Data acquisition method, device and equipment, readable storage medium and program product

Abstract

The invention relates to a data acquisition method and device, equipment, a readable storage medium and a program product. The method comprises the following steps: acquiring a plurality of initial sample data, and performing temperature interval extraction on a first temperature interval and a second temperature interval according to a preset sample expansion multiple for one initial sample data, obtaining a plurality of first temperature subintervals corresponding to the first temperature interval and a plurality of second temperature subintervals corresponding to the second temperature interval, carrying out sample expansion processing on the input data sequence according to the plurality of first temperature subintervals, and carrying out sample expansion processing on the tag data sequence according to the plurality of second temperature subintervals; the initial sample data comprises an input data sequence and a label data sequence corresponding to the input data sequence, the input data sequence corresponds to the first temperature interval, and the label data sequence corresponds to the second temperature interval. By adopting the method, the sample data acquisition cost can be reduced.

Description

Technical Field

[0001] The present application relates to the field of thermal technology, and in particular to a data acquisition method, device, equipment, readable storage medium and program product. Background Art

[0002] With the development of artificial intelligence technology, artificial intelligence models have been widely used in various industries. For example, in the field of thermal technology, thermal properties can be described by thermal property models. During the development process of the model, it is necessary to train with a large amount of sample data to obtain a mature and usable model.

[0003] The traditional method of obtaining sample data is usually for researchers to conduct experimental tests under different preset conditions to obtain the sample data required for the model training process.

[0004] However, the above sample data acquisition method has the problem of high acquisition cost. Summary of the invention

[0005] Based on this, it is necessary to provide a data acquisition method, device, equipment, readable storage medium and program product that can reduce the cost of acquiring sample data in response to the above technical problems.

[0006] In a first aspect, the present application provides a data acquisition method, comprising:

[0007] Acquire a plurality of initial sample data, the initial sample data comprising an input data sequence and a label data sequence corresponding to the input data sequence, the input data sequence corresponds to a first temperature interval, and the label data sequence corresponds to a second temperature interval;

[0008] For an initial sample data, temperature interval extraction is performed on the first temperature interval and the second temperature interval respectively according to a preset sample expansion multiple, to obtain a plurality of first temperature sub-intervals corresponding to the first temperature interval and a plurality of second temperature sub-intervals corresponding to the second temperature interval;

[0009] Sample expansion processing is performed on the input data sequence according to the multiple first temperature sub-intervals, and sample expansion processing is performed on the label data sequence according to the multiple second temperature sub-intervals to obtain multiple target sample data.

[0010] In one embodiment, temperature interval extraction is performed on the first temperature interval and the second temperature interval respectively according to a preset sample expansion multiple to obtain a plurality of first temperature sub-intervals corresponding to the first temperature interval and a plurality of second temperature sub-intervals corresponding to the second temperature interval, including:

[0011] Determine a plurality of first temperature points according to the sample expansion multiple and the first interval conversion condition;

[0012] Determine a plurality of second temperature points according to the sample expansion multiple and the second interval conversion condition;

[0013] A temperature interval is extracted from the first temperature interval according to each first temperature point to obtain a plurality of first temperature sub-intervals, and a temperature interval is extracted from the second temperature interval according to each second temperature point to obtain a plurality of second temperature sub-intervals.

[0014] In one embodiment, a temperature interval is extracted from a first temperature interval according to each first temperature point to obtain a plurality of first temperature sub-intervals, and a temperature interval is extracted from a second temperature interval according to each second temperature point to obtain a plurality of second temperature sub-intervals, including:

[0015] Obtaining a first starting point corresponding to the first temperature interval and a second starting point corresponding to the second temperature interval;

[0016] For a first temperature point, determining a temperature interval between a first starting point and a first temperature point as a first temperature sub-interval;

[0017] For a second temperature point, a temperature interval between the second starting point and the second temperature point is determined as a second temperature sub-interval.

[0018] In one embodiment, sample expansion processing is performed on the input data sequence according to a plurality of first temperature sub-intervals, and sample expansion processing is performed on the label data sequence according to a plurality of second temperature sub-intervals to obtain a plurality of target sample data, including:

[0019] For a first temperature sub-interval, according to the number of first sampling points and the first sampling conversion condition, a plurality of first sampling temperature points are determined, each of the first sampling temperature points being within the first temperature sub-interval;

[0020] For the second temperature sub-interval corresponding to the first temperature sub-interval, according to the number of second sampling points and the second sampling conversion condition, a plurality of second sampling temperature points are determined, each of which is within the second temperature sub-interval;

[0021] The target sample data within the first temperature sub-interval and the second temperature sub-interval corresponding to the first temperature sub-interval are determined according to each first sampling temperature point, each second sampling temperature point, the input data sequence fragment corresponding to the first temperature sub-interval and the label data sequence fragment corresponding to the second temperature sub-interval.

[0022] In one embodiment, determining target sample data within the first temperature sub-interval according to each first sampling temperature point, each second sampling temperature point, an input data sequence segment corresponding to the first temperature sub-interval, and a label data sequence segment corresponding to the second temperature sub-interval includes:

[0023] Determine the sampled input data corresponding to each first sampling temperature point according to the input data sequence fragment corresponding to the first temperature sub-interval and the first interpolation condition;

[0024] Determine the sampling label data corresponding to each second sampling temperature point according to the label data sequence fragment corresponding to the second temperature sub-interval and the second preset interpolation condition;

[0025] The target sample data corresponding to the first temperature sub-interval is determined according to each sample input data and each sample label data.

[0026] In one embodiment, determining the target sample data corresponding to the first temperature sub-interval according to each sample input data and each sample tag data includes:

[0027] The target sampling label data in each sampling label data is used as the target label data, and the order of the second sampling temperature points corresponding to the target sampling label data in the second temperature sub-interval meets the preset condition;

[0028] Taking each sample input data as target input data;

[0029] The target input data and the target label data are used as the target sample data.

[0030] In a second aspect, the present application also provides a model training method, comprising:

[0031] Acquire a plurality of target sample data using the data acquisition method provided in any one of the embodiments of the first aspect;

[0032] The initial model is iteratively trained according to multiple target sample data to obtain a target model, which is used to determine label data corresponding to the input data according to the input data.

[0033] In a third aspect, the present application further provides a data acquisition device, comprising:

[0034] An acquisition module, used to acquire a plurality of initial sample data, the initial sample data including an input data sequence and a label data sequence corresponding to the input data sequence, the input data sequence corresponds to the first temperature interval, and the label data sequence corresponds to the second temperature interval;

[0035] A temperature interval extraction module is used to extract the temperature interval of the first temperature interval and the second temperature interval according to a preset sample expansion multiple for an initial sample data, and obtain a plurality of first temperature sub-intervals corresponding to the first temperature interval and a plurality of second temperature sub-intervals corresponding to the second temperature interval;

[0036] The target module is used to perform sample expansion processing on the input data sequence according to multiple first temperature sub-intervals, and perform sample expansion processing on the label data sequence according to multiple second temperature sub-intervals to obtain multiple target sample data.

[0037] In a fourth aspect, an embodiment of the present application provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method of the first aspect when executing the computer program.

[0038] In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the method of the first aspect described above are implemented.

[0039] In a sixth aspect, the present application further provides a computer program product, wherein the computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the method in the first aspect are implemented.

[0040] The above-mentioned data acquisition method, device, equipment, readable storage medium and program product, by acquiring multiple initial sample data, can, for one initial sample data, perform temperature interval extraction on the first temperature interval and the second temperature interval respectively according to a preset sample expansion multiple, to obtain multiple first temperature sub-intervals corresponding to the first temperature interval and multiple second temperature sub-intervals corresponding to the second temperature interval, and then perform sample expansion processing on the input data sequence according to the multiple first temperature sub-intervals, and perform sample expansion processing on the label data sequence according to the multiple second temperature sub-intervals to obtain multiple target sample data, wherein the initial sample data includes an input data sequence and a label data sequence corresponding to the input data sequence, the input data sequence corresponds to the first temperature interval, and the label data sequence corresponds to the second temperature interval. In this way, temperature interval extraction is performed on the first temperature interval corresponding to the input data sequence used to input the model in the initial sample data according to the sample expansion multiple that the sample data is to be expanded, and temperature interval extraction is performed on the second temperature interval to which the label data sequence corresponding to the input data sequence belongs, to obtain multiple groups of first temperature sub-intervals and second temperature sub-intervals, and based on each of the first temperature sub-intervals and the second temperature sub-intervals, multiple expanded target sample data are obtained, and the number of target sample data obtained is the product of the original initial sample data and the sample expansion multiple, which greatly increases the scale of the sample data and reduces the cost of obtaining the sample data. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 An application environment diagram of a data acquisition method in an embodiment;

[0043] Figure 2 A schematic diagram of a flow chart of a data acquisition method in an embodiment;

[0044] Figure 3 A schematic diagram of initial sample data obtained from a set of experiments in another embodiment;

[0045] Figure 4 is a flow chart of step 202 in another embodiment;

[0046] Figure 5 is a schematic flow chart of step 203 in another embodiment;

[0047] Figure 6 is a flow chart of an exemplary data acquisition method in another embodiment;

[0048] Figure 7 A schematic diagram of a flow chart of a model training method in one embodiment;

[0049] Figure 8 A schematic diagram of the input and output dimensions of a model in another embodiment;

[0050] Fig. 9 A schematic diagram of a flow chart of a model training process in another embodiment;

[0051] Fig.10 is a schematic diagram of the change of loss value with the number of iterations in another embodiment;

[0052] Fig.11 A schematic diagram showing a comparison between average interpolation and test loss values ​​in another embodiment;

[0053] Fig.12 A schematic diagram of comparison between predicted label data and target label data in another embodiment;

[0054] Fig.13 is a schematic diagram of second predicted label data and a second loss value in another embodiment;

[0055] Fig.14 is a structural block diagram of a data acquisition device in one embodiment;

[0056] Fig.15 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0058] With the development of artificial intelligence technology, artificial intelligence models have been widely used in various industries, such as the field of thermal technology. In the field of thermal technology, there are still many conversion reaction mechanisms and problems that have not been deeply understood. Using artificial intelligence methods, researchers can establish a "black box" model based on sufficient experimental data to accurately describe or predict the characteristic data of samples. During the development process of the model, it is necessary to train with a large amount of sample data to obtain a mature and usable model. The data scale of the sample data is usually more than one million.

[0059] Compared with application scenarios such as image recognition and language processing, it is very difficult to obtain data samples in the field of engineering research. The traditional method of obtaining sample data in the field of thermal technology is usually for researchers to conduct experimental tests under different preset conditions to obtain the sample data required for the model training process.

[0060] However, the cost of experiments is very high. If you want to obtain a large amount of sample data through experiments, you will have to pay a very high cost.

[0061] In addition, sample data from different sources (such as different researchers or different experimental environments resulting in different sources of sample data) are not reliable when training models.

[0062] In view of this, the present application provides a data acquisition method, apparatus, device, readable storage medium and program product. By acquiring multiple initial sample data, temperature interval extraction can be performed on the first temperature interval and the second temperature interval respectively according to a preset sample expansion multiple for one initial sample data, so as to obtain multiple first temperature sub-intervals corresponding to the first temperature interval and multiple second temperature sub-intervals corresponding to the second temperature interval, and then sample expansion processing is performed on the input data sequence according to the multiple first temperature sub-intervals, and sample expansion processing is performed on the label data sequence according to the multiple second temperature sub-intervals to obtain multiple target sample data, wherein the initial sample data includes an input data sequence and a label data sequence corresponding to the input data sequence, the input data sequence corresponds to the first temperature interval, and the label data sequence corresponds to the second temperature interval. In this way, temperature interval extraction is performed on the first temperature interval corresponding to the input data sequence used to input the model in the initial sample data according to the sample expansion multiple that the sample data is to be expanded, and temperature interval extraction is performed on the second temperature interval to which the label data sequence corresponding to the input data sequence belongs, to obtain multiple groups of first temperature sub-intervals and second temperature sub-intervals, and based on each of the first temperature sub-intervals and the second temperature sub-intervals, multiple expanded target sample data are obtained, and the number of target sample data obtained is the product of the original initial sample data and the sample expansion multiple, which greatly increases the scale of the sample data and reduces the cost of obtaining the sample data.

[0063] The data acquisition method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Among them, the data storage system can store the data that the server 101 needs to process. The data storage system can be integrated on the server 101, or it can be placed on the cloud or other network servers. The server 101 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides cloud computing services.

[0064] In an exemplary embodiment, Figure 2 As shown, a data acquisition method is provided, which is applied to Figure 1 The method is described by taking the server 101 in the example as an example. It is understandable that the method can also be applied to a terminal, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. The method includes the following steps 201 to 203. Among them:

[0065] Step 201, obtaining a plurality of initial sample data.

[0066] The initial sample data may be sample data obtained after at least one experiment on a thermal sample (such as a battery), wherein the initial sample data includes an input data sequence and a label data sequence corresponding to the input data sequence.

[0067] Let the input data sequence be X (k) , represents the kth input data sequence, and the label data sequence corresponding to the input data sequence is Y (k) , represents the kth label data sequence, then the kth initial sample data can be recorded as (X (k) , Y (k) ),k={1,2,……,K}.

[0068] The input data sequence can be temperature-related data used to input into the artificial intelligence model. In an embodiment of the present application, the input data sequence can characterize certain thermal properties of the thermal engineering sample at different temperatures, such as the heat generation power characteristics of the positive and negative electrodes of the battery and the electrolyte material.

[0069] Optionally, the input data sequence may include only one type of thermal characteristic data; optionally, the input data sequence may include multiple types of thermal characteristic data.

[0070] The label input sequence can be data related to temperature and affected by the input data sequence. It is also the data that the artificial intelligence model hopes to obtain after analyzing the input data sequence. The label input sequence can also characterize certain thermal properties of thermal samples at different temperatures, such as the temperature rise rate characteristics of thermal samples at different temperatures. The temperature rise rate is affected by the heat generation power characteristics of the positive and negative electrodes of the battery and the electrolyte materials.

[0071] Optionally, the input data sequence may include only one type of thermal characteristic data; optionally, the input data sequence may include multiple types of thermal characteristic data.

[0072] The input data sequence corresponds to the first temperature interval, and the label data sequence corresponds to the second temperature interval. In the embodiment of the present application, the first temperature interval may be the same as the second temperature interval, or may be different from the second temperature interval.

[0073] It can be understood that both the first temperature interval and the second temperature interval are monotonically increasing or monotonically decreasing. For example, the first temperature interval may be 0-300°C, and the second temperature interval may be 600-300°C.

[0074] Taking the thermal engineering sample as an example, the battery is composed of a positive electrode, a negative electrode, an electrolyte and a current collector, a diaphragm, a tab, a battery shell, etc. through a certain structure. In the process of the battery generating heat from self-heat to the battery thermal runaway and complete scrapping, complex thermochemical reactions occur between the positive electrode, negative electrode, electrolyte and other materials inside the battery, resulting in a continuous increase in the battery temperature. In the embodiment of the present application, the model can be used to determine the relationship between the heat generation power characteristics of the positive and negative electrodes and the electrolyte materials of the battery and the temperature rise rate, that is, the corresponding temperature rise rate is determined according to the heat generation power characteristics of the positive and negative electrodes and the electrolyte materials of the battery. The input data sequence may include the heat generation power of the positive electrode material, the heat generation power of the negative electrode material, the heat generation power of the positive electrode + negative electrode material, the heat generation power of the positive electrode + electrolyte, the heat generation power of the negative electrode + electrolyte, etc., and the label data sequence may include the temperature rise rate characteristics.

[0075] Regarding the process of acquiring initial sample data, in one possible implementation, the server can determine the specific data included in the input data sequence and the label data sequence based on the model input and output goals, and then conduct experiments under preset conditions based on the specific data included in the input data sequence and the label data sequence to obtain multiple initial sample data.

[0076] In an embodiment of the present application, if the input data sequence includes the heat generation power of the positive electrode material, the heat generation power of the negative electrode material, the heat generation power of the positive electrode + negative electrode material, the heat generation power of the positive electrode + electrolyte, the heat generation power of the negative electrode + electrolyte, etc., a calorimetric instrument (such as a differential scanning calorimeter) can be used in the experiment to test batteries of different materials or different material combinations under linear temperature rising conditions to obtain the heat generation power curves corresponding to the batteries of different materials or different material combinations, thereby obtaining data such as the heat generation power of the positive electrode material, the heat generation power of the negative electrode material, the heat generation power of the positive electrode + negative electrode material, the heat generation power of the positive electrode + electrolyte, and the heat generation power of the negative electrode + electrolyte; if the output data sequence includes temperature rise rate data, an adiabatic accelerating calorimeter can be used in the experiment to test batteries of different materials or different material combinations to obtain a temperature rise rate curve corresponding to the input data sequence, with reference to Figure 3 , is a set of initial sample data obtained from experiments, including.

[0077] Optionally, the server may obtain the initial sample data input by the user through an external input device; optionally, the server may directly obtain the stored initial sample data from a database.

[0078] Step 202 , for an initial sample data, temperature interval extraction is performed on the first temperature interval and the second temperature interval respectively according to a preset sample expansion multiple to obtain a plurality of first temperature sub-intervals corresponding to the first temperature interval and a plurality of second temperature sub-intervals corresponding to the second temperature interval.

[0079] The cost of a single experiment is too high. For example, the differential scanning calorimeter for testing the input data sequences such as the heat generation power of the positive electrode material, the heat generation power of the negative electrode material, the heat generation power of the positive electrode + negative electrode material, the heat generation power of the positive electrode + electrolyte, and the heat generation power of the negative electrode + electrolyte takes 3 hours and costs 1,000 yuan. The adiabatic accelerating calorimeter required to detect the label data sequence including the temperature rise rate takes 48 hours and costs 50,000 yuan. If you want to obtain large-scale sample data, the cost is too high. In order to save costs, in an embodiment of the present application, after obtaining multiple groups of initial sample data, the server can obtain a preset sample expansion multiple, and expand the initial sample data according to the preset sample expansion multiple, so as to obtain the target sample data of the preset data scale.

[0080] Among them, the preset sample expansion multiple is the multiple by which the data volume of the initial sample data is expected to be finally expanded. In the embodiment of the present application, the sample expansion multiple can be determined based on the model parameters and the data volume of the initial sample data. Referring to formula (1), the sample expansion multiple is λ, K is the data volume of the initial sample data, and ε is the number of model parameters.

[0081]

[0082] Then the amount of target sample data obtained after expansion is N, where N=λK.

[0083] After the server obtains the sample expansion multiple and the initial sample data, it can expand the initial sample data. In an embodiment of the present application, an idea of ​​temperature interval extraction is introduced. The server can perform temperature interval extraction on the first temperature interval corresponding to the input data sequence and the second temperature interval corresponding to the label data sequence according to the sample expansion multiple, and obtain input data sequence fragments corresponding to multiple first temperature sub-intervals and label data sequence fragments corresponding to multiple second temperature sub-intervals, thereby cutting the initial sample data into multiple sample data and increasing the number of sample data.

[0084] In a possible implementation, an idea of ​​temperature interval extraction based on reaction kinetics is introduced. In engineering applications such as thermal engineering, the thermal characteristics represented by the sample data are usually determined by a kinetic process within a temperature range, such as a thermal characteristic-temperature curve. The thermal characteristics of the sample (such as a battery) at a certain moment are related to the historical temperature change process. Therefore, in an embodiment of the present application, the server can obtain a plurality of first temperature points for temperature interval extraction corresponding to the first temperature interval and a plurality of second temperature points for temperature interval extraction corresponding to the second temperature interval according to the sample expansion multiple, and then use the starting point of the first temperature interval as the starting point of each first temperature sub-interval, the starting point of the second temperature interval as the starting point of each second temperature sub-interval, and each temperature point as the end point of each temperature sub-interval, so as to obtain a plurality of first temperature sub-intervals and a plurality of second temperature sub-intervals. For example, if the first temperature interval is 0-300°C, the first temperature point includes 100°C and 150°C, then the first temperature sub-interval includes 0-100°C, 0-150°C and 0-300°C.

[0085] Step 203 , performing sample expansion processing on the input data sequence according to the multiple first temperature sub-intervals, and performing sample expansion processing on the label data sequence according to the multiple second temperature sub-intervals, so as to obtain multiple target sample data.

[0086] After obtaining multiple first temperature sub-intervals and second temperature sub-intervals based on the first temperature interval and the second temperature interval, the input data sequence and the label data sequence are naturally divided into multiple input data sequence fragments corresponding to each first temperature sub-interval and multiple label data sequence fragments corresponding to each second temperature sub-interval. At this time, the server can expand each input data sequence fragment and each label data sequence fragment to obtain multiple target sample data.

[0087] Optionally, the server may input each input data sequence segment and each label data sequence segment into a trained expansion model respectively, thereby obtaining input data and label data after expansion of each input data sequence segment and each label data sequence segment, and the server may use these expanded input data and label data as target sample data.

[0088] Optionally, the input data sequence segment includes multiple input data, and the label data sequence segment includes multiple label data. The server can interpolate the input sequence segment according to the input data in the input data sequence segment to obtain an expanded input data sequence segment. The server can use the expanded input data sequence segment as the target input data. The same expansion method can be used for the label data sequence segment to obtain an expanded label data sequence segment. The server can use the expanded label data sequence segment as the target label data. Finally, the target input data and the target label data are used as target sample data.

[0089] In a possible implementation, after obtaining large-scale target sample data, the server may use the target sample data for model training.

[0090] In the above embodiment, the server performs temperature interval extraction on the first temperature interval corresponding to the input data sequence used for inputting the model in the initial sample data according to the sample expansion multiple by which the sample data is to be expanded, and performs temperature interval extraction on the second temperature interval to which the label data sequence corresponding to the input data sequence belongs, to obtain multiple groups of first temperature sub-intervals and second temperature sub-intervals, and obtains multiple expanded target sample data based on each of the first temperature sub-intervals and the second temperature sub-intervals, the number of the obtained target sample data is the product of the original initial sample data and the sample expansion multiple, which greatly increases the scale of the sample data and reduces the cost of obtaining the sample data.

[0091] In one embodiment, based on the above Figure 2 The embodiment shown, see Figure 4 This embodiment involves a process of extracting temperature intervals from a first temperature interval and a second temperature interval according to a preset sample expansion multiple, to obtain a plurality of first temperature sub-intervals corresponding to the first temperature interval and a plurality of second temperature sub-intervals corresponding to the second temperature interval. Figure 4 As shown, step 202 may include steps 401 to 403 .

[0092] Step 401, determining a plurality of first temperature points according to the sample expansion multiple and the first interval conversion condition.

[0093] Step 402, determining a plurality of second temperature points according to the sample expansion multiple and the second interval conversion condition.

[0094] Step 403 , extracting the temperature interval of the first temperature interval according to each first temperature point to obtain a plurality of first temperature sub-intervals, and extracting the temperature interval of the second temperature interval according to each second temperature point to obtain a plurality of second temperature sub-intervals.

[0095] For the first temperature interval, the server can first obtain multiple first temperature points for cutting. Regarding the method of obtaining the first temperature point, in an embodiment of the present application, the server can determine the number of temperature sub-intervals extracted from the first temperature interval based on the sample expansion multiple, thereby determining the number of first temperature points. For example, the first temperature interval is 0-300°C, and the sample expansion multiple is 3, which means that the first temperature interval needs to be generated into 3 first temperature sub-intervals, then the number of first temperature points is 3, and the last point is the end point of the first temperature interval, that is, 300°C.

[0096] After determining the number of first temperature points, the server can generate a rational number sequence corresponding to each first temperature point, denoted as Z λ (i)=i,{i=1,2,...,λ}.

[0097] Then, the server may convert the rational number sequence corresponding to each first temperature point according to the first interval conversion condition to obtain multiple first temperature points. The conversion process may refer to the following formula:

[0098] T X (i) = f TX (Z λ (i)), i=1,2,…,λ (2)

[0099] Among them, T X (i) represents the first temperature point of the i-th node, X represents the input data, and f TX () is the conversion formula corresponding to the first interval conversion condition. The conversion formula can be determined according to the first interval conversion condition. The first interval conversion condition can be linear conversion, polynomial conversion, exponential conversion, logarithmic conversion, etc.

[0100] Similarly, for the second temperature interval, the server can obtain multiple second temperature points according to the sample expansion multiple and the second interval conversion condition. The conversion process can refer to the following formula:

[0101] T Y (i) = f TY (Z λ (i)), i=1,2,…,λ (3)

[0102] Among them, T Y (i) represents the i-th second temperature point, Y represents the label data, and f TY () is the conversion formula corresponding to the second interval conversion condition. Regarding this conversion formula, it can be determined according to the second interval conversion condition. The first interval conversion condition can be linear conversion, polynomial transformation, exponential transformation, logarithmic transformation, etc.

[0103] It is understandable that f TX () and fTY () In the conversion interval Z λ are all monotonic functions.

[0104] After the server obtains the first temperature point corresponding to the first temperature interval and the second temperature point corresponding to the second temperature interval, the server performs temperature interval extraction processing on each temperature interval. Regarding the temperature interval extraction processing process, in a possible implementation, the server can obtain the first starting point corresponding to the first temperature interval and the second starting point corresponding to the second temperature interval, and then, for a first temperature point, determine the temperature interval between the first starting point and a first temperature point as a first temperature sub-interval, and for a second temperature point, determine the temperature interval between the second starting point and a second temperature point as a second temperature sub-interval.

[0105] In the above embodiment, the first starting point is the temperature point at which the first temperature interval starts, and the second starting point is the temperature point at which the second temperature interval starts. For example, if the first temperature interval is 0-300°C, the first starting point is 0°C, and if the second temperature interval is 600-200°C, the second starting point is 600°C.

[0106] For a first temperature point, the temperature interval between the first starting point and the first temperature point is determined to be a first temperature sub-interval, and for a second temperature point, the temperature interval between the second starting point and the second temperature point is determined to be a second temperature sub-interval. For example, the first starting point is 0°C, the second starting point is 600°C, the first temperature points are 10°C, 100°C and 200°C, and the second temperature points are 550°C, 450°C and 350°C. Then the first temperature sub-interval is 0-10°C, 0-100°C and 0-200°C, and the second temperature sub-interval is 600-550°C, 600-450°C and 600-350°C.

[0107] Regarding the process of obtaining each first temperature sub-interval and each second temperature sub-interval, the following formula may be referred to:

[0108]

[0109] Among them, T X1 Represents the first starting point, T Y1 Represents the second starting point, range(T X,i ) represents the first temperature sub-interval corresponding to the i-th first temperature point, range(T Y,i ) represents the second temperature sub-interval corresponding to the i-th second temperature point.

[0110] In one embodiment, based on the above Figure 2 The embodiment shown, see Figure 5This embodiment relates to a process of performing sample expansion processing on an input data sequence according to a plurality of first temperature sub-intervals, and performing sample expansion processing on a label data sequence according to a plurality of second temperature sub-intervals, so as to obtain a plurality of target sample data. Figure 5 As shown, step 203 may include steps 501 to 503 .

[0111] Step 501 : for a first temperature sub-interval, a plurality of first sampling temperature points are determined according to the number of first sampling points and a first sampling conversion condition.

[0112] Wherein, each first sampling temperature point is within the first temperature sub-interval.

[0113] Step 502 : for a second temperature sub-interval corresponding to the first temperature sub-interval, a plurality of second sampling temperature points are determined according to the number of second sampling points and a second sampling conversion condition.

[0114] Wherein, each second sampling temperature point is within the second temperature sub-interval.

[0115] Step 503, determine the target sample data within the first temperature sub-interval and the second temperature sub-interval corresponding to the first temperature sub-interval according to each first sampling temperature point, each second sampling temperature point, the input data sequence fragment corresponding to the first temperature sub-interval and the label data sequence fragment corresponding to the second temperature sub-interval.

[0116] The first number of sampling points is the number of sampling points that are expected to be included in the first temperature sub-interval. Similarly, the second number of sampling points is the number of sampling points that are expected to be included in the second temperature sub-interval. In the embodiment of the present application, the number of sampling points can be selected by researchers based on actual application scenarios, and the value is generally between 10-10000, wherein the first number of sampling points and the second number of sampling points may be different.

[0117] For a first temperature sub-interval, the server can generate a first sampling point sequence corresponding to the first temperature sub-interval according to the number of first sampling points and the number of second sampling points, denoted as Zγ x (j x )=j,j x ={1,2,…,γx}, where γ x is the number of the first sampling points. Similarly, the server can generate a second sampling point sequence corresponding to the second temperature sub-interval, denoted as Zγ Y (j Y )=j,j Y ={1,2,……,γ Y}, where γ Y is the number of the second sampling points.

[0118] For a first temperature sub-interval, the server may determine multiple first sampling temperature points according to the sampling point sequence and the first sampling conversion condition. The conversion process may refer to the following formula:

[0119] T X,i (j X ) = f TX,i (Z γX (j X )),j X =1,2,…,γ X (5)

[0120] Among them, T x,i (j x ) represents the jth temperature in the i-th first temperature sub-interval x The first sampling temperature point, f TX,i () represents the first sampling conversion condition corresponding to the first temperature sub-interval. The first sampling conversion condition can be a simple prior conversion relationship or a polynomial transformation, exponential transformation, logarithmic transformation, etc.

[0121] Similarly, for a second temperature sub-interval, the server may determine a plurality of second sampling temperature points according to the sampling point sequence and the second sampling conversion condition corresponding to the second temperature sub-interval. The conversion process may refer to the following formula:

[0122] T Y,i (j Y ) = f TY,i (Z γY (j Y )),j Y =1,2,…,γ Y (6)

[0123] Among them, T Y,j (j Y ) represents the jth temperature in the i-th second temperature sub-interval Y The second sampling temperature point, f TY,i () represents the second sampling conversion condition corresponding to the second temperature sub-interval. The second sampling conversion condition can be a simple prior conversion relationship or a polynomial transformation, exponential transformation, logarithmic transformation, etc.

[0124] After obtaining multiple first sampling temperature points and multiple second sampling temperature points, the server can determine the target sample data within the first temperature sub-interval based on the first sampling temperature points, the second sampling temperature points, the input data sequence fragments corresponding to the first temperature sub-interval, and the label data sequence fragments corresponding to the second temperature sub-interval.

[0125] In a possible implementation, the process may include the following steps: the server determines the sampled input data corresponding to each first sampling temperature point based on the input data sequence fragment corresponding to the first temperature sub-interval and the first interpolation condition; the server determines the sampled label data corresponding to each second sampling temperature point based on the label data sequence fragment corresponding to the second temperature sub-interval and the second preset interpolation condition; the server determines the target sample data within the first temperature sub-interval and the second temperature sub-interval corresponding to the first temperature sub-interval based on each sampled input data and each sampled label data.

[0126] It can be understood that the input data sequence fragments corresponding to the first temperature sub-interval and the label data sequence fragments corresponding to the second temperature sub-interval are both discrete data. Therefore, the input data corresponding to each first sampling temperature point and the label data corresponding to each second sampling temperature point do not necessarily exist. In order to obtain the data corresponding to each sampling temperature point, the server can perform interpolation processing based on each sampling temperature point.

[0127] For the multiple first sampling temperature points included in the first temperature sub-interval, the server can perform interpolation processing according to the input data sequence fragment corresponding to the first temperature sub-interval and the first interpolation condition, so as to obtain the sampled input data corresponding to each first sampling temperature point. Regarding the first interpolation condition, the server can select it according to the actual application conditions, such as linear interpolation, polynomial interpolation, spline curve interpolation, etc. The interpolation process can refer to the following formula:

[0128] X i =interpX(T X ,X,T X,i ), i=1,2,…,λ (7)

[0129] T X represents the first temperature sub-interval, X represents the input data, T X,i Indicates that the first temperature sub-interval is the i-th, X i The server can use X to represent the thermal characteristic curve corresponding to the first temperature sub-interval. i A set of sampled input data corresponding to each first sampled temperature point is determined.

[0130] Similarly, the server can perform interpolation processing based on the tag data sequence fragment corresponding to the second temperature sub-interval and the second interpolation condition, so as to obtain the sampled input data corresponding to each second sampling temperature point. Regarding the second interpolation condition, the server can select it according to the actual application conditions, such as linear interpolation, polynomial interpolation, spline curve interpolation, etc. The interpolation process can refer to the following formula:

[0131] Y i =interpY(T Y,Y,T Y,i ), i=1,2,…,λ (8)

[0132] T Y represents the second temperature sub-interval, Y represents the label data, T Y,i Indicates that the second temperature sub-interval is the i-th, Y i Represents the thermal characteristic curve corresponding to the second temperature sub-interval. The server can i A set of sampling label data corresponding to each first sampling temperature point is determined.

[0133] In a possible implementation, after obtaining each sample input data and each sample label data, the server can directly use the sample input data corresponding to each first temperature sub-interval as the target input data, and use the sample label data corresponding to the second temperature sub-interval corresponding to each first temperature sub-interval as the target label data. Finally, the target input data and the target label data are used as the target sample data. The target sample data can be recorded as It can be seen that dividing each initial sample data into λ parts greatly increases the amount of sample data.

[0134] In another possible implementation, since the label data may be the label of each sampled input data in the first temperature sub-interval, for a first temperature sub-interval, the server may assume that the label data corresponding to each sampled input data is consistent, and the server may set the number of second sampling points to 1. Accordingly, there is only one sampled label data corresponding to the second temperature sub-interval, and the server may use the sampled label data as the target label data, thereby reducing the waste of computing resources.

[0135] In an embodiment of the present application, the process of the server determining the target sample data corresponding to the first temperature sub-interval based on each sampling input data and each sampling label data may include the following steps: the server uses the target sampling label data in each sampling label data as the target label data; the server uses each sampling input data as the target input data; the server uses the target input data and the target label data as the target sample data.

[0136] It is understandable that for a first temperature sub-interval, the server can assume that the label data corresponding to each sampled input data is consistent. Therefore, the server can select a target sampled label data from multiple sampled label data as the target label data, wherein the order of the second sampled temperature points corresponding to the target sampled label data in the second temperature sub-interval meets the preset conditions. For example, the second sampled temperature point is the last point in the second temperature sub-interval. At this time, the sampled label data can be regarded as the final state of the second temperature sub-interval. Then the target label data Yi corresponding to the second temperature sub-interval can be determined by referring to the following formula:

[0137] Y i =Y i,interp (λ) (9)

[0138] After determining the target input data corresponding to the first temperature interval and the target label data corresponding to the second temperature interval, the server may use the target input data and the target label data as the target sample data in the first temperature interval and the second temperature sub-interval corresponding to the first temperature sub-interval.

[0139] In one embodiment, referring to Figure 6 , provides an exemplary data acquisition method, which can be applied to Figure 1 In the implementation environment shown.

[0140] Step 601, obtaining a plurality of initial sample data.

[0141] The initial sample data includes an input data sequence and a label data sequence corresponding to the input data sequence, the input data sequence corresponds to the first temperature interval, and the label data sequence corresponds to the second temperature interval.

[0142] Step 602, for an initial sample data, determine a plurality of first temperature points according to the sample expansion multiple and the first interval conversion condition.

[0143] Step 603, determining a plurality of second temperature points according to the sample expansion multiple and the second interval conversion condition.

[0144] Step 604: Acquire a first starting point corresponding to the first temperature interval and a second starting point corresponding to the second temperature interval.

[0145] Step 605: for a first temperature point, determine a temperature interval between a first starting point and a first temperature point as a first temperature sub-interval.

[0146] Step 606: for a second temperature point, determine a temperature interval between the second starting point and the second temperature point as a second temperature sub-interval.

[0147] Step 607 : for a first temperature sub-interval, a plurality of first sampling temperature points are determined according to the number of first sampling points and the first sampling conversion condition.

[0148] Wherein, each first sampling temperature point is within the first temperature sub-interval.

[0149] Step 608 : for the second temperature sub-interval corresponding to the first temperature sub-interval, determine a plurality of second sampling temperature points according to the number of second sampling points and the second sampling conversion condition.

[0150] Wherein, each second sampling temperature point is within the second temperature sub-interval.

[0151] Step 609 : Determine the sampled input data corresponding to each first sampling temperature point according to the input data sequence fragment corresponding to the first temperature sub-interval and the first interpolation condition.

[0152] Step 610: Determine the sampled label data corresponding to each second sampling temperature point according to the label data sequence fragment corresponding to the second temperature sub-interval and the second preset interpolation condition.

[0153] Step 611: Use target sampled label data in each sampled label data as target label data.

[0154] The order of the second sampling temperature points corresponding to the target sampling tag data in the second temperature sub-interval meets a preset condition.

[0155] Step 612: taking each sample input data as target input data.

[0156] Step 613: Use the target input data and the target label data as target sample data.

[0157] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0158] The present application also provides a model training method, referring to Figure 7 , the method comprising:

[0159] Step 701: Acquire a plurality of target sample data using any of the data acquisition methods in the above embodiments.

[0160] Step 702: iteratively train the initial model according to the multiple target sample data to obtain a target model.

[0161] The target model is used to determine label data corresponding to the input data based on the input data.

[0162] Through the data acquisition method in the above embodiment, the server can obtain the expanded target sample data, and then use the target sample data to train the initial model to obtain the target model.

[0163] The initial model can be selected according to the actual application. It can be a model using machine learning, deep learning, etc., or a combined model composed of different models. In the embodiment of the present application, according to the prediction requirements, such as the input data is the heat generation power of the battery material or material combination, and the predicted label data is the temperature rise rate, then a deep learning convolutional neural network and an artificial neural network combination can be selected to establish a correspondence between the input data and the label data, thereby obtaining the corresponding initial model.

[0164] Before establishing the initial model, you can first determine the input dimension and output dimension of the initial model, refer to Figure 8 , the input dimension is the dimension of the input data, and the output dimension is the dimension of the label data.

[0165] In an embodiment of the present application, if the input data is the heat generation power of the positive electrode material, the heat generation power of the negative electrode material, the heat generation power of the positive electrode + negative electrode material, the heat generation power of the positive electrode + electrolyte, and the heat generation power of the negative electrode + electrolyte, the input dimension is 5. If the label data is the temperature rise rate, the output dimension is 1.

[0166] In a possible implementation, the server may also use the temperature sequence corresponding to the target sample data as an input data, and the input dimension is the dimension of the input data plus 1.

[0167] After determining the input dimension and output dimension of the initial model, the server can build the initial model according to the input dimension, output dimension, selected model, input data, and label data. The network layer of the initial model can be determined according to the actual application requirements. In an embodiment of the present application, the initial model can be composed of three convolutional layers (C1, C2, C3), three pooling layers (P1, P2, P3) and three fully connected layers (D1, D2, D3). The arrangement order of the network layers of the initial model is C1-P1-C2-P2-C3-P3-D1-D2-D3.

[0168] After determining the network layer of the initial model, the server may also set different network parameters. For example, referring to Table 1, the following parameters may be set for the network layer determined above:

[0169]

[0170] Table 1

[0171] It is understandable that the number of network parameters should not be more than the data volume of the target sample data.

[0172] After the initial model is released, the server can iteratively train the initial model based on the target sample data.

[0173] In an embodiment of the present application, the server may first divide the target sample data into a test set and a training set according to a preset ratio. In the embodiment of the present application, the preset ratio may be 90%, wherein the target sample data in the training set is used to train the initial model, and the target sample data in the test set can be used to evaluate the trained model to determine whether the model has converged.

[0174] The server can iteratively train the initial model based on the target sample data in the test set. For one iteration of training, the training process can refer to Fig. 9 , the server can input the target input data in the target sample data into the model to obtain the predicted label data output by the model. The server can calculate the loss value of this iterative training process according to the preset loss function, the predicted label data, and the target label data, and adjust the network parameters in the model according to the loss value. It can be understood that the loss value of each round needs to gradually decrease with the change of the number of iterations. In the embodiment of the present application, reference can be made to Fig.10 is the change of loss value with the number of iterations.

[0175] When adjusting the network parameter values ​​according to the loss value, adjustment measures such as regularization penalty terms can be added to avoid the impact of erroneous target sample data on the network parameter update process.

[0176] In an embodiment of the present application, after the model has been iteratively trained with the target sample data in the test set, a trained intermediate model can be obtained. At this time, the server can verify the intermediate model through the target sample data in the test set to determine whether the intermediate model has converged. If it has converged, the intermediate model is determined to be the target model, otherwise the model training continues.

[0177] Regarding the verification process, the server can input each target input data in the test set into the intermediate model to obtain the predicted label data corresponding to each target input data, and then calculate the loss value corresponding to each target input data based on each predicted label data, the target label data corresponding to each target input data, and the loss function. If the loss value is greater than the preset threshold, the intermediate model is continued to be iteratively trained using each target sample data. If the loss value is less than the preset threshold, it can be determined that the intermediate model has converged, and the intermediate model is determined to be the target model.

[0178] In one possible implementation, the initial model can be trained and evaluated by a five-fold cross-validation method. The process of training the initial model by the five-fold cross-validation method can refer to the following steps: the server divides the training set into five groups, and uses the five training sets to iteratively train the model; for a set of target sample data in the training set, the server uses each target sample data to iteratively train the model to obtain an intermediate model. At this time, the server can verify the intermediate model through the target sample data in the test set to obtain the loss value corresponding to the training set, and continue to iteratively train the intermediate model with the training sets of other groups. By analogy, the target sample data in the five training sets have all trained the model, and the server has obtained the loss values ​​corresponding to the five training sets. The server can calculate the average of the loss values ​​of the five training sets to obtain the average difference. At this time, the server can also use the intermediate model trained by the five training sets to verify the target sample data in the test set to obtain the corresponding test loss value. Fig.11 , the average difference (left, 0.14) and the test loss value (right, 0.134) are similar in value (the difference is less than the threshold), which indicates that the intermediate model has converged and the intermediate model is used as the target model.

[0179] Reference Fig.12 , in order to use the target model to process the target input data in part of the target sample data, the obtained predicted label data and the corresponding loss value, it can be seen that the predicted label data is basically consistent with the target label data corresponding to the target input data.

[0180] After obtaining the target model, the server can also use supplementary test data to perform performance evaluation on the target model. For example, the server can obtain second initial sample data obtained from a new experiment, and then use the data acquisition method in the above embodiment to expand the second initial sample data to obtain second target sample data. The server can use each second target sample data to verify the target model to obtain a performance evaluation result of the target model.

[0181] The results of this process can be found in Fig.13, it can be seen that for the newly acquired second target sample data, the target model analyzes the second target input data in the second target sample data, and the result obtained is the second predicted label data corresponding to the second target input data ( Fig.13 Left), in addition, the second loss value corresponding to the second target sample data is obtained ( Fig.13 Right), it can be seen that the target model has a small error in processing the second target input data and has a high prediction accuracy.

[0182] In the subsequent process, the server may further expand the initial sample data obtained by the re-experiment according to the data acquisition method in the above embodiment, and the obtained target sample data may be used to update the target model.

[0183] Based on the same inventive concept, the embodiment of the present application also provides a data acquisition device for implementing the data acquisition method involved above. The implementation solution provided by the device to solve the problem is similar to the implementation solution recorded in the above method, so the specific limitations in one or more data acquisition device embodiments provided below can refer to the limitations on the data acquisition method above, and will not be repeated here.

[0184] In an exemplary embodiment, Fig.14 As shown, a data acquisition device is provided, including: an acquisition module 1401, a temperature range extraction module 1402 and a target module 1403, wherein:

[0185] An acquisition module 1401 is used to acquire a plurality of initial sample data, the initial sample data including an input data sequence and a label data sequence corresponding to the input data sequence, the input data sequence corresponds to a first temperature interval, and the label data sequence corresponds to a second temperature interval;

[0186] The temperature interval extraction module 1402 is used to extract the temperature interval of the first temperature interval and the second temperature interval according to a preset sample expansion multiple for an initial sample data, and obtain a plurality of first temperature sub-intervals corresponding to the first temperature interval and a plurality of second temperature sub-intervals corresponding to the second temperature interval;

[0187] The target module 1403 is used to perform sample expansion processing on the input data sequence according to multiple first temperature sub-intervals, and perform sample expansion processing on the label data sequence according to multiple second temperature sub-intervals to obtain multiple target sample data.

[0188] In one embodiment, the temperature interval extraction module 1402 includes:

[0189] A first interval conversion unit, used to determine a plurality of first temperature points according to the sample expansion multiple and the first interval conversion condition;

[0190] A second interval conversion unit, used to determine a plurality of second temperature points according to the sample expansion multiple and the second interval conversion condition;

[0191] The temperature interval extraction unit is used to extract the temperature interval of the first temperature interval according to each first temperature point to obtain multiple first temperature sub-intervals, and to extract the temperature interval of the second temperature interval according to each second temperature point to obtain multiple second temperature sub-intervals.

[0192] In one embodiment, the temperature interval extraction unit is specifically used to perform:

[0193] Obtaining a first starting point corresponding to the first temperature interval and a second starting point corresponding to the second temperature interval;

[0194] For a first temperature point, determining a temperature interval between a first starting point and a first temperature point as a first temperature sub-interval;

[0195] For a second temperature point, a temperature interval from the second starting point to the second temperature point is determined as a second temperature sub-interval.

[0196] In one embodiment, the target module 1403 includes:

[0197] A first sampling conversion unit, for determining, for a first temperature sub-interval, a plurality of first sampling temperature points according to the number of first sampling points and a first sampling conversion condition, each of the first sampling temperature points being within the first temperature sub-interval;

[0198] A second sampling conversion unit, for determining, for a second temperature subinterval corresponding to the first temperature subinterval, a plurality of second sampling temperature points according to the number of second sampling points and a second sampling conversion condition, each second sampling temperature point being within the second temperature subinterval;

[0199] A target determination unit is used to determine target sample data within the first temperature sub-interval and the second temperature sub-interval corresponding to the first temperature sub-interval based on each first sampling temperature point, each second sampling temperature point, the input data sequence fragment corresponding to the first temperature sub-interval, and the label data sequence fragment corresponding to the second temperature sub-interval.

[0200] In one embodiment, the target determination unit is specifically configured to perform:

[0201] Determine the sampled input data corresponding to each first sampling temperature point according to the input data sequence fragment corresponding to the first temperature sub-interval and the first interpolation condition;

[0202] Determine the sampling label data corresponding to each second sampling temperature point according to the label data sequence fragment corresponding to the second temperature sub-interval and the second preset interpolation condition;

[0203] The target sample data corresponding to the first temperature sub-interval is determined according to each sample input data and each sample label data.

[0204] In one embodiment, the target determination unit is further configured to perform:

[0205] The target sampling label data in each sampling label data is used as the target label data, and the order of the second sampling temperature points corresponding to the target sampling label data in the second temperature sub-interval meets the preset condition;

[0206] Taking each sample input data as target input data;

[0207] The target input data and the target label data are used as the target sample data.

[0208] Each module in the above data acquisition device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0209] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Fig.15 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data and obtain data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a data acquisition method is implemented.

[0210] Those skilled in the art will understand that Fig.15 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0211] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0212] Acquire a plurality of initial sample data, wherein the initial sample data includes an input data sequence and a label data sequence corresponding to the input data sequence, wherein the input data sequence corresponds to a first temperature interval, and the label data sequence corresponds to a second temperature interval;

[0213] For one of the initial sample data, temperature interval extraction is performed on the first temperature interval and the second temperature interval respectively according to a preset sample expansion multiple to obtain a plurality of first temperature sub-intervals corresponding to the first temperature interval and a plurality of second temperature sub-intervals corresponding to the second temperature interval;

[0214] The input data sequence is subjected to sample expansion processing according to the multiple first temperature sub-intervals, and the label data sequence is subjected to sample expansion processing according to the multiple second temperature sub-intervals, so as to obtain multiple target sample data.

[0215] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0216] Determining a plurality of first temperature points according to the sample expansion multiple and the first interval conversion condition;

[0217] Determining a plurality of second temperature points according to the sample expansion multiple and the second interval conversion condition;

[0218] The first temperature interval is subjected to temperature interval extraction according to each of the first temperature points to obtain a plurality of first temperature sub-intervals, and the second temperature interval is subjected to temperature interval extraction according to each of the second temperature points to obtain a plurality of second temperature sub-intervals.

[0219] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0220] Acquire a first starting point corresponding to the first temperature interval and a second starting point corresponding to the second temperature interval;

[0221] For one of the first temperature points, determining a temperature interval between the first starting point and one of the first temperature points as a first temperature sub-interval;

[0222] For one of the second temperature points, a temperature interval between the second starting point and one of the second temperature points is determined as one of the second temperature sub-intervals.

[0223] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0224] For one of the first temperature sub-intervals, determining a plurality of first sampling temperature points according to the number of first sampling points and a first sampling conversion condition, each of the first sampling temperature points being within the first temperature sub-interval;

[0225] For a second temperature sub-interval corresponding to the first temperature sub-interval, determining a plurality of second sampling temperature points according to the number of second sampling points and a second sampling conversion condition, each of the second sampling temperature points being within the second temperature sub-interval;

[0226] Determine target sample data within the first temperature sub-interval and the second temperature sub-interval corresponding to the first temperature sub-interval based on each of the first sampling temperature points, each of the second sampling temperature points, the input data sequence fragment corresponding to the first temperature sub-interval, and the label data sequence fragment corresponding to the second temperature sub-interval.

[0227] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0228] Determine, according to the input data sequence fragment corresponding to the first temperature subinterval and the first interpolation condition, the sampled input data corresponding to each of the first sampling temperature points;

[0229] Determine the sampling label data corresponding to each of the second sampling temperature points according to the label data sequence fragment corresponding to the second temperature sub-interval and the second preset interpolation condition;

[0230] The target sample data corresponding to the first temperature sub-interval is determined according to each of the sample input data and each of the sample label data.

[0231] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0232] Using target sampling label data in each of the sampling label data as target label data, the order of the second sampling temperature points corresponding to the target sampling label data in the second temperature sub-interval meets a preset condition;

[0233] Taking each of the sampled input data as target input data;

[0234] The target input data and the target label data are used as the target sample data.

[0235] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0236] Acquire a plurality of initial sample data, wherein the initial sample data includes an input data sequence and a label data sequence corresponding to the input data sequence, wherein the input data sequence corresponds to a first temperature interval, and the label data sequence corresponds to a second temperature interval;

[0237] For one of the initial sample data, temperature interval extraction is performed on the first temperature interval and the second temperature interval respectively according to a preset sample expansion multiple to obtain a plurality of first temperature sub-intervals corresponding to the first temperature interval and a plurality of second temperature sub-intervals corresponding to the second temperature interval;

[0238] The input data sequence is subjected to sample expansion processing according to the multiple first temperature sub-intervals, and the label data sequence is subjected to sample expansion processing according to the multiple second temperature sub-intervals, so as to obtain multiple target sample data.

[0239] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0240] Determining a plurality of first temperature points according to the sample expansion multiple and the first interval conversion condition;

[0241] Determining a plurality of second temperature points according to the sample expansion multiple and the second interval conversion condition;

[0242] The first temperature interval is subjected to temperature interval extraction according to each of the first temperature points to obtain a plurality of first temperature sub-intervals, and the second temperature interval is subjected to temperature interval extraction according to each of the second temperature points to obtain a plurality of second temperature sub-intervals.

[0243] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0244] Acquire a first starting point corresponding to the first temperature interval and a second starting point corresponding to the second temperature interval;

[0245] For one of the first temperature points, determining a temperature interval between the first starting point and one of the first temperature points as a first temperature sub-interval;

[0246] For one of the second temperature points, a temperature interval between the second starting point and one of the second temperature points is determined as one of the second temperature sub-intervals.

[0247] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0248] For one of the first temperature sub-intervals, determining a plurality of first sampling temperature points according to the number of first sampling points and a first sampling conversion condition, each of the first sampling temperature points being within the first temperature sub-interval;

[0249] For a second temperature sub-interval corresponding to the first temperature sub-interval, determining a plurality of second sampling temperature points according to the number of second sampling points and a second sampling conversion condition, each of the second sampling temperature points being within the second temperature sub-interval;

[0250] Determine target sample data within the first temperature sub-interval and the second temperature sub-interval corresponding to the first temperature sub-interval based on each of the first sampling temperature points, each of the second sampling temperature points, the input data sequence fragment corresponding to the first temperature sub-interval, and the label data sequence fragment corresponding to the second temperature sub-interval.

[0251] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0252] Determine, according to the input data sequence fragment corresponding to the first temperature subinterval and the first interpolation condition, the sampled input data corresponding to each of the first sampling temperature points;

[0253] Determine the sampling label data corresponding to each of the second sampling temperature points according to the label data sequence fragment corresponding to the second temperature sub-interval and the second preset interpolation condition;

[0254] The target sample data corresponding to the first temperature sub-interval is determined according to each of the sample input data and each of the sample label data.

[0255] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0256] Using target sampling label data in each of the sampling label data as target label data, the order of the second sampling temperature points corresponding to the target sampling label data in the second temperature sub-interval meets a preset condition;

[0257] Taking each of the sampled input data as target input data;

[0258] The target input data and the target label data are used as the target sample data.

[0259] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:

[0260] Acquire a plurality of initial sample data, wherein the initial sample data includes an input data sequence and a label data sequence corresponding to the input data sequence, wherein the input data sequence corresponds to a first temperature interval, and the label data sequence corresponds to a second temperature interval;

[0261] For one of the initial sample data, temperature interval extraction is performed on the first temperature interval and the second temperature interval respectively according to a preset sample expansion multiple to obtain a plurality of first temperature sub-intervals corresponding to the first temperature interval and a plurality of second temperature sub-intervals corresponding to the second temperature interval;

[0262] The input data sequence is subjected to sample expansion processing according to the multiple first temperature sub-intervals, and the label data sequence is subjected to sample expansion processing according to the multiple second temperature sub-intervals, so as to obtain multiple target sample data.

[0263] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0264] Determining a plurality of first temperature points according to the sample expansion multiple and the first interval conversion condition;

[0265] Determining a plurality of second temperature points according to the sample expansion multiple and the second interval conversion condition;

[0266] The first temperature interval is subjected to temperature interval extraction according to each of the first temperature points to obtain a plurality of first temperature sub-intervals, and the second temperature interval is subjected to temperature interval extraction according to each of the second temperature points to obtain a plurality of second temperature sub-intervals.

[0267] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0268] Acquire a first starting point corresponding to the first temperature interval and a second starting point corresponding to the second temperature interval;

[0269] For one of the first temperature points, determining a temperature interval between the first starting point and one of the first temperature points as a first temperature sub-interval;

[0270] For one of the second temperature points, a temperature interval between the second starting point and one of the second temperature points is determined as one of the second temperature sub-intervals.

[0271] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0272] For one of the first temperature sub-intervals, determining a plurality of first sampling temperature points according to the number of first sampling points and a first sampling conversion condition, each of the first sampling temperature points being within the first temperature sub-interval;

[0273] For a second temperature sub-interval corresponding to the first temperature sub-interval, determining a plurality of second sampling temperature points according to the number of second sampling points and a second sampling conversion condition, each of the second sampling temperature points being within the second temperature sub-interval;

[0274] Determine target sample data within the first temperature sub-interval and the second temperature sub-interval corresponding to the first temperature sub-interval based on each of the first sampling temperature points, each of the second sampling temperature points, the input data sequence fragment corresponding to the first temperature sub-interval, and the label data sequence fragment corresponding to the second temperature sub-interval.

[0275] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0276] Determine, according to the input data sequence fragment corresponding to the first temperature subinterval and the first interpolation condition, the sampled input data corresponding to each of the first sampling temperature points;

[0277] Determine the sampling label data corresponding to each of the second sampling temperature points according to the label data sequence fragment corresponding to the second temperature sub-interval and the second preset interpolation condition;

[0278] The target sample data corresponding to the first temperature sub-interval is determined according to each of the sample input data and each of the sample label data.

[0279] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0280] Using target sampling label data in each of the sampling label data as target label data, the order of the second sampling temperature points corresponding to the target sampling label data in the second temperature sub-interval meets a preset condition;

[0281] Taking each of the sampled input data as target input data;

[0282] The target input data and the target label data are used as the target sample data.

[0283] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0284] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0285] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0286] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A data acquisition method, characterized in that: The method comprises: Acquire a plurality of initial sample data, wherein the initial sample data includes an input data sequence and a label data sequence corresponding to the input data sequence, wherein the input data sequence corresponds to a first temperature interval, and the label data sequence corresponds to a second temperature interval; For one of the initial sample data, temperature interval extraction is performed on the first temperature interval and the second temperature interval respectively according to a preset sample expansion multiple to obtain a plurality of first temperature sub-intervals corresponding to the first temperature interval and a plurality of second temperature sub-intervals corresponding to the second temperature interval; The input data sequence is subjected to sample expansion processing according to the multiple first temperature sub-intervals, and the label data sequence is subjected to sample expansion processing according to the multiple second temperature sub-intervals, so as to obtain multiple target sample data.

2. The method according to claim 1, characterized in that: The step of extracting the temperature intervals of the first temperature interval and the second temperature interval according to a preset sample expansion multiple to obtain a plurality of first temperature sub-intervals corresponding to the first temperature interval and a plurality of second temperature sub-intervals corresponding to the second temperature interval includes: Determining a plurality of first temperature points according to the sample expansion multiple and the first interval conversion condition; Determining a plurality of second temperature points according to the sample expansion multiple and the second interval conversion condition; The first temperature interval is subjected to temperature interval extraction according to each of the first temperature points to obtain a plurality of first temperature sub-intervals, and the second temperature interval is subjected to temperature interval extraction according to each of the second temperature points to obtain a plurality of second temperature sub-intervals.

3. The method according to claim 2, characterized in that The step of extracting the temperature interval of the first temperature interval according to each of the first temperature points to obtain a plurality of first temperature sub-intervals, and extracting the temperature interval of the second temperature interval according to each of the second temperature points to obtain a plurality of second temperature sub-intervals includes: Acquire a first starting point corresponding to the first temperature interval and a second starting point corresponding to the second temperature interval; For one of the first temperature points, determining a temperature interval between the first starting point and one of the first temperature points as a first temperature sub-interval; For one of the second temperature points, a temperature interval between the second starting point and one of the second temperature points is determined as one of the second temperature sub-intervals.

4. The method according to claim 1, characterized in that: The step of performing sample expansion processing on the input data sequence according to the plurality of first temperature sub-intervals and performing sample expansion processing on the label data sequence according to the plurality of second temperature sub-intervals to obtain a plurality of target sample data includes: For one of the first temperature sub-intervals, determining a plurality of first sampling temperature points according to the number of first sampling points and a first sampling conversion condition, each of the first sampling temperature points being within the first temperature sub-interval; For a second temperature sub-interval corresponding to the first temperature sub-interval, determining a plurality of second sampling temperature points according to the number of second sampling points and a second sampling conversion condition, each of the second sampling temperature points being within the second temperature sub-interval; Determine target sample data within the first temperature sub-interval and the second temperature sub-interval corresponding to the first temperature sub-interval based on each of the first sampling temperature points, each of the second sampling temperature points, the input data sequence fragment corresponding to the first temperature sub-interval, and the label data sequence fragment corresponding to the second temperature sub-interval.

5. The method according to claim 4, characterized in that The determining the target sample data within the first temperature sub-interval according to each of the first sampling temperature points, each of the second sampling temperature points, the input data sequence fragment corresponding to the first temperature sub-interval, and the label data sequence fragment corresponding to the second temperature sub-interval comprises: Determine, according to the input data sequence fragment corresponding to the first temperature subinterval and the first interpolation condition, the sampled input data corresponding to each of the first sampling temperature points; Determine the sampling label data corresponding to each of the second sampling temperature points according to the label data sequence fragment corresponding to the second temperature sub-interval and the second preset interpolation condition; The target sample data corresponding to the first temperature sub-interval is determined according to each of the sample input data and each of the sample label data.

6. The method according to claim 5, characterized in that The determining the target sample data corresponding to the first temperature sub-interval according to each of the sample input data and each of the sample label data includes: Using target sampling label data in each of the sampling label data as target label data, the order of the second sampling temperature points corresponding to the target sampling label data in the second temperature sub-interval meets a preset condition; Taking each of the sampled input data as target input data; The target input data and the target label data are used as the target sample data.

7. A model training method, characterized in that: The method comprises: Acquire multiple target sample data using any data acquisition method of claims 1-6; The initial model is iteratively trained according to the multiple target sample data to obtain a target model, and the target model is used to determine label data corresponding to the input data according to the input data.

8. A data acquisition device, characterized in that: The device comprises: An acquisition module, configured to acquire a plurality of initial sample data, wherein the initial sample data includes an input data sequence and a label data sequence corresponding to the input data sequence, wherein the input data sequence corresponds to a first temperature interval, and the label data sequence corresponds to a second temperature interval; a temperature interval extraction module, configured to extract the first temperature interval and the second temperature interval according to a preset sample expansion multiple for the initial sample data, and obtain a plurality of first temperature sub-intervals corresponding to the first temperature interval and a plurality of second temperature sub-intervals corresponding to the second temperature interval; The target module is used to perform sample expansion processing on the input data sequence according to the multiple first temperature sub-intervals, and perform sample expansion processing on the label data sequence according to the multiple second temperature sub-intervals to obtain multiple target sample data.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

11. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images