Management Method for Artificial Intelligence Product Solutions of Power Systems Based on Knowledge Graphs

Abstract

An embodiment of the present invention provides a management method for an artificial intelligence product solution based on a knowledge graph in the power system, belonging to the technical field of artificial intelligence product solutions in the power system. The management method includes: obtaining historical solution data of artificial intelligence products in the power system; constructing a product solution knowledge graph according to the historical solution data. The management method for the artificial intelligence product solution based on the knowledge graph provided by the present invention summarizes the historical solution data of artificial intelligence products at each node in the power system, constructs a product solution knowledge graph according to the historical solution data, then obtains the current search request, and matches the search request with the product solution knowledge graph to obtain a solution to the artificial intelligence product failure, thereby timely solving the corresponding failure and ensuring the safe and reliable operation of the power grid system.

Description

Technical Field

[0001] The present invention relates to the technical field of artificial intelligence product solutions for power systems, and particularly to a management method for artificial intelligence product solutions for power systems based on a knowledge graph. Background Art

[0002] In order to ensure the operation of the power grid, multiple artificial intelligence devices are set in the power system to assist the safe and reliable operation of the power grid. For example, safety monitoring devices, intelligent inspection devices, intelligent operation and maintenance devices, etc. The above devices play a great role in the operation of the power grid. Therefore, when a failure occurs in the above devices, the timeliness of maintenance is of great significance.

[0003] Currently, when a device fails, there is generally a common device maintenance method. However, for the maintenance of such special devices, there are different maintenance methods among the various nodes in the power grid system, and there are also differences in the maintenance effect and maintenance time. For nodes with relatively scarce technology, it is impossible to obtain a suitable maintenance method, which will affect the safe and stable operation of the power grid at that node.

[0004] The inventors of the present application found in the process of implementing the present invention that the above-mentioned solutions of the prior art have the defect that untimely maintenance of artificial intelligence devices by some nodes affects the safe and stable operation of the power grid. Summary of the Invention

[0005] The purpose of the embodiments of the present invention is to provide a management method for artificial intelligence product solutions for power systems based on a knowledge graph, and this management method for artificial intelligence product solutions for power systems based on a knowledge graph can ensure the function of the safe and stable operation of the power grid at each node in the power system.

[0006] In order to achieve the above purpose, the embodiments of the present invention provide a management method for artificial intelligence product solutions for power systems based on a knowledge graph, including:

[0007] Obtain historical solution data of artificial intelligence products in the power system;

[0008] Construct a product solution knowledge graph according to the historical solution data;

[0009] Obtain the current search request;

[0010] Obtain the solution of the artificial intelligence product according to the search request and the product solution knowledge graph;

[0011] Apply the solution to the corresponding artificial intelligence product to solve the corresponding fault problem.

[0012] Optionally, the historical solution data includes a fault problem, a fault time, a solution, and an effectiveness score.

[0013] Optionally, constructing a product solution knowledge graph based on the historical solution data includes:

[0014] Performing knowledge extraction on the historical solution data;

[0015] Performing knowledge fusion on the extracted knowledge;

[0016] Processing the fused knowledge to eliminate ambiguity between entities;

[0017] Constructing a product solution knowledge graph.

[0018] Optionally, obtaining a solution for the artificial intelligence product according to the search request and the product solution knowledge graph includes:

[0019] Obtaining multiple keywords in the search request;

[0020] Respectively matching the multiple keywords with the product solution knowledge graph;

[0021] Calculating the relevance between each keyword and the content in the product solution knowledge graph according to formula (1),

[0022]

[0023] where γ(i) is the relevance between the i-th keyword and the content in the product solution knowledge graph, α1 and α2 are weight coefficients and are preset values, Λ(i) is the TF*IDF value of the i-th keyword, Λ msx (i) is the maximum TF*IDF value of the i-th keyword, Γ(i) is the number of links of the i-th keyword, is the average value of the number of links of all keywords, and i is an integer number.

[0024] Optionally, obtaining a solution for the artificial intelligence product according to the search request and the product solution knowledge graph further includes:

[0025] Obtaining the relevance weight of each keyword;

[0026] Calculating the relevance between the search request and the content in the product solution knowledge graph according to formula (2),

[0027]

[0028] where Υ is the relevance between the search request and the content in the product solution knowledge graph, i ≤ m, m is the number of keywords, βi is the relevance weight for the i-th keyword.

[0029] Optionally, obtaining the relevance weight of each of the keywords includes:

[0030] obtaining the relationship of the relevance weights of each of the keywords according to formula (3),

[0031]

[0032] where β i is directly proportional to γ(i).

[0033] Optionally, obtaining the solution of the artificial intelligence product according to the search request and the product solution knowledge graph further includes:

[0034] presetting the push quantity of the product solution knowledge graph;

[0035] obtaining the push quantity of solution data according to the relevance between the search request and the content in the product solution knowledge graph;

[0036] obtaining the failure time and effect score in the push solution data;

[0037] performing normalization processing on the failure time and effect score in the push solution data;

[0038] obtaining the importance of the push solution data according to the failure time and effect score in the push solution data after normalization processing;

[0039] performing sorting and pushing according to the importance of the push solution data.

[0040] Optionally, performing normalization processing on the failure time and effect score in the push solution data includes:

[0041] normalizing the failure time according to formula (4),

[0042] T j =(t j -t min ) / (t max -t min ), (4)

[0043] where, T j is the value after normalizing the failure time of the j-th push solution data, t j is the failure time of the j-th push solution data, t max is the maximum value of all the failure times, t min is the minimum value of all the failure times, and j is an integer number;

[0044] Normalize the effect score according to formula (5).

[0045] Ψ j =(μ j -μ min ) / (μ max -μ min ), (5)

[0046] where Ψ j is the value after normalizing the effect score of the j-th push scheme data, μ j is the effect score of the j-th push scheme data, μ max is the maximum value of all the effect scores, and μ min is the minimum value of all the effect scores.

[0047] Optionally, obtaining the importance of the push scheme data according to the failure time and the effect score in the push scheme data after normalization includes:

[0048] Calculating the importance of each push scheme data according to formula (6).

[0049]

[0050] where λ j is the importance of the j-th push scheme data, τ1 and τ2 are the adjustment coefficients of the effect score and the failure time respectively, and both τ1 and τ2 are greater than 1.

[0051] On the other hand, the present invention also provides a computer-readable storage medium storing instructions for being read by a machine to cause the machine to execute any one of the management methods described above.

[0052] Through the above technical solutions, the management method of the artificial intelligence product solution for the power system based on the knowledge graph provided by the present invention summarizes the historical solution data of the artificial intelligence products of each node in the power system, constructs a product solution knowledge graph according to the historical solution data, then obtains the current search request, and matches the search request with the product solution knowledge graph to obtain a solution to the artificial intelligence product failure, thereby timely solving the corresponding failure and ensuring the safe and reliable operation of the power grid system.

[0053] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent specific implementation part. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the embodiments of the present invention, but do not constitute a limitation to the embodiments of the present invention. In the accompanying drawings:

[0055] Figure 1 is a flowchart of a management method for an artificial intelligence product solution of a power system based on a knowledge graph according to an embodiment of the present invention;

[0056] Figure 2 is a flowchart of constructing a knowledge graph of product solutions in the management method for an artificial intelligence product solution of a power system based on a knowledge graph according to an embodiment of the present invention;

[0057] Figure 3 is a flowchart of obtaining a solution in the management method for an artificial intelligence product solution of a power system based on a knowledge graph according to an embodiment of the present invention;

[0058] Figure 4 is a flowchart of normalizing fault time and effect evaluation in the management method for an artificial intelligence product solution of a power system based on a knowledge graph according to an embodiment of the present invention. Specific Embodiments

[0059] The following will detail the specific embodiments of the embodiments of the present invention with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to illustrate and explain the embodiments of the present invention and do not limit the embodiments of the present invention.

[0060] Figure 1 is a flowchart of a management method for an artificial intelligence product solution of a power system based on a knowledge graph according to an embodiment of the present invention. In Figure 1 this, the management method may include:

[0061] In step S10, historical solution data of artificial intelligence products in the power system is obtained. Among them, the historical solution data may include fault problems, fault times, solution methods, and effect scores of all artificial intelligence products, etc.

[0062] In step S11, a knowledge graph of product solutions is constructed according to the historical solution data.

[0063] In step S12, a current search request is obtained. Among them, the search request may include text search, voice search, etc., not limited to the form, but finally input in the form of text.

[0064] In step S13, according to the search request and the product solution knowledge graph, the solution of the artificial intelligence product is obtained. Among them, by matching the search request with the product solution knowledge graph, multiple relevant product solutions can be obtained.

[0065] In step S14, the solution is applied to the corresponding artificial intelligence product to solve the corresponding fault problem.

[0066] In steps S10 to S14, first, historical data such as all fault problems, fault times, solutions, and effect scores of artificial intelligence products in the power system are obtained, and a product solution knowledge graph is constructed based on the above historical data. After a fault occurs in an artificial intelligence product at a certain current node, a search request is determined, and the search request is matched and searched with the product solution knowledge graph, and then relevant solutions can be obtained. According to this solution, the artificial intelligence product can be repaired in a timely and effective manner.

[0067] In the traditional power system, for the repair of such special equipment, there are different repair methods among the various nodes in the power grid system, and there are also differences in the repair effects and repair times. For nodes with relatively scarce technology, it is impossible to obtain appropriate repair methods, which will affect the safe and stable operation of the power grid at that node. In this embodiment of the present invention, by constructing a product solution knowledge graph, each node in the power system can obtain the solution of the artificial intelligence product in a timely and accurate manner, and then the corresponding faults can be solved in a timely manner, ensuring the safe and reliable operation of the power grid system.

[0068] In this embodiment of the present invention, when constructing the product solution knowledge graph, historical solution data also needs to be processed. The specific processing steps can Figure 2 be as shown. Specifically, in Figure 2 , the management method may further include:

[0069] In step S110, knowledge extraction is performed on the historical solution data.

[0070] In step S111, knowledge fusion is performed on the extracted knowledge.

[0071] In step S112, the fused knowledge is processed to eliminate the ambiguity between entities.

[0072] In step S113, a product solution knowledge graph is constructed.

[0073] In steps S110 to S113, the construction of the product solution knowledge graph can be steps known to those skilled in the art. Specifically, the node types to be constructed, the attributes of each type of node, the relationship types between nodes, and the attribute definitions of each node can be preset, and then knowledge extraction is performed on the historical solution data, that is, entity extraction and relationship extraction. After the extraction is completed, the knowledge obtained from different channels is fused, processed and integrated to eliminate contradictions and ambiguities. Finally, the quality of the fused and processed new knowledge is evaluated, and the qualified part is added to the knowledge base, thereby forming a product solution knowledge graph.

[0074] In this embodiment of the present invention, after obtaining a search request for a certain node in the power system, it is also necessary to match the search request with the product solution knowledge graph to obtain relevant solutions. Specifically, the matching process can be as Figure 3 shown. Specifically, in Figure 3 this, the management method may further include:

[0075] In step S130, multiple keywords in the search request are obtained. Among them, the extraction methods for keywords can be word frequency, tf-idf, position features, inherent attributes of words, etc., known to those skilled in the art.

[0076] In step S131, the multiple keywords are respectively matched with the product solution knowledge graph.

[0077] In step S132, the relevance between each keyword and the content in the product solution knowledge graph is calculated according to formula (1),

[0078]

[0079] where γ(i) is the relevance between the i-th keyword and the content in the product solution knowledge graph, α1 and α2 are weight coefficients and are preset values, Λ(i) is the TF*IDF value of the i-th keyword, Λ max (i) is the maximum TF*IDF value of the i-th keyword, Γ(i) is the number of links of the i-th keyword, is the average value of the number of links of all keywords, and i is an integer number. Specifically, the average value of the number of links of all keywords is the average of the sum of the number of links of all keywords.

[0080] In step S133, the relevance weight of each keyword is obtained. Among them, the obtaining of the relevance weight of each keyword can be as shown in formula (3). Specifically, the relationship of the relevance weight of each keyword is obtained according to formula (3),

[0081]

[0082] Among them, β i is directly proportional to γ(i). Specifically, β i being directly proportional to γ(i) means that the greater the relevance of the keyword to the content in the product solution knowledge graph, the greater the relevance weight, that is, the proportion of the relevance between the keyword with high relevance and the content in the search request is amplified here.

[0083] In step S134, calculate the relevance between the search request and the content in the product solution knowledge graph according to formula (2),

[0084]

[0085] where Υ is the relevance between the search request and the content in the product solution knowledge graph, i ≤ m, β i is the relevance weight of the i-th keyword, and m is the number of keywords.

[0086] In step S135, preset the push quantity of the product solution knowledge graph.

[0087] In step S136, obtain the push quantity of solution data according to the relevance between the search request and the content in the product solution knowledge graph.

[0088] In step S137, obtain the failure time and effect score in the push solution data.

[0089] In step S138, perform normalization processing on the failure time and effect score in the push solution data. Among them, the method of normalizing the failure time and effect evaluation in the push solution data can include the method of the sigmoid function well-known to those skilled in the art, etc. In a preferred example of the present invention, it can include as Figure 4 shown in the steps. Specifically, in Figure 4 , this normalization step may include:

[0090] In step S1380, normalize the failure time according to formula (4),

[0091] T j =(t j -t min ) / (t max -t min ), (4)

[0092] where T j is the value after normalizing the failure time of the j-th push solution data, t j is the failure time of the j-th push solution data, t max is the maximum value of all failure times, t minis the minimum value of all failure times, and j is an integer numbering.

[0093] In step S1381, the effect score is normalized according to formula (5).

[0094] Ψ j = (μ j - μ min ) / (μ max - μ min ), (5)

[0095] where Ψ j is the value after normalizing the effect score of the j-th push scheme data, μ j is the effect score of the j-th push scheme data, μ max is the maximum value of all effect scores, and μ min is the minimum value of all effect scores.

[0096] In step S139, the importance of the push scheme data is obtained based on the failure time and effect score in the normalized push scheme data. Among them, the calculation of the importance of the push scheme data can be as shown in formula (6). Specifically, the importance of each push scheme data is calculated according to formula (6).

[0097]

[0098] where λ j is the importance of the j-th push scheme data, τ1 and τ2 are the adjustment coefficients of the effect score and the failure time respectively, and both τ1 and τ2 are greater than 1.

[0099] In step S140, push is performed according to the importance of the push scheme data. Among them, the push scheme data with the greatest importance is used as the first push, the one with the second greatest importance is used as the second push, and so on.

[0100] In steps S130 to S140, the search request is first disassembled. For the subsequent multiple keywords, the multiple keywords are respectively matched with the content in the product solution knowledge graph, and the corresponding relevance is obtained. Then, according to the relevance of each keyword to the content, the relevance of the search request to the content is calculated. According to the preset number of pushes, the corresponding number of scheme data is obtained, and the failure time and effect score in the obtained scheme data are normalized. Finally, the importance of each scheme data of the number of pushes is calculated, and push is performed according to this importance. By using this push method, the time limitation of the scheme and the effect evaluation are synchronously considered, and the most suitable scheme can be effectively and accurately pushed, which is more intelligent and convenient.

[0101] Through the above technical solution, the management method of the artificial intelligence product solution for the power system based on the knowledge graph provided by the present invention summarizes the historical solution data of the artificial intelligence products of each node in the power system, constructs a product solution knowledge graph based on the historical solution data, then obtains the current search request, and matches the search request with the product solution knowledge graph to obtain the solution to the artificial intelligence product failure, thereby timely solving the corresponding failure and ensuring the safe and reliable operation of the power grid system.

[0102] On the other hand, the present invention also provides a computer-readable storage medium storing instructions for being read by a machine to cause the machine to execute the management method as described in any one of the above.

[0103] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0104] The present application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the specified functions in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0105] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the specified functions in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0106] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 steps of the functions specified in one block or multiple blocks.

[0107] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and a memory.

[0108] The memory may include non-permanent memory in the computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.

[0109] Computer-readable media includes permanent and non-permanent, removable and non-removable media and can be implemented by any method or technology for information storage. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0110] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, commodity or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, commodity or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, commodity or device comprising the element.

[0111] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the scope of the claims of the present application.

Claims

1. A management method for an artificial intelligence product solution of a power system based on a knowledge graph, characterized in that, Including: Obtaining historical solution data of artificial intelligence products in a power system; Constructing a product solution knowledge graph based on the historical solution data; Obtaining a current search request; Obtaining a solution of the artificial intelligence product according to the search request and the product solution knowledge graph; Applying the solution to the corresponding artificial intelligence product to solve the corresponding fault problem; The historical solution data includes fault problems, fault times, solutions, and effect scores; Obtaining a solution of the artificial intelligence product according to the search request and the product solution knowledge graph includes: Obtaining multiple keywords in the search request; Matching the multiple keywords with the product solution knowledge graph respectively; Calculating the relevance between each keyword and the content in the product solution knowledge graph according to formula (1); Among them, γ(i) is the relevance between the i-th keyword and the content in the product solution knowledge graph, α1 and α2 are weight coefficients and are preset values, Λ(i) is the TF*IDF value of the i-th keyword, and Λ max (i) is the maximum TF*IDF value of the i-th keyword, Γ(i) is the number of links of the i-th keyword, is the average value of the number of links of all keywords, and i is an integer number; Obtaining the relevance weight of each keyword; Calculating the relevance between the search request and the content in the product solution knowledge graph according to formula (2); Among them, Υ is the relevance between the search request and the content in the product solution knowledge graph, i ≤ m, where m is the number of keywords, and β i is the relevance weight of the i-th keyword; Obtaining the relevance weight of each keyword includes: Obtaining the relationship of the relevance weights of each keyword according to formula (3); where β i is directly proportional to γ(i); Presetting the push quantity of the product solution knowledge graph; Obtaining the push quantity of solution data according to the relevance between the search request and the content in the product solution knowledge graph; Obtaining the fault time and effect score in the push solution data; Performing normalization processing on the fault time and effect score in the push solution data; Obtaining the importance of the push solution data according to the fault time and effect score in the push solution data after normalization processing; Performing sorting and pushing according to the importance of the push solution data; Obtaining the importance of the push solution data according to the fault time and effect score in the push solution data after normalization processing includes: Calculating the importance of each push solution data according to formula (6); where λ j is the importance of the j-th push scheme data, τ1 and τ2 are the adjustment coefficients of the effect score and the failure time respectively, and both τ1 and τ2 are greater than 1.

2. The management method according to claim 1, wherein Constructing a product solution knowledge graph based on the historical solution data includes: Performing knowledge extraction on the historical solution data; Performing knowledge fusion on the extracted knowledge; Processing the fused knowledge to eliminate the ambiguity between entities; Constructing a product solution knowledge graph.

3. The management method according to claim 1, characterized in that Performing normalization processing on the fault time and effect score in the push solution data includes: Normalizing the fault time according to formula (4); T j = (t j - t min ) / (t max - t min ), (4) Among them, T j is the value after normalizing the failure time of the j-th said push scheme data, t j is the failure time of the j-th said push scheme data, t max is the maximum value of all said failure times, t min is the minimum value of all said failure times, and j is an integer number; Normalizing the effect score according to formula (5); Ψ j = (μ j - μ min ) / (μ max - μ min ), (5) Among them, Ψ j is the value after normalizing the effectiveness score of the j-th said push scheme data, μ j is the effectiveness score of the j-th said push scheme data, μ max is the maximum value of all said effectiveness scores, μ min is the minimum value of all said effectiveness scores.

4. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions, and the instructions are used to be read by a machine so that the machine executes the management method according to any one of claims 1 to 3.

Images