Reaction network simplification method and apparatus for molecular-level refining procedure, and device

Abstract

The present disclosure relates to the technical field of petrochemical engineering. Provided are a reaction network simplification method and apparatus for a molecular-level refining procedure, and a device. The method comprises: determining a reaction network topology graph corresponding to a target refining procedure; generating a transition probability matrix for nodes in the reaction network topology graph, and on the basis of the transition probability matrix, determining node importance in the reaction network topology graph; on the basis of the node importance of each node in the reaction network topology graph, executing an iterative node-pruning procedure, and stopping iteration until a first iteration stop condition is met; and outputting as a target reaction network a corresponding reaction network topology graph when the first iteration stop condition is met. By means of the embodiments of the present disclosure, the computational load of simulating a reaction procedure in a refining process can be reduced, and the hardware resource requirements for a computer device can be lowered.

Description

Methods, apparatus and equipment for simplifying reaction networks in molecular-level refining processes

[0001] Related applications

[0002] This application claims priority to Chinese Patent Application No. 202411613215.0, filed on November 12, 2024, and incorporates the disclosure of the aforementioned patent application as part of this application. Technical Field

[0003] This disclosure relates to the field of petrochemical technology, and in particular to a method, apparatus and equipment for simplifying reaction networks in molecular-level refining processes. Background Technology

[0004] In the field of petroleum processing, there are increasingly higher requirements for the yield and energy efficiency of molecular refining processes, which necessitates continuous iterative optimization of production line structure and production processes.

[0005] In realizing the concept disclosed herein, the inventors discovered at least the following technical problems in the related technologies: In the related technologies, numerical simulation models corresponding to molecular refining processes are constructed to simulate the processes of various production devices and production lines. During the construction of the production model corresponding to the production device, it is required that the constructed production model can reflect the molecular reaction process and correspond to the actual production process and production data. Therefore, some solutions construct reaction networks based on reaction mechanisms. For example, the reaction network for the hydrocracking process of vapor gas oil (VGO) is constructed based on a reaction mechanism model, using molecular reactions as the basic reactions. The constructed reaction network includes all molecules participating in the reaction process, and the mechanism model includes most of the details of the reaction process, closely resembling the actual chemical reaction process. However, since the actual refining process involves hundreds or thousands, or even more, the reaction network constructed in this way is enormous in scale, has low computational efficiency, and places high demands on the hardware resources of computer equipment when simulating the reaction process of the corresponding refining process. The computational load is huge, and it is prone to overload and crashes. Summary of the Invention

[0006] The purpose of this disclosure is to provide a method, apparatus, and equipment for simplifying molecular-level refining process reaction networks, so as to reduce the computational load and hardware resource requirements for simulating refining process reaction processes.

[0007] To achieve the above objectives, in one aspect, embodiments of this disclosure provide a method for simplifying the reaction network of a molecular-level refining process, including:

[0008] Determine the reaction network topology corresponding to the target refining process. The reaction network topology includes reactant molecule nodes, reaction nodes, and product molecule nodes. The edges between reactant molecule nodes and reaction nodes point from reactant molecule nodes to reaction nodes; the edges between reaction nodes and product molecule nodes point from reaction nodes to product molecule nodes.

[0009] Generate the transition probability matrix between each node in the reaction network topology graph, and determine the importance of the nodes in the reaction network topology graph based on the transition probability matrix;

[0010] Based on the importance of each node in the reaction network topology, an iterative process of node reduction is executed until the first iteration stopping condition is met and the iteration stops.

[0011] The topology of the reaction network corresponding to the first iteration stopping condition is used as the target reaction network output.

[0012] In the molecular-level refining process reaction network simplification method of this disclosure, a transition probability matrix between each node in the reaction network topology graph is generated, and the importance of the nodes in the reaction network topology graph is determined based on the transition probability matrix, including:

[0013] Obtain the connection relationships between nodes in the reactive network topology graph;

[0014] Construct an adjacency matrix A based on the connection relationships to represent the connections between nodes; the elements A in the adjacency matrix A are... ij Used to represent the directional connection state from node i to node j; element A ij The element values ​​are associated with the reaction parameter information corresponding to the two nodes;

[0015] The transition probability matrix is ​​determined based on the adjacency matrix to represent the transition probabilities between each node.

[0016] The proportion of each molecule at the start of the reaction in the target refining process is used as the initial value of the node importance of the corresponding molecule node. The node importance of each node is iteratively calculated according to the transition probability matrix until the second iteration stopping condition is met and the iteration stops.

[0017] Obtain the node importance of each node in the reaction network topology graph when the second iteration stopping condition is met.

[0018] In the molecular-level refining process reaction network simplification method of this disclosure, an iterative process of node reduction is performed based on the importance of each node in the reaction network topology graph, including:

[0019] Based on the importance ranking of each node in the reaction network topology graph, the K nodes with the lowest importance in the ranking are selected as candidate nodes.

[0020] S nodes are selected from K candidate nodes for deletion to minimize the deviation between the molecular-level refining simulation results and the actual refining results corresponding to the reduced reaction network topology; K and S are both positive integers, and K≥S;

[0021] Determine whether the reaction parameters need to be updated in the reaction network topology after the deletion;

[0022] If reaction parameters need to be updated, the reaction parameters of the reduced reaction network topology diagram are updated according to the actual product composition data corresponding to the reduced reaction network topology diagram.

[0023] The updated reaction network topology is used as the target for the next round of node reduction, so as to continue the next round of node reduction iteration.

[0024] In the molecular-level refining process reaction network simplification method of this disclosure, the first iteration stopping condition includes:

[0025] The current reaction network topology has reached the preset deletion range.

[0026] In the molecular-level refining process reaction network simplification method of this disclosure, the preset deletion range is a set deletion ratio or a set deletion quantity; the set deletion ratio refers to the ratio of the total number of deleted nodes to the total number of all nodes in the initial reaction network topology.

[0027] In the molecular-level refining process reaction network simplification method of this disclosure, the initial reaction network topology is pre-constructed in the following manner:

[0028] To obtain the various reactions involved in the target refining process at the molecular level;

[0029] Based on the reaction rules corresponding to each reaction and the interconversion relationships between each molecule, a path model for the formation of product molecules from reactant molecules based on the reaction rules is constructed.

[0030] Based on the Petri network form, a network topology diagram is constructed for the path model to obtain the initial reaction network topology diagram corresponding to the target refining process.

[0031] In the molecular-level refining process reaction network simplification method of this disclosure, the construction of a network topology graph based on the Petri network form of the path model includes:

[0032] For the first logical relationship from reactant molecule to first reaction in the path model, the reactant molecule is regarded as the place node in the reaction network topology graph in the form of Petri network, and the first reaction it points to is regarded as the transition node in the reaction network topology graph in the form of Petri network. The first logical relationship is regarded as the first directed edge in the reaction network topology graph in the form of Petri network.

[0033] For the second logical relation pointing from the second reaction to the product molecule in the path model, the second reaction is regarded as a place node in the reaction network topology graph in the form of a Petri network, the product molecule it points to is regarded as a transition node in the reaction network topology graph in the form of a Petri network, and the second logical relation is regarded as the second directed edge in the reaction network topology graph in the form of a Petri network.

[0034] In the molecular-level refining process reaction network simplification method of this disclosure embodiment, the edges in the reaction network topology graph are assigned edge weights; the edge weight value is determined according to the formula w = v × k; where w is the edge weight value, v is the reaction rate constant of the reaction node in the two nodes connected by the edge, and k is the stoichiometric coefficient of the reactant molecule node or product molecule node connected by the edge.

[0035] In the molecular-level refining process reaction network simplification method of this disclosure, determining the transition probability matrix to represent the transition probabilities between each node based on the adjacency matrix includes:

[0036] For each row in the adjacency matrix, calculate the sum of the values ​​of all elements in that row to obtain the sum W of the outgoing chain weights of node i corresponding to the row index. i ;

[0037] Each element A in the adjacency matrix ij The sum of the outgoing weights of the corresponding node i, W i Calculate the ratio to obtain the transition probability A of randomly jumping from node i to node j. ij / W i The transition probability A ij / W i As elements of the transition probability matrix.

[0038] In the molecular-level refining process reaction network simplification method of this disclosure, the second iteration stopping condition includes:

[0039] The number of iterations reaches a set threshold, or the node importance converges to the target value.

[0040] On the other hand, embodiments of this disclosure also provide a simplified apparatus for molecular-level refining process reaction networks, comprising:

[0041] The topology graph determination module is used to determine the reaction network topology graph corresponding to the target refining process. The reaction network topology graph includes reactant molecule nodes, reaction nodes, and product molecule nodes. The edges between reactant molecule nodes and reaction nodes point from reactant molecule nodes to reaction nodes; the edges between reaction nodes and product molecule nodes point from reaction nodes to product molecule nodes.

[0042] The importance determination module is used to generate the transition probability matrix between each node in the reaction network topology graph, and to determine the importance of the nodes in the reaction network topology graph based on the transition probability matrix.

[0043] The iterative reduction module is used to perform an iterative process of node reduction based on the importance of each node in the reaction network topology graph, until the iteration stops when the first iteration stopping condition is met;

[0044] The data output module is used to output the topology of the reaction network when the first iteration stopping condition is met as the target reaction network.

[0045] On the other hand, this disclosure also provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the computer program, when run by the processor, executes the instructions of the above-described method.

[0046] On the other hand, this disclosure also provides a computer storage medium storing a computer program thereon, which, when run by the processor of a computer device, executes instructions for the above-described method.

[0047] On the other hand, this disclosure also provides a computer program product, which includes a computer program that, when run by the processor of a computer device, executes instructions for the above-described method.

[0048] As can be seen from the technical solutions provided in the above embodiments of this disclosure, in each round of node reduction iteration, nodes in the current reaction network topology can be reduced based on the importance of nodes in the current reaction network topology. In this way, through multiple rounds of node reduction, the computational load and hardware resource requirements of the refining and chemical process reaction simulation can be significantly reduced, thereby avoiding or reducing the risk of computer equipment crashing due to computational overload, improving the simulation efficiency of the refining and chemical process simulation, and reducing the impact of reaction network simplification on simulation accuracy. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0050] Figure 1 shows a structural block diagram of a simplified system for a molecular-level refining process reaction network in some embodiments of this disclosure;

[0051] Figure 2 shows a flowchart of a method for simplifying the reaction network of a molecular-level refining process in some embodiments of this disclosure;

[0052] Figure 3 shows a flowchart of the method shown in Figure 2 for generating the transition probability matrix between each node in the reaction network topology graph and determining the importance of the nodes in the reaction network topology graph based on the transition probability matrix.

[0053] Figure 4 shows a flowchart of the iterative process of node reduction based on the importance of each node in the reaction network topology graph in the method shown in Figure 2.

[0054] Figure 5 shows a flowchart of a method for simplifying the reaction network of a molecular-level refining process in some other embodiments of this disclosure;

[0055] Figure 6 shows a schematic diagram of the reactive network topology in an exemplary embodiment of this disclosure;

[0056] Figure 7 shows a structural block diagram of a simplified reaction network apparatus for molecular-level refining processes in some embodiments of this disclosure;

[0057] Figure 8 shows a structural block diagram of a computer device in some embodiments of this disclosure.

[0058] [Explanation of reference numerals in the attached diagram] 10. Client; 20. Server; 71. Topology determination module; 72. Importance determination module; 73. Iterative reduction module; 74. Data output module; 802. Computer equipment; 804. Processor; 806. Memory; 808. Drive mechanism; 810. Input / output interface; 812. Input device; 814. Output device; 816. Presentation device; 818. Graphical user interface; 820. Network interface; 822. Communication link; 824. Communication bus. Detailed Implementation

[0059] To enable those skilled in the art to better understand the technical solutions in this disclosure, the technical solutions in the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this disclosure.

[0060] It should be noted that in the embodiments of this disclosure, the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved are all information and data authorized and agreed upon by the user and fully authorized by all parties. That is, the acquisition, transmission, storage, use, and processing of data in the technical solution of this disclosure all comply with the relevant provisions of national laws and regulations.

[0061] It should be noted that in the embodiments disclosed herein, certain software, components, models, and other existing solutions in the industry may be mentioned. These should be considered as exemplary and are intended only to illustrate the feasibility of implementing the technical solutions disclosed herein. However, they do not mean that the applicant has used or necessarily used such solutions.

[0062] Figure 1 illustrates an application environment diagram of some embodiments of this disclosure; this application environment includes a client 10 and a server 20. The client 10 can initiate a simplification of the reaction network topology corresponding to the target refining process to the server; the server 20 can determine the reaction network topology corresponding to the target refining process, which includes reactant molecule nodes, reaction nodes, and product molecule nodes. The edges between reactant molecule nodes and reaction nodes point from reactant molecule nodes to reaction nodes; the edges between reaction nodes and product molecule nodes point from reaction nodes to product molecule nodes; a transition probability matrix is ​​generated between each node in the reaction network topology, and the importance of nodes in the reaction network topology is determined based on the transition probability matrix; an iterative process of node reduction is executed based on the importance of each node in the reaction network topology until the first iteration stopping condition is met; the reaction network topology corresponding to the first iteration stopping condition is output as the target reaction network. Through the embodiments of this disclosure, the computational load and hardware resource requirements for simulating refining process reactions can be reduced, simulation efficiency can be improved, and the impact of simplification on simulation accuracy can be reduced.

[0063] In some embodiments of this disclosure, client 10 can be a self-service terminal device, a mobile terminal (i.e., a smartphone), a display, a desktop computer, a tablet computer, a laptop computer, a digital assistant, or a smart wearable device, etc. Smart wearable devices can include smart bracelets, smartwatches, smart glasses, or smart helmets, etc. Of course, client 10 is not limited to the aforementioned physical electronic devices; it can also be software running on the aforementioned electronic devices.

[0064] In some embodiments of this disclosure, the server 20 may be an electronic device with computing and network interaction functions; or it may be software running on the electronic device that provides business logic for data processing and network interaction.

[0065] Furthermore, it should be noted that Figure 1 shows only one application environment provided by this disclosure. In actual applications, there can be multiple clients 10 and multiple servers 20. This disclosure does not impose any restrictions.

[0066] This disclosure provides a method for simplifying the reaction network of a molecular-level refining process, which can be applied to the server side described above. Referring to Figure 2, in some embodiments of this disclosure, the method for simplifying the reaction network of a molecular-level refining process may include the following steps:

[0067] Step 201: Determine the reaction network topology corresponding to the target refining process.

[0068] Step 202: Generate the transition probability matrix between each node in the reaction network topology graph, and determine the importance of the nodes in the reaction network topology graph based on the transition probability matrix.

[0069] Step 203: Based on the importance of each node in the reaction network topology graph, perform an iterative process of node reduction until the first iteration stopping condition is met and the iteration stops.

[0070] Step 204: The topology of the reaction network corresponding to the first iteration stopping condition is taken as the target reaction network output.

[0071] In this embodiment of the disclosure, in each round of node reduction iteration, nodes in the current reaction network topology can be reduced based on the importance of the nodes in the current reaction network topology. In this way, through multiple rounds of node reduction, the computational load and hardware resource requirements of the refining and chemical process reaction simulation can be significantly reduced, thereby avoiding or reducing the risk of computer equipment crashing due to computational overload, improving the simulation efficiency of the refining and chemical process simulation, and reducing the impact of reaction network simplification on simulation accuracy.

[0072] In some embodiments of this disclosure, the target refining process refers to a petroleum refining process at the molecular level, which can employ any suitable refining technology (e.g., naphtha refining). The reaction network topology diagram is a schematic diagram of the network topology structure of the numerical simulation network for the target refining process. The reaction network topology diagram includes reactant molecule nodes (reactant molecule nodes represent the input petroleum molecules), reaction nodes (reaction nodes represent the refining process of petroleum molecules), and product molecule nodes (product molecule nodes represent the molecules of the final output product). Edges between reactant molecule nodes and reaction nodes point from the reactant molecule node to the reaction node; edges between reaction nodes and product molecule nodes point from the reaction node to the product molecule node. Therefore, the reaction network topology diagram is a directed graph.

[0073] In some embodiments of this disclosure, determining the reaction network topology corresponding to the target refining process may include: obtaining the initial reaction network topology during the first node reduction iteration; and obtaining the reaction network topology after the (R-1)th node reduction iteration during the Rth (R≥2 and is a positive integer)th node reduction iteration.

[0074] In some embodiments of this disclosure, the initial reaction network topology can be pre-constructed in the following manner:

[0075] (1) Obtain the various reactions involved in the target refining process at the molecular level.

[0076] For example, taking naphtha refining as an example, the reaction can include, but is not limited to, atmospheric and vacuum distillation, catalytic cracking, hydrocracking and other reaction processes.

[0077] (2) Based on the reaction rules corresponding to each reaction and the interconversion relationship between each molecule, construct a path model for reactant molecules to generate product molecules based on the reaction rules.

[0078] The reaction rules can include those for converting reactants into intermediates and those for converting intermediates into final products. Path models are simplified versions of reaction mechanisms and actual reaction processes, representing the molecular-level transformation relationships of petroleum during refining (the reaction network constructed by the path model includes molecules and their interconversion relationships). The reactions that molecules can undergo can be given by the reaction rules; therefore, path models are more focused and computationally simpler.

[0079] (3) Constructing a network topology diagram of the path model based on the Petri-net form to obtain the initial reaction network topology diagram corresponding to the target refining process. The construction of the network topology diagram of the path model based on the Petri-net form may include:

[0080] For the first logical relationship from reactant molecule to first reaction in the path model, the reactant molecule is regarded as the place node in the reaction network topology graph in the form of Petri network, and the first reaction it points to is regarded as the transition node in the reaction network topology graph in the form of Petri network. The first logical relationship is regarded as the first directed edge in the reaction network topology graph in the form of Petri network.

[0081] For the second logical relation pointing from the second reaction to the product molecule in the path model, the second reaction is regarded as a place node in the reaction network topology graph in the form of a Petri network, the product molecule it points to is regarded as a transition node in the reaction network topology graph in the form of a Petri network, and the second logical relation is regarded as the second directed edge in the reaction network topology graph in the form of a Petri network.

[0082] In the embodiments of this disclosure, the reactive network topology graph is constructed based on the Petri network, which has significant advantages in terms of flexibility (scalability), formal verification, concurrency control, and ease of maintenance.

[0083] Referring to Figure 3, in some embodiments of this disclosure, generating a transition probability matrix between nodes in the reaction network topology graph and determining the importance of nodes in the reaction network topology graph based on the transition probability matrix may include the following steps:

[0084] Step 301: Obtain the connection relationships between nodes in the reaction network topology graph.

[0085] Step 302: Construct an adjacency matrix to represent the connection relationships between nodes based on the connection relationships.

[0086] Element A in adjacency matrix A ij Used to represent the directional connection state from node i to node j; element A ij The element values ​​are associated with the reaction parameter information corresponding to the two nodes (i.e., node i and node j). i and j both represent the node index, with values ​​ranging from 1 to N, where N is the total number of nodes.

[0087] In some embodiments of this disclosure, in the adjacency matrix A, the diagonal elements are zero; the elements corresponding to two nodes that are not directly connected by an edge are zero; the elements corresponding to nodes that are directly connected by an edge but whose index order is inconsistent with the edge pointing order are zero; each element in the same row represents the connection state from node i corresponding to the row number to each node j = 1 to N in the above reaction network topology diagram.

[0088] In some embodiments of this disclosure, the above-mentioned reaction parameter information includes: the reaction rate constant corresponding to the reaction, and the stoichiometric coefficients corresponding to the reactant molecules or product molecules; wherein, the connection state value of two nodes with a connection relationship is equal to the corresponding edge weight value, that is, the edges in the reaction network topology graph have edge weights; the edge weight value of the edge is determined according to the formula w = v × k; wherein w is the edge weight value of the edge, v is the reaction rate constant of the reaction node in the two nodes connected by the edge, and k is the stoichiometric coefficient of the reactant molecule node or product molecule node connected by the edge.

[0089] For example, in the reaction network topology shown in Figure 6, the total number of nodes N is equal to the sum of the number of molecules M (M=4) and the number of reactions U (U=1). As a simplified example, Figure 6 illustrates a reaction network topology with 4 molecular nodes and 1 reaction node. The molecular nodes are J1 to J4, and the corresponding reaction node is T1. The edge from molecular node J1 to reaction node T1 is LB1, the edge from molecular node J2 to reaction node T1 is LB2, the edge from reaction node T1 to molecular node J3 is LB3, and the edge from reaction node T1 to molecular node J4 is LB4. The edge weight of the above-mentioned connection LB1 is equal to the reaction rate constant of reaction node T1 multiplied by the stoichiometric coefficient of reactant molecule node J1; the edge weight of the above-mentioned connection LB2 is equal to the reaction rate constant of reaction node T1 multiplied by the stoichiometric coefficient of reactant molecule node J2; the edge weight of the above-mentioned connection LB3 is equal to the reaction rate constant of reaction node T1 multiplied by the stoichiometric coefficient of product molecule node J3; the edge weight of the above-mentioned connection LB4 is equal to the reaction rate constant of reaction node T1 multiplied by the stoichiometric coefficient of product molecule node J4.

[0090] For example, taking Figure 6 as an example, the adjacency matrix A of the reaction network topology containing 5 nodes (molecular nodes J1~J4 and reaction node T1) is 5×5 in size; the diagonal elements A 11 A 22 A 33 A 44 A 55 The values ​​of all elements are zero. Referring to Figure 6, the two nodes directly connected without any edges correspond to (in order): node 1 and node 2; node 1 and node 3; node 1 and node 4; node 2 and node 1; node 2 and node 3; node 2 and node 4; node 3 and node 1; node 3 and node 2; node 3 and node 4; node 4 and node 1; node 4 and node 2; node 4 and node 3. Therefore, the following elements in the adjacency matrix A have values ​​of zero: A 12A 13 A 14 A 21 A 23 A 24 A 31 A 32 A 34 A 41 A 42 A 43 .

[0091] Additionally, for nodes directly connected by an edge but whose element indexes are out of order with respect to the edge, the following elements all have a value of zero: A 35 A 45 A 51 A 52 For two directly connected nodes that are directional to each other, the following elements have non-zero values: A 15 A 25 A 53 A 54 The specific element values ​​are associated with the reaction parameter information corresponding to these two nodes. For example, A 15 = The reaction rate constant of reaction node T1 multiplied by the stoichiometric coefficient of reactant molecule node J1 (i.e., the edge weight value of edge LB1); A 25 = The reaction rate constant of reaction node T1 multiplied by the stoichiometric coefficient of reactant molecule node J2 (i.e., the edge weight value of edge LB2); A 53 = The reaction rate constant of reaction node T1 multiplied by the stoichiometric coefficient of product molecule node J3 (i.e., the edge weight value of edge LB3); A 54 = The reaction rate constant of reaction node T1 multiplied by the stoichiometric coefficient of product molecule node J4 (which is the edge weight value of edge LB4).

[0092] Step 303: Determine the transition probability matrix to represent the transition probabilities between each node based on the adjacency matrix.

[0093] In some embodiments of this disclosure, determining a transition probability matrix representing the transition probabilities between nodes based on the adjacency matrix may include:

[0094] (1) For each row of the adjacency matrix, calculate the sum of the values ​​of each element in that row to obtain the sum of the outgoing chain weights W of the node i corresponding to the row number. i ;

[0095] (2) Each element A in the adjacency matrix ij The sum of the outgoing weights of the corresponding node i, W iCalculate the ratio to obtain the transition probability A of randomly jumping from node i to node j. ij / W i The transition probability A ij / W i As elements of the transition probability matrix.

[0096] Wherein, the elements P of the transition probability matrix ij Satisfy the following expression:

[0097] (3) The process of iteratively calculating the node importance value of each node based on the above transition probability matrix P can be based on the following expression:

[0098] Where I represents the node importance vector corresponding to each element of the column vector for all nodes (e.g., N = 5 nodes here), E is a vector in which all elements are 1, and α is a random jump factor. Setting the random jump factor α solves problems such as "infinite loops" and "spider traps" (circular link structures) that exist during the iteration process. For example, in an exemplary embodiment of this disclosure, the value of α can be set to 0.85.

[0099] Step 304: Take the proportion of each molecule at the start of the reaction in the target refining process as the initial value of the node importance of the corresponding molecular node, and iteratively calculate the node importance of each node according to the transition probability matrix until the second iteration stopping condition is reached and the iteration stops.

[0100] Step 305: Obtain the node importance of each node in the reaction network topology graph corresponding to the second iteration stopping condition.

[0101] In some embodiments of this disclosure, the second iteration stopping condition may include: the number of iterations reaching a set threshold, or the node importance converging to a target value.

[0102] Referring to Figure 4, the iterative process of node reduction based on the importance of each node in the reaction network topology graph can include the following steps:

[0103] Step 401: Sort the nodes according to their importance in the reaction network topology graph, and select the K nodes with the lowest importance in the sort as candidate nodes.

[0104] Step 402: Select S nodes from the K candidate nodes for deletion, so that the deviation between the molecular-level refining simulation results and the actual refining results corresponding to the deleted reaction network topology is minimized; K and S are both positive integers, and K≥S.

[0105] In some embodiments of this disclosure, the values ​​of K and S can be preset by the user or generated according to a preset reduction range, that is, both K and S can vary. For example, if K is 8, then the value of S can be any one of {1,2,3,4,5,6,7,8}.

[0106] In some embodiments of this disclosure, a series of candidate reduction ratios can be generated using S / K, and for different candidate reduction ratios, the deviation between the molecular-level refining simulation results corresponding to the reduced reaction network topology and the actual refining results can be compared, so as to minimize the deviation between the molecular-level refining simulation results corresponding to the reduced reaction network topology and the actual refining results.

[0107] In some embodiments of this disclosure, when the number of candidate nodes is determined, multiple preset ratio options for candidate deletion ratios can be generated according to a system preset algorithm (which may be user settings or randomly generated by the computing system of an electronic device), and a target preset ratio option can be selected subsequently. The specific selection logic may be to make differentiated selections based on the relative size of each preset ratio option and the set candidate deletion ratio threshold.

[0108] This reduction method can significantly decrease the computational load and hardware resource requirements for simulating refining and chemical reaction processes, while also mitigating the impact of reaction network simplification on simulation accuracy.

[0109] Step 403: Determine whether the reaction parameters need to be updated after the deletion of the reaction network topology; if the reaction parameters need to be updated, proceed to step 404; otherwise, proceed to step 406.

[0110] For a reaction node model constructed based on the complete reaction network (also known as the full reaction network), the reaction parameters it contains are optimized based on actual production data for the complete reaction network. Considering that during the process of deleting nodes from the reaction network, the original reaction rate parameters will not be perfectly matched with the simplified reaction network, by setting logic to determine whether the reaction parameters are adapted for updating and logic to update the reaction parameters and use the updated target reaction network topology as the reaction network topology for the next round, the reaction network topology during the iteration process can be made more accurate, and simplification helps maintain the simulation accuracy of the target refining process. Of course, after simplifying the reaction network, when using the simplified reaction network to simulate the refining process, the reaction parameters can still be further fitted and optimized to obtain better reaction parameters (such as reaction rates).

[0111] Step 404: Update the reaction parameters of the reduced reaction network topology diagram based on the actual product composition data corresponding to the reduced reaction network topology diagram.

[0112] Referring to Figure 5, in some other embodiments of this disclosure, taking a complete reaction network as an example, the first-layer optimization model can be the reaction network optimization and reduction based on node importance ranking, and the second-layer optimization model can be the optimization of reaction rate parameters. The optimization of reaction rate parameters is incorporated into the reaction network simplification process, simultaneously achieving reaction network simplification and reaction parameter optimization (the specific calculation process is shown in Figure 5). That is, according to the set frequency of reaction parameter optimization, reaction rate parameter optimization is performed once after a certain number of reactions are removed. The parameter optimization process involves optimizing and updating the reaction rate parameters based on the removed reaction network and the actual product composition data, and then substituting the optimized reaction rate parameters into the reaction network.

[0113] Step 405: Use the updated reaction network topology as the target for the next round of node deletion.

[0114] Step 406: Determine whether the first iteration stopping condition has been met; if the first iteration stopping condition has not been met, proceed to step 407; if the first iteration stopping condition has been met, the process ends (stopping the iteration by deleting nodes), thus completing the simplification of the reaction network for the molecular-level refining process. The reaction network obtained at this point can be used for subsequent molecular-level refining numerical simulations.

[0115] In some embodiments of this disclosure, the first iteration stopping condition may include: the current reaction network topology graph reaches a preset reduction range; the preset reduction range is a set reduction ratio or a set reduction quantity; the set reduction ratio refers to the ratio of the total number of nodes removed to the total number of all nodes in the initial reaction network topology graph.

[0116] Step 407: Continue with the next round of node reduction iteration.

[0117] In some embodiments of this disclosure, if the value of S to be removed each time is small, the number of removals will be too large, and the calculation will take a long time; if the value of S to be removed each time is large, the optimal removal ratio range may be too rough and inaccurate results cannot be obtained. Taking all factors into consideration, the value of S to be removed each time can be about 1% of the total number of reactions.

[0118] In other embodiments of this disclosure, the K candidate nodes and S nodes to be deleted can be determined through the following steps:

[0119] (1) Obtain the preset ratio options for setting the sorting selection quantity and the candidate deletion ratio; the above-mentioned sorting selection quantity is used as the first quantity K of candidate nodes in this round.

[0120] The above-mentioned sorting selection quantity is the given sorting selection quantity when sorting according to the node importance.

[0121] In some embodiments of this disclosure, with a fixed number of candidate nodes, multiple preset ratio options for candidate deletion ratios can be generated according to a system preset algorithm (which may be user-set or randomly generated by the system), and a target preset ratio option can be selected subsequently. The specific selection logic may be to make differentiated selections based on the relative size of each preset ratio option and the set candidate deletion ratio threshold.

[0122] The aforementioned candidate deletion ratio threshold is used to indicate the degree of influence of changes in the preset ratio option on the deviation corresponding to the molecular-level refining calculation results after node deletion (this deviation refers to the deviation between the molecular-level refining calculation results after node deletion and the actual refining results). It corresponds to a turning point in the impact; when the threshold is less than the aforementioned candidate deletion ratio threshold, different preset ratio options have significantly different effects on the deviation corresponding to the molecular-level refining calculation results after node deletion. In other words, using different preset ratio options results in significantly different effects on the deviation between the molecular-level refining calculation results after node deletion and the actual refining results. Therefore, the comparison and selection logic of step (31) is adopted; after the preset ratio option is greater than the set candidate deletion ratio threshold, the difference in the influence of different preset ratio options on the corresponding deviation of the molecular level refining calculation results after the above node deletion is very small, that is, the difference in the influence of different preset ratio options on the deviation between the molecular level refining calculation results after the above node deletion and the actual refining results is very small. Therefore, the random selection logic of step (32) is adopted; for the case that includes preset ratio options that are both less than the threshold and higher than the threshold, the comparison and selection logic and the random selection logic are combined and the comparison and selection are performed a second time to determine the second quantity S (step (33) is adopted).

[0123] (2) Determine the relative size of each preset ratio option and the set candidate deletion ratio threshold L.

[0124] (31) When all preset ratio options are less than the set candidate deletion ratio threshold L, for different preset ratio options of candidate deletion ratio, calculate the deviation between the molecular level refining calculation result after node deletion in the reaction network topology and the actual refining result, and compare them. The second quantity corresponding to the target preset ratio option with the smallest deviation is determined as the second quantity S of the nodes to be deleted in this round.

[0125] (32) When all preset ratio options are higher than the set candidate deletion ratio threshold L, randomly select the second quantity corresponding to the target preset ratio option as the second quantity S of the nodes to be deleted in this round.

[0126] (33) When there is a first preset ratio option that is less than the set candidate deletion ratio threshold L and a second preset ratio option that is higher than the set candidate deletion ratio threshold L, select the first target preset ratio option with the smallest deviation from the first preset ratio option according to the deviation between the molecular level refining calculation result after node deletion and the actual refining result; randomly select a second target preset ratio option from the second preset ratio option; compare the deviation between the molecular level refining calculation result after node deletion and the actual refining result corresponding to the first target preset ratio option and the second target preset ratio option, and determine the second quantity corresponding to the preset ratio option with the smaller deviation as the second quantity S of the nodes to be deleted in this round.

[0127] Based on the above steps, given K, different processing logics are determined according to the relative size of the preset ratio option of the candidate deletion ratio and the set candidate deletion ratio threshold L. Thus, the value of S is determined according to the preset ratio option corresponding to the minimum deviation between the molecular-level refining calculation result after node deletion and the actual refining result. Based on the above S and K, the selection of the number of subsequent candidate nodes and the number of reaction results to be deleted can simplify the reaction network while ensuring the simulation accuracy of the simplified reaction network, thereby greatly reducing the simulation computation of the subsequent molecular refining process.

[0128] Although the process described above includes multiple operations that occur in a specific order, it should be clearly understood that these processes may include more or fewer operations, which may be executed sequentially or in parallel (e.g., using parallel processors or a multithreaded environment).

[0129] Corresponding to the above-described method for simplifying the reaction network of a molecular-level refining process, this disclosure also provides a device for simplifying the reaction network of a molecular-level refining process, which can be configured on the aforementioned server. Referring to Figure 7, in some embodiments of this disclosure, the device for simplifying the reaction network of a molecular-level refining process may include:

[0130] The topology graph determination module 71 is used to determine the reaction network topology graph corresponding to the target refining process. The reaction network topology graph includes reactant molecule nodes, reaction nodes, and product molecule nodes. The edges between reactant molecule nodes and reaction nodes point from reactant molecule nodes to reaction nodes; the edges between reaction nodes and product molecule nodes point from reaction nodes to product molecule nodes.

[0131] Importance determination module 72 is used to generate the transition probability matrix between each node in the reaction network topology graph, and to determine the importance of the nodes in the reaction network topology graph based on the transition probability matrix.

[0132] The iterative reduction module 73 is used to perform an iterative process of node reduction based on the importance of each node in the reaction network topology diagram, until the iteration stops when the first iteration stopping condition is met.

[0133] The data output module 74 is used to output the topology of the reaction network corresponding to the first iteration stopping condition as the target reaction network.

[0134] For ease of description, the above apparatus is described by dividing it into various functional units. Of course, in implementing this disclosure, the functions of each unit can be implemented in one or more software and / or hardware.

[0135] Embodiments of this disclosure also provide a computer device. As shown in FIG8, in some embodiments of this disclosure, the computer device 802 may include one or more processors 804, such as one or more central processing units (CPUs) or graphics processing units (GPUs), each processing unit implementing one or more hardware threads. The computer device 802 may also include any memory 806 for storing any kind of information such as code, settings, data, etc. In one specific embodiment, a computer program is stored on the memory 806 and can run on the processor 804. When the computer program is run by the processor 804, it can execute instructions of the molecular-level refining process reaction network simplification method of any of the above embodiments. Without limitation, for example, the memory 806 may include any type of RAM, any type of ROM, flash memory device, hard disk, optical disk, etc. More generally, any memory can use any technology to store information. Further, any memory can provide volatile or non-volatile retention of information. Further, any memory can represent a fixed or removable component of the computer device 802. In one scenario, when processor 804 executes associated instructions stored in any memory or combination of memories, computer device 802 can perform any operation of the associated instructions. Computer device 802 also includes one or more drive mechanisms 808 for interacting with any memory, such as hard disk drive mechanisms, optical disk drive mechanisms, etc.

[0136] Computer device 802 may also include an input / output interface 810 (I / O) for receiving various inputs (via input device 812) and providing various outputs (via output device 814). A specific output mechanism may include a presentation device 816 and an associated graphical user interface 818 (GUI). In other embodiments, the input / output interface 810 (I / O), input device 812, and output device 814 may be omitted, and the device may function solely as a computer device within a network. Computer device 802 may also include one or more network interfaces 820 for exchanging data with other devices via one or more communication links 822. One or more communication buses 824 couple the components described above together.

[0137] Communication link 822 can be implemented in any way, such as via a local area network, a wide area network (e.g., the Internet), a point-to-point connection, or any combination thereof. Communication link 822 may include any combination of hardwired links, wireless links, routers, gateway functions, name servers, etc., governed by any protocol or combination of protocols.

[0138] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), computer-readable storage media, and computer program products according to some embodiments of this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processor to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processor, create means for implementing the functions specified in one or more flowchart illustrations and / or one or more block diagrams.

[0139] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processor to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0140] These computer program instructions may also be loaded onto a computer or other programmable data processor to cause a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable device, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0141] In a typical configuration, a computer device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0142] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0143] Computer-readable media, including both permanent and non-permanent, removable and non-removable media, can store information using any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by computer equipment. As defined in this disclosure, computer-readable media does not include transient media, such as modulated data signals and carrier waves.

[0144] Those skilled in the art will understand that embodiments of this disclosure can be provided as methods, systems, or computer program products. Therefore, embodiments of this disclosure can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, embodiments of this disclosure can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0145] Embodiments of this disclosure can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. Embodiments of this disclosure can also be practiced in distributed computing environments where tasks are performed by remote processors connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0146] It should also be understood that, in the embodiments of this disclosure, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Furthermore, the character " / " in this disclosure generally indicates that the preceding and following related objects have an "or" relationship.

[0147] The various embodiments in this disclosure are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0148] In the description of this disclosure, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the embodiments of this disclosure. In this disclosure, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in a suitable manner in any one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this disclosure, as well as the features of different embodiments or examples.

[0149] The above description is merely an embodiment of this disclosure and is not intended to limit the scope of this disclosure. Various modifications and variations can be made to this disclosure by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the scope of the claims of this disclosure.

Claims

1. A method for simplifying a reaction network of a molecular level refining process, characterized by, include: Determine the reaction network topology corresponding to the target refining process. The reaction network topology includes reactant molecule nodes, reaction nodes, and product molecule nodes. The edges between reactant molecule nodes and reaction nodes point from reactant molecule nodes to reaction nodes. The edges between reaction nodes and product molecule nodes point from reaction nodes to product molecule nodes. Generate the transition probability matrix between each node in the reaction network topology graph, and determine the importance of the nodes in the reaction network topology graph based on the transition probability matrix; Based on the importance of each node in the reaction network topology, an iterative process of node reduction is executed until the first iteration stopping condition is met and the iteration stops. The topology of the reaction network corresponding to the first iteration stopping condition is used as the target reaction network output.

2. The method of claim 1, wherein the molecular level refining process reaction network is simplified by: The step of generating the transition probability matrix between each node in the reaction network topology graph, and determining the importance of nodes in the reaction network topology graph based on the transition probability matrix, includes: Obtain the connection relationships between nodes in the reaction network topology graph; According to the connection relationship, an adjacency matrix A for representing the connection relationship between each node is constructed; an element A ij of the adjacency matrix A represents the connection state of the node i to the node j with directionality; the element A ij of the element value is associated with the reaction parameter information corresponding to the two nodes. Based on the adjacency matrix, a transition probability matrix is ​​determined to represent the transition probability between each node; The proportion of each molecule at the start of the reaction in the target refining process is used as the initial value of the node importance of the corresponding molecule node. The node importance of each node is iteratively calculated according to the transition probability matrix until the second iteration stopping condition is met and the iteration stops. Obtain the node importance of each node in the reaction network topology graph corresponding to the second iteration stopping condition.

3. The method of claim 1, wherein the molecular level refining process reaction network is simplified by: The iterative process of removing nodes based on their importance in the reaction network topology includes: Based on the node importance ranking of each node in the reaction network topology graph, the K nodes with the lowest node importance in the ranking are selected as candidate nodes. S nodes are selected from K candidate nodes for deletion to minimize the deviation between the molecular-level refining simulation results and the actual refining results corresponding to the reduced reaction network topology; K and S are both positive integers, and K≥S; Determine whether the reaction parameters need to be updated in the reaction network topology after the deletion; If reaction parameter updates are required, the reaction parameters of the reduced reaction network topology are updated according to the actual product composition data corresponding to the reduced reaction network topology. The updated reaction network topology is used as the target for the next round of node reduction, so as to continue the next round of node reduction iteration.

4. The method of claim 1, wherein the molecular level refining process reaction network simplification method is characterized by, The first iteration stopping condition includes: The current reaction network topology has reached the preset deletion range.

5. The method of claim 4, wherein the molecular level refining process reaction network is simplified by: The preset deletion range is a set deletion ratio or a set deletion quantity; the set deletion ratio refers to the ratio of the total number of nodes deleted to the total number of nodes in the initial reaction network topology graph.

6. The method of claim 1, wherein the molecular level refining process reaction network simplification method is characterized by, The initial reaction network topology is constructed in advance using the following method: To obtain the various reactions involved in the target refining process at the molecular level; Based on the reaction rules corresponding to each reaction and the interconversion relationships between each molecule, a path model for the formation of product molecules from reactant molecules based on the reaction rules is constructed. The path model is constructed using the Petri network form to obtain the initial reaction network topology corresponding to the target refining process.

7. The method of claim 6, wherein the molecular level refining process reaction network is simplified by: The construction of the network topology graph of the path model based on the Petri network form includes: For the first logical relationship from reactant molecule to first reaction in the path model, the reactant molecule is regarded as a place node in the reaction network topology graph in the form of Petri network, the first reaction is regarded as a transition node in the reaction network topology graph in the form of Petri network, and the first logical relationship is regarded as the first directed edge in the reaction network topology graph in the form of Petri network. For the second logical relationship pointing from the second reaction to the product molecule in the path model, the second reaction is regarded as a place node in the reaction network topology graph in the form of a Petri network, the product molecule it points to is regarded as a transition node in the reaction network topology graph in the form of a Petri network, and the second logical relationship is regarded as a second directed edge in the reaction network topology graph in the form of a Petri network.

8. The method of claim 1, wherein the molecular level refining process reaction network simplification method is characterized by, The edges in the reaction network topology graph are assigned edge weights; the edge weight values ​​are determined according to the formula w = v × k; where w is the edge weight value, v is the reaction rate constant of the reaction node in the two nodes connected by the edge, and k is the stoichiometric coefficient of the reactant molecule node or product molecule node connected by the edge.

9. The method of claim 2, wherein the molecular level refining process reaction network is simplified by: The step of determining the transition probability matrix representing the transition probabilities between each node based on the adjacency matrix includes: For each row element in the adjacency matrix, the element value sum of each element of the row is calculated to obtain the out-link weight sum W of the node i corresponding to the row number of the row i ; Each element A ij of the adjacency matrix is the sum of the out-link weights W i of the nodes i that are adjacent to the node j. ij The transition probability A i from node i to node j is calculated as the ratio of the sum of the out-link weights W ij of the nodes i that are adjacent to the node j to the sum of the out-link weights W i of all nodes i.

10. The method of claim 2, wherein the molecular level refining process reaction network is simplified by: The second iteration stopping condition includes: The number of iterations reaches a set threshold, or the node importance converges to the target value.

11. A simplified reaction network device for a molecular-level refining process, characterized in that, include: The topology graph determination module is used to determine the reaction network topology graph corresponding to the target refining process. The reaction network topology graph includes reactant molecule nodes, reaction nodes, and product molecule nodes. The edges between reactant molecule nodes and reaction nodes point from the reactant molecule nodes to the reaction nodes; the edges between reaction nodes and product molecule nodes point from the reaction nodes to the product molecule nodes. The importance determination module is used to generate a transition probability matrix between each node in the reaction network topology graph, and to determine the importance of the nodes in the reaction network topology graph based on the transition probability matrix. The iterative reduction module is used to perform an iterative process of node reduction based on the importance of each node in the reaction network topology graph, until the iteration stops when the first iteration stopping condition is met. The data output module is used to output the reaction network topology diagram corresponding to the first iteration stopping condition as the target reaction network.

12. A computer device comprising a memory, a processor, and a computer program stored on the memory, wherein, When the computer program is run by the processor, it executes the instructions of the method according to any one of claims 1-10.

13. A computer storage medium having stored thereon a computer program, characterized in that When the computer program is run by the processor of the computer device, it executes the instructions of the method according to any one of claims 1-10.

14. A computer program product, characterised in that, The computer program product includes a computer program that, when run by the processor of a computer device, executes instructions according to any one of claims 1-10.

Images