Intelligent auditing method and system for power-cut plan of power grid based on large model technology

Abstract

The invention discloses a power grid power-cut plan intelligent auditing method and system based on a large model technology, and the method comprises the steps: collecting historical power-cut plan data, carrying out the cleaning and standardization processing of the data, and obtaining the processed data; obtaining a state grid power failure operation guide rule, preprocessing the guide rule, and training the AI large model by using the preprocessed guide rule; using the trained AI large model to identify and analyze a power failure line to obtain a final power failure item; and comparing the final power failure item with the reported power failure plan, judging whether the power failure plan has a missing item by using a data matching algorithm, and if yes, generating prompt information and sending the prompt information to related personnel for supplementation. According to the invention, by optimizing and calibrating the power failure period, the accuracy of the power failure plan is ensured, the tedious work and time cost of manual auditing are reduced, and the reliability and continuity of power supply are ensured.

Description

Technical Field

[0001] This invention relates to the field of power grid outage planning management technology, specifically to a method and system for intelligent review of power grid outage plans based on large-scale model technology. Background Technology

[0002] Currently, when power grid companies formulate and execute power outage plans, the actual outage time often deviates from the specifications in the technical documents due to various factors such as equipment condition, weather changes, and staffing. Traditional plan formulation and review rely heavily on manual judgment by dispatchers, which is inefficient and prone to errors.

[0003] In recent years, with the development of information technology, attempts have begun to explore the use of data analysis and artificial intelligence to assist in the optimization of power outage plans. However, most existing technologies are limited to rule engines or simple statistical analysis, lacking a deep understanding of complex power grid structures, equipment status, and the influence of various factors, thus failing to achieve accurate calibration and intelligent auditing. Furthermore, traditional methods struggle to fully utilize the potential patterns in historical data, making it difficult to effectively predict and correct plan deviations.

[0004] Therefore, there is an urgent need for an innovative method based on advanced artificial intelligence technology, especially large models (such as large-scale deep learning models), to intelligently identify, review, and optimize power grid outage plans, so as to improve the accuracy, scientific nature, and intelligence of the plans. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for intelligent review of power grid outage plans based on large model technology. This system can automatically analyze and identify potential problems in outage plans through deep learning models, calibrate outage schedules, ensure the scientific validity and rationality of the plans, thereby improving power grid operating efficiency, reducing unnecessary outage time, and enhancing user experience.

[0006] This invention adopts the following technical solution: a method for intelligent review of power grid outage plans based on large model technology, comprising the following steps:

[0007] S1. Collect historical power outage plan data, clean and standardize the data to obtain processed data.

[0008] S2. Obtain the State Grid's power outage operation guidelines, preprocess the guidelines, and use the preprocessed guidelines to train the AI ​​large model.

[0009] S3. Use the trained AI model to identify and analyze the power outage lines to obtain the final power outage information.

[0010] S4. Compare the final power outage items with the reported power outage plan, and use a data matching algorithm to determine if there are any omissions in the power outage plan. If so, generate a prompt message and send it to the relevant personnel for supplementation.

[0011] Furthermore, in step S1, the historical power outage plan data includes both paper documents and electronic data; the historical power outage plan data includes the outage time, the equipment affected, the reason for the outage, and the construction procedures.

[0012] Cleaning historical power outage plan data includes removing duplicate data and handling missing values ​​using mean or median filling.

[0013] Standardized historical power outage planning data includes a unified data format and units.

[0014] Furthermore, in step S2, the State Grid Power Outage Operation Guidelines are parsed to extract the rules and standards from the guidelines and transformed into a knowledge representation. Natural language processing technology is then used to preprocess this knowledge representation.

[0015] Based on the preprocessed knowledge representation, a supervised learning approach is adopted, with power outage duration as the target variable and the rules and standards of the guidelines as input variables. The AI ​​large model learns the relationship between the target variable and the input variables, calculates the deviation between the target variable and the input variables, adjusts and optimizes the parameters of the AI ​​large model according to the deviation, and combines the gradient descent algorithm to obtain the trained AI large model.

[0016] Based on the adjusted and optimized parameters of the AI ​​large model, the set power outage duration and the extreme power outage duration are determined.

[0017] Furthermore, in step S3, the power outage line includes the main wiring diagram, cross-sectional diagram, and plan view.

[0018] The main wiring diagram, cross-sectional view, and plan view are preprocessed. The preprocessed diagrams are then input into a trained AI model to extract text, symbols, and graphics from the three diagrams. Based on the project's construction type, character recognition and semantic analysis technologies are used to determine the final power outage items. Specifically, in the main wiring diagram, a visual recognition model is used to analyze the image content, identify the names of the switches under construction, and obtain the equipment connected to those switches. A visual inference model is used to obtain the outgoing lines of the main wiring diagram and determine the power outage lines. In the cross-sectional view, combined with the spatial direction information of the outgoing lines in the main wiring diagram, a visual recognition model is used to determine the crossing lines in the construction area. In the plan view, power outage equipment less than 8.5m in height is determined based on the construction area.

[0019] Among them, a visual deduction model is constructed based on the topology of the main wiring diagram and related electrical principles.

[0020] Further preprocessing includes image enhancement and noise reduction.

[0021] Furthermore, it also includes: using the extreme power outage duration to determine whether the power outage duration in the reported power outage plan is reasonable; if the power outage duration exceeds the extreme power outage duration, an early warning will be issued.

[0022] Furthermore, this invention also proposes an intelligent review system for power grid outage plans based on large model technology, comprising:

[0023] The data processing module is used to collect historical power outage plan data, clean and standardize the data to obtain processed data.

[0024] The model training module is used to obtain the State Grid's power outage operation guidelines, preprocess the guidelines, and use the preprocessed guidelines to train the AI ​​large model.

[0025] The identification module is used to identify and analyze power outage lines using a trained AI model to obtain the final power outage information.

[0026] The comparison module is used to compare the final power outage items with the reported power outage plan. It uses a data matching algorithm to determine whether there are any omissions in the power outage plan. If so, it generates a prompt message and sends it to the relevant personnel for supplementation.

[0027] Furthermore, the present invention also proposes an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the intelligent auditing method for power grid outage plans based on large model technology.

[0028] Furthermore, the present invention also proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the aforementioned intelligent auditing method for power grid outage plans based on large model technology.

[0029] Compared with the prior art, the present invention, employing the above technical solution, has the following technical effects:

[0030] This invention, through in-depth analysis of historical data and model training, can accurately identify the relationship between power outage duration and various influencing factors, thereby optimizing and calibrating power outage schedules and ensuring the accuracy of power outage plans.

[0031] This invention can not only significantly improve work efficiency by automatically identifying power outages, comparing and reviewing omissions, and judging the reasonableness of power outage duration, but also reduce the tedious work and time costs of manual review, and ensure the reliability and continuity of power supply.

[0032] This invention can reduce unnecessary power outage time, minimize the impact of power outages on users, and thus improve the stability of power supply and user satisfaction. Furthermore, reasonable power outage planning helps reduce resource waste and additional costs, thereby lowering operating costs and laying the foundation for intelligent power grid management, promoting the development of power grid management towards greater intelligence and automation. Attached Figure Description

[0033] Figure 1 This is a flowchart illustrating the overall implementation of the present invention.

[0034] Figure 2 This is a switch identification result diagram of the main wiring diagram in an embodiment of the present invention.

[0035] Figure 3 This is a diagram showing the power outage line result of the main wiring diagram in an embodiment of the present invention.

[0036] Figure 4 This is a cross-sectional view of the line crossing result in an embodiment of the present invention.

[0037] Figure 5 This is a diagram showing the results of the omission check in an embodiment of the present invention.

[0038] Figure 6 This is a diagram showing the power outage plan review results in an embodiment of the present invention. Detailed Implementation

[0039] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0040] To achieve the above objectives, this invention proposes an intelligent review method for power grid outage plans based on large-scale model technology, such as... Figure 1 As shown, the specific steps are as follows:

[0041] S1. Collect historical power outage planning data from the past 10 years, clean and standardize this data to obtain processed data. Specifically:

[0042] Historical power outage plan data includes both paper documents and electronic data; historical power outage plan data includes outage time, outage equipment, outage reason and construction procedures, etc.

[0043] Cleaning historical power outage plan data involves removing duplicate data and handling missing values ​​using means or median imputation. Standardizing historical power outage plan data includes unifying the data format and units. Ensuring data consistency and accuracy provides a high-quality data foundation for subsequent model training.

[0044] S2. Obtain the State Grid's power outage operation guidelines, preprocess the guidelines, and use the preprocessed guidelines to train the AI ​​large-scale model. Specifically:

[0045] The State Grid's guidelines for power outage operations are parsed to extract the rules and standards, which are then transformed into a knowledge representation. Natural language processing techniques (such as lexical analysis and syntactic analysis) are used to preprocess this knowledge representation.

[0046] Based on the preprocessed knowledge representation, and through deep learning and data mining techniques, a supervised learning approach is adopted. With power outage duration as the target variable and the rules and standards of the guidelines as input variables, the AI ​​model (such as a neural network model in deep learning) learns the relationship between the target variable and the input variables, identifies patterns and rules, calculates the deviation between the target variable and the input variables, and adjusts and optimizes the parameters of the AI ​​model according to the deviation to calibrate the power outage period of the State Grid guidelines. Combined with the gradient descent algorithm, the AI ​​model is continuously iterated until the performance of the model reaches a satisfactory level, resulting in the trained AI model.

[0047] Based on the adjusted and optimized parameters of the AI ​​large model, the set power outage duration and the extreme power outage duration are determined, and these durations are then incorporated into the trained AI large model.

[0048] S3. Utilize the trained AI model to identify and analyze the power outage lines to obtain the final power outage details. Specifically:

[0049] The power outage circuit includes the main wiring diagram, cross-sectional diagram, and plan view.

[0050] The main wiring diagram, cross-sectional view, and plan view are preprocessed, including image enhancement and noise reduction. The preprocessed images are then input into a trained AI model to extract text, symbols, and graphic information from the three images. Based on the project's construction type (new construction, renovation, expansion), character recognition and semantic analysis technologies are used to determine the final power outage items. Specifically, in the main wiring diagram, the switches to be constructed are identified (generally indicated in red), and the equipment connected to these switches, including bus I and bus II, is obtained. In the cross-sectional view, the lines to be crossed are determined based on the construction area. In the plan view, power outage equipment less than 8.5m away is determined based on the construction area.

[0051] S4. Compare the final power outage items with the reported power outage plan, and use a data matching algorithm (such as a string matching algorithm) to determine if there are any omissions in the power outage plan. If so, automatically generate a prompt message and send it to the relevant personnel for supplementation.

[0052] S5. Using data from the State Grid's power outage operation guidelines and historical experience, determine whether the power outage duration in the reported power outage plan is reasonable by using the extreme power outage duration. If the power outage duration exceeds the extreme power outage duration, issue an early warning and provide adjustment suggestions.

[0053] S6. Utilize a trained AI model to intelligently review reported power outage plans and provide experience-based guidance. When reviewing new power outage plans, the model can provide staff with reference opinions based on historical data and learned patterns, helping them formulate more reasonable power outage plans. For example, when encountering similar power outage scenarios, the trained AI model can recommend previously successful power outage plan arrangements to help staff develop more reasonable power outage plans.

[0054] Example:

[0055] like Figure 2 As shown, the names of the switches in the construction area are determined through the main wiring diagram (generally indicated in red). A visual recognition model is then applied to analyze the image content of the uploaded main wiring diagram to identify the switch names marked in red. Subsequently, a summary model is used to output and organize the identified switch names according to a predetermined format to ensure information standardization and ease of subsequent processing.

[0056] like Figure 3 As shown, based on the construction type selected by the user and the identified switch name, the I bus or II bus connected to that switch is determined. The outgoing line names of the main wiring diagram are obtained through a visual deduction model, thereby identifying the power outage line. The visual deduction model is constructed based on the topology of the main wiring diagram and related electrical principles.

[0057] like Figure 4 As shown, the direction of the corresponding outgoing lines in the main wiring diagram is identified on the cross-sectional diagram. Combined with the spatial direction information of the lines, a visual recognition model is used to determine whether the lines cross the construction area, thereby identifying lines that cross the construction area and providing a basis for subsequent schedule assessment and risk control.

[0058] like Figure 5 As shown, check if there are any omissions in the reported power outage plan:

[0059] (a) Using a summary model, extract the specific content information of the power outage plan from the reported power outage plan documents;

[0060] (b) Based on the final power outage details (including information on outage lines and crossing lines), use the summary model to analyze the reported power outage plan documents, determine whether there are any missing lines or construction items, output the detection results and display the missing items information.

[0061] like Figure 6As shown, based on the construction content in the power outage plan, the construction type is matched according to the State Grid power outage operation guidelines to identify the construction period and compare the project duration:

[0062] The process iterates through the construction content in the reported power outage plan documents. Based on the type of construction content, a summary model is used, referencing the power outage guidelines knowledge base, to match the corresponding construction period. Then, by combining this with the extreme construction periods from historical power outage plan data, the reasonableness of the planned construction period is determined. In this step, the summary model can search for the corresponding standard construction period from the power outage guidelines knowledge base based on the type of construction content, and combine this with the extreme construction periods from historical power outage plan data to comprehensively evaluate and judge the construction period of the current power outage plan.

[0063] This invention also proposes an intelligent power grid outage plan review system based on large model technology, including a data processing module, a model training module, an identification module, a comparison module, and a computer program that can run on a processor. It should be noted that each module in the above system corresponds to a specific step of the method provided in this invention embodiment, possessing the corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in this embodiment can be found in the method provided in this invention embodiment.

[0064] This invention also proposes an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. It should be noted that when the processor executes the computer program, it corresponds to the specific steps of the method provided in this invention, possessing the corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in this embodiment can be found in the method provided in this invention.

[0065] This invention also proposes a computer-readable storage medium storing a computer program. It should be noted that when the computer program is executed by a processor, it corresponds to the specific steps of the method provided in this invention, possessing the corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in this embodiment can be found in the method provided in this invention.

[0066] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for intelligent review of power grid outage plans based on large model technology, characterized in that, include: S1. Collect historical power outage plan data, clean and standardize the data to obtain processed data; S2. Obtain the State Grid's power outage operation guidelines, preprocess the guidelines, and use the preprocessed guidelines to train the AI ​​large model. S3. Use the trained AI model to identify and analyze the power outage lines to obtain the final power outage information; S4. Compare the final power outage items with the reported power outage plan, and use a data matching algorithm to determine if there are any omissions in the power outage plan. If so, generate a prompt message and send it to the relevant personnel for supplementation.

2. The intelligent review method for power grid outage plans based on large model technology according to claim 1, characterized in that, In step S1, the historical power outage plan data includes both paper documents and electronic data; the historical power outage plan data includes the outage time, the equipment affected, the reason for the outage, and the construction procedures. Cleaning historical power outage plan data includes removing duplicate data and handling missing values ​​using mean or median filling. Standardized historical power outage planning data includes a unified data format and units.

3. The intelligent review method for power grid outage plans based on large model technology according to claim 1, characterized in that, In step S2, the State Grid Power Outage Operation Guidelines are parsed to extract the rules and standards from the guidelines and transformed into a knowledge representation. Natural language processing technology is then used to preprocess this knowledge representation. Based on the preprocessed knowledge representation, a supervised learning approach is adopted, with the power outage duration as the target variable and the rules and standards of the guidelines as the input variables. The AI ​​large model learns the relationship between the target variable and the input variables, calculates the deviation between the target variable and the input variables, adjusts and optimizes the parameters of the AI ​​large model according to the deviation, and combines the gradient descent algorithm to obtain the trained AI large model. Based on the adjusted and optimized parameters of the AI ​​large model, the set power outage duration and the extreme power outage duration are determined.

4. The intelligent review method for power grid outage plans based on large model technology according to claim 1, characterized in that, In step S3, the power outage line includes the main wiring diagram, cross-sectional view, and plan view; The main wiring diagram, cross-sectional view, and plan view are preprocessed. The preprocessed diagrams are then input into a trained AI model to extract text, symbols, and graphics from the three diagrams. Based on the project's construction type, character recognition and semantic analysis technologies are used to determine the final power outage items. Specifically, in the main wiring diagram, a visual recognition model is used to analyze the image content, identify the names of the switches under construction, and obtain the equipment connected to those switches. A visual inference model is used to obtain the outgoing lines of the main wiring diagram and determine the power outage lines. In the cross-sectional view, combined with the spatial direction information of the outgoing lines in the main wiring diagram, a visual recognition model is used to determine the crossing lines in the construction area. In the plan view, power outage equipment less than 8.5m in height is determined based on the construction area. Among them, a visual deduction model is constructed based on the topology of the main wiring diagram and related electrical principles.

5. The intelligent review method for power grid outage plans based on large model technology according to claim 4, characterized in that, Preprocessing includes image enhancement and noise reduction.

6. The intelligent review method for power grid outage plans based on large model technology according to claim 3, characterized in that, Also includes: The maximum outage duration is used to determine whether the outage duration in the reported outage plan is reasonable. If the outage duration exceeds the maximum outage duration, an early warning is issued.

7. A system applied to the intelligent review method for power grid outage plans based on large model technology as described in any one of claims 1-6, characterized in that, include: The data processing module is used to collect historical power outage plan data, clean and standardize the data to obtain processed data; The model training module is used to obtain the State Grid's power outage operation guidelines, preprocess the guidelines, and use the preprocessed guidelines to train the AI ​​large model. The identification module is used to identify and analyze the power outage lines using a trained AI model to obtain the final power outage information. The comparison module is used to compare the final power outage items with the reported power outage plan. It uses a data matching algorithm to determine whether there are any omissions in the power outage plan. If so, it generates a prompt message and sends it to the relevant personnel for supplementation.

8. An electronic device 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, it implements the steps of the intelligent review method for power grid outage plans based on large model technology as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, The computer program, when executed by the processor, performs the intelligent auditing method for power grid outage plans based on large model technology, as described in any one of claims 1 to 6.

Images