Transformer area load decomposition method based on user grouping and related device

Abstract

The invention provides a transformer area load decomposition method based on user grouping and a related device, and the method comprises the steps: obtaining the power utilization archive information of all users of a target transformer area, and obtaining a plurality of pieces of power utilization archive information; classifying all users of the target transformer area according to the plurality of pieces of power utilization archive information to obtain a plurality of classification groups; executing a preset first calculation operation for each classification group in the plurality of classification groups to obtain a plurality of contribution weight coefficients; determining an optimization interval corresponding to each contribution weight coefficient in the plurality of contribution weight coefficients according to a preset numerical value floating ratio to obtain a plurality of optimization intervals; randomly generating a weight coefficient vector according to each optimization interval in the plurality of optimization intervals to obtain a plurality of initial weight coefficient vectors; and executing a preset second calculation operation for each initial weight coefficient vector in the plurality of initial weight coefficient vectors to obtain a plurality of optimal weight coefficient vectors. The reliability and the accuracy of the district load decomposition result can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of power grid planning, and in particular to a method for decomposing loads in a substation based on user grouping and related devices. Background Art

[0002] At present, the method of classifying users only by the electricity consumption category of the metering point and directly using the electricity proportion of each category as the weight coefficient is too simple, which may lead to unreliable weight coefficients when decomposing loads. When adjusting the weight coefficients according to the existing typical load curve feature library, if the value range of the weight coefficients is not restricted, the model may fall into a local optimal solution and fail to reach a better global optimal solution.

[0003] Therefore, how to improve the reliability and accuracy of the load decomposition results in the substation area needs to be solved urgently. Summary of the invention

[0004] The embodiment of the present application provides a method and related device for substation load decomposition based on user grouping, which can group users according to their electricity consumption information and optimize weight coefficients according to electricity consumption behavior data to improve the reliability and accuracy of the substation load decomposition results.

[0005] In a first aspect, an embodiment of the present application provides a method for decomposing load in a station area based on user grouping, the method comprising:

[0006] Obtaining electricity usage file information of all users in the target area to obtain multiple electricity usage file information; the multiple electricity usage file information is used to characterize the area load of the target area;

[0007] Classify all users in the target area according to the plurality of electricity usage archive information to obtain a plurality of classification groups; the classification groups include any one of the following: residential life, charging service, public service, and store operation;

[0008] Performing a preset first calculation operation on each of the plurality of classification groups to obtain a plurality of contribution weight coefficients; each classification group corresponds to a contribution weight coefficient;

[0009] Determine the optimization interval corresponding to each contribution weight coefficient in the plurality of contribution weight coefficients according to a preset numerical floating ratio, and obtain a plurality of optimization intervals;

[0010] Randomly generate a weight coefficient vector according to each of the multiple optimization intervals to obtain multiple initial weight coefficient vectors; the initial weight coefficient vectors correspond to the optimization intervals one by one;

[0011] A preset second calculation operation is performed on each of the multiple initial weight coefficient vectors to obtain multiple optimal weight coefficient vectors.

[0012] In a second aspect, an embodiment of the present application provides a device for decomposing load in a substation based on user grouping, the device comprising:

[0013] An acquisition module, used to acquire the electricity usage file information of all users in the target area, and obtain a plurality of electricity usage file information; the plurality of electricity usage file information is used to characterize the area load of the target area;

[0014] A classification module, used to classify all users in the target area according to the plurality of electricity usage archive information to obtain a plurality of classification groups; the classification groups include any one of the following: residential life, charging service, public service, and store operation;

[0015] A first calculation module, configured to perform a preset first calculation operation on each of the plurality of classification groups to obtain a plurality of contribution weight coefficients; each classification group corresponds to a contribution weight coefficient;

[0016] A determination module, used to determine the optimization interval corresponding to each contribution weight coefficient in the plurality of contribution weight coefficients according to a preset numerical floating ratio, to obtain a plurality of optimization intervals;

[0017] A generating module, configured to randomly generate a weight coefficient vector according to each of the plurality of optimization intervals to obtain a plurality of initial weight coefficient vectors; the initial weight coefficient vectors correspond to the optimization intervals one by one;

[0018] The second calculation module is used to perform a preset second calculation operation on each of the multiple initial weight coefficient vectors to obtain multiple optimal weight coefficient vectors.

[0019] In a third aspect, an embodiment of the present application provides an electronic device, comprising a processor, a memory, a communication interface, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the processor, and the program includes instructions for executing the steps of any method in the first aspect of the embodiment of the present application.

[0020] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program for electronic data exchange, wherein the computer program enables a computer to execute part or all of the steps described in any method of the first aspect of the embodiment of the present application.

[0021] In a fifth aspect, an embodiment of the present application provides a computer program product, wherein the computer program product includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to cause a computer to execute some or all of the steps described in any method of the first aspect of the embodiment of the present application. The computer program product may be a software installation package.

[0022] By implementing the embodiments of the present application, users can be grouped according to their electricity usage information, and weight coefficients can be optimized according to electricity usage behavior data to improve the reliability and accuracy of the load decomposition results of the substation. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0024] Figure 1 This is a schematic diagram of a classification process of a user group in a station area provided in an embodiment of the present application;

[0025] Figure 2 is a structural schematic diagram of an electronic device provided in an embodiment of the present application;

[0026] Figure 3 It is a flow chart of a method for decomposing load in a substation based on user grouping provided in an embodiment of the present application;

[0027] Figure 4 It is a block diagram of the functional modules of a substation load decomposition device based on user grouping provided in an embodiment of the present application. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0029] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices.

[0030] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article indicates that the associated objects before and after are in an "or" relationship. The "plurality" appearing in the embodiments of the present application refers to two or more.

[0031] In the embodiments of the present application, "at least one item" or similar expressions refer to any combination of these items, including any combination of single items or plural items, and refer to one or more, and multiple refers to two or more. For example, at least one item of a, b, or c can represent the following seven situations: a, b, c, a and b, a and c, b and c, a, b, and c. Among them, each of a, b, and c can be an element or a set containing one or more elements.

[0032] The "connection" that appears in the embodiments of the present application refers to various connection methods such as direct connection or indirect connection to achieve communication between devices, and the embodiments of the present application do not impose any limitations on this.

[0033] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0034] The following is an explanation of the relevant terms involved in this application, as follows:

[0035] Substation: refers to a collection of power equipment and lines within the power supply service area, usually consisting of one or several distribution transformers and the users they supply.

[0036] Electricity usage category at the metering point: refers to the electricity usage category of the user or equipment, which can be divided and classified according to the user's electricity demand characteristics, usage purpose and billing needs.

[0037] At present, the method of classifying users only by the electricity consumption category of the metering point and directly using the electricity proportion of each category as the weight coefficient is too simple, which may lead to unreliable weight coefficients when decomposing loads. When adjusting the weight coefficients according to the existing typical load curve feature library, if the value range of the weight coefficients is not restricted, the model may fall into a local optimal solution and fail to achieve a better global optimal solution. Therefore, how to improve the reliability and accuracy of the load decomposition results in the substation area needs to be solved urgently.

[0038] To solve the above problems, an embodiment of the present application provides a method for decomposing substation load based on user grouping and a related device. First, the electricity usage file information of all users in the target substation is obtained to obtain multiple electricity usage file information; the multiple electricity usage file information is used to characterize the substation load of the target substation; all users in the target substation are classified according to the multiple electricity usage file information to obtain multiple classification groups; the classification groups include any of the following: residential life, charging service, public service, and shop operation; then, a preset first calculation operation is performed for each of the multiple classification groups to obtain multiple contribution weight coefficients; each classification group corresponds to a contribution weight coefficient; the optimization interval corresponding to each of the multiple contribution weight coefficients is determined according to a preset numerical floating ratio to obtain multiple optimization intervals; a weight coefficient vector is randomly generated according to each of the multiple optimization intervals to obtain multiple initial weight coefficient vectors; the initial weight coefficient vector corresponds to the optimization interval one by one; finally, a preset second calculation operation is performed for each of the multiple initial weight coefficient vectors to obtain multiple optimal weight coefficient vectors. Users can be grouped according to their electricity consumption information, and weight coefficients can be optimized according to electricity consumption behavior data to improve the reliability and accuracy of the load decomposition results in the substation.

[0039] Combine the following Figure 1 A classification process of a user group in a station area in an embodiment of the present application is described. Figure 1 It is a schematic diagram of the classification process of a substation user group provided in an embodiment of the present application, wherein the user classification groups include residential life, charging service, public service, and store operation.

[0040] Among them, the electricity consumption file information of all users in the target area can be obtained, and the file information includes the electricity consumption category and electricity consumption address information of the metering point, that is, the users can be classified according to the electricity consumption category and electricity consumption address information of the metering point. First, all users are preliminarily classified according to the electricity consumption category field of the metering point, and the electricity consumption category field of the metering point is used to indicate the electricity consumption category of the user. Among them, users in the residential life category can be divided into residential life groups, users in the commercial category can be divided into shop operation groups, and users in other categories are divided into public service groups. Then, the residential life group is further classified according to the user's electricity address information. Among them, for all users in the residential life group, users whose electricity consumption address information contains public service keywords can be classified into public service groups, and users whose electricity consumption address information contains charging service keywords can be classified into charging service groups, and users in other situations are not processed. Finally, all users in the target area can be divided into four classification groups: residential life, charging service, public service, and shop operation, which is convenient for subsequent data collection of representative users in each classification group, thereby reducing the cost and complexity of data collection.

[0041] It should be noted that public service keywords include but are not limited to lighting, elevators, fire protection, pumps, power distribution, passenger elevators, and security, and charging service keywords include but are not limited to charging piles and charging stations.

[0042] The characteristics of the four classification groups are shown in the following table:

[0043]

[0044] In a possible embodiment, the user's metering point electricity usage category is residential life, but the electricity usage address information is "xx Province xx City xx District xx Road xx Community xx Building Elevator". Because the electricity usage address information contains the public service keyword "elevator", the user can be classified into the public service category group.

[0045] It can be seen that by grouping users, it is convenient to collect data from representative users of each classification group, which greatly reduces the cost and complexity of data collection and reflects the power load characteristics of users in the target substation more comprehensively and accurately.

[0046] Combine the following Figure 2 The electronic device in the embodiment of the present application is described. Figure 2 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, the electronic device includes one or more processors, a memory, a communication interface and one or more programs, and the processor is communicatively connected with the memory and the communication interface via an internal communication bus.

[0047] Among them, the processor is mainly used for:

[0048] Obtaining electricity usage file information of all users in the target substation to obtain multiple electricity usage file information; the multiple electricity usage file information is used to characterize the substation load of the target substation;

[0049] Classify all users in the target area according to multiple power consumption archive information to obtain multiple classification groups; the classification groups include any of the following: residential life, charging service, public service, and store operation;

[0050] Performing a preset first calculation operation on each of the multiple classification groups to obtain multiple contribution weight coefficients; each classification group corresponds to a contribution weight coefficient;

[0051] Determine the optimization interval corresponding to each contribution weight coefficient in the plurality of contribution weight coefficients according to a preset numerical floating ratio, and obtain a plurality of optimization intervals;

[0052] A weight coefficient vector is randomly generated according to each optimization interval in the multiple optimization intervals to obtain multiple initial weight coefficient vectors; the initial weight coefficient vectors correspond to the optimization intervals one by one;

[0053] A preset second calculation operation is performed on each of the multiple initial weight coefficient vectors to obtain multiple optimal weight coefficient vectors.

[0054] The one or more programs are stored in the above-mentioned memory and are configured to be executed by the above-mentioned processor, and the one or more programs include instructions for executing any step in the above-mentioned method embodiment.

[0055] Among them, the processor can be, for example, a central processing unit (CPU), a general processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It can implement or execute various exemplary logic blocks, units and circuits described in conjunction with the disclosure of this application. The processor can also be a combination that implements a computing function, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, and the like. The communication unit can be a communication interface, a transceiver, a transceiver circuit, etc., and the storage unit can be a memory.

[0056] The memory may be a volatile memory or a nonvolatile memory, or may include both volatile and nonvolatile memories. Among them, the nonvolatile memory may be a read-only memory (ROM), a programmable ROM (PROM), an erasable programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM bus random access memory (DR RAM).

[0057] It is understood that the electronic device may include more or fewer structural elements than those in the above structural block diagram, for example, including a power module, a physical button, a Wi-Fi module, a speaker, a Bluetooth module, a sensor, a display module, etc., which are not limited here. It is understood that the electronic device can be used to perform the following operations: Figure 1 The classification process described.

[0058] After understanding the software and hardware architecture of this application, Figure 3 A method for decomposing load in a substation based on user grouping in an embodiment of the present application is described. Figure 3 A flow chart of a method for decomposing load in a substation based on user grouping provided in an embodiment of the present application specifically includes the following steps:

[0059] Step S301, obtaining the electricity usage file information of all users in the target area, and obtaining a plurality of electricity usage file information.

[0060] The plurality of electricity usage archive information are used to characterize the substation load of the target substation.

[0061] Specifically, the electricity usage file information of all users in the target substation can be obtained from the power company or the power grid management system to obtain multiple electricity usage file information, which includes the electricity usage category and electricity usage address information of the metering point. Among them, the electricity usage category of the metering point includes residential category, commercial category, and other categories, and the electricity usage address information includes the specific electricity usage address of the user, which can be used to identify the location and electricity usage characteristics of each user. The multiple electricity usage file information is not only used to characterize the substation load situation of the target substation, but also provides the necessary data support and decision-making basis for the power grid operation and management of the target substation.

[0062] Step S302: classify all users in the target area according to the plurality of electricity usage file information to obtain a plurality of classification groups.

[0063] Among them, the classification groups include any of the following: residents' life, charging services, public services, and shop operations.

[0064] Specifically, users whose electricity usage category at the metering point is residential life can be classified as residential life, users in the commercial category can be classified as shop operations, and users in other categories can be classified as public services. Users in the residential life classification group can be further classified, and users whose electricity usage address information contains public service keywords can be classified as public services, and users whose electricity usage address information contains charging service keywords can be classified as charging service groups. Users in other situations will not be processed. In this way, all users in the target area can be divided into multiple classification groups, and the classification groups include any of the following: residential life, charging services, public services, and shop operations.

[0065] Step S303 : performing a preset first calculation operation on each of the plurality of classification groups to obtain a plurality of contribution weight coefficients.

[0066] Specifically, each classification group corresponds to a contribution weight coefficient, and a preset first calculation operation is performed on each classification group in the multiple classification groups to obtain multiple contribution weight coefficients. The specific steps include: selecting p typical users in the first classification group, and collecting load data of the p typical users according to a preset collection strategy to obtain a load data set; p is a positive integer; the first classification group is any one of the multiple classification groups; the load data set is calculated according to a preset clustering algorithm to obtain a first typical load curve; the horizontal axis of the first typical load curve is time, and the vertical axis is load data; the contribution weight coefficient of the first classification group is determined according to the first typical load curve.

[0067] Among them, p typical users are selected in the first classification group, and the load data of the p typical users are collected according to a preset collection strategy to obtain a load data set. The specific steps include: obtaining the daily frozen power of each user in the first classification group to obtain multiple first-day frozen power; sorting the users in the first classification group from large to small according to the size of the daily frozen power according to the multiple first-day frozen power to obtain a first user sorting list; accumulating and summing the daily frozen power of the first p-1 users and the first p users in the first user sorting list respectively to obtain the second-day frozen power and the third-day frozen power; determining half of the total daily frozen power of the first classification group as the reference power threshold; if the second-day frozen power is less than the reference power threshold, and the third-day frozen power is greater than or equal to the reference power threshold, then selecting the p users as the p typical users; collecting the p typical users according to the preset collection frequency and collection duration to obtain the load data set.

[0068] Specifically, first collect the daily frozen power of each user in the first classification group to obtain multiple first-day frozen power, that is, the power consumption of each user in one day. Then sort the users from large to small according to the daily frozen amount to obtain the first user sorting list. The daily frozen power of the first p-1 users in the first user sorting list is cumulatively summed to obtain the second-day frozen power, and then the daily frozen power of the first p users is cumulatively summed to obtain the third-day frozen power. In order to select a representative user in the first classification group, the total daily frozen power of the first classification group is obtained, and 50% of the total daily frozen power is used as the reference power threshold. Among them, if the second-day frozen power is less than the reference power threshold, and the third-day frozen power is greater than or equal to the reference power threshold, it can be determined that the p users are p typical users of the first classification group. Finally, collect p typical users according to the preset collection frequency and collection duration to obtain a load data set. Among them, the collection frequency can be collected every 15 minutes, and the collection duration can be one day, that is, 24 hours, which is not specifically limited here. It should be noted that there are 24 hours in a day and 4 15-minute time periods in each hour, so each user has 96 data points in a day.

[0069] The load data set includes p load data subsets; the load data set is calculated according to a preset clustering algorithm to obtain a first typical load curve, and the specific steps include:

[0070] A1. Determine p load curves according to the p load data sets;

[0071] The load data set includes p load data sets, each of which corresponds to a typical user. Based on the p load data sets, p load curves corresponding to the p typical users are generated, and each load curve represents the electricity consumption of a typical user. It should be noted that the load curve refers to the curve formed by the load data corresponding to 96 time points of a typical user in a day. The horizontal axis of the curve is the time point of the day, and the vertical axis is the load size corresponding to each time point.

[0072] A2. Randomly select k load curves from the p load curves, and use the k load curves as k first center points of k clusters respectively; k is a positive integer;

[0073] Among them, k load curves are randomly selected from the p load curves, and then k clusters are created, and each load curve is used as the center point corresponding to each cluster to obtain k first center points. It should be noted that the silhouette coefficient can be used as an evaluation indicator of the k value, and is calculated according to the preset silhouette coefficient calculation formula to determine the size of the optimal k value. The silhouette coefficient calculation formula is as follows:

[0074]

[0075] Among them, s(i) is the silhouette coefficient corresponding to load curve i, a(i) is the average distance between load curve i and other load curves in its cluster, and b(i) is the minimum value of the average distance between load curve i and all load curves in other clusters.

[0076] Specifically, the value range of the silhouette coefficient s(i) is between -1 and 1. If s(i) is close to 1, it means that the load curve i is correctly clustered, and the distance with other load curves in the same cluster is much smaller than the distance with load curves in other clusters; if s(i) is close to -1, it means that the load curve i may be incorrectly clustered to an adjacent cluster; if s(i) is close to 0, it means that the load curve i is at the boundary of two clusters. Then, the average silhouette coefficient of all load curves is taken as the average, and the average silhouette coefficient is used to represent the overall clustering quality evaluation index, where the closer the average silhouette coefficient is to 1, the better the clustering result. Therefore, the average silhouette coefficient corresponding to different k values ​​can be calculated, and the k value that makes the average silhouette coefficient closest to 1 is selected as the optimal k value.

[0077] A3. Calculate the distance between each of the p load curves and the k first center points, and assign it to the cluster corresponding to the first center point closest to it;

[0078] Among them, the distance between each of the p load curves and each first center point can be calculated, and it can be assigned to the cluster corresponding to the first center point closest to it, that is, each load curve is assigned to the nearest cluster. It should be noted that each load curve corresponds to the load data collected at 96 time points, which can be regarded as a 96-dimensional vector, and each center point can also be regarded as a 96-dimensional vector. Therefore, the distance between the load curve and the first center point can be calculated using the Euclidean distance formula.

[0079] A4. After all the p load curves are distributed, take the average of all the load curves of each of the k clusters to obtain k second center points; the k second center points correspond to the k clusters one by one;

[0080] Among them, after the p load curves are allocated to the nearest cluster, the average of all load curves in each cluster is taken to obtain the second center point corresponding to each cluster, that is, k second center points corresponding to k clusters.

[0081] A5, repeating the steps A3 and A4 until the center point of each of the k clusters no longer changes, and then going to step A6;

[0082] Wherein, step A3 and step A4 are repeatedly performed to facilitate determining the best center point corresponding to each cluster. When the center point of each of the k clusters no longer changes, it is determined that the clustering results of each cluster converge to a stable state.

[0083] A6. Determine the center point of the cluster with the largest number of load curves among the k clusters as the first typical load curve.

[0084] Among them, the cluster with the largest number of load curves among the k clusters is the most representative, and its center point is taken as the first typical load curve corresponding to the first classification group.

[0085] It should be noted that the typical load curve of each classification group can be aggregated by using the k-means clustering algorithm, or by using clustering algorithms such as hierarchical clustering and DBSCAN, which are not specifically limited here.

[0086] The step of determining the contribution weight coefficient of the first classification group according to the first typical load curve specifically includes:

[0087] The first typical load curve is calculated according to a preset first calculation formula to obtain the contribution weight coefficient; the first typical load curve is [α1, α2, α3, …, α n ];

[0088] The first calculation formula is as follows:

[0089]

[0090] Among them, d α is the total daily frozen electricity of the first classification group; α is the contribution weight coefficient of the first classification group; n It is the load data collected at the nth time point in the first typical load curve; n is an integer greater than 1, and c is a positive integer.

[0091] In a possible embodiment, n can be 96, then c is 4. α1+α2+α3+…+α n It represents the sum of the load data collected every 15 minutes, that is, the load data is collected 4 times per hour. When calculating the total daily frozen power of the first classification group, the load data of the first classification group within 24 hours can be collected and summed to obtain the total daily frozen power d α . The collection process requires 96 data collections, and the power unit is kilowatt-hour, while the load unit is kilowatt. Therefore, the sum of the load data collected 96 times is divided by 4 to get the power data corresponding to 24 hours. It can be seen that the total daily frozen power d of the first classification group is α The real data is obtained through actual collection, and the only unknown quantity in the first calculation formula is ω α Therefore, the contribution weight coefficient ω of the first classification group can be calculated by the first calculation formula: α .

[0092] Step S304, determining the optimization interval corresponding to each contribution weight coefficient in the plurality of contribution weight coefficients according to a preset numerical floating ratio, and obtaining a plurality of optimization intervals.

[0093] The numerical floating ratio can be 10%, which is not specifically limited here. Therefore, each contribution weight coefficient can be floated by 10% before and after, and used as the optimization interval corresponding to the contribution weight coefficient, to obtain multiple optimization intervals. For example, the contribution weight coefficient is 0.8, and after floating by 10% before and after, its corresponding optimization interval can be obtained as [0.72, 0.88].

[0094] Step S305 , randomly generating a weight coefficient vector according to each of the multiple optimization intervals to obtain multiple initial weight coefficient vectors.

[0095] The initial weight coefficient vector corresponds to the optimization interval one by one. A weight coefficient vector is randomly generated for each optimization interval to obtain multiple initial weight coefficient vectors, and the initial weight coefficient vectors are all 96-dimensional vectors.

[0096] Step S306: performing a preset second calculation operation on each of the multiple initial weight coefficient vectors to obtain multiple optimal weight coefficient vectors.

[0097] The step of performing a preset second calculation operation on each of the multiple initial weight coefficient vectors to obtain multiple optimal weight coefficient vectors specifically includes:

[0098] B1. Calculate the first weight coefficient vector w according to a preset second calculation formula to obtain a reference root mean square error value; the first weight coefficient vector w is any initial weight coefficient vector among the multiple initial weight coefficient vectors; the first weight coefficient vector w is [ω1, ω2, ω3, …, ω n ];

[0099] The second calculation formula is as follows:

[0100]

[0101] Wherein, L(w) is the reference root mean square error value; h(x i ) is the reference load curve obtained by weighted summation of the weight coefficient vector and the typical load curve; x i is the typical load curve of the multiple classification groups on the i-th day; i is the load curve of the area on the i-th day of actual collection; m is the total number of sample days; i is a positive integer;

[0102] It should be noted that the reference area load curve h(x i ), where the weighted calculation formula is as follows:

[0103] A*X=h(x i )

[0104] Where A=[w1,w2,w3,…,w N ] is used to represent the weight coefficient vector corresponding to the typical load curve of M classification groups, and w M is the weight coefficient vector corresponding to the typical load curve of the Mth classification group; X = [X1, X2, X3, …, X M ] T A matrix representing the typical load curves of M classification groups, and X N is the typical load curve of the Mth classification group; h(x i )=[β1,β2,β3,…,β 96 ] T It is used to represent the reference area load curve recorded every 15 minutes, then β 96is the load data of the substation corresponding to the 96th time point, and the interval between any two time points is 15 minutes. Where M is a positive integer.

[0105] Among them, m is the total number of sample days, that is, the load data of one or more days can be used as samples, and there is no specific limitation here. i is the typical load curve of multiple classification groups on day i, that is, x i is the typical load curve of M classification groups on day i, and y i It is the real data of the load curve of the area collected on the i-th day.

[0106] B2. Calculate a first gradient of the first weight coefficient vector w, and iteratively update the first weight coefficient vector w according to the first gradient;

[0107] B3. Repeat step B2 until L(w) does not decrease after five consecutive iterative updates, then stop updating the first weight coefficient vector w and obtain the optimal weight coefficient vector.

[0108] Among them, step B2 is continuously repeated, and each iteration will update to obtain a smaller L(w) until the convergence condition of the iterative update is met, the update of the first weight coefficient vector w is stopped, and the latest first weight coefficient vector w is determined to be the optimal weight coefficient vector. It should be noted that the convergence condition is that the weight coefficient vector after 5 consecutive iterative updates can no longer obtain a smaller L(w), that is, L(w) after 5 consecutive iterative updates no longer decreases, and it is determined that the algorithm has completed convergence. Finally, the preset second calculation operation is performed on each initial weight coefficient vector, and multiple optimal weight coefficient vectors can be obtained, and the multiple optimal weight coefficient vectors are the load decomposition results of the substation.

[0109] The step of calculating the first gradient of the first weight coefficient vector w and iteratively updating the first weight coefficient vector w according to the first gradient specifically comprises:

[0110] The first gradient of the first weight coefficient vector w is calculated according to a preset gradient calculation formula; the gradient calculation formula is as follows:

[0111]

[0112] Wherein, ΔL(w) is a first gradient obtained by derivation of the reference root mean square error value with respect to the first weight coefficient vector w; is the gradient value corresponding to the j-th weight coefficient value in the first weight coefficient vector w; ij The j-th load data in the typical load curve of the multiple classification groups on the i-th day;

[0113] The first weight coefficient vector w is iteratively updated according to a preset weight coefficient update formula; the weight coefficient update formula is as follows:

[0114]

[0115] Among them, ω j is the jth weight coefficient value in the first weight coefficient vector w; ε is the step size of iterative update.

[0116] It should be noted that ε is the step size of the iterative update, that is, the learning rate ε can be 0.001. The learning rate ε refers to the parameter that controls the step size during the iterative update process, which can be used to adjust the size of each update to ensure that the parameter update is not too large during the update process and the optimal solution is not missed. In addition, iterative convergence can be performed through the gradient descent optimization algorithm, or optimization algorithms such as the least squares method, Newton method, or maximum likelihood estimation can be used, which are not specifically limited here.

[0117] It can be seen that through the above method, users can be grouped according to the electricity consumption information of users in the substation area, and representative typical users can be selected in each classification group to collect the electricity consumption behavior data of typical users, that is, load curve data, which effectively reduces the cost of data collection. The load curve data is analyzed by the clustering algorithm to obtain the typical load curve of each classification group, and the typical load curve can be used to more comprehensively and accurately describe the load characteristics of the users in the substation area, which helps to improve the credibility and accuracy of load decomposition. Randomly generating weight coefficient vectors according to the optimization interval can find the optimal weight coefficient faster and more accurately, and reduce the cost of iterative convergence to a certain extent. Iteratively adjusting the weight coefficient vector according to the gradient descent method helps to minimize the error between the predicted load curve and the actual load curve, and improve the accuracy and reliability of load decomposition.

[0118] The above mainly introduces the scheme of the embodiment of the present application from the perspective of the execution process on the method side. It is understandable that, in order to realize the above functions, the electronic device includes a hardware structure and / or software module corresponding to the execution of each function. It should be easily appreciated by those skilled in the art that, in combination with the units and algorithm steps of each example described in the embodiments provided herein, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present application.

[0119] The embodiment of the present application can divide the electronic device into functional units according to the above method example. For example, each functional unit can be divided according to each function, or two or more functions can be integrated into one processing unit. The above integrated unit can be implemented in the form of hardware or in the form of software functional units. It should be noted that the division of units in the embodiment of the present application is schematic and is only a logical function division. There may be other division methods in actual implementation.

[0120] In the case of dividing each functional module into corresponding functional modules, Figure 4 The functional module composition block diagram of a load decomposition device for a substation based on user grouping provided in an embodiment of the present application, the load decomposition device for a substation based on user grouping 400 comprises:

[0121] The acquisition module 410 is used to acquire the power usage file information of all users in the target area, and obtain a plurality of power usage file information; the plurality of power usage file information is used to characterize the area load of the target area;

[0122] The classification module 420 is used to classify all users in the target area according to the plurality of electricity usage archive information to obtain a plurality of classification groups; the classification groups include any one of the following: residential life, charging service, public service, and store operation;

[0123] A first calculation module 430 is used to perform a preset first calculation operation on each of the plurality of classification groups to obtain a plurality of contribution weight coefficients; each classification group corresponds to a contribution weight coefficient;

[0124] A determination module 440 is used to determine the optimization interval corresponding to each contribution weight coefficient in the plurality of contribution weight coefficients according to a preset numerical floating ratio, so as to obtain a plurality of optimization intervals;

[0125] A generating module 450 is used to randomly generate a weight coefficient vector according to each of the plurality of optimization intervals to obtain a plurality of initial weight coefficient vectors; the initial weight coefficient vectors correspond to the optimization intervals one by one;

[0126] The second calculation module 460 is used to perform a preset second calculation operation on each of the multiple initial weight coefficient vectors to obtain multiple optimal weight coefficient vectors.

[0127] Optionally, in the aspect of performing a preset first calculation operation on each of the plurality of classification groups to obtain a plurality of contribution weight coefficients, the first calculation module 430 is specifically configured to:

[0128] Selecting p typical users from the first classification group, and collecting the load data of the p typical users according to a preset collection strategy to obtain a load data set; p is a positive integer; the first classification group is any one of the multiple classification groups;

[0129] The load data set is calculated according to a preset clustering algorithm to obtain a first typical load curve; the horizontal axis of the first typical load curve is time, and the vertical axis is load data;

[0130] A contribution weight coefficient of the first classification group is determined according to the first typical load curve.

[0131] Optionally, in terms of selecting p typical users from the first classification group and collecting load data of the p typical users according to a preset collection strategy to obtain a load data set, the first calculation module 430 is further specifically used to:

[0132] Obtaining the daily frozen power of each user in the first classification group to obtain multiple first-day frozen power;

[0133] According to the plurality of first-day frozen power quantities, the users in the first classification group are sorted from large to small according to the size of the daily frozen power quantities to obtain a first user sorting list;

[0134] The daily frozen power of the first p-1 users and the first p users in the first user ranking list are respectively accumulated and summed to obtain the second day's frozen power and the third day's frozen power;

[0135] Determine half of the total daily frozen power of the first classification group as a reference power threshold;

[0136] If the second-day frozen electricity is less than the reference electricity threshold, and the third-day frozen electricity is greater than or equal to the reference electricity threshold, selecting the p users as the p typical users;

[0137] The load data set is obtained by collecting data from the p typical users according to a preset collection frequency and collection duration.

[0138] Optionally, in the case where the load data set includes p load data subsets; and in the case where the load data set is calculated according to a preset clustering algorithm to obtain a first typical load curve, the first calculation module 430 is further specifically used to:

[0139] A1. Determine p load curves according to the p load data sets;

[0140] A2. Randomly select k load curves from the p load curves, and use the k load curves as k first center points of k clusters respectively; k is a positive integer;

[0141] A3. Calculate the distance between each of the p load curves and the k first center points, and assign it to the cluster corresponding to the first center point closest to it;

[0142] A4. After all the p load curves are distributed, take the average of all the load curves of each of the k clusters to obtain k second center points; the k second center points correspond to the k clusters one by one;

[0143] A5, repeating the steps A3 and A4 until the center point of each of the k clusters no longer changes, and then going to step A6;

[0144] A6. Determine the center point of the cluster with the largest number of load curves among the k clusters as the first typical load curve.

[0145] Optionally, in determining the contribution weight coefficient of the first classification group according to the first typical load curve, the first calculation module 430 is further specifically configured to:

[0146] The first typical load curve is calculated according to a preset first calculation formula to obtain the contribution weight coefficient; the first typical load curve is [α1, α2, α3, …, α n ];

[0147] The first calculation formula is as follows:

[0148]

[0149] Among them, d α is the total daily frozen electricity of the first classification group; α is the contribution weight coefficient of the first classification group; n It is the load data collected at the nth time point in the first typical load curve; n is an integer greater than 1, and c is a positive integer.

[0150] Optionally, in the aspect of performing a preset second calculation operation on each of the multiple initial weight coefficient vectors to obtain multiple optimal weight coefficient vectors, the second calculation module 460 is specifically used to:

[0151] B1. Calculate the first weight coefficient vector w according to a preset second calculation formula to obtain a reference root mean square error value; the first weight coefficient vector w is any initial weight coefficient vector among the multiple initial weight coefficient vectors; the first weight coefficient vector w is [ω1, ω2, ω3, …, ω n ];

[0152] The second calculation formula is as follows:

[0153]

[0154] Wherein, L(w) is the reference root mean square error value; h(x i ) is the reference load curve obtained by weighted summation of the weight coefficient vector and the typical load curve; x i is the typical load curve of the multiple classification groups on the i-th day; i is the load curve of the area on the i-th day of actual collection; m is the total number of sample days; i is a positive integer;

[0155] B2. Calculate a first gradient of the first weight coefficient vector w, and iteratively update the first weight coefficient vector w according to the first gradient;

[0156] B3. Repeat step B2 until L(w) does not decrease after five consecutive iterative updates, then stop updating the first weight coefficient vector w and obtain the optimal weight coefficient vector.

[0157] Optionally, in the aspect of calculating the first gradient of the first weight coefficient vector w and iteratively updating the first weight coefficient vector w according to the first gradient, the second calculation module 460 is further specifically configured to:

[0158] The first gradient of the first weight coefficient vector w is calculated according to a preset gradient calculation formula; the gradient calculation formula is as follows:

[0159]

[0160] Wherein, ΔL(w) is a first gradient obtained by derivation of the reference root mean square error value with respect to the first weight coefficient vector w; is the gradient value corresponding to the j-th weight coefficient value in the first weight coefficient vector w; ij The j-th load data in the typical load curve of the multiple classification groups on the i-th day;

[0161] The first weight coefficient vector w is iteratively updated according to a preset weight coefficient update formula; the weight coefficient update formula is as follows:

[0162]

[0163] Among them, ω j is the jth weight coefficient value in the first weight coefficient vector w; ε is the step size of iterative update.

[0164] It can be seen that grouping users according to their electricity consumption information and optimizing weight coefficients based on electricity consumption behavior data can improve the reliability and accuracy of the load decomposition results in the substation.

[0165] It should be noted that the specific implementation of each operation can adopt the corresponding description of the method embodiment shown above, and the station load decomposition device 400 based on user grouping can be used to execute the above method embodiment of the present application, which will not be described in detail.

[0166] An embodiment of the present application also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program for electronic data exchange, wherein the computer program enables a computer to execute part or all of the steps of any method recorded in the above method embodiments, and the above computer includes an electronic device.

[0167] The embodiment of the present application also provides a computer program product, the computer program product includes a non-transitory computer-readable storage medium storing a computer program, the computer program is operable to cause a computer to execute some or all of the steps of any method described in the method embodiment. The computer program product may be a software installation package, and the computer includes an electronic device.

[0168] It should be noted that, for the above-mentioned various embodiments, for the sake of simple description, they are all expressed as a series of action combinations. Those skilled in the art should be aware that the present application is not limited by the described order of actions, because some steps in the embodiments of the present application can be performed in other orders or simultaneously. In addition, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions, steps, modules or units involved are not necessarily required by the embodiments of the present application.

[0169] In the above embodiments, the embodiments of the present application have different focuses on the description of each embodiment. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0170] Those skilled in the art can understand that to implement all or part of the processes in the above-mentioned embodiments, the processes can be completed by computer programs to instruct related hardware, and the programs can be stored in computer-readable storage media. When the programs are executed, they can include the processes of the above-mentioned method embodiments. The aforementioned storage media include: ROM or random access memory RAM, magnetic disk or optical disk and other media that can store program codes.

[0171] The steps of the method or algorithm described in the embodiments of the present application can be implemented in hardware or by executing software instructions by a processor. The software instructions can be composed of corresponding software modules, and the software modules can be stored in RAM, flash memory, ROM, EPROM, electrically erasable programmable read-only memory (electrically EPROM, EEPROM), registers, hard disks, mobile hard disks, read-only compact disks (CD-ROMs) or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an ASIC. In addition, the ASIC can be located in a terminal device or a management device. Of course, the processor and the storage medium can also be present in a terminal device or a management device as discrete components.

[0172] Those skilled in the art should be aware that in one or more of the above examples, the functions described in the embodiments of the present application can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiments of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server, or data center to another website site, computer, server, or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (digital subscriber line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrations. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a digital video disc (DVD)), or a semiconductor medium (eg, a solid state disk (SSD)).

[0173] The modules / units included in the devices and products described in the above embodiments may be software modules / units or hardware modules / units, or may be partially software modules / units and partially hardware modules / units. For example, for the devices and products applied to or integrated in the chip, the modules / units included therein may all be implemented in the form of hardware such as circuits, or at least some of the modules / units may be implemented in the form of software programs, which run on the processor integrated inside the chip, and the remaining (if any) modules / units may be implemented in the form of hardware such as circuits; for the devices and products applied to or integrated in the chip module, the modules / units included therein may all be implemented in the form of hardware such as circuits, and different modules / units may be located in the same component (such as a chip, circuit module, etc.) or in different components of the chip module, or at least some of the modules / units may be implemented in the form of software programs. The software programs run on the processor integrated inside the chip, and the remaining (if any) modules / units may be implemented in the form of hardware such as circuits. It is implemented in the form of a software program, which runs on a processor integrated inside the chip module, and the remaining (if any) modules / units can be implemented in hardware such as circuits; for various devices and products applied to or integrated in the terminal equipment, the various modules / units contained therein can be implemented in hardware such as circuits, and different modules / units can be located in the same component (for example, chip, circuit module, etc.) or in different components in the terminal equipment, or, at least some modules / units can be implemented in the form of a software program, which runs on a processor integrated inside the terminal equipment, and the remaining (if any) modules / units can be implemented in hardware such as circuits.

[0174] The specific implementation methods described above further illustrate the purpose, technical solutions and beneficial effects of the embodiments of the present application. It should be understood that the above description is only the specific implementation method of the embodiments of the present application and is not intended to limit the protection scope of the embodiments of the present application. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solutions of the embodiments of the present application should be included in the protection scope of the embodiments of the present application.

Claims

1. A method for decomposing load in a substation area based on user grouping, characterized in that: include: Obtaining the electricity usage file information of all users in the target area, and obtaining multiple electricity usage file information; The plurality of electricity usage archive information is used to characterize the load of the target substation; Classify all users in the target area according to the plurality of electricity usage archive information to obtain a plurality of classification groups; the classification groups include any one of the following: residential life, charging service, public service, and store operation; Performing a preset first calculation operation on each of the plurality of classification groups to obtain a plurality of contribution weight coefficients; each classification group corresponds to a contribution weight coefficient; Determine the optimization interval corresponding to each contribution weight coefficient in the plurality of contribution weight coefficients according to a preset numerical floating ratio, and obtain a plurality of optimization intervals; Randomly generate a weight coefficient vector according to each of the multiple optimization intervals to obtain multiple initial weight coefficient vectors; the initial weight coefficient vectors correspond to the optimization intervals one by one; A preset second calculation operation is performed on each of the multiple initial weight coefficient vectors to obtain multiple optimal weight coefficient vectors.

2. The method according to claim 1, characterized in that The performing of a preset first calculation operation on each of the plurality of classification groups to obtain a plurality of contribution weight coefficients includes: Selecting p typical users from the first classification group, and collecting the load data of the p typical users according to a preset collection strategy to obtain a load data set; p is a positive integer; the first classification group is any one of the multiple classification groups; The load data set is calculated according to a preset clustering algorithm to obtain a first typical load curve; the horizontal axis of the first typical load curve is time, and the vertical axis is load data; A contribution weight coefficient of the first classification group is determined according to the first typical load curve.

3. The method according to claim 2, characterized in that The step of selecting p typical users from the first classification group and collecting load data of the p typical users according to a preset collection strategy to obtain a load data set includes: Obtaining the daily frozen power of each user in the first classification group to obtain multiple first-day frozen power; According to the plurality of first-day frozen power quantities, the users in the first classification group are sorted from large to small according to the size of the daily frozen power quantities to obtain a first user sorting list; The daily frozen power of the first p-1 users and the first p users in the first user ranking list are respectively accumulated and summed to obtain the second day's frozen power and the third day's frozen power; Determine half of the total daily frozen power of the first classification group as a reference power threshold; If the second-day frozen electricity is less than the reference electricity threshold, and the third-day frozen electricity is greater than or equal to the reference electricity threshold, selecting the p users as the p typical users; The load data set is obtained by collecting data from the p typical users according to a preset collection frequency and collection duration.

4. The method according to claim 2, characterized in that The load data set includes p load data subsets; the load data set is calculated according to a preset clustering algorithm to obtain a first typical load curve, including: A1. Determine p load curves according to the p load data sets; A2. Randomly select k load curves from the p load curves, and use the k load curves as k first center points of k clusters respectively; k is a positive integer; A3. Calculate the distance between each of the p load curves and the k first center points, and assign it to the cluster corresponding to the first center point closest to it; A4. After all the p load curves are distributed, take the average of all the load curves of each of the k clusters to obtain k second center points; the k second center points correspond to the k clusters one by one; A5, repeating the steps A3 and A4 until the center point of each of the k clusters no longer changes, and then going to step A6; A6. Determine the center point of the cluster with the largest number of load curves among the k clusters as the first typical load curve.

5. The method according to claim 2, characterized in that The determining the contribution weight coefficient of the first classification group according to the first typical load curve includes: The first typical load curve is calculated according to a preset first calculation formula to obtain the contribution weight coefficient; the first typical load curve is [α1, α2, α3, …, α n ]; The first calculation formula is as follows: Among them, d α is the total daily frozen electricity of the first classification group; α is the contribution weight coefficient of the first classification group; n It is the load data collected at the nth time point in the first typical load curve; n is an integer greater than 1, and c is a positive integer.

6. The method according to claim 5, characterized in that The step of performing a preset second calculation operation on each of the plurality of initial weight coefficient vectors to obtain a plurality of optimal weight coefficient vectors includes: B1. Calculate the first weight coefficient vector w according to a preset second calculation formula to obtain a reference root mean square error value; the first weight coefficient vector w is any initial weight coefficient vector among the multiple initial weight coefficient vectors; the first weight coefficient vector w is [ω1, ω2, ω3, …, ω n ]; The second calculation formula is as follows: Wherein, L(w) is the reference root mean square error value; h(x i ) is the reference load curve obtained by weighted summation of the weight coefficient vector and the typical load curve; x i is the typical load curve of the multiple classification groups on the i-th day; i is the load curve of the area on the i-th day of actual collection; m is the total number of sample days; i is a positive integer; B2. Calculate a first gradient of the first weight coefficient vector w, and iteratively update the first weight coefficient vector w according to the first gradient; B3. Repeat step B2 until L(w) does not decrease after five consecutive iterative updates, then stop updating the first weight coefficient vector w and obtain the optimal weight coefficient vector.

7. The method according to claim 6, characterized in that The calculating a first gradient of the first weight coefficient vector w, and iteratively updating the first weight coefficient vector w according to the first gradient, comprises: The first gradient of the first weight coefficient vector w is calculated according to a preset gradient calculation formula; the gradient calculation formula is as follows: Wherein, ΔL(w) is a first gradient obtained by derivation of the reference root mean square error value with respect to the first weight coefficient vector w; is the gradient value corresponding to the j-th weight coefficient value in the first weight coefficient vector w; ij The j-th load data in the typical load curve of the multiple classification groups on the i-th day; The first weight coefficient vector w is iteratively updated according to a preset weight coefficient update formula; the weight coefficient update formula is as follows: Among them, ω j is the jth weight coefficient value in the first weight coefficient vector w; ε is the step size of iterative update.

8. A load decomposition device for a substation based on user grouping, characterized in that: It includes an acquisition module, a classification module, a first calculation module, a determination module, a generation module, and a second calculation module, wherein: The acquisition module is used to acquire the electricity usage file information of all users in the target area, and obtain multiple electricity usage file information; the multiple electricity usage file information is used to characterize the area load of the target area; The classification module is used to classify all users in the target area according to the plurality of electricity usage archive information to obtain a plurality of classification groups; the classification groups include any one of the following: residential life, charging service, public service, and store operation; The first calculation module is used to perform a preset first calculation operation on each of the multiple classification groups to obtain a plurality of contribution weight coefficients; each classification group corresponds to a contribution weight coefficient; The determination module is used to determine the optimization interval corresponding to each contribution weight coefficient in the multiple contribution weight coefficients according to a preset numerical floating ratio to obtain multiple optimization intervals; The generating module is used to randomly generate a weight coefficient vector according to each of the multiple optimization intervals to obtain multiple initial weight coefficient vectors; the initial weight coefficient vectors correspond to the optimization intervals one by one; The second calculation module is used to perform a preset second calculation operation on each of the multiple initial weight coefficient vectors to obtain multiple optimal weight coefficient vectors.

9. An electronic device, characterized in that: include: A processor, a memory, a communication interface, and one or more programs; The one or more programs are stored in the memory and configured to be executed by the processor, the programs comprising instructions for executing the steps in the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor is caused to perform the method according to any one of claims 1 to 7.

Images