Oxidation kinetic index calculation method and device, electronic equipment and storage medium

Abstract

The application discloses an oxidation kinetic index calculation method and device, electronic equipment and a storage medium, and relates to the technical field of alloys. An oxidation characteristic tree is acquired, the oxidation characteristic tree is a binary tree, includes nodes and node edges, the nodes include a root node and end nodes, and the node edges store alloy data values. Then, nodes are divided from the alloy data according to alloy characteristic data, a target node edge is determined one by one from the root node according to a comparison result of the alloy characteristic data and alloy characteristic values on the node edges, until the end nodes are reached, and the end nodes are determined as target end nodes. Finally, an oxidation kinetic index in the target end nodes is taken as the oxidation kinetic index of the alloy data. Thus, according to the pre-constructed oxidation characteristic tree, the oxidation kinetic index and the oxidation kinetic equation can be accurately calculated and acquired, which is of great significance to tasks such as oxidation process description, oxidation-resistant material mechanism research and oxidation-resistant design, thereby promoting the development of related research.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of alloys, in particular to an oxidation kinetic index calculation method and device, electronic equipment and storage medium. BACKGROUND

[0002] Oxidation of alloys is an important oxidation reaction process, which has an important influence on the performance and life of materials. Understanding the kinetic behavior of alloy oxidation, i.e. the rate and variation of weight gain during oxidation, is of great significance for mechanism research and oxidation resistance design of oxidation-resistant materials. By establishing an oxidation kinetic equation, the oxidation weight gain process of alloys can be better described and predicted, and the relationship between the oxidation rate and test conditions such as time, temperature, and pressure can be understood.

[0003] The oxidation kinetic index is one of the important parameters for establishing the oxidation kinetic equation. In related technologies, the oxidation kinetic index is usually a fixed value, which is used to describe the oxidation law of a specific alloy under specific oxidation conditions. However, this method cannot study the quantitative relationship between the oxidation kinetic index itself and the alloy composition and test conditions, nor can it comprehensively describe the oxidation characteristics of multiple alloys under different test conditions. Therefore, it is difficult to calculate the oxidation kinetic index and the corresponding oxidation kinetic equation for different alloy data, hindering the further development of related research. SUMMARY

[0004] The present application aims to at least solve one of the technical problems existing in the prior art. To this end, the present application provides an oxidation kinetic index calculation method and device, electronic equipment and storage medium, which can calculate the oxidation kinetic index for different alloy data and promote the development of related research.

[0005] In a first aspect, the present application provides an oxidation kinetic index calculation method, comprising:

[0006] Obtaining a pre-constructed oxidation feature tree, the oxidation feature tree being a binary tree, the binary tree comprising a plurality of nodes and node edges, the nodes comprising a root node and at least two end nodes, and the node edges storing alloy feature values, the alloy feature values being boundaries for alloy data division;

[0007] From the alloy data, the nodes are divided according to alloy feature data, starting from the root node, and the target node edges are determined one by one according to the comparison results of the alloy feature data and the alloy feature values on the node edges, until the end nodes are reached, and the end nodes are determined as target end nodes;

[0008] The oxidation kinetic index stored in the target end node is taken as the oxidation kinetic index of the alloy data.

[0009] In some embodiments of the present application, the constructing the oxidation feature tree comprises:

[0010] obtaining an oxidation dataset; wherein the oxidation dataset comprises a plurality of alloy samples, each of the alloy samples comprises a plurality of kinetic features and corresponding kinetic feature values;

[0011] performing a kinetic feature differentiation process based on the oxidation dataset to construct the oxidation feature tree; the kinetic feature differentiation process comprises:

[0012] constructing a parent node; the parent node stores the alloy samples; the initial value of the parent node is the root node, and the root node is constructed based on the oxidation feature set;

[0013] traversing the kinetic features and the kinetic feature values of the alloy samples in the parent node to select target kinetic features and target kinetic feature values therefrom;

[0014] splitting the alloy samples in the parent node according to the target kinetic features and the target kinetic feature values to obtain a first child node and a second child node of the parent node, and storing the target kinetic feature values in corresponding node edges;

[0015] determining that the first child node and / or the second child node do not satisfy a stop differentiation condition, taking the first child node and / or the second child node as the parent node, and performing the kinetic feature differentiation process.

[0016] In some embodiments of the present application, the kinetic features of the alloy samples comprise oxidation time and alloy composition, and the alloy samples further comprise oxidation weight gain; after the obtaining of the oxidation dataset, the method further comprises:

[0017] performing logarithmic preprocessing on the oxidation time and the oxidation weight gain;

[0018] performing combined calculation on the alloy composition to generate a preset number of secondary alloy features;

[0019] adding the secondary alloy features to the oxidation dataset.

[0020] In some embodiments of the present application, the traversing the kinetic features and the kinetic feature values of the alloy samples in the parent node to select target kinetic features and target kinetic feature values therefrom comprises:

[0021] selecting a division parameter from the kinetic features; the division parameter comprises a candidate feature and a candidate feature value, and the candidate feature value is selected from the kinetic feature values of the candidate feature in the alloy samples;

[0022] dividing the alloy samples in the parent node into a first candidate set and a second candidate set according to numerical values of the candidate characteristic values;

[0023] obtaining a first fitting result of the first candidate set and a second fitting result of the second candidate set by using an oxidation kinetics equation;

[0024] calculating a linear gain value of the division parameter according to a first linearity value of the first fitting result, a second linearity value of the second fitting result and a total linearity value of the parent node;

[0025] updating the candidate characteristic values by traversing the kinetic characteristic values and updating the candidate characteristics by traversing the kinetic characteristics, and obtaining the target kinetic characteristic and the target kinetic characteristic value according to the division parameter with the largest linear gain value.

[0026] In some embodiments of the present application, the calculation of the linear gain value of the division parameter according to the first linearity value of the first fitting result, the second linearity value of the second fitting result and the total linearity value of the parent node comprises:

[0027] calculating the first linearity value according to the first fitting result and the alloy samples in the first candidate set;

[0028] calculating the second linearity value according to the second fitting result and the alloy samples in the second candidate set;

[0029] obtaining a total sample number of the alloy samples in the parent node, a first sample number of the alloy samples in the first candidate set and a second sample number of the alloy samples in the second candidate set;

[0030] calculating a first sample ratio according to the first sample number and the total sample number, and calculating a second sample ratio according to the second sample number and the total sample number;

[0031] multiplying the first sample ratio and the first linearity value to obtain a first linearity index, and multiplying the second sample ratio and the second linearity value to obtain a second linearity index;

[0032] adding the first linearity index and the second linearity index and subtracting the total linearity value to obtain the linear gain value of the division parameter.

[0033] In some embodiments of the present application, the first fitting result is an oxidation kinetics fitting equation, the kinetic characteristic value includes an oxidation time value, and the alloy sample further includes an oxidation weight gain value; the first linearity value is calculated according to the first fitting result and the alloy sample in the first candidate set, including:

[0034] The oxidation time value of the alloy sample in the first candidate set is substituted into the oxidation kinetics fitting equation to obtain an oxidation weight gain fitting value;

[0035] The covariance of the oxidation weight gain value and the oxidation weight gain fitting value is calculated to obtain a first intermediate result;

[0036] The standard deviation of the oxidation weight gain value is calculated to obtain a first standard deviation, the standard deviation of the oxidation weight gain fitting value is calculated to obtain a second standard deviation, and the first standard deviation and the second standard deviation are multiplied to obtain a second intermediate result;

[0037] The first intermediate result is divided by the second intermediate result to obtain the first linearity value;

[0038] Or,

[0039] The error sum of squares of the oxidation weight gain value and the oxidation weight gain fitting value is calculated to obtain a third intermediate result;

[0040] The total sum of squares of the oxidation weight gain value is calculated to obtain a fourth intermediate result;

[0041] The third intermediate result is divided by the fourth intermediate result to obtain a fifth intermediate result, and 1 is subtracted from the fifth intermediate result to obtain the first linearity value.

[0042] In some embodiments of the present application, according to the target kinetic characteristic and the target kinetic characteristic value, the alloy sample in the parent node is split to obtain a first child node and a second child node of the parent node, including:

[0043] According to the target kinetic characteristic, the numerical value of the kinetic characteristic value corresponding to the alloy sample in the parent node and the target kinetic characteristic value is compared;

[0044] If the kinetic characteristic value is less than or equal to the target kinetic characteristic value, the alloy sample is divided into the first child node;

[0045] If the kinetic characteristic value is greater than the target kinetic characteristic value, the alloy sample is divided into the second child node.

[0046] In some embodiments of the present application, the stop differentiation condition includes a linearity threshold value and / or a single degree threshold value; the method further includes:

[0047] if the first linearity value of the first sub-node meets the linearity threshold, the first sub-node is taken as the end node;

[0048] if the second linearity value of the second sub-node meets the linearity threshold, the second sub-node is taken as the end node;

[0049] and / or,

[0050] if the oxidation time of the alloy sample in the first sub-node meets the single-degree threshold, the parent node of the first sub-node is taken as the end node;

[0051] if the oxidation time of the alloy sample in the second sub-node meets the single-degree threshold, the parent node of the second sub-node is taken as the end node.

[0052] In some embodiments of the present application, the end node includes a plurality of alloy samples, an oxidation kinetics fitting equation of the alloy samples is obtained by using an oxidation kinetics equation, the oxidation kinetics fitting equation includes the oxidation kinetics index; the method further includes:

[0053] inputting the oxidation kinetics index of the end node and the alloy samples into a symbolic regression model;

[0054] regressing the alloy samples in the kinetics feature as a regression target to obtain an analytical equation of the oxidation kinetics index.

[0055] In a second aspect, the embodiments of the present application further provide an oxidation kinetics index calculation device, which implements the oxidation kinetics index calculation method as described in the first aspect of the present application, and includes:

[0056] a first acquisition module, configured to acquire a pre-constructed oxidation feature tree, the oxidation feature tree being a binary tree, the binary tree including a plurality of nodes and node edges, the nodes including a root node and at least two end nodes, and the node edges storing alloy feature values, the alloy feature values being boundary values divided according to alloy feature data;

[0057] a data comparison module, configured to divide the nodes according to the alloy feature data from alloy data, and configured to determine target node edges one by one from the root node according to comparison results of the alloy feature data and the alloy feature values on the node edges, until the end nodes are reached, and determine the end nodes as target end nodes;

[0058] an index acquisition module, configured to take the oxidation kinetics index stored in the target end nodes as the oxidation kinetics index of the alloy data.

[0059] In some embodiments of the present application, the device further comprises:

[0060] a second obtaining module, configured to obtain an oxidation dataset; wherein the oxidation dataset comprises a plurality of alloy samples, and each of the alloy samples comprises a plurality of kinetic characteristics and corresponding kinetic characteristic values;

[0061] a differentiation module, configured to perform a kinetic characteristic differentiation process based on the oxidation dataset to construct the oxidation characteristic tree; the kinetic characteristic differentiation process comprises:

[0062] constructing a parent node; the parent node stores the alloy samples; an initial value of the parent node is the root node, and the root node is constructed based on the oxidation characteristic set;

[0063] traversing the kinetic characteristics and the kinetic characteristic values of the alloy samples in the parent node to select a target kinetic characteristic and a target kinetic characteristic value therefrom;

[0064] splitting the alloy samples in the parent node according to the target kinetic characteristic and the target kinetic characteristic value to obtain a first child node and a second child node of the parent node, and storing the target kinetic characteristic value in a corresponding node edge;

[0065] determining that the first child node and / or the second child node do not satisfy a stop differentiation condition, taking the first child node and / or the second child node as the parent node, and performing the kinetic characteristic differentiation process.

[0066] In some embodiments of the present application, the device further comprises:

[0067] a parsing module, configured to input the oxidation kinetic index of the end node and the alloy sample into a symbolic regression model; perform symbolic regression on the kinetic characteristics in the alloy sample with the oxidation kinetic index as a regression target to obtain an analytical equation of the oxidation kinetic index.

[0068] In a third aspect, the embodiments of the present application further provide an electronic device, including a memory and a processor, the memory stores a computer program, and the processor implements the oxidation kinetic index calculation method according to the embodiments of the first aspect of the present application when executing the computer program.

[0069] In a fourth aspect, the embodiments of the present application further provide a computer readable storage medium, the storage medium stores a program, and the program is executed by a processor to implement the oxidation kinetic index calculation method according to the embodiments of the first aspect of the present application.

[0070] The embodiments of this application include at least the following beneficial effects:

[0071] This application provides a method, apparatus, electronic device, and storage medium for calculating oxidation kinetics indices. The method involves acquiring a pre-constructed oxidation feature tree, which is a binary tree comprising multiple nodes and edges. Each node includes a root node and at least two end nodes. The edges store alloy data values, where alloy feature values ​​define the boundaries of the alloy data. Nodes are then defined from the alloy data based on the alloy feature data. Specifically, starting from the root node of the oxidation feature tree, target node edges are determined one by one based on the comparison between the alloy feature data and the alloy feature values ​​on the node edges, until an end node is reached and designated as the target end node. Finally, the oxidation kinetics index stored in the target end node is used as the oxidation kinetics index for that alloy data. Thus, by storing oxidation kinetic indices corresponding to different alloy compositions in end nodes based on the pre-constructed oxidation feature tree, and by progressively determining target end nodes by comparing the alloy feature data in the alloy data with the edges of the oxidation feature tree nodes, the oxidation kinetics index and oxidation kinetic equation can be accurately calculated. This method is of great significance for tasks such as oxidation process description, oxidation-resistant material mechanism research, and oxidation resistance design, thereby promoting related research development.

[0072] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0073] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0074] Figure 1 This is a flowchart illustrating a method for calculating oxidation kinetic index according to an embodiment of this application;

[0075] Figure 2 This is a flowchart illustrating a method for calculating the oxidation kinetic index provided in another embodiment of this application;

[0076] Figure 3 yes Figure 2 A flowchart illustrating step S302;

[0077] Figure 4 yes Figure 3 A flowchart illustrating step S404;

[0078] Figure 5 yes Figure 4 A flowchart illustrating step S501;

[0079] Figure 6 yesFigure 2 Flowchart of step S303;

[0080] Figure 7 Figure 2 Flowchart of step S304;

[0081] Figure 8 Figure 2 Another flowchart of step S304;

[0082] Figure 9 Flowchart of the method for calculating the oxidation kinetic index provided by another embodiment of the present application;

[0083] Figure 10 Oxidation characteristic tree diagram provided by an embodiment of the present application;

[0084] Figure 11 Module diagram of the device for calculating the oxidation kinetic index provided by an embodiment of the present application;

[0085] Figure 12 Module diagram of the device for calculating the oxidation kinetic index provided by another embodiment of the present application;

[0086] Figure 13 Structure diagram of the electronic device provided by an embodiment of the present application.

[0087] The reference signs: first acquisition module 100, data comparison module 200, index acquisition module 300, second acquisition module 400, differentiation module 500, analysis module 600, electronic device 1000, processor 1001, memory 1002. DETAILED DESCRIPTION

[0088] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application.

[0089] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the drawings, in which the same or similar reference signs represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the drawings are exemplary and are only used to explain the present application and should not be understood as limiting the present application.

[0090] ​​In the description of the present application, it should be understood that the orientation description, such as the orientation or position relationship indicated by up, down, front, back, left, right and the like, is based on the orientation or position relationship shown in the drawings, and is only for the convenience of describing the present application and simplifying the description, and does not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application.

[0091] In the description of the present application, the meaning of several is one or more, the meaning of multiple is more than two, greater than, less than, more than, etc. are understood as not including the number, above, below, etc. are understood as including the number. If the first, second is described, it is only for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features or the order of indicated technical features.

[0092] In the description of the present application, unless otherwise explicitly limited, the words such as setting, installing, connecting, etc. should be broadly understood, and those skilled in the art can reasonably determine the specific meaning of the above words in the present application in combination with the specific content of the technical solution.

[0093] Oxidation of alloys is an important oxidation process that has a significant impact on the performance and life of materials. Understanding the kinetic behavior of alloy oxidation, i.e. the rate and variation of weight gain during oxidation, is of great importance for mechanism research and oxidation-resistant material design. By establishing an oxidation kinetic equation, the oxidation weight gain process of alloys can be better described and predicted, and the relationship between oxidation rate and test conditions such as time, temperature, and pressure can be understood.

[0094] The oxidation kinetic exponent is one of the important parameters for establishing the oxidation kinetic equation. In related technologies, the oxidation kinetic exponent is usually a fixed value, used to describe the oxidation law of a specific alloy under specific oxidation conditions. However, this method cannot study the quantitative relationship between the oxidation kinetic exponent itself and the alloy composition and test conditions, nor can it comprehensively describe the oxidation characteristics of multiple alloys under different test conditions. Therefore, it is difficult to accurately calculate the oxidation kinetic exponent of different alloy data and establish the corresponding oxidation kinetic equation, hindering the further development of related research.

[0095] Based on this, the embodiment of the present application provides an oxidation kinetic index calculation method and device, electronic equipment and storage medium. The oxidation kinetic index corresponding to different alloy components can be stored in the end node according to the pre-constructed oxidation feature tree, and the target end node can be gradually determined by comparing the alloy feature data in the alloy data and the oxidation feature tree node edge, so that the oxidation kinetic index and the oxidation kinetic equation can be accurately calculated and obtained, which is of great significance to the oxidation process description, the mechanism research of oxidation-resistant materials and the oxidation-resistant design, thereby promoting the development of related research.

[0096] The embodiment of the present application provides an oxidation kinetic index calculation method and device, electronic equipment and storage medium, which is specifically described by the following embodiments. First, the oxidation kinetic index calculation method in the embodiment of the present application is described.

[0097] The oxidation kinetic index calculation method provided by the embodiment of the present application relates to the field of material technology, especially to the field of alloy technology. The oxidation kinetic index calculation method provided by the embodiment of the present application can be applied in a terminal, can be applied in a server, and can also be a computer program running in a terminal or a server. For example, the computer program can be a native program or a software module in an operating system; it can be a native application program (APP), that is, a program that needs to be installed in an operating system to run, such as a client supporting oxidation kinetic index calculation, that is, a program that can run only by being downloaded into a browser environment. In short, the above computer program can be any form of application program, module or plug-in. The terminal communicates with the server through a network. The oxidation kinetic index calculation method can be executed by the terminal or the server, or cooperatively executed by the terminal and the server.

[0098] In some embodiments, the terminal can be a smartphone, a tablet computer, a notebook computer, a desktop computer, a smart watch, or the like. The server can be a standalone server, or a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and basic cloud computing services such as big data and artificial intelligence platforms; or a service node in a blockchain system, the service nodes in the blockchain system form a peer-to-peer (P2P) network, and the P2P protocol is an application layer protocol running on a transmission control protocol (TCP) protocol. The server can install a server of an oxidation kinetic index calculation system, and the server can interact with the terminal, for example, the server installs corresponding software, the software can be an application implementing the xxxxx method, but is not limited to the above forms. The terminal and the server can be connected through a communication connection mode such as Bluetooth, a universal serial bus (USB), or a network, and the present embodiment is not limited herein.

[0099] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and the like. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, and the like that perform specific tasks or implement specific abstract data types. The present application can also be practiced in a distributed computing environment, in which tasks are performed by remote processing devices connected by a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media, including storage devices.

[0100] The oxidation kinetic index calculation method in the embodiments of the present application is described below.

[0101] Referring to Figure 1 As shown in the figure, the embodiments of the present application provide an oxidation kinetic index calculation method, which includes but is not limited to the following steps S101 to S103.

[0102] Step S101, obtaining a pre-constructed oxidation feature tree.

[0103] In some embodiments, the oxidation feature tree is a binary tree, and the nodes include a root node and at least two end nodes. It can be understood that a binary tree is a tree data structure, and the root node is the top node of the binary tree, which is the starting point of the tree. There is only one root node in a binary tree. Each node of the binary tree has at most two child nodes, and a node without child nodes is an end node, also known as a leaf node.

[0104] The nodes are connected by node edges between the nodes and their superior and inferior nodes. In some embodiments, the node edges of the oxidation feature tree store alloy feature values, which are boundaries for dividing alloy data to different nodes, and can be the mass percentage of a certain alloy component in a multi-component alloy. The alloy feature values can be set according to actual needs by those skilled in the art, and the embodiments are not limited in this regard.

[0105] In step S102, the nodes are divided from the alloy data according to the alloy feature data. Starting from the root node, the target node edge is determined one by one according to the comparison result of the alloy feature data and the alloy feature value on the node edge, until the end node is reached, and the end node is determined as the target end node.

[0106] The alloy data includes but is not limited to component data, physical property data, and chemical property data, etc. For example, the component data includes each alloy component and its mass percentage, the physical property data includes the density, thermal conductivity, electrical conductivity, melting point and hardness of the alloy, and the chemical property data includes chemical reactivity and corrosion resistance, etc. In some embodiments, the nodes are divided from the alloy data according to the alloy feature data, and specifically, the alloy components of the multi-component alloy and the mass percentage of each alloy component are taken as the alloy feature data. The embodiments are not limited in this regard.

[0107] In some embodiments, starting from the root node, the target node edge is determined by comparing the alloy feature data and the alloy feature value on the node edge. For example, the root node has two child nodes, which are the left child node and the right child node. The node edge from the root node to the left child node is that the mass percentage of iron element is less than or equal to 30wt.%, and the node edge from the root node to the right child node is that the mass percentage of iron element is greater than 30wt.%. Therefore, when the mass percentage of iron element in the alloy component is less than or equal to 30wt.%, it meets the alloy feature value on the node edge between the root node and the left child node, otherwise it meets the alloy feature value on the node edge between the root node and the right child node. Thus, the target node edge can be determined by comparing the alloy feature data and the alloy feature value on the node edge, until the end node is reached, and then the end node is determined as the target end node of the alloy data.

[0108] In step S103, the oxidation kinetic index stored in the target end node is taken as the oxidation kinetic index of the alloy data.

[0109] In some embodiments, the oxidation feature tree is constructed in advance according to the alloy samples in the oxidation dataset, and the oxidation kinetic index corresponding to different alloy samples is stored in each end node of the oxidation feature tree. The oxidation kinetic index stored in the target end node is taken as the oxidation kinetic index of the alloy data, and thus the oxidation kinetic index and the oxidation kinetic equation can be accurately calculated by comparing the alloy feature data in the alloy data with the node edges of the oxidation feature tree step by step to determine the target end node.

[0110] Referring to Figure 2 In some embodiments of the present application, constructing the oxidation feature tree can further include, but is not limited to, the following steps S201 to S202.

[0111] In step S201, an oxidation dataset is obtained.

[0112] In some embodiments, the oxidation dataset includes a plurality of alloy samples, each alloy sample including a plurality of kinetic features and corresponding kinetic feature values, so that each kinetic feature in the oxidation dataset has a plurality of kinetic feature values. Each alloy sample further includes an oxidation weight gain and a corresponding oxidation weight gain value. It can be understood that the oxidation dataset of the multi-element alloy can be accumulated by consulting literature and experimental synthesis. The oxidation dataset is input into a table file and stored in the form of a csv file.

[0113] For example, there are 95 supercritical CO2 alloy samples in total, and the kinetic features of each alloy sample include alloy components such as iron, nickel, chromium, aluminum, cobalt, manganese, silicon, titanium, copper, carbon, bismuth, etc., and the corresponding kinetic feature values are the mass percentages of the alloy components. The kinetic features further include test conditions such as temperature, pressure, and oxidation time. Referring to Table 1 shown below, the table file includes the kinetic features of the alloy samples and the regression targets such as the oxidation weight gain. Specifically, each row in Table 1 represents an alloy sample, each column represents a kinetic feature, and the last column represents the oxidation weight gain. In the kinetic features, the oxidation time is placed in the second-to-last column of the table, which is not limited in the present embodiment.

[0114] Table 1

[0115]

[0116] In some embodiments, the oxidation dataset is preprocessed. Specifically, the oxidation time value corresponding to the oxidation time in the kinetic features is logarithmically processed, and the unit of the oxidation weight gain value corresponding to the oxidation weight gain is uniformly converted to milligrams per square decimeter (mg / dm2) and logarithmically processed to obtain a table file as shown in Table 2.

[0117] Table 2

[0118]

[0119] In some embodiments, the alloy components in the kinetic features are randomly combined by addition, subtraction and multiplication to generate a preset number of secondary alloy features. For example, iron + nickel, cobalt-aluminum + manganese, (iron-cobalt) x silicon, etc. are randomly generated secondary alloy features. Specifically, the number of secondary alloy features can be generated one time the number of alloy component features, and then added to the oxidation data set, so that better nonlinear relationships can be represented. The secondary alloy features and the original alloy components are the same level of kinetic features, and both belong to alloy features. The formed table file is shown in Table 3, and the secondary alloy features are placed before the second-to-last column corresponding to the oxidation time, or can be placed before the alloy components, which is not limited in the present embodiment.

[0120] Table 3

[0121]

[0122] Step S202, performing a kinetic feature differentiation process based on the oxidation data set to construct an oxidation feature tree.

[0123] In some embodiments, the kinetic feature differentiation process is performed according to the kinetic feature values of the alloy samples in the oxidation data set, so as to construct the oxidation feature tree. Specifically, the specific process of constructing the oxidation feature tree can include but is not limited to the following steps S301 to S304.

[0124] Step S301, constructing a parent node.

[0125] In some embodiments, the parent node stores the alloy samples in the oxidation data set, and the initial value of the parent node is the root node, which is constructed based on the oxidation data set. Therefore, the root node stores all the alloy samples in the oxidation data set. It can be understood that in the binary tree structure, except for the root node, there is no parent node, and each node has and only has one parent node.

[0126] Step S302, traversing the kinetic features and kinetic feature values of the alloy samples in the parent node, and selecting target kinetic features and target kinetic feature values therefrom.

[0127] In some embodiments, the dynamic characteristics and corresponding dynamic characteristic values of the alloy samples in the parent node are traversed, and a target dynamic characteristic and a corresponding target dynamic characteristic value are selected therefrom. For the same dynamic characteristic, different alloy samples correspond to different dynamic characteristic values. For example, when the dynamic characteristic is iron element, the corresponding dynamic characteristic value is the mass percentage of iron element, and the dynamic characteristic value in alloy sample 1 is 0.07wt.%, and the dynamic characteristic value in alloy sample 95 is 3wt.%. Therefore, after selecting a dynamic characteristic as the target dynamic characteristic, a dynamic characteristic value needs to be further selected as the target dynamic characteristic value, which provides a branch condition for constructing the oxidation characteristic tree.

[0128] In step S303, the alloy samples in the parent node are split according to the target dynamic characteristic and the target dynamic characteristic value, to obtain a first child node and a second child node of the parent node, and the target dynamic characteristic value is stored in the corresponding node edge.

[0129] In some embodiments, the alloy samples in the parent node are split according to the target dynamic characteristic and the target dynamic characteristic value, to obtain a first child node and a second child node of the parent node. Specifically, the alloy samples in the parent node are compared with the target dynamic characteristic and the target dynamic characteristic, and the alloy samples are divided into two parts according to the comparison result, one part as the first child node and the other part as the second child node. For example, there are 95 alloy samples in the parent node, and according to the target dynamic characteristic and the target dynamic characteristic value, the 95 alloy samples in the parent node are split to obtain a first child node containing 60 alloy samples and a second child node containing 35 alloy samples.

[0130] In some embodiments, after obtaining the first child node and the second child node of the parent node, the parent node and the first child node are connected using a node edge, and the parent node and the second child node are connected using a node edge, and the target dynamic characteristic value is stored in the node edge. It can be understood that the target dynamic characteristic is also stored in the node edge, which is not limited in this embodiment.

[0131] In step S304, it is determined that the first child node and / or the second child node does not satisfy the stop differentiation condition, and the first child node and / or the second child node is taken as the parent node, and the dynamic characteristic differentiation process is executed.

[0132] In some embodiments, if the first child node satisfies the stop differentiation condition, the first child node is taken as an end node and the differentiation is stopped, otherwise the first child node is taken as a parent node, and the dynamic feature differentiation process is iteratively executed. Similarly, if the second child node satisfies the stop differentiation condition, the second child node is taken as an end node and the differentiation is stopped, otherwise the second child node is taken as a parent node, and the dynamic feature differentiation process is iteratively executed, thereby constructing the oxidation feature tree based on the oxidation data set.

[0133] Referring to Figure 3 In some embodiments of the present application, the step S302 can further include, but is not limited to, the following steps S401 to S405.

[0134] Step S401, selecting a division parameter from the dynamic features.

[0135] In some embodiments, the division parameter includes a candidate feature and a candidate feature value, and the candidate feature value is selected from the dynamic feature values of the candidate feature in the alloy samples. Specifically, any one of the dynamic features is taken as a candidate feature, and then a dynamic feature value corresponding to the candidate feature in the alloy samples is selected as a candidate feature value. For example, the iron element in the dynamic features is taken as a candidate feature, and there are 95 dynamic feature values of the iron element in the parent node containing 95 alloy samples. The dynamic feature value of the iron element in the 22nd alloy sample is 49.74wt.%, which can be taken as a candidate feature value. The present embodiment does not limit this.

[0136] Step S402, dividing the alloy samples in the parent node into a first candidate set and a second candidate set according to the numerical value of the candidate feature value.

[0137] In some embodiments, the dynamic feature value corresponding to the candidate feature in the alloy sample in the parent node is compared with the candidate feature value according to the numerical value of the candidate feature value. If the dynamic feature value of the alloy sample is less than or equal to the candidate feature value, the alloy sample is divided into the first candidate set; if the dynamic feature value of the alloy sample is greater than the candidate feature value, the alloy sample is divided into the second candidate set, thereby dividing the alloy samples in the parent node into the first candidate set and the second candidate set.

[0138] For example, the candidate feature is the iron element in the kinetic feature, and the candidate feature value is the mass percentage of the iron element in the 22nd alloy sample, which is 49.74 wt.%. According to the numerical size of the candidate feature value, the 95 alloy samples in the parent node are divided. If the kinetic feature value of the iron element in the alloy sample is less than or equal to 49.74 wt.%, the alloy sample is divided into the first candidate set; if the kinetic feature value of the iron element in the alloy sample is greater than 49.74 wt.%, the alloy sample is divided into the second candidate set. Finally, the first candidate set containing 65 alloy samples and the second candidate set containing 30 alloy samples can be obtained, which are not limited in the embodiment.

[0139] In step S403, the first fitting result of the first candidate set and the second fitting result of the second candidate set are obtained by using the oxidation kinetic equation.

[0140] The oxidation rule of the alloy satisfies the oxidation kinetic equation y = kt n where y is the oxidation weight gain value, k is the oxidation kinetic constant, t is the oxidation time value, and n is the oxidation kinetic index. In some embodiments, in order to conveniently and quickly fit the alloy samples by using the oxidation kinetic equation, the oxidation time value and the oxidation weight gain value in the kinetic feature are logarithmically processed, and the oxidation kinetic equation is also logarithmically processed. That is, ln(y) = n·ln(t) + ln(k) is obtained, where ln(·) represents logarithmic processing, and thus the expression form of the oxidation kinetic method in the logarithmic space is a linear relationship. Therefore, in the embodiment of the present application, the oxidation weight gain value corresponds to ln(y), the oxidation time value corresponds to ln(t), and the fitting constant corresponds to ln(k).

[0141] In some embodiments, the alloy samples in the first candidate set are fitted by using the oxidation kinetic equation after logarithmic processing to obtain the first fitting result; and the alloy samples in the second candidate set are fitted by using the oxidation kinetic equation after logarithmic processing to obtain the second fitting result.

[0142] Specifically, the kinetic characteristics related to the oxidation kinetic equation are the oxidation time and the oxidation weight gain of the alloy sample. Since the expression of the oxidation kinetic method in the logarithmic space is a linear relationship, the present embodiment takes the oxidation weight gain as the regression target and takes the oxidation time as the variable to fit the candidate set, to obtain the corresponding fitting result, i.e., the oxidation kinetic fitting equation, in the form of ln(y') = n·ln(t) + ln(k), where ln(y') is the oxidation weight gain fitting value, ln(k) is the oxidation kinetic constant, ln(t) is the oxidation time value, and n is the oxidation kinetic index. By calculating the Euclidean distance error between the oxidation weight gain fitting value ln(y') and the oxidation weight gain value ln(y) of the alloy sample, the size of the oxidation kinetic index n and the fitting constant ln(k) can be calculated, so that the other kinetic characteristics of the alloy sample are taken as the branch basis, which is not limited in the present embodiment.

[0143] In step S404, the linear gain value of the division parameter is calculated according to the first linearity value of the first fitting result, the second linearity value of the second fitting result, and the total linearity value of the parent node.

[0144] In some embodiments, the linearity value is used to measure the data fitting accuracy of the alloy samples in the first candidate set and the second candidate set. For example, the Pearson correlation coefficient is used to calculate the linearity value. It can be understood that the Pearson correlation coefficient has a value range of -1 to 1, where -1 represents complete negative correlation, 1 represents complete positive correlation, and 0 represents no linear correlation. Therefore, the closer the linearity is to 1 or -1, the closer the alloy sample is to the linear relationship, and the better the linear fitting effect is; and the closer the linearity is to 0, the less linear relationship the alloy sample has, and the poorer the linear fitting effect is.

[0145] In some embodiments, the first linearity value of the first fitting result is calculated according to the alloy samples in the first candidate set, the second linearity value of the second fitting result is calculated according to the alloy samples in the second candidate set, and then the linear gain value of the differentiation result corresponding to the division parameter is calculated according to the first linearity value, the second linearity value, and the total linearity value. It can be understood that when the parent node is the root node, the alloy samples in the oxidation data set are fitted by using the oxidation kinetic equation after logarithmic processing to obtain the total fitting result, and then the total linearity value of the total fitting result is calculated according to the alloy samples in the oxidation data set.

[0146] In step S405, the candidate feature values are updated by traversing the kinetic characteristic values, and the candidate features are updated by traversing the kinetic characteristics, and the target kinetic feature and the target kinetic feature value are obtained according to the division parameter with the maximum linear gain value.

[0147] One dynamic characteristic corresponds to multiple dynamic characteristic values. For example, the parent node stores 95 alloy samples, and thus each dynamic characteristic corresponds to 95 dynamic characteristic values. The candidate characteristic values are updated by traversing the dynamic characteristic values. Specifically, when the candidate characteristic is iron, the mass percentage of iron in the 95 alloy samples is sequentially taken as the candidate characteristic value. The alloy samples in the parent node are split to obtain the first candidate set and the second candidate set of 95 different division parameters, and the linear gain value of each division parameter is calculated.

[0148] In some embodiments, after the 95 mass percentages of iron are traversed, the candidate characteristic is updated by traversing the dynamic characteristic. Specifically, the next dynamic characteristic of nickel in Table 3, for example, nickel, is taken as the candidate characteristic, and the mass percentages of nickel in the 95 alloy samples are traversed as the candidate characteristic value to split the alloy samples in the parent node to obtain the first candidate set and the second candidate set. The alloy samples in the parent node are split to obtain the first candidate set and the second candidate set of 95 different division parameters, and the linear gain value of each division parameter is calculated. Here, it is assumed that nickel has 95 different values, which is not limited in this embodiment.

[0149] In some embodiments, the dynamic characteristic values in the alloy sample can be traversed in units of dynamic characteristics, or each dynamic characteristic can be traversed in units of dynamic characteristic values in the alloy sample until each dynamic characteristic and each dynamic characteristic value in the parent node are taken as the division parameter. Finally, according to the division parameter with the largest linear gain, the candidate characteristic in the division parameter is taken as the target dynamic characteristic value, and the candidate characteristic value in the division parameter is taken as the target dynamic characteristic value, thereby obtaining the target dynamic characteristic and the target dynamic characteristic value.

[0150] Referring to Figure 4 In some embodiments of the present application, the above step S404 can further include but is not limited to the following steps S501 to S506.

[0151] Step S501, calculating a first linearity value according to the first fitting result and the alloy samples in the first candidate set.

[0152] In some embodiments, the first linearity value of the first fitting result is calculated according to the oxidation kinetics fitting equation of the first fitting result and the alloy samples in the first candidate set. It can be understood that the first linearity value is used to measure the accuracy of the first fitting result. Specifically, the oxidation time value of the alloy samples in the first candidate set is substituted into the oxidation kinetics fitting equation, and the oxidation weight gain fitting value can be calculated by using the oxidation kinetics fitting equation according to the known fitting constant and the oxidation kinetics index. Finally, the first linearity value is calculated according to the real oxidation weight gain value of the alloy samples and the calculated oxidation weight gain fitting value, which is not limited in the present embodiment.

[0153] In step S502, the second linearity value is calculated according to the second fitting result and the alloy samples in the second candidate set.

[0154] In some embodiments, the second linearity value of the second fitting result is calculated according to the oxidation kinetics fitting equation of the second fitting result and the alloy samples in the second candidate set. It can be understood that the second linearity value is used to measure the accuracy of the second fitting result. Specifically, the oxidation time value of the alloy samples in the second candidate set is substituted into the oxidation kinetics fitting equation, and the oxidation weight gain fitting value can be calculated by using the oxidation kinetics fitting equation according to the known fitting constant and the oxidation kinetics index. Finally, the second linearity value is calculated according to the real oxidation weight gain value of the alloy samples and the calculated oxidation weight gain fitting value, which is not limited in the present embodiment.

[0155] In step S503, the total sample number of the alloy samples in the parent node, the first sample number of the alloy samples in the first candidate set, and the second sample number of the alloy samples in the second candidate set are obtained.

[0156] In some embodiments, the total sample number of the alloy samples in the parent node is obtained, and the first sample number of the alloy samples in the first candidate set and the second sample number of the alloy samples in the second candidate set are obtained. It can be understood that the total sample number of the parent node is the sum of the first sample number and the second sample number. For example, there are 60 alloy samples in the first candidate set and 35 alloy samples in the second candidate set, and there are 60+35=95 alloy samples in the parent node.

[0157] In step S504, the first sample ratio is calculated according to the first sample number and the total sample number, and the second sample ratio is calculated according to the second sample number and the total sample number.

[0158] In some embodiments, the first sample ratio corresponding to the first candidate set can be calculated by dividing the first sample number by the total sample number, and the second sample ratio corresponding to the second candidate set can be calculated by dividing the second sample number by the total sample number.

[0159] Step S505: Multiply the first sample proportion and the first linearity value to obtain the first linearity index, and multiply the second sample proportion and the second linearity value to obtain the second linearity index.

[0160] In some embodiments, a first linearity index is obtained by multiplying a first sample proportion and a first linearity value, and a second linearity index is obtained by multiplying a second sample proportion and a second linearity value. Specifically, K represents the total number of samples, k1 represents the first sample size, k2 represents the second sample size, and LG... l LG represents the first linearity value. r This represents the second linearity value. Therefore, the proportion of the first sample is k1 / K, the proportion of the second sample is k2 / K, and the first linearity index is LG. l ·k1 / K, the first linearity index is LG r ·k2 / K.

[0161] Step S506: Add the first linearity index and the second linearity index together and subtract the bus linearity value to obtain the linear gain value of the partitioning parameter.

[0162] In some embodiments, the first linearity index and the second linearity index are added together, i.e., LG l ·k1 / K+LG r ·k2 / K, then subtract the bus performance value LG p This yields the linear gain value ΔLG of the partitioning parameters. Therefore, ΔLG i,j =LG l ·k1 / K+LG r ·k2 / K-LG p Where i represents the i-th kinetic feature, and j represents the j-th kinetic feature value, i.e., the kinetic feature value in the j-th alloy sample. Therefore, we can iterate through each kinetic feature and kinetic feature value of the alloy samples in the parent node and use them as partitioning parameters to split the alloy samples in the parent node and calculate the linear gain value corresponding to the partitioning parameters.

[0163] Reference Figure 5 As shown, in some embodiments of this application, step S501 may also include, but is not limited to, steps S601 to S607.

[0164] Step S601: Substitute the oxidation time values ​​of the alloy samples in the first candidate set into the oxidation kinetics fitting equation to calculate the oxidation weight gain fitting value.

[0165] In some embodiments, the oxidation time value of each alloy sample in the first candidate set is substituted into the oxidation kinetics fitting equation ln(y') = n ln(t) + ln(k), and the oxidation weight gain fitting value can be calculated using the oxidation kinetics equation according to the known fitting constant and the oxidation kinetics index. For example, there are 60 alloy samples in the first candidate set, and 60 oxidation weight gain fitting values can be calculated according to the 60 oxidation time values thereof.

[0166] In step S602, the covariance of the oxidation weight gain value and the oxidation weight gain fitting value is calculated to obtain a first intermediate result.

[0167] In some embodiments, the oxidation weight gain value of the alloy sample in the first candidate set is obtained, the covariance of the oxidation weight gain value and the corresponding oxidation weight gain fitting value is calculated, and the covariance is taken as the first intermediate result. It can be understood that the oxidation weight gain value in the present embodiment is y, and the oxidation weight gain fitting value is y'. Those skilled in the art can also select ln(y) as the oxidation weight gain value and ln(y') as the oxidation weight gain fitting value according to actual needs, and the present embodiment does not limit this.

[0168] In step S603, the standard deviation of the oxidation weight gain value is calculated to obtain a first standard deviation, the standard deviation of the oxidation weight gain fitting value is calculated to obtain a second standard deviation, and the first standard deviation and the second standard deviation are multiplied to obtain a second intermediate result.

[0169] In some embodiments, the first standard deviation is calculated according to the oxidation weight gain value of the alloy sample in the first candidate set. Specifically, the average value of the oxidation weight gain of the alloy sample is calculated first, then the oxidation weight gain value of each alloy sample is subtracted from the average value of the oxidation weight gain to obtain a difference result, the square operation is performed on the difference result to obtain the square result of the alloy sample. Finally, the square results of all alloy samples are accumulated and square root operation is performed to obtain the first standard deviation. Similarly, the second standard deviation is calculated according to the oxidation weight gain fitting value. Finally, the first standard deviation and the second standard deviation are multiplied to obtain the second intermediate result, and the present embodiment does not limit this.

[0170] In step S604, the first intermediate result is divided by the second intermediate result to obtain a first linearity value.

[0171] In some embodiments, the first intermediate result is divided by the second intermediate result to obtain the first linearity value of the first fitting result. Specifically, the first linearity value satisfies the following calculation formula:

[0172]

[0173] wherein k1 represents the first sample number of the alloy sample in the first candidate set, y represents the oxidation weight gain value, represents the average value of the oxidation weight gain, y' represents the fitting value of the oxidation weight gain, represents the fitting average value of the oxidation weight gain.

[0174] It can be understood that the specific implementation of the second linearity value calculation process is basically consistent with the first linearity value, and satisfies the following calculation formula:

[0175]

[0176] wherein k2 represents the second sample quantity of the alloy sample in the second candidate set, y represents the oxidation weight gain value, represents the average value of the oxidation weight gain, y' represents the fitting value of the oxidation weight gain, represents the fitting average value of the oxidation weight gain.

[0177] In some embodiments, the calculation of the first linearity can further include but is not limited to the following steps S704 to S707.

[0178] Step S605, calculating the error sum of squares of the oxidation weight gain value and the oxidation weight gain fitting value to obtain a third intermediate result.

[0179] In some embodiments, the error sum of squares of the oxidation weight gain value and the oxidation weight gain fitting value is calculated. Specifically, for each alloy sample in the first candidate set, the oxidation weight gain value is subtracted from the oxidation weight gain fitting value to obtain a fitting difference value, then the fitting difference value is squared to obtain a square result, and finally all square results are accumulated to obtain the error sum of squares as the third intermediate result.

[0180] Step S606, calculating the total sum of squares of the oxidation weight gain value to obtain a fourth intermediate result.

[0181] In some embodiments, the average value of the oxidation weight gain can be calculated according to the oxidation weight gain value of the alloy sample in the first candidate set. For each alloy sample, the oxidation weight gain value is subtracted from the average value of the oxidation weight gain to obtain a difference result, the difference result is squared to obtain a square result, and finally all square results are accumulated to obtain the total sum of squares of the oxidation weight gain value as the fourth intermediate result.

[0182] Step S607, dividing the third intermediate result by the fourth intermediate result to obtain a fifth intermediate result, and subtracting 1 from the fifth intermediate result to obtain the first linearity value.

[0183] In some embodiments, the third intermediate result is divided by the fourth intermediate result to obtain the fifth intermediate result, and finally 1 is subtracted from the fifth intermediate result to calculate the first linearity value. Specifically, the first linearity value satisfies the following formula:

[0184]

[0185] wherein, k1 represents the first sample number of the alloy samples in the first candidate set, y represents the oxidation weight gain value, represents the oxidation weight gain average value, and y' represents the oxidation weight gain fitting value.

[0186] It can be understood that the specific implementation of the second linearity value calculation process is basically consistent with the first linearity value, and satisfies the following calculation formula:

[0187]

[0188] wherein, k2 represents the second sample number of the alloy samples in the second candidate set, y represents the oxidation weight gain value, represents the oxidation weight gain average value, and y' represents the oxidation weight gain fitting value.

[0189] Referring to Figure 6 It can be understood that the specific implementation of the second linearity value calculation process is basically consistent with the first linearity value, and satisfies the following calculation formula:

[0190] Step S701, according to the target kinetic characteristic, comparing the numerical size of the kinetic characteristic value corresponding to the alloy sample in the parent node and the target kinetic characteristic value.

[0191] In some embodiments, according to the target kinetic characteristic, the numerical size of the kinetic characteristic value corresponding to the alloy sample in the parent node and the target kinetic characteristic value is compared one by one. For example, when the target kinetic characteristic is iron element and the target kinetic characteristic value is 49.74wt.%, the mass percentage of iron element in the 95 alloy samples in the parent node and 49.74wt.% are compared in numerical size.

[0192] Step S702, if the kinetic characteristic value is less than or equal to the target kinetic characteristic value, the alloy sample is divided into the first sub-node.

[0193] In some embodiments, if the kinetic characteristic value is less than or equal to the target kinetic characteristic value, for example, the mass percentage of iron element in the alloy sample is less than or equal to 49.74wt.%, the alloy sample is divided into the first sub-node. It can be understood that at this time, the first sub-node corresponds to the division parameter, which is the iron element and the mass percentage of 49.74wt.% for the first candidate set, so the linearity value of the first sub-node is the first linearity value of the first candidate set. Similarly, the oxidation kinetic fitting equation of the first sub-node is the first fitting result of the first candidate set.

[0194] Step S703, if the kinetic characteristic value is greater than the target kinetic characteristic value, the alloy sample is divided into the second sub-node.

[0195] In some embodiments, if the kinetic characteristic value is greater than the target kinetic characteristic value, for example, the mass percentage of iron element in the alloy sample is greater than 49.74wt.%, the alloy sample is divided into the second sub-node. It can be understood that at this time, the second sub-node corresponds to the division parameter of 49.74wt.% of the mass percentage of iron element, and the second candidate set is obtained by dividing, and thus the linearity value of the second sub-node is the second linearity value of the second candidate set. Similarly, the oxidation kinetic fitting equation of the second sub-node is the second fitting result of the second candidate set.

[0196] Referring to Figure 7 and Figure 8 As shown in FIGS. 8, 9 and 10, in some embodiments of the present application, the step S304 can further include, but is not limited to, the following steps S801-S802 and / or steps S901-S902.

[0197] In step S801, if the first linearity value of the first sub-node meets the linearity threshold, the first sub-node is taken as an end node.

[0198] In some embodiments, the stop branch condition includes a linearity threshold, for example, the linearity threshold is 0.9. If the first linearity value of the first sub-node meets the linearity threshold, for example, the first linearity value is greater than 0.9, the first sub-node is taken as an end node. It can be understood that if the first linearity value does not meet the linearity threshold, the first sub-node is taken as a parent node to execute the kinetic characteristic differentiation process, and the first linearity value of the first sub-node is taken as the total linearity value of the parent node. The present embodiment does not limit this.

[0199] In step S802, if the second linearity value of the second sub-node meets the linearity threshold, the second sub-node is taken as an end node.

[0200] In some embodiments, if the second linearity value of the second sub-node meets the linearity threshold, for example, the second linearity value is greater than 0.9, the second sub-node is taken as an end node. It can be understood that if the second linearity value does not meet the linearity threshold, the second sub-node is taken as a parent node to execute the kinetic characteristic differentiation process, and the second linearity value of the second sub-node is taken as the total linearity value of the parent node. The present embodiment does not limit this.

[0201] In step S901, if the oxidation time of the alloy sample in the first sub-node meets the single-degree threshold, the parent node of the first sub-node is taken as an end node.

[0202] In some embodiments, the stop branch condition can further include, but is not limited to, a single degree threshold, for example, the single degree threshold is 2. If the oxidation time value corresponding to the oxidation time of the alloy sample in the first child node satisfies the single degree threshold, for example, the oxidation time value only has two different values of 500h and 1200h, that is, the oxidation time value in the alloy sample is less than or equal to two cases, the parent node of the first child node is taken as an end node. It can be understood that since the oxidation time value of the alloy sample is less than or equal to two cases, it belongs to obvious over-fitting linearity, which will cause the regression of the oxidation kinetics equation to be unstable, so this node is not allowed to exist, and the parent node thereof needs to be taken as an end node.

[0203] In step S902, if the oxidation time of the alloy sample in the second child node satisfies the single degree threshold, the parent node of the second child node is taken as an end node.

[0204] In some embodiments, if the oxidation time value corresponding to the oxidation time of the alloy sample in the second child node satisfies the single degree threshold, for example, the oxidation time value only has two different values of 100h and 300h, that is, the oxidation time value in the alloy sample is less than or equal to two cases, the parent node of the second child node is taken as an end node. It can be understood that since the oxidation time value of the alloy sample is less than or equal to two cases, it belongs to obvious over-fitting linearity, which will cause the regression of the oxidation kinetics equation to be unstable, so this node is not allowed to exist, and the parent node thereof needs to be taken as an end node.

[0205] It can be understood that the stop branch condition can also be that the linearity value of the child node is compared with the linearity value of the parent node, for example, the linearity value of the child node is less than the linearity value of the parent node, and the corresponding parent node is taken as an end node. Those skilled in the art can set different stop branch conditions according to actual needs, and the present embodiment does not limit this.

[0206] Referring to FIG. 10, Figure 9 In some embodiments of the present application, the oxidation kinetics index calculation method can further include, but is not limited to, the following steps S1001 to S1002.

[0207] In step S1001, the oxidation kinetics index of the end node and the alloy sample are input into the symbolic regression model.

[0208] In some embodiments, the end node includes at least one alloy sample, and an oxidation kinetics fitting equation of the alloy sample is obtained by using the oxidation kinetics equation, and the oxidation kinetics fitting equation includes the oxidation kinetics index. Therefore, each alloy sample in the end node is described as one oxidation kinetics fitting equation, and these alloy samples have the same oxidation kinetics index.

[0209] In some embodiments, after constructing the oxidation feature tree, multiple end nodes can be obtained, and the oxidation kinetic index of the end node and the corresponding alloy sample are input into a symbolic regression model. For example, the symbolic regression model is constructed based on an open source regression algorithm, such as gplearn or SISSO, and the present embodiment does not limit this.

[0210] In step S1002, the oxidation kinetic index is taken as a regression target, and the kinetic features in the alloy sample are subjected to symbolic regression to obtain an analytical equation of the oxidation kinetic index.

[0211] In some embodiments, the oxidation kinetic index of the end node is taken as a regression target, and the kinetic features of the alloy sample in the end node are subjected to symbolic regression, and an analytical expression between the oxidation kinetic index and the kinetic features can be obtained, i.e., an analytical equation of the oxidation kinetic index. For example, n = f(kinetic feature 1, kinetic feature 2,..., kinetic feature m), where f represents an analytical function, and n represents the oxidation kinetic index.

[0212] It can be understood that the obtained function of n is substituted into the oxidation kinetic equation y' = kt n The analytical equation of the oxidation weight gain can be obtained, and thus the final oxidation kinetic equation satisfies the kinetic process and is completely analytical, effectively analyzing the oxidation mechanism of the multi-component alloy, facilitating the design of new materials more resistant to oxidation, and promoting the development of related research.

[0213] The present application is described below through a complete embodiment:

[0214] Referring to Figure 10 The oxidation feature tree diagram shown in FIG. 1, the present example takes the oxidation mechanism of supercritical CO2 alloy as the research object. The obtained oxidation data set has 95 supercritical CO2 alloy samples, and the kinetic features of each alloy sample include alloy components such as iron, nickel, chromium, aluminum, cobalt, manganese, silicon, titanium, copper, carbon, and bismuth, and the corresponding kinetic feature values are the mass fractions of the alloy components. The kinetic features also include temperature, pressure, and oxidation time.

[0215] It can be understood that different alloys have different compositions, and the oxidation data obtained by exposing them to various test conditions will result in very complex changes in oxidation weight gain, which brings great difficulties to data analysis and oxidation research. Since the experimental cost of the oxidation sample is very high, especially under special environmental conditions such as supercritical water, supercritical CO2, etc., it is not realistic to conduct a large number of control experiments. The present embodiment studies the analytical relationship between the oxidation kinetic index and the kinetic features in the alloy sample by the linear regression tree classifier oxidation kinetic index calculation method, which has the advantages of simplicity, speed, and accuracy.

[0216] Firstly, the root node numbered #95 is constructed based on the oxidation dataset, and the alloy samples are fitted by using the oxidation kinetics equation after logarithmic processing, and the linear degree value corresponding to the node is calculated. It can be understood that the oxidation dataset also includes secondary alloy features generated by randomly combining alloy components by addition, subtraction and multiplication. The final fitting obtains the oxidation kinetic index n=0.585 of the root node, the fitting constant ln(k)=-1.224, and the calculated linear degree value LG=0.284, which is less than the linear degree threshold 0.9, so it does not meet the stop branching condition, and the kinetic feature differentiation process needs to be performed.

[0217] Specifically, the kinetic features and kinetic feature values of the alloy samples in the root node are traversed, such as (iron+nickel), (chromium-aluminum+manganese), …, iron, nickel, chromium, aluminum, cobalt, manganese, silicon, titanium, copper, carbon, bismuth, temperature, pressure, and the like, and the kinetic feature values thereof are used as division parameters to split the alloy samples in the root node to obtain a first candidate set and a second candidate set. The first candidate set and the second candidate set are fitted by using the oxidation kinetics equation to obtain the corresponding fitting results, i.e., the oxidation kinetics fitting equation. Finally, the linear gain value corresponding to the division parameter is calculated according to the first linear degree value of the first fitting result, the second linear degree value of the second fitting result, and the total linear degree value of the parent node.

[0218] Through traversal comparison, it is found that when the kinetic feature is iron element and the kinetic target feature value is 49.74wt.%, the linear gain value for differentiating the parent node is the largest. From Figure 10 It can be known that at this time, the total number of samples is 95, the first number of samples is 65, the second number of samples is 35, the first linear degree value is 0.667, and the second linear degree value is 0.612, so the calculation linear gain value corresponds to Therefore, the iron element is taken as the target kinetic feature and its mass percentage 49.74wt.% is taken as the target kinetic feature value to split the alloy samples in the parent node, and 95 alloy samples are split into a first sub-node numbered #60 and a second sub-node numbered #34, wherein the numerical value of the node number represents the number of alloy samples in the node.

[0219] Since the first sub-node and the second sub-node do not meet the stop differentiation condition, they are taken as the parent node to perform the kinetic feature differentiation process, and the branches continue to grow to the end nodes, and finally the oxidation feature tree of nine end nodes is obtained. In some embodiments, the end nodes are numbered from left to right starting from NO_0. From Figure 11It can be seen that the linearity LG of the oxidation kinetic equation is the smallest at the end node No_7, which is 0.76, and the largest at the leaf No_5, which is 0.96. A quadratic alloy feature (Co-(Al-Mn)) is found to be used in the differentiation process, which has a significant impact on the oxidation kinetic index n.

[0220] Thus, 95 alloy samples are divided into 9 end nodes, and each alloy sample is marked with its corresponding oxidation kinetic index value, as shown in Table 4 below. By using the alloy samples in Table 4 as the regression target of the oxidation kinetic index, the analytical equation of the oxidation kinetic index, i.e., n = f(iron, nickel, chromium, aluminum, cobalt, manganese, silicon, titanium, copper, carbon, bismuth, temperature, pressure), can be obtained, which is not limited in the embodiment.

[0221] Table 4: Data pairs of features and oxidation kinetic index

[0222]

[0223] In the field of material technology, it is of great significance to combine machine learning with expert knowledge mechanism analysis. Machine learning can extract the correlation between material properties and performance through large-scale data analysis, accelerating the material discovery and design process. For example, by analyzing data such as the composition, structure, and processing conditions of materials, machine learning can predict the properties and performance of materials. Expert knowledge mechanism analysis can provide a deep understanding of the basic principles and behavior of materials, providing targeted guidance and interpretation capabilities for machine learning algorithms.

[0224] The linear regression tree classifier oxidation kinetic index calculation method is applied to the oxidation data of multi-component alloys, and a loss function that meets the kinetic equation is constructed based on the correlation between features and targets. Thus, expert knowledge can be easily introduced. In a huge feature space, relying on the consistency of the kinetic equation in describing data, data that meet different physical mechanisms can be distinguished. The main driving factors that affect the physical mechanism can be identified, and the mapping relationship between features and mechanisms can be constructed, and then the definition space of the features can be continuously. In turn, combined with symbolic regression, interpretive algorithms, etc., the physical process of metal oxidation is described analytically under the kinetic framework. This is of great significance for the discovery and revision of the oxidation kinetic equation, the description of the oxidation process, the mechanism research of oxidation-resistant materials, and the design of oxidation-resistant materials.

[0225] It can be understood that the embodiments of the present application are not limited to the study of the oxidation kinetic equation, but can also be directly applied to the activation energy analysis of the thermal activation process in thermodynamics. For example, in the node, the thermal activation process that meets the thermodynamic equation fitting data, where Q is the activation energy constant, R is the ideal gas constant, and T is the temperature, which is not limited in the embodiment.

[0226] This invention also provides an oxidation kinetic index calculation device, which can implement the above-mentioned oxidation kinetic index calculation method, referring to... Figure 11 As shown, in some embodiments of this application, the oxidation kinetic index calculation device includes:

[0227] The first acquisition module 100 is used to acquire a pre-constructed oxidation feature tree. The oxidation feature tree is a binary tree, which includes multiple nodes and node edges. Each node includes a root node and at least two end nodes. The node edges store alloy feature values, which are the boundaries for dividing alloy data.

[0228] The data comparison module 200 is used to divide nodes from the alloy data based on the alloy feature data. Starting from the root node, it determines the target node edge one by one according to the comparison result between the alloy feature data and the alloy feature value on the node edge, until the end node is reached and the end node is determined as the target end node.

[0229] The index acquisition module 300 is used to use the oxidation kinetic index stored in the target end node as the oxidation kinetic index of the alloy data.

[0230] Reference Figure 12 As shown, in some embodiments of this application, the oxidation kinetic index calculation device further includes:

[0231] The second acquisition module 400 is used to acquire an oxidation dataset; wherein the oxidation dataset includes multiple alloy samples, and each alloy sample includes multiple kinetic features and corresponding kinetic feature values;

[0232] Differentiation module 500 is used to perform a dynamic feature differentiation process based on the oxidation dataset to construct an oxidation feature tree; the dynamic feature differentiation process includes:

[0233] Construct a parent node; the parent node stores alloy samples; the initial value of the parent node is the root node, which is constructed based on the oxidation feature set;

[0234] Traverse the dynamic characteristics and dynamic characteristic values ​​of alloy samples in the parent node, and select the target dynamic characteristics and target dynamic characteristic values ​​from them;

[0235] Based on the target dynamic characteristics and target dynamic characteristic values, the alloy sample in the parent node is split to obtain the first child node and the second child node of the parent node, and the target dynamic characteristic values ​​are stored in the corresponding node edges;

[0236] If the first child node and / or the second child node do not meet the differentiation stop condition, the first child node and / or the second child node are taken as parent nodes, and the dynamic feature differentiation process is executed.

[0237] The analysis module 600 is configured to input the oxidation kinetic index of the end node and the alloy sample into the symbolic regression model; perform symbolic regression on the kinetic characteristics in the alloy sample with the oxidation kinetic index as the regression target, to obtain an analytical equation of the oxidation kinetic index.

[0238] The specific implementation of the oxidation kinetic index calculation device in the embodiment is basically the same as that of the oxidation kinetic index calculation method described above, and will not be repeated here.

[0239] Figure 13 An electronic device 1000 provided by an embodiment of the present application is shown. The electronic device 1000 includes a processor 1001, a memory 1002, and a computer program stored in the memory 1002 and executable on the processor 1001, and the computer program is configured to execute the oxidation kinetic index calculation method described above when executed.

[0240] The processor 1001 and the memory 1002 can be connected by a bus or other means.

[0241] The memory 1002, as a non-transitory computer readable storage medium, can be used to store non-transitory software programs and non-transitory computer executable programs, such as the oxidation kinetic index calculation method described in the embodiments of the present application. The processor 1001 executes the non-transitory software programs and instructions stored in the memory 1002, thereby implementing the oxidation kinetic index calculation method described above.

[0242] The memory 1002 can include a program storage area and a data storage area, wherein the program storage area can store an operating system and at least one application required by a function; the data storage area can store the oxidation kinetic index calculation method described above. In addition, the memory 1002 can include a high-speed random access memory 1002, and can also include a non-transitory memory 1002, such as at least one storage device memory device, a flash memory device or other non-transitory solid-state memory device. In some embodiments, the memory 1002 can optionally include a memory 1002 remotely arranged with respect to the processor 1001, and these remote memories 1002 can be connected to the electronic device 1000 through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0243] The non-transitory software programs and instructions required to implement the oxidation kinetic index calculation method described above are stored in the memory 1002, and when executed by one or more processors 1001, the oxidation kinetic index calculation method described above is executed, for example, the method steps S101 to S104 in Figure 1 , the method steps S201 to S202 and steps S301 to S304 in Figure 2 ,Figure 3 Method steps S401 to S405 Figure 4 Method steps S501 to S506 in the above method Figure 5 Method steps S601 to S607, Figure 6 Method steps S701 to S703, Figure 7 Method steps S801 to S802, Figure 8 Method steps S901 to S902, Figure 9 The method steps S1001 to S1002.

[0244] This application also provides a storage medium, which is a computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned oxidation kinetic index calculation method. The memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, the memory may include high-speed random access memory and non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0245] The oxidation kinetic index calculation method, apparatus, electronic device, and storage medium provided in this application obtain a pre-constructed oxidation feature tree. The oxidation feature tree is a binary tree, including multiple nodes and edges. Each node includes a root node and at least two end nodes, and the edges store alloy data values. Then, nodes are divided from the alloy data based on alloy feature data. Starting from the root node of the oxidation feature tree, target node edges are determined one by one based on the comparison results between the alloy feature data and the alloy feature values ​​on the node edges, until an end node is reached and designated as the target end node. Finally, the oxidation kinetic index stored in the target end node is used as the oxidation kinetic index of the alloy data. Thus, based on the pre-constructed oxidation feature tree, oxidation kinetic indices corresponding to different alloy compositions are stored in end nodes. By comparing the alloy feature data in the alloy data with the nodes and edges of the oxidation feature tree to progressively determine the target end node, the oxidation kinetic index and oxidation kinetic equation can be accurately calculated and obtained. This is of great significance for tasks such as oxidation process description, oxidation-resistant material mechanism research, and oxidation resistance design, thereby promoting related research development.

[0246] The present application continuously divides the oxidation data of alloy samples in the feature space by maximizing the linearity gain, so that the oxidation data of alloy samples on the same end node can be described by the same oxidation kinetics equation. After obtaining all the end node alloy samples and their oxidation kinetics indexes, the oxidation kinetics indexes obtained by all the end nodes are expressed as an analytical equation of the studied features by the symbolic regression algorithm gplearn, and then a unified and analytical oxidation kinetics equation is used to describe the oxidation of different alloys under different conditions. Thus, expert knowledge is introduced, and in the huge feature space, the consistency of the data to the kinetic equation description is relied on to distinguish the data satisfying different physical mechanisms. The main driving factors affecting the physical mechanism can be identified, the mapping relationship between the features and the mechanism is constructed, and then the definition space is continuously. Combined with symbolic regression, explanatory algorithms and other algorithms, machine learning is used to accelerate the cognitive physical process. The present application can be used in, for example, multicomponent alloy performance optimization, discovery and correction of oxidation kinetics equation, oxidation process description, oxidation-resistant material mechanism research and oxidation-resistant design and other tasks. The present application is a machine learning data analysis method based on oxidation kinetics relationship, and the feasibility of the method for discovering and strengthening the correlation between data, analyzing the oxidation mechanism of alloys is verified in related data evaluation.

[0247] The above-described embodiments are merely illustrative for describing the embodiments and the units described as separate components can or can not be physically separated, that is, can be located in one place, or can be distributed to multiple network units. According to actual needs, part or all of the modules can be selected to achieve the purpose of the embodiments.

[0248] As will be appreciated by one of ordinary skill in the art, all or some of the steps, systems, and techniques disclosed herein can be embodied in software, firmware, hardware, and / or suitable combination thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a micro-processing unit, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on computer readable media, which can comprise computer storage media (or non-transitory media), and communication media (or transitory media). As is well known to those of ordinary skill in the art, the term computer storage media includes both volatile and non-volatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by a computer. Further, as will be appreciated by one skilled in the art, communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal, such as carrier waves or other transport mechanisms, and includes any information delivery media.

[0249] It should also be appreciated that various embodiments provided by the present application can be arbitrarily combined to achieve different technical effects. The above is a specific description of the preferred embodiments of the present application, but the present application is not limited to the above embodiments, and those skilled in the art can make various equivalent modifications or replacements without departing from the spirit of the present application.

Claims

1. An oxidative kinetic index calculation method characterized by, The method comprises the following steps: obtaining a pre-constructed oxidation feature tree, the oxidation feature tree being a binary tree, the binary tree comprising a plurality of nodes and node edges, the nodes comprising at least a root node and at least two end nodes, and the node edges storing alloy feature values which are boundaries of alloy data division; dividing the nodes from the alloy data according to alloy feature data, starting from the root node, determining target node edges one by one according to comparison results of the alloy feature data and the alloy feature values on the node edges, and reaching the end nodes until the end nodes are determined as target end nodes; taking the oxidation kinetic index stored in the target end nodes as the oxidation kinetic index of the alloy data; constructing the oxidation feature tree, comprising: obtaining an oxidation data set; wherein the oxidation data set comprises a plurality of alloy samples, each of the alloy samples comprising a plurality of kinetic features and corresponding kinetic feature values; performing a kinetic feature differentiation process based on the oxidation data set to construct the oxidation feature tree; the kinetic feature differentiation process comprises: constructing a parent node; the parent node stores the alloy samples; the initial value of the parent node is the root node, and the root node is constructed based on the oxidation data set; traversing the kinetic features and the kinetic feature values of the alloy samples in the parent node to select target kinetic features and target kinetic feature values therefrom; splitting the alloy samples in the parent node according to the target kinetic features and the target kinetic feature values to obtain a first child node and a second child node of the parent node, and storing the target kinetic feature values in corresponding node edges; determining that the first child node and / or the second child node do not satisfy a stop differentiation condition, taking the first child node and / or the second child node as the parent node, and performing the kinetic feature differentiation process.

2. The method according to claim 1, characterized by, The kinetic features of the alloy samples comprise oxidation time and alloy composition, and the alloy samples further comprise oxidation weight gain; after obtaining the oxidation data set, the following steps are further included: performing logarithmic preprocessing on the oxidation time and the oxidation weight gain; performing combined calculation on the alloy composition to generate a preset number of secondary alloy features; adding the secondary alloy features to the oxidation data set.

3. The method according to claim 1, characterized by, The step of traversing the kinetic features and the kinetic feature values of the alloy samples in the parent node to select target kinetic features and target kinetic feature values therefrom comprises the following steps: selecting a division parameter from the kinetic features; the division parameter comprises a candidate feature and a candidate feature value, and the candidate feature value is obtained from the kinetic feature values of the candidate feature in the alloy samples; dividing the alloy samples in the parent node into a first candidate set and a second candidate set according to the numerical values of the candidate feature values; obtaining a first fitting result of the first candidate set and a second fitting result of the second candidate set by using an oxidation kinetic equation; calculating a linear gain value of the division parameter according to a first linearity value of the first fitting result, a second linearity value of the second fitting result, and a total linearity value of the parent node; updating the candidate characteristic value by traversing the dynamic characteristic value, and updating the candidate characteristic by traversing the dynamic characteristic, and obtaining the target dynamic characteristic and the target dynamic characteristic value according to the division parameter with the largest linear gain value.

4. The method according to claim 3, wherein the method is characterized by, The calculating a linear gain value of the division parameter according to a first linearity value of the first fitting result, a second linearity value of the second fitting result, and a total linearity value of the parent node comprises: calculating the first linearity value according to the first fitting result and the alloy sample in the first candidate set; calculating the second linearity value according to the second fitting result and the alloy sample in the second candidate set; obtaining a total sample quantity of the alloy sample in the parent node, a first sample quantity of the alloy sample in the first candidate set, and a second sample quantity of the alloy sample in the second candidate set; calculating a first sample proportion according to the first sample quantity and the total sample quantity, and calculating a second sample proportion according to the second sample quantity and the total sample quantity; multiplying the first sample proportion and the first linearity value to obtain a first linearity index, and multiplying the second sample proportion and the second linearity value to obtain a second linearity index; adding the first linearity index and the second linearity index and subtracting the total linearity value to obtain the linear gain value of the division parameter.

5. The method according to claim 4, wherein the method is characterized by, The first fitting result is an oxidation dynamic fitting equation, the dynamic characteristic value comprises an oxidation time value, and the alloy sample further comprises an oxidation weight gain value; the calculating the first linearity value according to the first fitting result and the alloy sample in the first candidate set comprises: substituting the oxidation time value of the alloy sample in the first candidate set into the oxidation dynamic fitting equation to obtain an oxidation weight gain fitting value; calculating a first intermediate result by calculating a covariance of the oxidation weight gain value and the oxidation weight gain fitting value; calculating a first standard deviation by calculating a standard deviation of the oxidation weight gain value, calculating a second standard deviation by calculating a standard deviation of the oxidation weight gain fitting value, and multiplying the first standard deviation and the second standard deviation to obtain a second intermediate result; dividing the first intermediate result by the second intermediate result to obtain the first linearity value; or calculating a third intermediate result by calculating a sum of squares of errors of the oxidation weight gain value and the oxidation weight gain fitting value; calculating a fourth intermediate result by calculating a total sum of squares of the oxidation weight gain value; dividing the third intermediate result by the fourth intermediate result to obtain a fifth intermediate result, and subtracting the fifth intermediate result from 1 to obtain the first linearity value. The splitting the alloy sample in the parent node according to the target dynamic characteristic and the target dynamic characteristic value to obtain a first child node and a second child node of the parent node comprises:

6. The method according to claim 1, wherein ​ comparing a numerical value of the kinetic characteristic value corresponding to the alloy sample in the parent node with the target kinetic characteristic value according to the target kinetic characteristic; if the kinetic characteristic value is less than or equal to the target kinetic characteristic value, the alloy sample is divided into the first sub-node; if the kinetic characteristic value is greater than the target kinetic characteristic value, the alloy sample is divided into the second sub-node.

7. The method according to claim 1, wherein the method is characterized by, The stop differentiation condition includes a linearity threshold value and / or a single degree threshold value; the method further includes: if a first linearity value of the first sub-node satisfies the linearity threshold value, the first sub-node is taken as the end node; if a second linearity value of the second sub-node satisfies the linearity threshold value, the second sub-node is taken as the end node; and / or, if an oxidation time of the alloy sample in the first sub-node satisfies the single degree threshold value, the parent node of the first sub-node is taken as the end node; if an oxidation time of the alloy sample in the second sub-node satisfies the single degree threshold value, the parent node of the second sub-node is taken as the end node.

8. The oxidation kinetic index calculation method according to any one of claims 1 to 7, characterized by, The end node includes a plurality of alloy samples, an oxidation kinetic fitting equation of the alloy sample is obtained by using an oxidation kinetic equation, the oxidation kinetic fitting equation includes the oxidation kinetic index; the method further includes: inputting the oxidation kinetic index of the end node and the alloy sample into a symbolic regression model; performing symbolic regression on the kinetic characteristics in the alloy sample with the oxidation kinetic index as a regression target to obtain an analytical equation of the oxidation kinetic index.

9. An oxidation kinetic index calculation device characterized by comprising: The application of the oxidation kinetic index calculation method according to any one of claims 1 to 8 includes: a first acquisition module, configured to acquire a pre-constructed oxidation characteristic tree, the oxidation characteristic tree being a binary tree, the binary tree including a plurality of nodes and node edges, the nodes including at least a root node and at least two end nodes, and the node edges storing alloy characteristic values, the alloy characteristic values being boundaries for dividing alloy data; a data comparison module, configured to divide the nodes from the alloy data according to alloy characteristic data, determine target node edges one by one from the root node according to comparison results of the alloy characteristic data and the alloy characteristic values on the node edges, and determine the end nodes as target end nodes until the end nodes are reached; an index acquisition module, configured to take the oxidation kinetic index stored in the target end node as an oxidation kinetic index of the alloy data; a second acquisition module, configured to acquire an oxidation data set; wherein the oxidation data set includes a plurality of alloy samples, and each alloy sample includes a plurality of kinetic characteristics and corresponding kinetic characteristic values. The differentiation module is configured to perform a kinetic feature differentiation process based on the oxidation dataset to construct the oxidation feature tree; the kinetic feature differentiation process comprises: constructing a parent node; the parent node stores the alloy sample; an initial value of the parent node is the root node, and the root node is constructed based on the oxidation dataset; traversing the kinetic feature and the kinetic feature value of the alloy sample in the parent node to select a target kinetic feature and a target kinetic feature value; splitting the alloy sample in the parent node according to the target kinetic feature and the target kinetic feature value to obtain a first child node and a second child node of the parent node, and storing the target kinetic feature value in a corresponding node edge; determining whether the first child node and / or the second child node satisfies a stop differentiation condition, taking the first child node and / or the second child node as the parent node, and performing the kinetic feature differentiation process.

10. The oxidation kinetic index calculation device according to claim 9, characterized by Further comprising: The analysis module is configured to input the oxidation kinetic index of the end node and the alloy sample into a symbolic regression model; regression, to obtain an analytical equation of the oxidation kinetic index.

11. An electronic device, comprising: The memory stores a computer program, and the processor executes the computer program to realize the oxidation kinetic index calculation method in any one of claims 1 to 8.

12. A computer-readable storage medium, characterized in that, The storage medium stores a program, and the program is executed by the processor to realize the oxidation kinetic index calculation method in any one of claims 1 to 8.

Images