file for describing the changes in atomic structure in a chemical simulation reaction, and then generating N molecular graphs of a multi-molecule system in a chemical simulation reaction based on the trajectory file.

[0079] The trajectory file is used to identify the atoms involved in the multi-molecule system and their three-dimensional positions at different reaction frames. As previously mentioned, a collection of reaction frames arranged in time sequence can be called a trajectory, and a trajectory file can be understood as a collection of the corresponding reaction frames in the multi-molecule system. The structure of each reaction frame in the trajectory file is derived from the structural evolution of the molecules involved in the multi-molecule system, that is, the changes in the atomic structure.

[0080] In the first method of obtaining the molecular graph mentioned above, it is necessary to use an additional software algorithm to obtain the bonding relationship between atoms in the reaction frame. Based on the acquired bonding relationship, the graph theory breadth-first search (BFS) algorithm is used to extract molecules through atomic connectivity to obtain the molecular graph. That is, it is necessary to obtain additional bond-level information to determine the bonding relationship between atoms.

[0081] Compared to the acquisition mode of the first molecular graph, in the embodiment of the present application, the acquisition of molecular graph is realized based on the simplest trajectory file, the participation of extra software algorithm is avoided, only by the atom identified in the trajectory file and the three-dimensional coordinate corresponding to the atom, the generation of molecular graph can be realized. The trajectory file in the embodiment of the present application can be in xyz format, and all includes the coordinate information relating to atom in each reaction frame of the trajectory file under this format, and the trajectory file is generally composed of the coordinate (x, y, z) of the three-dimensional position of atomic number and each atom. Based on the trajectory file, the acquisition of molecular graph can be realized, although the judgment of bond relationship is also related to in the process of acquisition, it is not necessary to utilize extra algorithm to carry out, but determine the atomic distance between each atom according to the coordinate of the three-dimensional position of atom, directly carry out the comparison with chemical bonding distance.

[0082] Through the above-mentioned method for obtaining a molecular graph, it is possible to obtain a molecular graph based on the simplest trajectory file. During the process, only atoms and their three-dimensional position information need to be used. The acquisition method is simple and does not require the help of other algorithms. It can save computing power while improving the efficiency of obtaining molecular graphs.

[0083] Since the trajectory file only contains atoms and their three-dimensional positions, it is necessary to establish bonding relationships for atoms during the generation of the molecular graph so that the atoms form molecules based on the bonding relationships. Of course, there is also a more special case, that is, the molecule is composed of a single atom. In this case, the atom will not establish a bonding relationship with other atoms. The following is an explanation of the case where the molecule is composed of multiple atoms. For the first and second atoms in the i-th reaction frame, the molecular graph generation process mentioned above for "generating N molecular graphs of a multi-molecule system in a chemical simulation reaction based on the trajectory file" is:

[0084] A1: Determine a chemical bonding distance between the first atom and the second atom according to the atomic types of the first atom and the second atom.

[0085] The chemical bonding distance refers to the average distance between two atoms when a chemical bond (i.e., a bonding relationship) is formed. There are many types of bonding relationships between atoms, such as covalent bonds, ionic bonds, and metallic bonds. The chemical bonding distance is correlated with the atom type, and the chemical bonding distance depends on the type of atom. Different atom types correspond to different chemical bonding distances. For example, assuming that there are atoms A, B, and C, and the three atoms correspond to different atomic types, the chemical bonding distance between atom A and atom B is approximately 0.11 to 0.13 nanometers, the chemical bonding distance between atom A and atom C is approximately 0.115 to 0.145 nanometers, and the chemical bonding distance between atom B and atom C is approximately 0.16 to 0.18 nanometers.

[0086] To determine whether a first atom and a second atom are bonded, we need to compare the atomic distance between the first and second atoms to see if it satisfies the chemical bonding distance between the first and second atoms. If the atomic distance satisfies the chemical bonding distance, it indicates that a bond exists between the first and second atoms; if the atomic distance does not meet the chemical bonding distance, it indicates that no bond exists between the first and second atoms.

[0087] A2: Determine the atomic distance between the first atom and the second atom according to the three-dimensional positions of the first atom and the second atom.

[0088] Atomic distance refers to the distance between two atoms in a multi-molecular system, for example, the distance between the first and second atoms. Different types of bonding relationships form with different atomic distances. Generally speaking, covalent bonds typically have shorter atomic distances, while ionic and metallic bonds have longer atomic distances.

[0089] In chemical simulations, new bonds are created and old bonds are broken, resulting in the formation of new molecules. The atomic distances between the atoms that make up the molecules change over the course of the reaction, so the same two atoms may have different atomic distances in different reaction frames.

[0090] In the embodiments of the present application, a molecular graph is generated using a trajectory file, which includes the three-dimensional positions of atoms and atoms in different reaction frames. In a single reaction frame, the atomic distance between a first atom and a second atom can be calculated using the three-dimensional position of the first atom and the three-dimensional position of the second atom.

[0091] A3: In response to the atomic distance being consistent with the chemical bonding distance, determining that the first atom and the second atom have a bonding relationship.

[0092] When atomic distances meet the chemical bonding distance, it means the distance between the first and second atoms satisfies the conditions for a chemical bond (i.e., a bonding relationship) between the two atoms. Atomic distances meeting the chemical bonding distance can generally be understood as the distance between the first and second atoms being within the chemical bonding distance range.

[0093] For example, assume that atoms A and B exist, and the chemical bonding distance between atoms A and B is approximately 0.11 to 0.13 nanometers. In the second reaction frame, the atomic distance between atoms A and B is 0.17 nanometers. This atomic distance does not meet the chemical bonding distance, so it is determined that there is no bond between atoms A and B. In the fifth reaction frame, the atomic distance between atoms A and B is 0.12 nanometers. This atomic distance meets the chemical bonding distance, so it is determined that there is a bond between atoms A and B.

[0094] A4: In the i-th molecule graph, add a line between the first atom and the second atom.

[0095] There is a corresponding relationship between the reaction frame and the molecular graph. For the i-th reaction frame, there is an i-th molecular graph corresponding to it. By determining the bonding relationship of atoms in the reaction frame, connecting lines are added between atoms in the i-th molecular graph to identify the bonding relationship between atoms.

[0096] When adding lines, different line forms can be determined based on the different bonding relationships to distinguish different bonding relationships. The same line form can also be used to schematically identify the bonding relationships between atoms.

[0097] By adding a line between the first atom and the second atom according to the determined bonding relationship in the molecular graph, the originally isolated first atom and the second atom are connected by the line, and the first atom and the second atom with the line will constitute an independent molecule.

[0098] Through the molecular graph generation process provided above, the chemical bonding distance of the atoms is first determined according to the atom type, and then the atomic distance between the atoms is determined based on the three-dimensional position provided in the trajectory file. Finally, the consistency between the atomic distance and the chemical bonding distance is judged to determine the bonding relationship between the atoms, and the construction of atoms to molecules is realized in the molecular graph, so that independent expression of molecules is realized in the molecular graph. The construction process is simple and does not require the help of other software. The bonding relationship between atoms can be determined only by judging the consistency between the atomic distance and the chemical bonding distance, which can achieve efficient molecular graph generation.

[0099] After completing the addition of the bond relationship between atoms, since the aforementioned method is only based on the relationship between the atomic distance and the chemical bonding distance between atoms when determining the bond relationship, the conditions are relatively simple, and the bond relationship determined may not conform to the laws of nature. Therefore, it is necessary to judge the rationality of the added bond relationship, and the judgment process needs to be carried out based on the charge of the atom itself. The accuracy of the formed molecular graph can be guaranteed to a certain extent by checking the bond relationship. Taking the aforementioned first atom as an example, after completing the determination of all the bond relationships of the first atom, it is necessary to check the bond relationship. In one possible implementation, the method of checking is specifically as follows:

[0100] B1: Determine a charge range of the first atom according to the atomic type of the first atom.

[0101] Different types of atoms have different charge ranges. An atom's charge can be negative, positive, or zero. The specific charge depends on the difference between the atom's nuclear charge and the number of its outer electrons. Therefore, an atom's charge is not a unique, fixed value, but rather a range, or charge range, that includes multiple charge ranges. The charge range of the first atom is associated with the total number of bond orders corresponding to the first atom's bonding relationship. If the initial bonding relationship is assumed to correspond to a single bond, and the number of atoms that can form a bonding relationship is less than the total number of bond orders allowed, this indicates that other types of bonding relationships exist around the atom or that the atom's charge does not match.

[0102] The range of atomic charge is associated with the total number of bond orders possessed by the atom (i.e., the number of all bond orders). The bond relationship is used to indicate that there is a bond relationship between the first atom and other atoms, and the bond order can be understood as a unit for describing the number of bonds formed between the first atom and other atoms based on the bond relationship (or possessed by the first atom itself). Different types of bond relationships correspond to different numbers of bond orders. For example, the types of bond relationships may include: single bonds, double bonds, and triple bonds. Among them, the number of bond orders corresponding to a single bond is 1, the number of bond orders corresponding to a double bond is 2, and the number of bond orders corresponding to a triple bond is 3. When the default bond relationship corresponds to a single bond in the initial state, assuming that the number of bond relationships possessed by the first atom is 4, the total number of bond orders corresponding to the first atom is 4; if the default bond relationship corresponds to a double bond in the initial state, assuming that the number of bond relationships possessed by the first atom is 4, the total number of bond orders corresponding to the first atom is 8.

[0103] Once the atomic type of the first atom is determined, the charge range of the first atom can be determined accordingly. The charge range of the first atom will affect the formation of bonding relationships between the first atom and other atoms. Different charges of the first atom correspond to different bonding relationships formed with other atoms. Based on the charge range of the first atom, the range of the number of bonding relationships that the first atom can form can be determined.

[0104] B2: Using the charge range as a bond relationship verification condition, perform a bond order quantity verification on all bond relationships of the first atom.

[0105] After determining all bond relationships of the first atom, the number of bond orders (i.e., the total number of bond orders) for all bond relationships of the first atom can be verified based on the charge range of the first atom. During the verification process, the range of the total number of bond orders that can be formed by the first atom within the charge range can be determined in combination with the charge range of the first atom.

[0106] For example, assume that the charge range of atom A indicates that under normal circumstances (which can be understood as when the charge of atom A is 0), the number of all bond levels that can be formed is 3. If the charge range corresponding to atom A is -1 to +1, then at this time, corresponding to this charge range, the number of all bond levels that atom A can form ranges from 2 to 4. In other words, when the charge range corresponding to atom A is -1, 0, and 1, respectively, the number of all bond levels corresponding to the bond relationship formed by atom A can be 2, 3, or 4. When the number of all bond levels corresponding to the bond relationship of atom A is verified, as long as the number can match its charge, it is proved that the number of all bond relationships of atom A meets the bond relationship verification conditions.

[0107] B3: In response to the number of all bond orders of the first atom satisfying the bond relationship verification condition, determining to retain the connection with the first atom.

[0108] When the total number of bond orders corresponding to the bonding relationship of the first atom meets the bonding relationship verification conditions within the aforementioned charge range, the original connection lines between the first atom and other atoms are retained, that is, the original bonding relationship between the first atom and other atoms is retained.

[0109] B4: In response to the number of all bond orders of the first atom not satisfying the bond relationship verification condition, redetermining the type of the bond relationship between the first atom and other atoms according to a cause leading to the non-satisfaction.

[0110] When the total number of bonding relationships of the first atom does not meet the aforementioned bonding relationship verification conditions, the types of bonding relationships between the first atom and other atoms or the charges of the related atoms may be adjusted based on the reasons for the non-compliance. Specifically, the reasons for the non-compliance may include: ① the bonding relationship type corresponding to the bonding relationship of the first atom does not match the total number of bond orders it should have; ② the bonding relationship of the first atom is inconsistent with the charge and bonding configuration of the first atom.

[0111] When the first reason is not satisfied, it means that the total number of bond levels formed by the bonding relationships between the first atom and other atoms does not match the total number of bond levels that the first atom itself can form. For example, assuming that the total number of bond levels that the first atom can form is 4 based on the range of charge, then if the first atom's bonding relationships are determined to be less than 4 or more than 4, it proves that the total number of bond levels formed corresponding to the type of bonding relationships around the first atom (i.e., the total number of bond levels) does not match the total number of bond levels that the first atom should have.

[0112] When the default bonding relationship is single-bonded, if there are fewer than four bonding relationships to the first atom, this means that fewer than four atoms in the trajectory file have atomic distances with the first atom that meet the chemical bonding distance requirement. In this case, you can consider adjusting the types of some bonding relationships to ensure that the total number of bond orders meets the requirement. For example, suppose all possible chemical bonding distances to the first atom are approximately 0.11-0.15 nm. However, in a reaction frame in the trajectory file, only three atoms have atomic distances with the first atom between 0.11-0.15 nm. This means that the total number of bond orders between the first atom and other atoms is now three. In this case, you can consider adjusting the bonding relationship type, such as changing one of the bonding relationships from a single bond to a double bond based on the reasonableness of the chemical bonding distance. This can increase the total number of bond orders without changing the bonding relationships.

[0113] If the first atom has more than four bonds, meaning that the number of atoms in the trajectory file with atomic distances that meet the chemical bonding distance requirement is greater than four, consider adjusting some of the bond types to smaller bond orders, such as downgrading triple bonds to double bonds or breaking bonds with unreasonable bonding angles to reduce the total number of bond orders to within the required range. For example, suppose the chemical bonding distance of the first atom is 0.11-0.15 nanometers, but in a reaction frame in the trajectory file, there are five atoms with atomic distances that meet the chemical bonding distance requirement. This means that the total number of bond orders between the first atom and other atoms is now five. Breaking the bonds with unreasonable bonding angles can reduce the total number of bond orders and remove atoms with unreasonable bonding directions from the first atom. The repulsive forces between electrons on different bonds around an atom create a regular spatial distribution of bonds around the atom, known as the bonding configuration. Bonding directions that do not conform to the typical bonding configuration around the atom warrant special attention. However, the bonding configuration often requires first determining which atoms are connected to it, so it needs to be verified and adjusted during the inspection stage.

[0114] That is, in one possible implementation, during the verification process, the number of all bond levels that the first atom itself can form can be first determined based on the indication of the charge range of the first atom. When it is determined that the number of all bond levels of the first atom does not meet the verification conditions of the bond relationship, it can be considered to determine whether there is an unreasonable bonding direction in the bond relationship formed between the first atom and other atoms based on the bonding configuration of the first atom, and then redetermine the bond relationship between the first atom and other atoms under normal circumstances.

[0115] It should be noted that the bonding configuration of an atom is used to describe the bonding relationship between a certain type of atom and other atoms, that is, the spatial arrangement of the atoms that form a molecule (including the bonding direction or angle between the atoms). The bonding configuration can generally be determined according to chemical specifications, and is used to determine whether the bonding relationship formed between the first atom and the other atoms conforms to the chemical specifications. When the number of all bond levels of the first atom is verified, if the number of all bond levels does not match the number of all bond levels that the first atom should have, the bonding relationship between the first atom and the other atoms under normal circumstances can also be re-determined based on the bonding configuration, such as disconnecting the bonds corresponding to the bonding relationships whose bonding angles do not conform to the chemical specifications.

[0116] If the second condition is not met, it means that the directions of the bond orders between the first atom and each of the atoms with which it is bonded are inconsistent with the bonding configuration of the first atom and the total number of bond orders that the first atom can normally form. In this case, it can be further determined whether the total number of bond orders surrounding the atom falls within the range of the total number of bond orders corresponding to the bond orders determined based on the bonding configuration and charge range. If it does not fall within this range, the total number of bond orders of the first atom does not meet the bond verification condition.

[0117] For example, assuming that the charge of the first atom ranges from +2 to -2, according to the bonding configuration of the first atom, and in the neutral state, the total number of bond levels possessed by the first atom is 4. If, after determining all the bonding relationships and types of bonding relationships of the first atom, it is found that the total number of bond levels possessed by the first atom is only 3, it proves that the charge of the first atom is inconsistent with the bonding configuration of the first atom. At this time, it is necessary to further judge based on the charge range. When the charge of the first atom can range from -2 to +2, the corresponding bonding configuration of the first atom corresponds to the total number of bond levels that can be formed in the range of 2 to 6. At this time, the charge of the first atom needs to be adjusted. Assuming that the original charge of the first atom is 0, its charge can be adjusted to -1 so that its charge meets the chemical specifications.

[0118] Every time the type of bonding relationship or the amount of charge of the first atom changes, it is necessary to iteratively check the other atoms connected to it synchronously to ensure that the bonding conditions around other atoms in the molecule after the change also meet the chemical specifications.

[0119] By determining the charge range corresponding to the atomic type of the first atom as a bond verification condition, the number of all bond relationships for the first atom is verified. This ensures the accuracy of the bond relationships of each atom in the resulting molecular graph, and also ensures the accuracy of the molecules in the subgraph formed based on the bond relationships. Furthermore, if the number of bond relationships consistently fails to meet the aforementioned bond verification conditions, the bond relationships of the first atom can be directly re-determined by scaling the chemical bonding distances. This simple adjustment method helps improve the efficiency of molecular graph generation.

[0120] As mentioned above, the bonding relationship between atoms is determined based on the atomic distance and the chemical bonding distance. After the bonding relationship is determined, the charge range corresponding to the atoms is used as the bonding relationship verification condition, and the bonding relationship of the determined atoms is quantitatively verified to ensure the accuracy of the bonding relationship of the atoms. Then, the molecular graph will include a large number of sets of atoms with bonding relationships. The atoms here are the aforementioned first nodes, and the bonding relationship of the atoms is the connection between the aforementioned first nodes. At this time, it is necessary to divide the molecular graph into subgraphs. In one possible implementation method, the subgraph division method is: first, the graph connectivity is calculated based on the connection between the first nodes in the molecular graph. Then, based on the results of the graph connectivity calculation, the molecular graph is divided into multiple subgraphs, and different subgraphs are independent of each other in terms of graph connectivity.

[0121] Graph connectivity refers to the connection between the first nodes in the molecular graph, that is, it describes whether there is a bond relationship (connection line) between the first nodes in the molecular graph. Figure 4 A schematic diagram of subgraph division provided in an embodiment of the present application is shown as follows: Figure 4 As shown, in Figure 4 (a) shows a set of first nodes with a key relationship. According to the calculation of graph connectivity, the key relationship can be traversed starting from a first node, and the calculation result is obtained after the traversal is completed. Figure 4 (b) shows multiple subgraphs obtained after calculating the connectivity of the graph, namely subgraph A, subgraph B, subgraph C and subgraph D.

[0122] Specifically, the process of subdividing a graph is the process of identifying independent molecules. Graph connectivity allows for the identification and partitioning of independent subgraphs within a molecular graph. Consider the atoms in a molecule as first nodes, and the lines connecting them as bonds (i.e., chemical bonds). Graph connectivity can be used to identify and partition which atoms (i.e., first nodes) belong to the same molecule. The result of this partitioning is the identification of independent molecules within the molecular graph.

[0123] It should be noted that during the chemical simulation reaction, the bonding relationship between atoms may change. Through graph connectivity, the molecular graphs corresponding to different reaction frames can be divided into subgraphs to determine the molecules existing in the corresponding molecular graphs, so as to obtain the changes in molecular structure in the chemical simulation reaction.

[0124] As pointed out in the previous introduction, a graph that is not fully connected refers to a molecular graph, and the connected components decomposed into refer to the subgraphs obtained by dividing the molecular graph, and the subgraphs are used to identify independent molecules. Full connectivity here means that there are connections between any pair of first nodes. In the molecular graph in the embodiment of the present application, any first node will only have connections with some of the first nodes. In the process of subgraph division, a molecule can be regarded as a connected component, and the role of graph connectivity is to identify the connected components in the molecular graph.

[0125] The proposed subgraph partitioning method can calculate graph connectivity based on the connections between the first nodes in a molecular graph. Based on the calculated results, the molecular graph can be partitioned into multiple independent subgraphs. The specific subgraph partitioning method is relatively simple, and the formation of independent molecules facilitates subsequent determination of molecular changes during a reaction based on the independent molecules in each molecular graph, thereby enabling the identification of reaction events based on these molecular changes.

[0126] By the aforementioned introduction, the division of the subgraphs of the molecular graph is realized, that is, the independent expression of the molecules in the molecular graphs corresponding to different reaction frames is realized. Below, the determination of the reaction event can be carried out according to the molecular situation in the different reaction frames. In order to determine the reaction event, the reactants and products can be first determined by the molecular situation of the different reaction frames. In an embodiment of the present application, the molecules in the 1st reaction frame are generally regarded as alternative reactants, and the molecules in other reaction frames except the 1st reaction frame are used as alternative products. After completing the determination of reactants and products, the reaction between reactants and products can be shown in the form of a reaction diagram. Therefore, in a possible implementation, for the first reactant in the reactant, the first product in the product, the method for generating a reaction diagram is:

[0127] C1: Determine the atomic repetitiveness in the first reactant and the first product.

[0128] Atomic reproducibility refers to whether the same atoms exist between two molecules. If the same atoms exist, the two molecules are considered to have molecular reproducibility. In the embodiment of the present application, to determine atomic reproducibility, each atom, i.e., the aforementioned first node, can be assigned an initial index. Different first nodes correspond to different initial indices, and the initial index does not change during the chemical simulation reaction.

[0129] The first reactant is in the first reaction frame, and the first product can be in any reaction frame except the first reaction frame. In the process of generating a reaction graph for the i-th reaction frame, it is first necessary to obtain the reactants from the first reaction frame and the products from the i-th reaction frame. The first reactant is determined in the first reaction frame, and the first product is determined in the i-th reaction frame. Then it is necessary to determine whether there is atomic repeatability between the aforementioned first reactant and the first product. When there is atomic repeatability between the first reactant and the first product, it can be determined that there is a corresponding relationship between the first reactant and the first product, that is, it is determined that the first product is generated by the first reactant.

[0130] C2: In response to the first product having the same atom as the first reactant, adding a connection line between the second node identifying the first reactant and the third node identifying the first product.

[0131] In the embodiment of the present application, the reaction graph includes a second node identifying a reactant and a third node identifying a product, and the line connecting the second node and the third node is used to identify a candidate reaction to be identified between the molecules corresponding to the second node and the third node, respectively. When it is determined that there is atomic repeatability between the first reactant and the first product, it is determined that there is a possibility of a reaction event between the first reactant and the first product.

[0132] Adding a line between the second node identifying the first reactant and the third node identifying the first product means determining that a candidate reaction to be identified exists between the molecule corresponding to the first reactant and the molecule corresponding to the first product.

[0133] C3: In response to the first product and the first reactant not having the same atoms, not adding a link between the second node identifying the first reactant and the third node identifying the first product.

[0134] When the first product and the first reactant have the same atoms, a set of connection relationships can be constructed between the second node corresponding to the first reactant and the third node corresponding to the first product, and the connection relationships are identified by connecting lines.

[0135] When the first product and the first reactant do not have the same atoms, it is determined that there is no connection relationship between the first reactant and the first product, that is, there is no possibility of a candidate reaction to be identified, so no connection line is added between the first product and the first reactant.

[0136] It should be noted that the reaction diagram may include both the first product corresponding to the first reactant and other products. The difference between different products lies in whether they are connected to the first reactant by a line, that is, whether they are products generated by the first reactant.

[0137] The proposed reaction graph generation method, based on the atomic repetitiveness of the first reactant and the first product, can determine the relationship between the products involved in any reaction frame and the reactants, and determine whether there is a potential candidate reaction between the reactants and products. This establishes a correlation between the reactants and products, facilitating the subsequent identification of reaction events in chemical simulation reactions.

[0138] The construction of the reaction graph is described in detail above. When the construction of the reaction graph is completed, the situation in which the products in different reaction frames and the reactants in the first reaction frame have candidate reactions to be identified is determined, and they are marked in the reaction graph by connecting lines. The following needs to determine the reaction event for the candidate reaction between the reactant and the product. In S203, it is mentioned that "the reaction event of obtaining the product corresponding to the target third node from the reactant corresponding to the target second node is determined by the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node". In one possible implementation, the method for determining the reaction event is: through the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node, the atomic consistency of the reactant corresponding to the target second node and the product corresponding to the target third node, and the three-dimensional geometric features of the subgraph of the target second node and the subgraph of the target third node, the reaction event of obtaining the product corresponding to the target third node from the reactant corresponding to the target second node is determined.

[0139] The target second node refers to the second node corresponding to a reactant in the first reaction frame determined in the reaction graph, and the target third node refers to the product that has atomic repeatability with the reactant in the i-th reaction frame. In the reaction graph, there is a connection between the target second node and the target third node, which is used to identify the candidate reaction to be identified between the target second node and the target third node. In the process of determining the reaction event based on the candidate reaction, it is necessary to determine the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node. The graph isomorphism can be used to determine the degree of structural similarity of the molecules between the reactant corresponding to the target second node and the product corresponding to the target third node. The aforementioned graph isomorphism measures the degree of similarity between the two subgraphs based on the isomorphic properties of the aforementioned graphs. In the measurement process, it is necessary to consider both the nodes and the connections between the nodes.

[0140] That is, in the process of determining the graph isomorphism between the subgraph of the second target node and the subgraph of the third target node, it is necessary to determine the degree of similarity in the number of second nodes in the subgraph of the second target node and the third nodes in the subgraph of the third target node. In addition, it is also necessary to determine the degree of similarity in the number of links between the second nodes in the subgraph of the second target node and the third nodes in the subgraph of the third target node.

[0141] In addition to determining the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node, it is also necessary to determine the atomic consistency of the reactant corresponding to the target second node and the product corresponding to the target third node. In other words, the degree of consistency between the atoms in the reactant corresponding to the target second node and the atoms in the product corresponding to the target third node is determined. Specifically, there are two situations, namely, the atoms in the target second node are completely consistent with the atoms in the target third node, and the atoms in the target second node are partially consistent with the atoms in the target third node.

[0142] After completing the determination of graph isomorphism and atomic consistency between the subgraph of the second target node and the subgraph of the third target node, it is also necessary to determine the three-dimensional geometric features of the subgraph of the second target node and the subgraph of the third target node. The three-dimensional geometric features can be understood as the characteristics of the spatial positions between the atoms involved in the molecules corresponding to the subgraphs in the three-dimensional space. For ease of understanding, Figure 5 A schematic diagram of a reaction graph provided for an embodiment of the present application, as shown in the figure, includes target second node and target third node.The number of atoms and the number of lines of the molecule between the product corresponding to the reactant corresponding to the target second node and the target third node are consistent, and meet graph isomorphism.And the atoms involved in the corresponding product of the reactant corresponding to the target second node and the target third node are consistent, and are 3 C atoms, 3 O atoms and 4 H atoms, so meet atomic consistency.It is necessary to judge the three-dimensional geometric feature between the product corresponding to the reactant corresponding to the target second node and the target third node below, it can be seen that in target second node, C atoms and O atoms are in the same geometric plane, and in target third node, there is an O atom and C atom that are not in the same geometric plane, and an angle of θ is formed with the geometric plane, so the three-dimensional geometric feature of the subgraph of target second node and the subgraph of target third node are inconsistent.

[0143] Through the above-mentioned method for determining reaction events, the reaction event from the second target node to the third target node is determined based on the graph isomorphism, atomic consistency, and three-dimensional geometric features between the second target node and the third target node. Determining reaction events through these three aspects can, to a certain extent, ensure the accuracy of the reaction event identification results. While involving multiple aspects, the specific reaction event identification process is simple and easy to execute, ensuring a high efficiency in determining reaction events.

[0144] The above three aspects are proposed for the identification of reaction events: graph isomorphism, atomic consistency, and three-dimensional geometric features. Based on the matching results of these three aspects with the event identification conditions, the reaction event can be determined.

[0145] The above-mentioned event identification conditions are based on the principle of chemical simulation reaction and are determined according to atomic consistency, graph isomorphism and three-dimensional geometric features. That is, the event identification conditions need to consider atomic consistency, graph isomorphism and three-dimensional geometric features at the same time. The specific consideration process has been specifically described above and will not be repeated here. The matching results of different event identification conditions correspond to different graph isomorphisms, atomic consistency and three-dimensional geometric features. The reaction event in which the reactant corresponding to the target second node obtains the product corresponding to the target third node can be determined by the matching results.

[0146] The aforementioned specific reaction events may include the following three types: chemical reaction events, ionization reaction events, and no-reaction events. According to the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node, the atomic consistency of the reactant corresponding to the target second node and the product corresponding to the target third node, and the three-dimensional geometric features of the subgraph of the target second node and the subgraph of the target third node, a matching result is obtained by matching with the event identification condition. In response to the matching result of the event identification condition, the aforementioned "determining the reaction event of obtaining the product corresponding to the target third node from the reactant corresponding to the target second node" is specifically determined as follows:

[0147] D1: In response to the fact that the atoms of the reactant corresponding to the target second node and the product corresponding to the target third node are not completely consistent, determining that the reaction event of obtaining the product corresponding to the target third node from the reactant corresponding to the target second node is a chemical reaction.

[0148] The following describes the event recognition conditions corresponding to chemical reaction events, ionization reaction events, and no-reaction events. These conditions consider atomic consistency, graph isomorphism, and three-dimensional geometric features.

[0149] There are two event recognition conditions corresponding to chemical reaction events: ① The atoms in the product corresponding to the second target node and the atoms in the reactant corresponding to the third target node do not satisfy atomic consistency; ② The atoms in the product corresponding to the second target node and the atoms in the reactant corresponding to the third target node satisfy atomic consistency, but do not satisfy graph isomorphism.

[0150] The event recognition conditions corresponding to the ionization reaction event are: the atoms in the product corresponding to the target second node and the atoms in the reactant corresponding to the target third node satisfy both atomic consistency and graph isomorphism, but the three-dimensional geometric features of the product corresponding to the target second node and the reactant corresponding to the target third node are inconsistent.

[0151] The event recognition condition corresponding to the no-reaction event is that the atoms in the product corresponding to the second target node are exactly the same as the atoms in the reactant corresponding to the third target node, that is, the number and type of atoms are the same, and the interatomic connections and three-dimensional geometric features are consistent.

[0152] A chemical reaction is a process in which the bonds between the original reactants change to produce new products. This bond change can include breaking the original bonds between atoms in the reactants corresponding to the target second node and forming new bonds. When the atoms between the reactants corresponding to the target second node and the products corresponding to the target third node are not completely identical, this indicates a bond change between the reactants corresponding to the target second node and the products corresponding to the target third node.

[0153] For example, suppose the reactant corresponding to the second target node is H2O, and the product corresponding to the third target node is H2. Based on the reactants and products, we can determine that the atoms of the reactant corresponding to the second target node and the product corresponding to the third target node are not completely identical. Therefore, by matching the event identification conditions, we can determine that the reaction event from the reactant corresponding to the second target node to the product corresponding to the third target node is a chemical reaction.

[0154] D2: In response to the fact that the atoms of the reactant corresponding to the target second node and the product corresponding to the target third node are completely identical, but the subgraph of the target second node and the subgraph of the target third node are not isomorphic, it is determined that the reaction event of obtaining the product corresponding to the target third node from the reactant corresponding to the target second node is a chemical reaction.

[0155] In the process of judging the reaction event, the atomic consistency of the target second node and the target third node can be judged first. When it is determined that the atoms of the reactant corresponding to the target second node and the product corresponding to the target third node are completely consistent, the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node can be further judged. If it is judged that the subgraph of the target second node and the subgraph of the target third node are not isomorphic, then matching with the event recognition condition can determine that the reaction event of obtaining the product corresponding to the target third node from the reactant corresponding to the target second node is a chemical reaction.

[0156] D3: In response to the fact that the atoms of the reactant corresponding to the target second node and the product corresponding to the target third node are completely consistent, and the subgraph of the target second node and the subgraph of the target third node are isomorphic, but the three-dimensional geometric features of the subgraph of the target second node and the subgraph of the target third node are inconsistent, it is determined that the reaction event of obtaining the product corresponding to the target third node from the reactant corresponding to the target second node is an ionization reaction.

[0157] After the graph isomorphism and atomic consistency of the reactant corresponding to the target second node and the product corresponding to the target third node have been determined, the three-dimensional geometric features of the subgraph of the target second node and the subgraph of the target third node can be determined to determine whether the three-dimensional geometric features of the two are consistent. If the three-dimensional geometric features of the subgraph of the target second node and the subgraph of the target third node are determined to be inconsistent, the reaction event corresponding to the target third node obtained from the reactant corresponding to the target second node can be matched with the event recognition condition to determine that it is an ionization reaction.

[0158] D4: In response to the reactant corresponding to the target second node and the product corresponding to the target third node being the same, determining that no reaction event occurs between the reactant corresponding to the target second node and the product corresponding to the target third node.

[0159] The reactant corresponding to the second node of the target is the same as the product corresponding to the third node of the target. The same here can be understood as the high consistency between the reactant and the product, that is, the reactant corresponding to the second node of the target and the product corresponding to the third node of the target simultaneously satisfy the consistency of atomic consistency, graph isomorphism and three-dimensional geometric features. At this time, matching with the event recognition condition can be considered to obtain the product corresponding to the third node of the target from the reactant corresponding to the second node of the target. No reaction event occurs between them. At this time, the line between the second node identifying the reactant and the third node identifying the product can be removed from the reaction graph.

[0160] Through the method for determining reaction events provided above, the determination process of three reaction events - chemical reaction events, ionization reaction events and no-reaction events is specifically explained. By matching the graph isomorphism, molecular consistency and three-dimensional geometric feature consistency between the reactant corresponding to the target second node and the product corresponding to the target third node with the event matching conditions, the type of reaction event occurring from the reactant corresponding to the target second node to the product corresponding to the target third node can be quickly and conveniently determined.

[0161] The above specifically describes the method for determining reaction events. In practical applications, there is a situation where the role of surface materials in the reaction needs to be analyzed. Therefore, the reaction events occurring at the surface materials need to be identified and extracted for subsequent analysis and use. The chemical simulation reaction simulates the reaction inside a battery. The battery has a reactive surface material inside. After determining that the product corresponding to the target third node obtained from the reactant corresponding to the target second node has a reaction event, the reaction event of the surface material is determined as follows: first, the atomic type of the atoms included in the product corresponding to the target third node is determined. Then, in response to the product corresponding to the target third node including atoms of the reactive surface material, the reaction event is determined as a reaction event occurring near the reactive surface material. In response to the product corresponding to the target third node not including atoms of the reactive surface material, the reaction event is determined as a reaction event far away from the reactive surface material.

[0162] The reaction surface material can be understood as a material attached to the surface of an object in a reaction environment, and the reaction surface material can appear in the form of a coating on the surface of the object. Specifically, the type of atoms involved in the target third node is first determined. The method for determining the reaction event near the reaction surface material is to determine whether the product corresponding to the target third node contains atoms of the reaction surface material. The atoms involved in the reaction surface material are pre-determined information. It is determined whether the target third node includes the atoms involved in the reaction surface material. When the target third node includes the atoms involved in the reaction surface material, the reaction event corresponding to the target third node can be determined as a reaction event occurring near the reaction surface material. If the target third node does not include atoms of the reaction surface material, the reaction event corresponding to the target third node can be determined as a reaction event far away from the reaction surface material.

[0163] For example, assuming the reaction surface material is aluminum, the atomic type of the atoms included in the product corresponding to the third target node is determined to determine whether aluminum atoms are included. If the product corresponding to the third target node includes aluminum atoms, the reaction event corresponding to the third target node is determined to occur near the reaction surface material. If the product corresponding to the third target node does not include aluminum atoms, the reaction event corresponding to the third target node is determined to occur far from the reaction surface material.

[0164] The above-mentioned method for identifying reaction events near a reactive surface material determines whether the reaction event corresponding to the target third node occurs near the reactive surface material by determining whether the atoms involved in the target third node include atoms corresponding to the reactive surface material. The specific identification process is relatively simple, requiring only a comparison of atoms in the target third node and those in the reactive surface material. This screening and identification process is simple and easy to implement.

[0165] The aforementioned method for identifying reaction events is specifically described. After completing the identification of all reaction events in the chemical simulation reaction, the identification results of all reaction events can be obtained. Equivalent reactions may occur in the reaction events. Equivalent reactions can be understood as having equivalence between the reactants and products involved in different reaction events. This equivalence needs to be judged according to graph isomorphism. After determining equivalent reactions, they can be merged. Therefore, in a possible manner, after determining all reaction events of the chemical simulation reaction by the reaction graphs of N reaction frames, the method for determining equivalent reactions is: first determine the graph isomorphism between the subgraphs of the reactants in different reaction events, and determine the graph isomorphism between the subgraphs of the products in different reaction events. Then, in response to the subgraph isomorphism of the reactants of the first reaction event and the second reaction event, and the subgraph isomorphism of the products of the first reaction event and the second reaction event, determine that the first reaction event and the second reaction event are equivalent reactions, and finally merge the equivalent reactions in all reaction events.

[0166] In an embodiment of the present application, before determining an equivalent reaction, the atomic consistency of the reactants and products involved in the reaction event can be judged first. When it is determined that the reactants in the first reaction event and the second reaction event have the same atoms, and the products in the first reaction event and the second reaction event also have the same atoms, the graph isomorphism between the subgraphs of the reactants in the first reaction event and the second reaction event, as well as the graph isomorphism between the subgraphs of the products in the first reaction event and the second reaction event can be further determined.

[0167] The method for determining graph isomorphism has been described in detail previously and will not be repeated here. Once the subgraphs of the reactants of the first and second reaction events are determined to be isomorphic, as well as the subgraphs of the products of the first and second reaction events are determined to be isomorphic, the first and second reaction events can be determined to be equivalent reactions. Finally, the equivalent reactions can be merged to avoid a large number of duplicate and redundant equivalent reactions within the determined reaction events.

[0168] The above-mentioned method for determining equivalent reactions can be used to determine whether a reaction event is an equivalent reaction by determining the graph isomorphism between the subgraphs of reactants and products of different reaction events. Once equivalent reactions are determined, the equivalent reactions in the determined reaction events can be merged, thereby avoiding redundant equivalent reactions in the final determined reaction event and improving the intuitiveness of the reaction event identification results.

[0169] There is also a special reaction in chemical simulation reactions - transient reactions. Transient reactions refer to the process in which a reactant is continuously transformed into other products in each reaction frame, and finally transformed back into a product consistent with the original reactant, that is, the product eventually returns to the initial reactant itself. In the process of identifying reaction events, the transient reaction can be identified, thereby achieving the distinction between real and effective chemical transformations and transient reactions in chemical simulation reactions. Therefore, in one possible implementation, for the second reactant in the reactants, after determining all reaction events of the chemical simulation reaction through the reaction graph of N reaction frames, the method for determining the transient reaction is: first determine the event sequence of the reaction events involved in the second reactant in N reaction frames, and then, in response to the second product involved in the event sequence being equivalent to the second reactant, collectively identify the jth reaction event of obtaining the second product to the first reaction event of the second reactant in the event sequence as a transient reaction, j>1.

[0170] Specifically, based on the atoms involved in the second reactant, reaction events in each reaction frame that share the same atoms as the second reactant can be extracted to form an event sequence of reaction events. When it is determined that the second product involved in the event sequence is equivalent to the second reactant, the entire event sequence from the jth reaction event that produces the second product to the first reaction event of the second reactant can be collectively identified as a transient reaction.

[0171] The aforementioned second product and second reactant are equivalent. This equivalence can be understood as possessing graphical isomorphism, atomic identity, and three-dimensional geometric consistency between the second product and the second reactant, meaning they are identical molecules. The reaction event in which the second product occurs can be determined from the event sequence, and the reaction events from the first reaction event to the second product are collectively labeled as transient reactions.

[0172] For example, assuming the second reactant is molecule A, and the event sequence of reaction events involving molecule A in the reaction frame includes 10 reaction events, when it is determined that the second product and the second reactant involved in the event sequence are equivalent, both are molecule A. In this case, the event sequence from the 10th reaction event that produces the second product to the 1st reaction event of the second reactant can be collectively identified as a transient reaction.

[0173] Through the transient reaction identification method provided above, the transient reactions involved in the chemical simulation reaction can be identified, thereby distinguishing the real and effective chemical transformations and transient reactions in the chemical simulation reaction, and then distinguishing the reaction events corresponding to the effective chemical transformations in the chemical simulation reaction, which is conducive to the identification of the effectiveness of reaction events in the chemical simulation reaction.

[0174] Figure 6 A flowchart of a reaction event determination provided in an embodiment of the present application is as follows: Figure 6 As shown, in the aforementioned description, a chemical simulation reaction can be completed by N reaction frames. In the embodiment of the present application, an xyz file can be obtained, and then the xyz file can be separated into single frames. The single frame separation process is the process of generating N molecular graphs of a multi-molecule system in a chemical simulation reaction. There is a one-to-one correspondence between the molecular graph and the reaction frame. The xyz file is a trajectory file, and the trajectory file is a collection file of corresponding reaction frames in a multi-molecule system. Single frame separation is the separation of reaction frames.

[0175] After completing single-frame separation, single-frame analysis is required, as shown in area A in the figure. The single-frame analysis process includes the following steps: ① Obtaining the adjacency relationship of atoms; ② Inferring bond order and charge; ③ Constructing a molecular graph; ④ Separating subgraphs.

[0176] The first step is to obtain the adjacency relationship of atoms. Since the xyz file includes the number of atoms and the coordinates (x, y, z) of the three-dimensional position of each atom, the adjacency relationship of each atom can be determined through these coordinates. The determination process can be obtained by using the open source chemical information processing Python library ASE (Atomistic Simulation Environment) based on the three-dimensional coordinate analysis in the file. The adjacency relationship can be determined based on 1.20 times the covalent radius of each atom as a reference value.

[0177] After obtaining the adjacency relationship of atoms, it is necessary to infer bond order and charge, and the bond order and charge corresponding to each connection relationship are inferred by the open source chemical information processing python library RDKit (chemical information software package). The bond order is mainly inferred by comparing the empirical statistics of atomic distances and chemical bonding distances, and the charge is inferred according to the bonding configuration of each atom. The inference method of bond order and charge enables the embodiment of the application to not rely on additional bond order information to determine the bond relationship.

[0178] When the inference bond order and charge amount are completed, and after determining the bond relationship between each atom, it is necessary to build a molecular graph.Molecular graph includes the first node that identifies the atom, and the line between the two first nodes is used to identify the atoms corresponding to the two first nodes with a bond relationship, and the i-th molecular graph is used to identify the chemical simulation reaction when the i-th reaction frame, the bond relationship between atoms in the multi-molecule system. The process of building a molecular graph in an embodiment of the present application can be realized based on the open source python library Networkx (Python software package for creating, operating and analyzing complex networks) of graph (Graph) and network (Network) information processing.

[0179] After completing the construction of the molecular graph, it is necessary to separate the subgraphs. The subgraphs in the molecular graph can realize the independent expression of the molecules. The separation process of the subgraphs can be based on the connected component analysis algorithm in graph theory (that is, based on the atomic connectivity - the bond relationship). The analysis of the connected components of the graph can be used to identify independent subgraphs in a graph. These subgraphs are independent of each other in the structure of the graph. Connected components refer to the subgraphs obtained by dividing the molecular graph. If the atoms in the molecule are regarded as vertices and the chemical bonds are regarded as edges, then a molecule can be regarded as a connected component. By separating the subgraphs, the independent expression of molecules can be achieved.

[0180] After completing the single-frame analysis, reactant extraction is required. The aforementioned division of the reaction frames is based on the change in the bond relationship and is arranged in time sequence. In the embodiment of the present application, the first molecular graph is used as the starting molecular graph, and the reactants of the chemical simulation reaction can be determined from the subgraphs of the first molecular graph.

[0181] After the reactants are extracted, molecular changes need to be determined. For details, see area B in the figure. Molecular change determination includes the following steps: ① atomic index following; ② graph isomorphism test; ③ 3D geometric feature comparison.

[0182] First, atom index tracking. To more precisely determine the changes in the molecules corresponding to each reactant, an index can be set for each atom in the reaction. This index does not change during the chemical simulation reaction. When it is determined that the subgraph of the i-th molecular graph contains atoms with the same index as the subgraph of the first molecular graph, the molecules with the same index are mapped into a pair of reactants and products.

[0183] Then comes the graph isomorphism test. When the reactants and products with the same index are obtained, the reactants and products will be further tested for graph isomorphism to determine whether there are differences between the reactants and products, that is, whether the reactants have changed during the reaction process.

[0184] Finally, after completing atomic index tracking and graph isomorphism testing, 3D geometric feature comparison is used to capture ionization reactions by setting specific bond angle and dihedral angle change thresholds. The products of such reactions are graph-isomorphic to the reactants, but the relative geometric positions of some atoms have changed.

[0185] After the molecular changes are determined, the reaction events need to be determined. For details, see area C in the figure. The reaction event determination specifically includes the following processes: ① Screening unreacted reactants; ② Following and arranging the changed molecules; ③ Obtaining a separated reaction map.

[0186] First, unreacted reactants are screened. Unreacted reactants can be understood as reactants that underwent a transient reaction and then reconverted to the reactant after the chemical simulation reaction, meaning that the reactant was not effectively converted. Unreacted reactants can also be molecules that did not undergo a change during the chemical simulation reaction, that is, molecules in the first reaction frame that did not participate in the chemical simulation reaction. After screening out these unreacted reactants, they can be removed.

[0187] After screening unreacted reactants, it's necessary to track and arrange the changing molecules. By indexing atoms, we can determine the changes in the atom corresponding to that index during the chemical simulation reaction. By tracking and arranging the molecules corresponding to that atom, we can determine the changes in the reactants containing that atom during the chemical simulation reaction. The screening process can be completed based on the reaction graph. Using the indices carried by the atoms, the molecules corresponding to atoms with the same index are connected to form a reaction graph.

[0188] After tracking and arranging the changing molecules, it is necessary to obtain a separate reaction map. The separate reaction map can be understood as the classification that completes the identification of reaction events. Candidate reactions that meet one of the following three conditions will be judged as reaction events: 1) The atomic indices contained in the product molecules / ions are not completely consistent with those of the reactants. Such reactions will be marked as chemical reactions; 2) The atomic indices contained in the product molecules / ions are completely consistent with those of the reactants, but the product structure is not consistent with the reactants. Such reactions will be marked as chemical reactions; 3) The product molecules / ions are isostructural with the reactant molecules / ions, but the three-dimensional geometric features are inconsistent. Such reactions will be specially marked as ionization reactions.

[0189] In addition to the chemical reactions and ionization reactions mentioned above, these also include equivalent reactions, transient reactions, and reaction events near reaction surface materials. Specifically, equivalent reactions can be determined by the graph isomorphism between the subgraphs of reactants and products. Transient reactions can be determined by determining whether the reactants and products are equivalent and, based on the determination, by including the transient reactions in the event sequence of the reactants. Reaction events near reaction surface materials can be determined by determining whether atoms in the products include atoms of the reaction surface material.

[0190] Figure 7 A flowchart of another reaction event determination provided in an embodiment of the present application is as follows: Figure 7 As shown, in one possible implementation, the process of reaction event determination can be divided into the following four steps: ① molecular graph construction; ② subgraph separation; ③ molecular change tracking; ④ reaction event identification.

[0191] The process of constructing a molecular graph includes: ① obtaining the three-dimensional position of atoms; ② determining the adjacency relationship; and ③ completing the molecular graph construction. First, it is necessary to obtain the three-dimensional position of each atom in the trajectory file, and then determine the adjacency relationship between each atom in different reaction frames based on the three-dimensional position. The adjacency relationship can be determined based on 1.20 times the covalent radius of each atom as a reference value. Finally, the molecular graph is constructed based on the adjacency relationship. The molecular graph construction process requires comparing the atomic distance between atoms with the chemical bonding distance to determine the bonding relationship between atoms with adjacency relationships, and then verifying the number of determined bonding relationships through the bonding configuration and charge range. After adjustment based on the verification results, the completed molecular graph is obtained.

[0192] After the molecular graph is constructed, subgraph separation is required. This process allows the molecules in the molecular graph to be expressed independently, and the subgraphs need to be divided according to their connectivity. Specifically, the graph connectivity is calculated based on the bond relationships between atoms in the molecular graph, and the subgraph separation is completed based on the calculated results.

[0193] After the subgraph separation is completed, molecular change tracking is required. The process of molecular change tracking includes: ① determining atomic repeatability; ② graph isomorphism test; ③ three-dimensional geometric feature comparison. Molecular change tracking means that during the chemical simulation reaction, the molecules corresponding to the reactants will be converted into other molecules after the reaction. It is necessary to follow the changes of the molecules to determine the corresponding reaction events when the reactants are converted into different products. In order to facilitate the tracking of molecular changes in chemical simulation reactions, each atom in the molecule can be given a unique index, and the index will not change during the chemical simulation reaction.

[0194] During molecular change tracking, it is necessary to determine whether there is atomic repetition between the reactants and products. If there is atomic repetition, the reactants and products can be further tested for graph isomorphism and compared with their three-dimensional geometric features. This process can determine whether the molecules corresponding to the reactants have a candidate reaction to be identified.

[0195] After completing molecular change tracking, reaction event identification is required. The reaction event identification process includes: ① screening unreacted reactants; ② tracking and arranging the changed molecules; ③ obtaining a separated reaction graph. First, it is necessary to screen unreacted reactants. After screening out the unreacted reactants, the reactants can be removed. After screening out the unreacted reactants, it is necessary to track and arrange the changed molecules. The products involved in the reactants in the reaction frame are tracked and arranged to obtain the transformation process of the reactants during the chemical simulation reaction. Finally, a separated reaction graph is obtained. Based on the separated reaction graph, the reaction event can be identified according to the event recognition conditions.

[0196] The functional differences between the reaction identification method proposed in the embodiment of the present application and the related art are shown in Table 1:

[0197] Table 1. Function comparison between the present application solution and related technical solutions

[0198]

[0199] As can be seen from Table 1, in the related art, it is necessary to rely on ReaxFF (Reactive Force Field, a molecular dynamics model) to provide bond order information, and ReaxFF is an empirical force field for simulating chemical reactions, which can describe the formation and fracture of chemical bonds (i.e., bond relationships). Compared with related art, the reaction identification method provided in the embodiment of the present application can complete bond order determination alone, that is, only by the atoms in the trajectory file and the three-dimensional position information of the atoms (the trajectory structure in the table), the bond relationship between each atom is inferred, and the division of the subgraph in the molecular graph is realized. And the determination of molecular equivalence can be carried out by graph isomorphism, while identification can be determined for ionization reaction, surface reaction event (i.e., reaction event occurring near the reaction surface material), etc., compared to related art, there is higher independent operation capability and scalability.

[0200] As mentioned above, in the embodiments of this application, the 3T algorithm can be used to generate an optimization trajectory. An optimization trajectory refers to a trajectory that uses an optimization method to find the stable configuration or lowest energy state of a multi-molecular system. A collection of reaction frames arranged in time sequence can be called a trajectory, which can reflect the changing process of the reactants in the chemical simulation reaction.

[0201] Taking the electrochemical reaction of lithium-ion electrolytes as an example, using the 3T algorithm, it is only necessary to put the component molecules into a box and call 50-100 DFT (Density Functional Theory) steps to generate a chemical simulation reaction, so that the experimentally observed lithium-ion electrolyte electrochemical reaction products can be studied and reproduced. The aforementioned DFT is a quantum mechanical method used to describe atomic, molecular or solid systems. In the embodiments of the present application, it can be used to determine multi-molecular systems, and VASP (Vienna Ab-initio Simulation Package, a chemical simulation software package for electronic structure calculations) can be used for density functional theory calculations.

[0202] In order to realize the chemical simulation reaction, a chemical simulation environment needs to be built. First, a negative electrode surface with lithiated graphite as the reaction surface material is established. The fully lithiated graphite 100 surface is selected to simulate the negative electrode of the battery. The edge adopts a zigzag structure terminated by =O and is exposed to a vacuum. In order to prevent the lithium embedded in the graphite from being attracted by the edge oxygen, an additional lithium equal to the oxygen is placed near the reaction surface material to coordinate with it. Then, the box is scaled to reserve space for solvent filling. Finally, a triclinic periodic box is created, and ethylene carbonate (EC), dimethyl carbonate (DMC), vinylene carbonate (VC), lithium cation (Li + ) and hexafluorophosphate anion After setting up the chemical simulation environment, five cycles of 3T-FF (3-point Thole-type Force Field, a molecular dynamics simulation method) were run, each lasting 200 steps, to disperse the molecules within the box while disabling chemical reactions. Three cycles of 3T-VASP (a molecular dynamics simulation method combining 3T and VASP) were then run, each lasting 50 steps, to further relax the molecules while allowing the electrochemical simulation to continue.

[0203] In a chemical simulation reaction, the 3T algorithm is first run to obtain an optimized trajectory. Then, based on the reaction identification method provided in this application, a large number of reaction events in the trajectory are automatically separated and statistically analyzed. Figure 8 A schematic diagram of a reaction event occurrence provided in an embodiment of the present application is shown in the figure. Figure 8 (a) in the figure shows the DFT step number when the reduction reaction occurs. Figure 8 (b) shows the DFT steps when reductive ionization occurs, showing the ionization reactions that occurred during the 150-step DFT calculation. These ionization reactions are the precursors to many reduction reactions. Some atoms accept charges, thereby changing the hybridization state of some of the atoms. The hybridization state refers to the mixing of atomic orbitals to form new hybrid orbitals with different shapes and energies during the formation of chemical bonds. This can be understood as changing the three-dimensional geometric characteristics of atoms in the molecule. The figure includes ethylene carbonate (EC), dimethyl carbonate (DMC), vinylene carbonate (VC), and hexafluorophosphate anions. The frequencies of respective reduction reactions and reductive ionization.

[0204] By detecting whether the chemical simulation reaction interacts with the reaction surface material, the reaction occurring near the reaction surface material can be isolated. Figure 9 A schematic diagram of another reaction event occurrence provided in an embodiment of the present application is shown in the figure, wherein Figure 9(a) shows the categories of reaction events involved near the surface of the lithiated graphite anode and in the electrolyte liquid phase observed using 3T-VASP. Figure 9 (b) shows the statistics of the number of reaction events occurring near the reaction surface material and in the electrolyte liquid phase. In (a), (ae) are reactions originating from EC molecules; (fg) are reactions originating from DMC molecules; (hj) are reactions originating from VC molecules; (kl) are reactions originating from Anion reactions (mo) originate from reactions involving EC and other reactants. EC-based reactions dominate electrochemical reduction reactions. Comparing the reaction events occurring near the surface material with those occurring in the electrolyte bulk phase reveals little difference in type, but significant differences in the probability of occurrence of different reaction events.

[0205] Figure 10 A schematic diagram of a chemical reaction impact statistics provided in an embodiment of the present application is shown in the figure, wherein Figure 10 (a) and (b) represent the results of the reaction in the electrolyte bulk and the reaction near the surface based on the statistics of the system's additional charge, respectively. Figure 10 (c) and (d) show the results of reactions occurring in the electrolyte bulk and near the surface, respectively, based on the number of free lithium ions. It can be seen that the total charge level and the number of free lithium ions have a significant impact on reactions occurring in the electrolyte bulk, but have a limited impact on reactions occurring near the surface.

[0206] Through the above example, we can perform statistical analysis on reaction events in chemical simulation reactions based on the optimization trajectory determined by the 3T algorithm, and determine the scientific meaning behind the reaction events through the statistical results.

[0207] In the aforementioned Figure 1-10 Based on the corresponding embodiments, Figure 11 A schematic diagram of a reaction identification device provided in an embodiment of the present application is shown, wherein the reaction identification device 1000 includes: an acquisition module 1001, a generation module 1002, and a determination module 1003;

[0208] The acquisition module 1001 is used to acquire N molecular graphs of a multi-molecule system in a chemical simulation reaction, where the chemical simulation reaction is completed through N reaction frames, the molecular graphs correspond to the reaction frames one-to-one, the molecular graphs include first nodes that identify atoms, and a line between two first nodes is used to identify that the atoms corresponding to the two first nodes have a bond relationship, the i-th molecular graph is used to identify the bond relationship between atoms in the multi-molecule system in the i-th reaction frame of the chemical simulation reaction, the molecular graph includes subgraphs, and the subgraphs are used to identify molecules, where N>1;

[0209] The generating module 1002 is configured to determine the reactants of the chemical simulation reaction based on the subgraph of the first molecular graph, determine the products of the chemical simulation reaction based on the subgraph of the i-th molecular graph, and generate a reaction graph of the chemical simulation reaction in the i-th reaction frame based on the reactants and the products, wherein the reaction graph includes a second node identifying the reactants and a third node identifying the products, and a line between the second node and the third node is used to identify that there is a candidate reaction to be identified between the molecules corresponding to the second node and the third node, respectively. <i≤N;

[0210] The determination module 1003 is used to determine, for the target second node and the target third node having the candidate reaction, a reaction event of obtaining the product corresponding to the target third node from the reactant corresponding to the target second node through the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node.

[0211] In a possible implementation, the acquisition module 1001 is specifically configured to:

[0212] Obtaining a trajectory file for describing atomic structure changes in the chemical simulation reaction, wherein the trajectory file is used to identify atoms involved in the multi-molecule system and the three-dimensional positions of the atoms in different reaction frames;

[0213] N molecular graphs of the multi-molecule system in the chemical simulation reaction are generated according to the trajectory file.

[0214] In a possible implementation, for the first atom and the second atom in the i-th reaction frame, the acquisition module 1001 is specifically configured to:

[0215] determining a chemical bonding distance between the first atom and the second atom according to the atomic types of the first atom and the second atom;

[0216] determining an atomic distance between the first atom and the second atom based on the three-dimensional positions of the first atom and the second atom;

[0217] In response to the atomic distance being consistent with the chemical bonding distance, determining that the first atom and the second atom have a bonding relationship;

[0218] In the i-th molecule graph, a line is added between the first atom and the second atom.

[0219] In a possible implementation, after determining all bonding relationships of the first atoms, the apparatus is specifically configured to:

[0220] determining a charge range of the first atom according to the atomic type of the first atom;

[0221] Using the charge range as a bond relationship verification condition, performing a bond order quantity verification on all bond relationships of the first atom;

[0222] In response to the number of all bond orders of the first atom satisfying the bond relationship verification condition, determining to retain the connection with the first atom;

[0223] In response to the number of all bond orders of the first atom not satisfying the bond relationship check condition, the type of the bond relationship between the first atom and other atoms is re-determined according to a cause causing the non-satisfaction.

[0224] In a possible implementation, after obtaining N molecular graphs of a multi-molecule system in a chemical simulation reaction, the apparatus is specifically configured to:

[0225] Performing graph connectivity calculation based on the connection between the first nodes in the molecular graph;

[0226] According to the result of the graph connectivity calculation, the molecular graph is divided into multiple subgraphs, and different subgraphs are independent of each other in terms of graph connectivity.

[0227] In a possible implementation, for the first reactant among the reactants and the first product among the products, the generating module 1002 is specifically configured to:

[0228] determining atomic repetitiveness in the first reactant and the first product;

[0229] In response to the first product having the same atom as the first reactant, adding a link between the second node identifying the first reactant and the third node identifying the first product;

[0230] In response to the first product and the first reactant not having the same atoms, no link is added between the second node identifying the first reactant and the third node identifying the first product.

[0231] In a possible implementation, the determining module 1003 is specifically configured to:

[0232] The reaction event of obtaining the product corresponding to the target third node from the reactant corresponding to the target second node is determined through the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node, the atomic consistency of the reactant corresponding to the target second node and the product corresponding to the target third node, and the three-dimensional geometric features of the subgraph of the target second node and the subgraph of the target third node.

[0233] In one possible implementation, based on the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node, the atomic consistency between the reactant corresponding to the target second node and the product corresponding to the target third node, and the three-dimensional geometric features of the subgraph of the target second node and the subgraph of the target third node, a matching result is obtained by matching with the event recognition condition. In response to the matching result of the event recognition condition, the determination module 1003 is specifically used to:

[0234] In response to the fact that atoms of the reactant corresponding to the target second node and the product corresponding to the target third node are not completely identical, determining that a reaction event in which the product corresponding to the target third node is obtained from the reactant corresponding to the target second node is a chemical reaction;

[0235] In response to the fact that atoms of the reactant corresponding to the target second node and the product corresponding to the target third node are completely identical, but the subgraph of the target second node and the subgraph of the target third node are not isomorphic, determining that a reaction event in which the product corresponding to the target third node is obtained from the reactant corresponding to the target second node is a chemical reaction;

[0236] In response to the fact that atoms of the reactant corresponding to the target second node and the product corresponding to the target third node are completely identical, and the subgraph of the target second node and the subgraph of the target third node are isomorphic, but the three-dimensional geometric features of the subgraph of the target second node and the subgraph of the target third node are inconsistent, determining that a reaction event in which the product corresponding to the target third node is obtained from the reactant corresponding to the target second node is an ionization reaction;

[0237] In response to the reactant corresponding to the target second node and the product corresponding to the target third node being the same, it is determined that no reaction event occurs between the reactant corresponding to the target second node and the product corresponding to the target third node.

[0238] In one possible implementation, the chemical simulation reaction simulates a reaction inside a battery having a reactive surface material. After determining that a reaction event occurs in which a product corresponding to the target third node is obtained from a reactant corresponding to the target second node, the device is specifically configured to:

[0239] Determining the atomic type of the atoms included in the product corresponding to the target third node;

[0240] In response to the product corresponding to the target third node including atoms of the reaction surface material, determining the reaction event as a reaction event occurring near the reaction surface material;

[0241] In response to the product corresponding to the target third node not including atoms of the reaction surface material, the reaction event is determined as a reaction event far away from the reaction surface material.

[0242] In a possible implementation, after determining all reaction events of the chemical simulation reaction through the reaction graphs of the N reaction frames, the apparatus is specifically configured to:

[0243] Determine graph isomorphism between subgraphs of reactants in different reaction events, and determine graph isomorphism between subgraphs of products in different reaction events;

[0244] In response to subgraph isomorphism between reactants of a first reaction event and a second reaction event, and subgraph isomorphism between products of the first reaction event and the second reaction event, determining that the first reaction event and the second reaction event are equivalent reactions;

[0245] Equivalent reactions were combined across all reaction events.

[0246] In a possible implementation, for the second reactant among the reactants, after determining all reaction events of the chemical simulation reaction through the reaction graphs of N reaction frames, the apparatus is specifically configured to:

[0247] determining an event sequence of reaction events involving the second reactant in the N reaction frames;

[0248] In response to the second product involved in the event sequence being equivalent to the second reactant, the j-th reaction event obtaining the second product to the 1st reaction event of the second reactant in the event sequence are collectively identified as transient reactions, j>1.

[0249] Through the reaction identification device provided above, for the N reaction frames that complete the chemical simulation reaction, a molecular graph is used to express the molecules involved in the multi-molecule system in each reaction frame. The specific expression method is: the first node is used to express the atoms in the molecular graph, and the line between the two first nodes is used to express the bond relationship between the two atoms. The molecular graph includes one or more subgraphs, and each individual subgraph is used to identify a molecule, that is, the independent expression of the molecule can be achieved in the molecular graph. In order to accurately identify the reaction events generated in the chemical simulation reaction, for the i-th reaction frame including the product, the subgraph in the molecular graph of the reaction frame and the subgraph of the first molecular graph constitute the reaction graph of the reaction frame. The reaction graph can accurately express the molecular changes from the first reaction frame to the i-th reaction frame from the molecular level. In the reaction graph, the second node is used to identify the reactant, and the third node is used to identify the product. The line between the second node and the third node identifies the candidate reaction to be identified between the molecules corresponding to the second node and the third node. Graph isomorphism can accurately determine whether changes have occurred between the reactant corresponding to the second node and the product corresponding to the third node, thereby determining whether a reaction event has actually occurred between the reactants and products associated with the candidate reaction. By subdividing the molecular graph into subgraphs to identify independent molecules and using graph isomorphism to identify the occurrence of reaction events between the molecules corresponding to reactants and products in the reaction graph, the reaction events generated in each reaction frame of a completed chemical simulation reaction can be quickly and automatically identified without the complex computational costs of related technologies, effectively improving the efficiency of reaction event identification in chemical simulation reactions.

[0250] The embodiment of the present application further provides a computer device, including a terminal device or a server, in which the aforementioned reaction recognition device can be configured. The computer device is described below with reference to the accompanying drawings.

[0251] If the computer device is a terminal device, see Figure 12 As shown, the embodiment of the present application provides a terminal device, taking a mobile phone as an example:

[0252] Figure 12 The block diagram shows a partial structure of the mobile phone provided by the embodiment of the present application. Figure 12 The mobile phone includes components such as a radio frequency (RF) circuit 1410, a memory 1420, an input unit 1430, a display unit 1440, a sensor 1450, an audio circuit 1460, a wireless fidelity (WiFi) module 1470, a processor 1480, and a power supply 1490. Those skilled in the art will understand that Figure 12The mobile phone structure shown in the figure does not constitute a limitation to the mobile phone, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0253] The following combination Figure 12 A detailed introduction to the various components of a mobile phone:

[0254] The RF circuit 1410 may be used for receiving and sending signals during information transmission or calls. In particular, after receiving downlink information from the base station, it is sent to the processor 1480 for processing. In addition, the designed uplink data is sent to the base station.

[0255] Memory 1420 can be used to store software programs and modules. Processor 1480 executes the various functional applications and data processing of the mobile phone by running the software programs and modules stored in memory 1420. Memory 1420 may mainly include a program storage area and a data storage area. The program storage area may store an operating system and at least one application required for a function (such as a sound playback function, an image playback function, etc.); the data storage area may store data created based on the use of the mobile phone (such as audio data, a phone book, etc.). In addition, memory 1420 may include high-speed random access memory and non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0256] The input unit 1430 may be configured to receive input digital or character information and generate key signal input related to user settings and function control of the mobile phone. Specifically, the input unit 1430 may include a touch panel 1431 and other input devices 1432 .

[0257] The display unit 1440 may be configured to display information input by the user or information provided to the user, as well as various menus of the mobile phone. The display unit 1440 may include a display panel 1441 .

[0258] The mobile phone may also include at least one sensor 1450, such as a light sensor, a motion sensor, and other sensors.

[0259] The audio circuit 1460 , the speaker 1461 , and the microphone 1462 can provide an audio interface between the user and the mobile phone.

[0260] WiFi is a short-range wireless transmission technology. The mobile phone can help users send and receive emails, browse web pages, and access streaming media through the WiFi module 1470, providing users with wireless broadband Internet access.

[0261] The processor 1480 is the control center of the mobile phone. It uses various interfaces and lines to connect various parts of the entire mobile phone. It executes various functions of the mobile phone and processes data by running or executing software programs and / or modules stored in the memory 1420 and calling data stored in the memory 1420.

[0262] The mobile phone also includes a power supply 1490 (such as a battery) for supplying power to various components.

[0263] In this embodiment, the processor 1480 included in the terminal device is also used to execute the steps in the methods of each embodiment of the present application.

[0264] If the computer device is a server, this embodiment of the application also provides a server, see Figure 13 As shown, Figure 13 The structural diagram of the server 1500 provided in the embodiment of the present application, the server 1500 may have relatively large differences due to different configurations or performances, and may include one or more central processing units (CPUs) 1522 (for example, one or more processors) and a memory 1532, and one or more storage media 1530 (for example, one or more mass storage devices) for storing application programs 1542 or data 1544. Among them, the memory 1532 and the storage medium 1530 can be temporary storage or permanent storage. The program stored in the storage medium 1530 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations on the server. Furthermore, the central processing unit 1522 can be configured to communicate with the storage medium 1530 to execute a series of instruction operations in the storage medium 1530 on the server 1500.

[0265] The server 1500 may also include one or more power supplies 1526, one or more wired or wireless network interfaces 1550, one or more input and output interfaces 1558, and / or one or more operating systems 1541, such as Windows Server 2003. TM , Mac OS X TM , Unix TM ,Linux TM , FreeBSD TM etc.

[0266] The steps performed by the server in the above embodiment can be based on Figure 13 The server structure shown.

[0267] In addition, an embodiment of the present application further provides a storage medium, which is used to store a computer program, and the computer program is used to execute the method provided by the above embodiment.

[0268] An embodiment of the present application further provides a computer program product including a computer program, which, when executed on a computer device, enables the computer device to execute the method provided in the above embodiment.

[0269] Those skilled in the art will understand that all or part of the steps of implementing the above-mentioned method embodiment can be completed by hardware related to program instructions, and the above-mentioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above-mentioned method embodiment; and the above-mentioned storage medium can be at least one of the following media: read-only memory (English: Read-only Memory, abbreviated: ROM), RAM, magnetic disk or optical disk, etc., various media that can store computer programs.

[0270] In the embodiments of the present application, the term "module" or "unit" refers to a computer program or a part of a computer program that has a predetermined function and works together with other related parts to achieve a predetermined goal, and can be implemented in whole or in part by using software, hardware (such as processing circuits or memories) or a combination thereof. Similarly, a processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be part of an overall module or unit that includes the function of the module or unit.

[0271] It should be noted that the various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device and system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments. The device and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of this embodiment. A person of ordinary skill in the art can understand and implement it without expending creative work.

[0272] The above is only one specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the scope of protection of the present application. Moreover, based on the implementation methods provided in the above aspects, the present application can also be further combined to provide more implementation methods. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A reaction identification method, characterized in that: The method comprises: Obtaining N molecular graphs of a multi-molecule system in a chemical simulation reaction, where the chemical simulation reaction is completed through N reaction frames, the molecular graphs corresponding one-to-one to the reaction frames, the molecular graphs including first nodes identifying atoms, a line between two first nodes used to identify a bond relationship between the atoms corresponding to the two first nodes, the i-th molecular graph used to identify the bond relationship between atoms in the multi-molecule system in the i-th reaction frame of the chemical simulation reaction, the molecular graphs including subgraphs used to identify molecules, and N>1; Determine the reactants of the chemical simulation reaction according to the subgraph of the first molecular graph, determine the products of the chemical simulation reaction according to the subgraph of the i-th molecular graph, and generate a reaction graph of the chemical simulation reaction in the i-th reaction frame based on the reactants and the products, wherein the reaction graph includes a second node identifying the reactants and a third node identifying the products, and a line between the second node and the third node is used to identify that there is a candidate reaction to be identified between the molecules corresponding to the second node and the third node, respectively. <i≤N; For the target second node and the target third node having the candidate reaction, the reaction event of obtaining the product corresponding to the target third node from the reactant corresponding to the target second node is determined through the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node.

2. The method according to claim 1, characterized in that The obtaining of N molecular graphs of a multi-molecule system in a chemical simulation reaction includes: Obtaining a trajectory file for describing atomic structure changes in the chemical simulation reaction, wherein the trajectory file is used to identify atoms involved in the multi-molecule system and the three-dimensional positions of the atoms in different reaction frames; N molecular graphs of the multi-molecule system in the chemical simulation reaction are generated according to the trajectory file.

3. The method according to claim 2, characterized in that For the first atom and the second atom in the i-th reaction frame, generating N molecular graphs of the multi-molecule system in the chemical simulation reaction according to the trajectory file includes: determining a chemical bonding distance between the first atom and the second atom according to the atomic types of the first atom and the second atom; determining an atomic distance between the first atom and the second atom based on the three-dimensional positions of the first atom and the second atom; In response to the atomic distance being consistent with the chemical bonding distance, determining that the first atom and the second atom have a bonding relationship; In the i-th molecule graph, a line is added between the first atom and the second atom.

4. The method according to claim 3, characterized in that After determining all bonding relationships of the first atom, the method further includes: determining a charge range of the first atom according to the atomic type of the first atom; Using the charge range as a bond relationship verification condition, performing a bond order quantity verification on all bond relationships of the first atom; In response to the number of all bond orders of the first atom satisfying the bond relationship verification condition, determining to retain the connection with the first atom; In response to the number of all bond orders of the first atom not satisfying the bond relationship check condition, the type of the bond relationship between the first atom and other atoms is re-determined according to a cause causing the non-satisfaction.

5. The method according to claim 1, characterized in that After obtaining N molecular graphs of a multi-molecule system in a chemical simulation reaction, the method further includes: Performing graph connectivity calculation based on the connection between the first nodes in the molecular graph; According to the result of the graph connectivity calculation, the molecular graph is divided into multiple subgraphs, and different subgraphs are independent of each other in terms of graph connectivity.

6. The method according to claim 1, wherein For a first reactant among the reactants and a first product among the products, generating a reaction graph of the chemical simulation reaction in an i-th reaction frame based on the reactants and the products includes: determining atomic repetitiveness in the first reactant and the first product; In response to the first product having the same atom as the first reactant, adding a link between the second node identifying the first reactant and the third node identifying the first product; In response to the first product and the first reactant not having the same atoms, no link is added between the second node identifying the first reactant and the third node identifying the first product.

7. The method according to claim 1, characterized in that The determining, by using graph isomorphism between a subgraph of the target second node and a subgraph of the target third node, a reaction event for obtaining a product corresponding to the target third node from a reactant corresponding to the target second node includes: The reaction event of obtaining the product corresponding to the target third node from the reactant corresponding to the target second node is determined through the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node, the atomic consistency of the reactant corresponding to the target second node and the product corresponding to the target third node, and the three-dimensional geometric features of the subgraph of the target second node and the subgraph of the target third node.

8. The method according to claim 7, characterized in that According to the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node, the atomic consistency between the reactant corresponding to the target second node and the product corresponding to the target third node, and the three-dimensional geometric features of the subgraph of the target second node and the subgraph of the target third node, a matching result is obtained by matching with an event identification condition. In response to the matching result of the event identification condition, determining a reaction event of obtaining the product corresponding to the target third node from the reactant corresponding to the target second node includes: In response to the fact that atoms of the reactant corresponding to the target second node and the product corresponding to the target third node are not completely identical, determining that a reaction event in which the product corresponding to the target third node is obtained from the reactant corresponding to the target second node is a chemical reaction; In response to the fact that atoms of the reactant corresponding to the target second node and the product corresponding to the target third node are completely identical, but the subgraph of the target second node and the subgraph of the target third node are not isomorphic, determining that a reaction event in which the product corresponding to the target third node is obtained from the reactant corresponding to the target second node is a chemical reaction; In response to the fact that atoms of the reactant corresponding to the target second node and the product corresponding to the target third node are completely identical, and the subgraph of the target second node and the subgraph of the target third node are isomorphic, but the three-dimensional geometric features of the subgraph of the target second node and the subgraph of the target third node are inconsistent, determining that a reaction event in which the product corresponding to the target third node is obtained from the reactant corresponding to the target second node is an ionization reaction; In response to the reactant corresponding to the target second node and the product corresponding to the target third node being the same, it is determined that no reaction event occurs between the reactant corresponding to the target second node and the product corresponding to the target third node.

9. The method according to any one of claims 1 to 8, characterized in that The chemical simulation reaction simulates a reaction inside a battery having a reaction surface material. After determining that a reaction event occurs in which a product corresponding to the target third node is obtained from a reactant corresponding to the target second node, the method further includes: Determining the atomic type of the atoms included in the product corresponding to the target third node; In response to the product corresponding to the target third node including atoms of the reaction surface material, determining the reaction event as a reaction event occurring near the reaction surface material; In response to the product corresponding to the target third node not including atoms of the reaction surface material, the reaction event is determined as a reaction event far away from the reaction surface material.

10. The method according to any one of claims 1 to 8, characterized in that: After determining all reaction events of the chemical simulation reaction through the reaction graphs of the N reaction frames, the method further includes: Determine graph isomorphism between subgraphs of reactants in different reaction events, and determine graph isomorphism between subgraphs of products in different reaction events; In response to subgraph isomorphism between reactants of a first reaction event and a second reaction event, and subgraph isomorphism between products of the first reaction event and the second reaction event, determining that the first reaction event and the second reaction event are equivalent reactions; Equivalent reactions were combined across all reaction events.

11. The method according to any one of claims 1 to 8, characterized in that: For a second reactant among the reactants, after determining all reaction events of the chemical simulation reaction through the reaction graphs of N reaction frames, the method further includes: determining an event sequence of reaction events involving the second reactant in the N reaction frames; In response to the second product involved in the event sequence being equivalent to the second reactant, the j-th reaction event obtaining the second product to the 1st reaction event of the second reactant in the event sequence are collectively identified as transient reactions, j>1.

12. A reaction recognition device, characterized in that: The device includes: an acquisition module, a generation module and a determination module; The acquisition module is used to acquire N molecular graphs of a multi-molecule system in a chemical simulation reaction, where the chemical simulation reaction is completed through N reaction frames, the molecular graphs correspond to the reaction frames one-to-one, the molecular graphs include first nodes that identify atoms, a line between two first nodes is used to identify that the atoms corresponding to the two first nodes have a bond relationship, the i-th molecular graph is used to identify the bond relationship between atoms in the multi-molecule system in the i-th reaction frame of the chemical simulation reaction, the molecular graph includes subgraphs, and the subgraphs are used to identify molecules, where N>1; The generation module is used to determine the reactants of the chemical simulation reaction according to the subgraph of the first molecular graph, determine the products of the chemical simulation reaction according to the subgraph of the i-th molecular graph, and generate a reaction graph of the chemical simulation reaction in the i-th reaction frame based on the reactants and the products, wherein the reaction graph includes a second node identifying the reactants and a third node identifying the products, and a line between the second node and the third node is used to identify that there is a candidate reaction to be identified between the molecules corresponding to the second node and the third node respectively. <i≤N; The determination module is used to determine, for the target second node and the target third node having the candidate reaction, a reaction event in which the product corresponding to the target third node is obtained from the reactant corresponding to the target second node through the graph isomorphism between the subgraph of the target second node and the subgraph of the target third node.

13. A computer device, characterized in that: The computer device includes a processor and a memory: The memory is used to store computer programs; The processor is configured to execute the method according to any one of claims 1 to 11 according to the computer program.

14. A computer-readable storage medium, characterized in that The computer-readable storage medium is used to store a computer program, and when the computer program is executed by a computer device, the computer program implements the method according to any one of claims 1 to 11.

15. A computer program product comprising a computer program, which, when run on a computer device, causes the computer device to perform the method according to any one of claims 1 to 11.

Images