Distribution transformer supplier evaluation method based on graph neural network and related equipment

Abstract

The invention relates to the field of power material supply management and intelligent purchasing, and discloses a distribution transformer supplier evaluation method based on a graph neural network and related equipment, and the method comprises the steps: building a distribution transformer supplier evaluation index system based on the comprehensive capability data of a distribution transformer supplier; building a distribution transformer supplier supply chain attribute quantitative evaluation index system based on the component material of the distribution transformer supplier cooperation; constructing a model based on a graph neural network, taking the distribution transformer supplier evaluation index system and the distribution transformer supplier supply chain attribute quantitative evaluation index system as input, and outputting a distribution transformer supplier evaluation result; and performing distribution transformer supplier evaluation based on the graph neural network according to the evaluation result to obtain an evaluation result. The method guarantees the accuracy of the evaluation result through the powerful learning capability of the graph neural network model, and can reflect the real level of the supplier.

Description

Technical Field

[0001] The present invention relates to the field of power material supply management and intelligent procurement, and specifically to a distribution transformer supplier evaluation method based on graph neural network and related equipment. Background Art

[0002] With the intensification of the global energy crisis, improving the energy efficiency of transformers has become an important development direction of the industry. In recent years, the rapid development of my country's economy and science and technology has promoted the development of the power system, and the loss in the huge power network has become a problem that cannot be ignored. As a vital power equipment in the power system, the loss and energy efficiency of the transformer are directly related to the efficiency of the entire power network, so high-quality green transformers need to be selected in the construction of the power grid. However, there are many transformer suppliers on the market, and the quality of the transformers they produce is also uneven, so it is very important to accurately evaluate these suppliers. Most of the existing studies only evaluate the quality of the transformer itself, without rating the comprehensive strength of the supplier itself, and do not increase the scientificity and objectivity of the evaluation. Combined with the guidance of actual operation and maintenance support analysis in recent years, in addition to considering various factors during production, it is also necessary to consider the comprehensive strength of the component material supplier cooperating with the supplier, and to reduce the subjectivity when evaluating the transformer. Therefore, it is very important to conduct an overall level survey on the distribution transformer industry, explore the deep impact of the relationship between suppliers on product quality, and analyze the quality level of distribution transformer products provided by each supplier, which is very important for improving the quality and efficiency of distribution transformer procurement, promoting the healthy development of the distribution supplier industry chain, and improving the objectivity of supplier evaluation. Summary of the invention

[0003] In order to overcome the defects of the above-mentioned prior art, the purpose of the present invention is to provide a distribution transformer supplier evaluation method and related equipment based on graph neural network, so as to solve the technical problem of how to establish a distribution transformer supplier evaluation system in the prior art.

[0004] The present invention is achieved through the following technical solutions: In a first aspect, the present invention provides a distribution transformer supplier evaluation method based on a graph neural network, comprising: Establish a distribution transformer supplier evaluation index system based on the comprehensive capability data of distribution transformer suppliers; Build a quantitative evaluation index system for the supply chain attributes of distribution transformer suppliers based on the component materials of distribution transformer suppliers; Construct a graph neural network model, take the distribution transformer supplier evaluation index system and the distribution transformer supplier supply chain attribute quantitative evaluation index system as input, and output the distribution transformer supplier evaluation result; According to the evaluation results, the distribution transformer supplier evaluation based on graph neural network is carried out to obtain the evaluation results.

[0005] Preferably, in the step of building a distribution transformer supplier evaluation index system based on the comprehensive data of distribution transformer suppliers, the construction process of the distribution transformer supplier evaluation index system is as follows: Obtain production information and equipment operation information of distribution transformers; Based on the production information and equipment operation information of distribution transformers, we can obtain the product performance data of distribution transformer suppliers, the equipment quantity data of data suppliers and the performance evaluation data of power grid enterprises; An evaluation index system for distribution transformer suppliers is established based on the product performance data of distribution transformer suppliers, the equipment quantity data of data suppliers and the performance evaluation data of power grid enterprises.

[0006] Preferably, in the step of building a quantitative evaluation index system for the supply chain attributes of distribution transformer suppliers based on the component materials of the distribution transformer suppliers, the construction process of the quantitative evaluation index system for the supply chain attributes of distribution transformer suppliers is as follows: Obtain comprehensive analysis of several supplier data, and extract distribution transformer supplier information from the comprehensive analysis of several supplier data; Based on the information of distribution transformer suppliers, the bidding times of distribution transformer suppliers and the third-party enterprise credit evaluation data are analyzed to determine the comprehensive capability data of component material suppliers; A quantitative evaluation index system for the supply chain attributes of distribution transformer suppliers is established based on the comprehensive capability data of component material suppliers.

[0007] Furthermore, the component materials include core-silicon steel sheets, electromagnetic wires, insulating oil, bushings, tap changers, fans, oil storage cabinets, coolers or radiators, temperature control equipment, and pressure relief valves.

[0008] Preferably, in the step of constructing a graph neural network model, a few-sample learning strategy is adopted to use a small portion of labeled data as a training set, and the remaining unlabeled data is used to verify and test the accuracy of the model.

[0009] Preferably, the distribution transformer supplier evaluation index system and the distribution transformer supplier supply chain attribute quantitative evaluation index system are used as inputs, and in the step of outputting the distribution transformer supplier evaluation results, the specific process is as follows: The graph neural network model is used to embed node and edge features into the latent feature space. The input node features are the comprehensive data of distribution transformer suppliers and the supplier's ability data to provide component materials. The edge features are the supply relationship between suppliers. The distribution transformer supplier evaluation results are output based on the feature information of suppliers cooperating with the input distribution transformer supplier.

[0010] In a second aspect, the present invention also provides a distribution transformer supplier evaluation system based on graph neural network, comprising: An evaluation index system building module is used to build an evaluation index system for distribution transformer suppliers based on the comprehensive capability data of distribution transformer suppliers; A module for building a quantitative attribute evaluation index system, which is used to build a quantitative attribute evaluation index system for the distribution transformer supplier supply chain based on the component materials that the distribution transformer supplier cooperates with; A graph neural network model processing module is used to construct a graph neural network model, take the distribution transformer supplier evaluation index system and the distribution transformer supplier supply chain attribute quantitative evaluation index system as input, and output the distribution transformer supplier evaluation result; The evaluation module evaluates the distribution transformer supplier based on the graph neural network and obtains the evaluation result according to the evaluation result.

[0011] In a third aspect, the present invention also provides a mobile terminal comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the distribution transformer supplier evaluation method based on graph neural network as described above when executing the computer program.

[0012] In a fourth aspect, the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the distribution transformer supplier evaluation method based on graph neural network as described above are implemented.

[0013] In a fifth aspect, the present invention also provides a computer program product, comprising computer instructions, wherein the computer instructions instruct a computing device to perform operations corresponding to the distribution transformer supplier evaluation method based on graph neural network as described above.

[0014] Compared with the prior art, the present invention has the following beneficial technical effects: The present invention provides a distribution transformer supplier evaluation method based on graph neural network. First, a comprehensive capability data evaluation index system of distribution transformer suppliers is established. At the same time, the component materials of supplier cooperation are also considered, and a supply chain attribute quantitative evaluation index system is established. By introducing the graph neural network model, complex relational data can be efficiently processed, such as the cooperation network between suppliers, the material flow in the supply chain, etc., which may be difficult to accurately capture in traditional evaluation methods. The graph neural network model can learn the deep-level features in the data, improve the accuracy and efficiency of the evaluation, and make the evaluation results closer to the actual situation. The use of a data-based quantitative evaluation method avoids the subjectivity caused by human factors, making the evaluation results more objective and fair. The accuracy of the evaluation results is guaranteed by the powerful learning ability of the graph neural network model, which can reflect the true level of the supplier.

[0015] Furthermore, the present invention is based on the CensNet graph neural network model, which effectively utilizes the joint embedding of nodes and edges in the model and can better mine the complex relationship characteristics between nodes and edges, which has certain guiding value for evaluating the level of distribution transformer suppliers and their position in the supply chain.

[0016] Furthermore, the present invention conducts supplier evaluation based on the CensNet graph neural network model, and combines the distribution transformer supplier level evaluation indicators and the distribution transformer supplier supply chain attribute quantitative evaluation indicators to achieve effective definition of the distribution transformer high-quality supplier domain; at the same time, it provides clear guidance and suggestions for optimizing key factors such as product quality and supply and marketing relationships of distribution transformer suppliers, and effectively assists in the intelligent rejection of the initial evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flow chart of a method for evaluating a distribution transformer supplier based on a graph neural network in an embodiment of the present invention; Figure 2 It is a neural network structure diagram in the embodiment of the present invention; Figure 3 A comparison chart of the system evaluation and actual evaluation of some transformer manufacturers in an embodiment of the present invention; Figure 4 It is a structural schematic diagram of a distribution transformer supplier evaluation system based on graph neural network in an embodiment of the present invention.

[0018] In the figure: 1. Evaluation index system building module; 2. Attribute quantification evaluation index system building module; 3. Graph neural network model processing module; 4. Evaluation module. DETAILED DESCRIPTION

[0019] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0020] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. 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 necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0021] The purpose of the present invention is to provide a distribution transformer supplier evaluation method and related equipment based on graph neural network to solve the technical problem of how to establish a distribution transformer supplier evaluation system in the prior art.

[0022] The present invention is further described in detail below in conjunction with the accompanying drawings: See also Figure 1 In one embodiment of the present invention, a method for evaluating a distribution transformer supplier based on a graph neural network is provided, comprising: Step 1: Building a distribution transformer supplier evaluation index system based on the comprehensive capability data of distribution transformer suppliers; Specifically, in the steps of building a distribution transformer supplier evaluation index system based on the comprehensive data of distribution transformer suppliers, the construction process of the distribution transformer supplier evaluation index system is as follows: Obtain production information and equipment operation information of distribution transformers; Based on the production information and equipment operation information of distribution transformers, we can obtain the product performance data of distribution transformer suppliers, the equipment quantity data of data suppliers and the performance evaluation data of power grid enterprises; An evaluation index system for distribution transformer suppliers is established based on the product performance data of distribution transformer suppliers, the equipment quantity data of data suppliers and the performance evaluation data of power grid enterprises.

[0023] Among them, the comprehensive strength of the supplier in this embodiment is first calculated according to the product performance of the distribution transformer supplier, according to the three categories of dry silicon steel sheet, oil-immersed silicon steel sheet, and amorphous alloy. Secondly, according to the number of all equipment owned by this supplier, the types of equipment include winding machines, drying furnaces, vacuum oil filters, foil winding machines, test transformers, DC resistance testers, insulation resistance testers, ratio testers, partial discharge testers, power analyzers, medium frequency generator sets or variable frequency power supplies, and voltage regulators, etc., as part of its comprehensive capabilities; power grid enterprise performance evaluation, the power grid company rates the performance of relevant power companies, and the ratings are divided into three categories: A, B, and C. The comprehensive strength of the supplier is obtained based on the performance judgment. The company's strength is judged based on the performance. Taking the evaluation of some suppliers as an example, different manufacturers are evaluated. The specific indicator evaluation is shown in Table 1.

[0024] Table 1 Evaluation index table of distribution transformers of some suppliers

[0025] Step 2: Build a quantitative evaluation index system for the supply chain attributes of distribution transformer suppliers based on the component materials that the distribution transformer suppliers cooperate with; Specifically, the process of building a quantitative evaluation index system for the supply chain attributes of distribution transformer suppliers is as follows: Obtain comprehensive analysis of several supplier data, and extract distribution transformer supplier information from the comprehensive analysis of several supplier data; Based on the information of distribution transformer suppliers, the bidding times of distribution transformer suppliers and the third-party enterprise credit evaluation data are analyzed to determine the comprehensive capability data of component material suppliers; A quantitative evaluation index system for the supply chain attributes of distribution transformer suppliers is established based on the comprehensive capability data of component material suppliers. Table 2 shows the quantitative evaluation indicators for the supply chain attributes of some distribution transformer suppliers.

[0026] Table 2 Quantitative evaluation indicators of supply chain attributes of some distribution transformer suppliers

[0027] Among them, component materials include iron core-silicon steel sheets, electromagnetic wires, insulating oil, bushings, tap changers, fans, oil storage cabinets, coolers or radiators, temperature control equipment and pressure relief valves.

[0028] Among them, the comprehensive strength of the component material suppliers cooperating with the distribution transformer supplier can be considered to indirectly reflect the level of this distribution transformer supplier in the industry. If the comprehensive level of the component material suppliers is high, it can reflect the level of this distribution transformer supplier to a certain extent.

[0029] Step 3: Build a graph neural network model, such as Figure 2As shown, the distribution transformer supplier evaluation index system and the distribution transformer supplier supply chain attribute quantitative evaluation index system are used as inputs, and the distribution transformer supplier evaluation results are output; Specifically, the CensNet model is trained using a few-sample learning strategy, that is, a small portion of labeled data is extracted as a training set, a small portion is used for manual comprehensive evaluation, and this portion of the results is input into the model for training. When using the remaining large amount of unlabeled data to test accuracy, the above-mentioned distribution transformer supplier level evaluation indicators are used as node features, and the supply relationship between suppliers is used as an edge feature. The CensNet model is used to embed the feature space for comprehensive evaluation, thereby achieving the purpose of optimizing model performance and verifying system accuracy.

[0030] Specifically, 30%, 40%, and 50% of the data were used for sample training, and then compared with the actual situation of the remaining test samples. It can be seen that the accuracy ranged from 0.71 to 0.79. The higher the proportion of training samples in the data set, the higher the model accuracy, as shown in Table 3.

[0031] Table 3 Data training accuracy

[0032] Step 4: Based on the evaluation results, a graph neural network-based distribution transformer supplier evaluation is performed to obtain the evaluation results.

[0033] The evaluation index of distribution transformers and the quantitative evaluation index of supplier attributes are used as input, and the CensNet graph neural network model is used to embed node and edge features into a latent feature space to capture and model the complex relationship between nodes and edges. Among them, in this embodiment, the input node features are the component materials and process level feature information of the supplier itself, and the edge features are the supply relationship between suppliers. Based on the supplier's own level and the feature information of suppliers cooperating with the input distribution transformer supplier, the supplier is comprehensively evaluated from the above two dimensions.

[0034] The CensNet graph neural network model uses approximate spectral graph convolution in hierarchical propagation and defines the normalized (Laplacian) node adjacency matrix with self-loops as follows: (1) in yes The diagonal matrix of is the identity matrix; In the CensNet graph neural network model of this embodiment, the propagation of the node layer is the core process of updating the node feature representation, which includes several key steps: first, neighbor aggregation is performed to collect the neighbor node feature information of each node, and then the aggregated neighbor features undergo one or more linear transformations. The linearly transformed features usually pass through a nonlinear activation function. Finally, these transformed features are combined with the original node features or the output of the previous layer to generate a new feature representation of each node, thereby completing the update of the node features; the layer-by-layer propagation of the node features in the (l+1)th layer is defined as: (2) The matrix is a binary transformation matrix, Represents edge Whether to connect node i. Given that each edge is formed by two nodes, each column of the matrix T has two elements that are 1 and all other elements are 0. P is a de dimensional vector, defined as the learnable weight of the edge feature vector. represents the diagonalization operation, which places the one-dimensional vectors on the diagonal of the square matrix. represents the Hadamard product or element-wise product. If there are no physical edges, the line graph will not have any effect, so we maintain the sparsity of the original graph, which brings significant computational advantages; Edge layer propagation is a key process for updating edge feature representation in the CensNet graph neural network model, and it works closely with the propagation rules of the node layer. This process first relies on the latest node features output by the node layer, which are then used to guide the update of edge features. The update operation usually involves combining the node features with the edge adjacency matrix, and adjusting the feature representation of the edge through specific graph convolution operations or attention mechanisms. In this process, the edge layer uses an attention mechanism to measure the importance of each node connected to the edge, and uses this measurement as the weight for updating the edge features. Finally, after being processed by the edge layer propagation rules, the edge features are updated. These updated feature representations can provide richer information for the graph neural network, and then be used to perform various graph-related downstream tasks such as link prediction, graph classification, etc.; similarly, the normalized (Laplacian) edge adjacency matrix is ​​defined as: (3) in yes The degree matrix of the edge feature propagation rule is defined as follows: (4) The T matrix is ​​the same as formula (2), but represents the learnable weights of nodes. In return, node features and adjacency matrix are used to improve edge embeddings. These two parts connect signals on nodes and edges and update the embeddings of nodes and edges alternately.

[0035] By updating the feature representation of nodes and edges, the CensNet graph neural network model is continuously optimized; by integrating the material process evaluation and centrality index analysis of transformer manufacturers, and adopting the advanced CensNet graph neural network model and few-sample learning strategy, an efficient and accurate transformer manufacturer evaluation system is successfully implemented.

[0036] Based on the previous comprehensive analysis of the distribution transformer evaluation indicators of various distribution transformer suppliers and the supplier attribute quantitative evaluation indicators, a detailed data table containing transformer manufacturer information is constructed. The table divides more than 600 distribution transformer manufacturers into four categories according to manufacturing equipment: dry silicon steel sheet transformer manufacturers, dry amorphous alloy transformer manufacturers, oil-immersed silicon steel sheet transformer manufacturers, and oil-immersed amorphous alloy transformer manufacturers. The overall strength analysis of distribution transformer suppliers is combined. Taking the 606 manufacturers of distribution transformers as an example, the comprehensive strength dimension covers: product performance, which reflects product quality through transformer product performance; the number of supplier equipment, which analyzes its comprehensive strength through the total number of equipment it owns; power grid enterprise performance evaluation, the power grid company rates the performance of related power companies and judges the company's strength based on performance; the comprehensive capability dimension of component material suppliers covers: the number of biddings of component material suppliers and third-party corporate credit evaluation (Sky Eye score), etc. The above information is input as node information, and then based on the embedded edge features, while considering the supplier's own level, the characteristic information of the suppliers cooperating with it is also taken into account. Finally, the above two dimensions are combined to comprehensively evaluate the manufacturer company, and the comprehensive evaluation gives three levels: A, B, and C. According to Table 4, the evaluation of several oil-immersed silicon steel sheet transformer manufacturers is listed.

[0037] Table 4 Evaluation of some distribution transformer manufacturers

[0038] In summary, the present invention provides a distribution transformer supplier evaluation method based on graph neural network. First, a comprehensive capability data evaluation index system for distribution transformer suppliers is established. At the same time, the component materials of supplier cooperation are also considered, and a supply chain attribute quantitative evaluation index system is established. By introducing the graph neural network model, it is able to efficiently process complex relational data, such as the cooperation network between suppliers, material flow in the supply chain, etc., which may be difficult to accurately capture in traditional evaluation methods. The graph neural network model can learn the deep-level features in the data, improve the accuracy and efficiency of the evaluation, and make the evaluation results closer to the actual situation. The use of a data-based quantitative evaluation method avoids the subjectivity caused by human factors, making the evaluation results more objective and fair. The powerful learning ability of the graph neural network model ensures the accuracy of the evaluation results, which can reflect the true level of the supplier. Figure 3 As shown, a comparison of the system evaluation and actual evaluation of some transformer manufacturers is shown.

[0039] Example 2 according to Figure 4 As shown, the present invention also provides a distribution transformer supplier evaluation system based on graph neural network, comprising: Evaluation index system building module 1 is used to build a distribution transformer supplier evaluation index system based on the comprehensive capability data of distribution transformer suppliers; Module 2 for building an attribute quantitative evaluation index system is used to build an attribute quantitative evaluation index system for the distribution transformer supplier supply chain based on the component materials of the distribution transformer supplier cooperation; A graph neural network model processing module 3 is used to construct a graph neural network model, take the distribution transformer supplier evaluation index system and the distribution transformer supplier supply chain attribute quantitative evaluation index system as input, and output the distribution transformer supplier evaluation result; Evaluation module 4 performs distribution transformer supplier evaluation based on graph neural network according to the evaluation results to obtain the evaluation results.

[0040] Example 3 The present invention also provides a mobile terminal, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, such as a distribution transformer supplier evaluation program based on a graph neural network.

[0041] When the processor executes the computer program, the steps of the above-mentioned method for evaluating distribution transformer suppliers based on graph neural network are implemented, for example: Establish a distribution transformer supplier evaluation index system based on the comprehensive capability data of distribution transformer suppliers; Build a quantitative evaluation index system for the supply chain attributes of distribution transformer suppliers based on the component materials of distribution transformer suppliers; Construct a graph neural network model, take the distribution transformer supplier evaluation index system and the distribution transformer supplier supply chain attribute quantitative evaluation index system as input, and output the distribution transformer supplier evaluation result; According to the evaluation results, the distribution transformer supplier evaluation based on graph neural network is carried out to obtain the evaluation results.

[0042] Alternatively, when the processor executes the computer program, the functions of each module in the above system are realized, for example: Evaluation index system building module 1 is used to build a distribution transformer supplier evaluation index system based on the comprehensive capability data of distribution transformer suppliers; Module 2 for building an attribute quantitative evaluation index system is used to build an attribute quantitative evaluation index system for the distribution transformer supplier supply chain based on the component materials of the distribution transformer supplier cooperation; A graph neural network model processing module 3 is used to construct a graph neural network model, take the distribution transformer supplier evaluation index system and the distribution transformer supplier supply chain attribute quantitative evaluation index system as input, and output the distribution transformer supplier evaluation result; Evaluation module 4 performs distribution transformer supplier evaluation based on graph neural network according to the evaluation results to obtain the evaluation results.

[0043] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, which are used to describe the execution process of the computer program in the mobile terminal.

[0044] For example, the computer program can be divided into an evaluation index system building module 1, an attribute quantification evaluation index system building module 2, a graph neural network model processing module 3 and an evaluation module 4; The specific functions of each module are as follows: Evaluation index system building module 1 is used to build a distribution transformer supplier evaluation index system based on the comprehensive capability data of distribution transformer suppliers; Module 2 for building an attribute quantitative evaluation index system is used to build an attribute quantitative evaluation index system for the distribution transformer supplier supply chain based on the component materials of the distribution transformer supplier cooperation; A graph neural network model processing module 3 is used to construct a graph neural network model, take the distribution transformer supplier evaluation index system and the distribution transformer supplier supply chain attribute quantitative evaluation index system as input, and output the distribution transformer supplier evaluation result; Evaluation module 4 performs distribution transformer supplier evaluation based on graph neural network according to the evaluation results to obtain the evaluation results.

[0045] The mobile terminal may be a computing device such as a desktop computer, a notebook, a palm computer, a cloud server, etc. The mobile terminal may include, but is not limited to, a processor and a memory.

[0046] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the mobile terminal, and uses various interfaces and lines to connect various parts of the entire mobile terminal.

[0047] The memory may be used to store the computer programs and / or modules, and the processor implements various functions of the mobile terminal by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory.

[0048] The memory may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (SmartMediaCard, SMC), a secure digital (SecureDigital, SD) card, a flash card (FlashCard), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0049] Example 4 The present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the steps of the distribution transformer supplier evaluation method based on graph neural network.

[0050] If the module / unit integrated in the mobile terminal is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium.

[0051] Based on this understanding, the present invention implements all or part of the processes in the above method, and can also be completed by instructing related hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and when the computer program is executed by a processor, the steps of the above-mentioned aggregate reinforcement learning resource scheduling method can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc.

[0052] The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0053] It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media does not include electrical carrier signals and telecommunication signals.

[0054] Example 5 A computer program product includes computer instructions, which instruct a computing device to perform operations corresponding to the distribution transformer supplier evaluation method based on graph neural network as described above.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A distribution transformer supplier evaluation method based on graph neural network, characterized in that: include: Establish a distribution transformer supplier evaluation index system based on the comprehensive capability data of distribution transformer suppliers; Build a quantitative evaluation index system for the supply chain attributes of distribution transformer suppliers based on the component materials of distribution transformer suppliers; Construct a graph neural network model, take the distribution transformer supplier evaluation index system and the distribution transformer supplier supply chain attribute quantitative evaluation index system as input, and output the distribution transformer supplier evaluation result; According to the evaluation results, the distribution transformer supplier evaluation based on graph neural network is carried out to obtain the evaluation results.

2. According to the method for evaluating distribution transformer suppliers based on graph neural network in claim 1, in the step of building a distribution transformer supplier evaluation index system based on the comprehensive data of distribution transformer suppliers, the construction process of the distribution transformer supplier evaluation index system is as follows: Obtain production information and equipment operation information of distribution transformers; Based on the production information and equipment operation information of distribution transformers, we can obtain the product performance data of distribution transformer suppliers, the equipment quantity data of data suppliers and the performance evaluation data of power grid enterprises; An evaluation index system for distribution transformer suppliers is established based on the product performance data of distribution transformer suppliers, the equipment quantity data of data suppliers and the performance evaluation data of power grid enterprises.

3. According to the method for evaluating distribution transformer suppliers based on graph neural network in claim 1, in the step of building a quantitative evaluation index system for the supply chain attributes of distribution transformer suppliers based on the component materials cooperated by the distribution transformer suppliers, the construction process of the quantitative evaluation index system for the supply chain attributes of distribution transformer suppliers is as follows: Obtain comprehensive analysis of several supplier data, and extract distribution transformer supplier information from the comprehensive analysis of several supplier data; Based on the information of distribution transformer suppliers, the bidding times of distribution transformer suppliers and the third-party enterprise credit evaluation data are analyzed to determine the comprehensive capability data of component material suppliers; A quantitative evaluation index system for the supply chain attributes of distribution transformer suppliers is established based on the comprehensive capability data of component material suppliers.

4. According to the graph neural network-based distribution transformer supplier evaluation method of claim 3, the component materials include iron core-silicon steel sheet, electromagnetic wire, insulating oil, bushing, tap changer, fan, oil storage cabinet, cooler or radiator, temperature control equipment and pressure relief valve.

5. According to the graph neural network-based distribution transformer supplier evaluation method described in claim 1, in the step of constructing a graph neural network model, a few-sample learning strategy is adopted to use a small part of labeled data as a training set, and the remaining unlabeled data is used to verify and test the accuracy of the model.

6. According to the method for evaluating distribution transformer suppliers based on graph neural network in claim 1, the distribution transformer supplier evaluation index system and the distribution transformer supplier supply chain attribute quantitative evaluation index system are used as inputs, and in the step of outputting the distribution transformer supplier evaluation results, the specific process is as follows: The graph neural network model is used to embed node and edge features into the latent feature space. The input node features are the comprehensive data of distribution transformer suppliers and the supplier's ability data to provide component materials. The edge features are the supply relationship between suppliers. The distribution transformer supplier evaluation results are output based on the feature information of suppliers cooperating with the input distribution transformer supplier.

7. A distribution transformer supplier evaluation system based on graph neural network, characterized in that: include An evaluation index system building module is used to build an evaluation index system for distribution transformer suppliers based on the comprehensive capability data of distribution transformer suppliers; A module for building a quantitative attribute evaluation index system, which is used to build a quantitative attribute evaluation index system for the distribution transformer supplier supply chain based on the component materials that the distribution transformer supplier cooperates with; A graph neural network model processing module is used to construct a graph neural network model, take the distribution transformer supplier evaluation index system and the distribution transformer supplier supply chain attribute quantitative evaluation index system as input, and output the distribution transformer supplier evaluation result; The evaluation module evaluates the distribution transformer supplier based on the graph neural network and obtains the evaluation result according to the evaluation result.

8. A mobile terminal comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the distribution transformer supplier evaluation method based on graph neural network as described in any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the distribution transformer supplier evaluation method based on graph neural network as described in any one of claims 1 to 6 are implemented.

10. A computer program product comprising computer instructions, characterized in that: The computer instructions instruct the computing device to perform operations corresponding to the distribution transformer supplier evaluation method based on graph neural network as described in any one of claims 1-6.

Images