A method and apparatus for generating power document data based on entity relationship extraction

Abstract

This application relates to a method and apparatus for generating power document data based on entity relationship extraction. The method includes: acquiring power grid operation text information, power grid construction progress information, power grid operation regulations information, and preset power grid relational structure data of a target power grid; extracting entity relationships from the power grid operation text information, power grid construction progress information, and power grid operation regulations information to obtain power grid entity relational structure data; determining an initial data processing method from argumentative reasoning and counterfactual reasoning of the target power grid based on a first difference value between the power grid entity relational structure data and the power grid preset relational structure data; using the initial data processing method as a guiding condition, retrieving argumentative reasoning and counterfactual reasoning, and performing spiral iterative processing on the power grid entity relational structure data and the power grid preset relational structure data to obtain the target power document data. This method can meet the needs of power grid operation and maintenance for efficient, accurate, and intelligent data processing.

Description

Technical Field

[0001] This application relates to the field of smart grid technology, and in particular to a method and apparatus for generating power document data based on entity relationship extraction. Background Technology

[0002] In the traditional process of generating power document data, manual compilation of power grid operation records, equipment maintenance manuals, and repair reports is the primary method. Template tools or basic text processing software are used for formatting, while some systems employ OCR technology or simple natural language processing algorithms to extract and transform document content. However, traditional methods of generating power document data suffer from low automation and insufficient data structuring capabilities, making it difficult to meet the demands of power grid operation and maintenance for efficient, accurate, and intelligent data processing. Summary of the Invention

[0003] Therefore, it is necessary to provide a method, apparatus, and computer equipment for generating power document data based on entity relationship extraction, which can meet the needs of power grid operation and maintenance for efficient, accurate, and intelligent data processing, in response to the above-mentioned technical problems.

[0004] Firstly, this application provides a method for generating power document data based on entity relationship extraction, including:

[0005] Acquire the target power grid's power grid operation text information, power grid construction progress information, power grid operation regulations information, and power grid preset relationship structure data;

[0006] The power grid operation text information, the power grid construction progress information, and the power grid operation regulations information are extracted to establish entity relationships, thereby obtaining power grid entity relationship structure data.

[0007] Based on the first difference value between the power grid entity relationship structure data and the power grid preset relationship structure data, the initial data processing method is determined from the argumentative arguments and counterfactual reasoning of the target power grid;

[0008] Using the initial data processing method as a guiding condition, the argumentative argument and the counterfactual reasoning are retrieved, and the power grid entity relationship structure data and the power grid preset relationship structure data are processed in a spiral loop iteration to obtain the target power document data.

[0009] The target power document data is used to generate power grid operation instructions and perform operation and maintenance operations on the target power grid.

[0010] Secondly, this application also provides an apparatus for generating power document data based on entity relationship extraction, comprising:

[0011] The power grid data acquisition module is used to acquire power grid operation text information, power grid construction progress information, power grid operation regulations information, and power grid preset relationship structure data of the target power grid;

[0012] The entity relationship extraction module is used to extract the power grid operation text information, the power grid construction progress information, and the power grid operation regulations information to obtain power grid entity relationship structure data.

[0013] The processing order determination module is used to determine the initial data processing method from the argumentative arguments and counterfactual reasoning of the target power grid based on the first difference value between the power grid entity relationship structure data and the power grid preset relationship structure data;

[0014] The document data generation module is used to use the initial data processing method as a guiding condition, retrieve the argumentative argument and the counterfactual reasoning, and perform spiral cyclic iterative processing on the power grid entity relationship structure data and the power grid preset relationship structure data to obtain the target power document data.

[0015] The target power document data is used to generate power grid operation instructions and perform operation and maintenance operations on the target power grid.

[0016] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement any step of a method for generating power document data based on entity relation extraction.

[0017] The aforementioned method, apparatus, and computer equipment for generating power document data based on entity relationship extraction systematically acquire power grid operation text information, power grid construction progress information, and power grid operation regulations information. Combined with pre-set power grid relationship structure data, it constructs comprehensive power grid entity relationship structure data, achieving an accurate depiction of the power grid's operational status. Furthermore, a method combining argumentative argumentation and counterfactual reasoning dynamically analyzes the first difference value between the power grid entity relationship structure data and the pre-set relationship data, intelligently determining the initial data processing method to ensure the scientific rigor and adaptability of data processing. Subsequently, in a spiral iterative process, the interactive optimization mechanism of argumentative argumentation and counterfactual reasoning continuously improves the power grid data processing method, gradually enhancing data quality and the accuracy of information reasoning. Ultimately, it generates high-precision target power document data, which can be directly used to generate power grid operation instructions. This meets the power grid operation and maintenance requirements for efficient, accurate, and intelligent data processing, realizing intelligent and automated power grid operation and maintenance, and significantly improving the accuracy of power grid dispatching, the real-time response, and the overall stability and security of operation. Meanwhile, this invention can reduce manual intervention, improve the efficiency and accuracy of power grid data processing, optimize power grid resource allocation, reduce operation and maintenance costs, enhance the power grid's adaptability to abnormal operating conditions and emergencies, and provide strong support for the intelligent upgrading of the power system. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is an application environment diagram of a power document data generation method based on entity relationship extraction in one embodiment;

[0020] Figure 2 This is a flowchart illustrating a method for generating power document data based on entity relationship extraction in one embodiment.

[0021] Figure 3 This is a flowchart illustrating the first method for obtaining target power document data in one embodiment;

[0022] Figure 4 This is a flowchart illustrating the second method for obtaining target power document data in one embodiment;

[0023] Figure 5 This is a flowchart illustrating a third method for obtaining target power document data in one embodiment;

[0024] Figure 6 This is a flowchart illustrating the fourth method for obtaining target power document data in one embodiment;

[0025] Figure 7 This is a flowchart illustrating the fifth method for obtaining target power document data in one embodiment;

[0026] Figure 8 This is a flowchart illustrating the sixth method for obtaining target power document data in one embodiment;

[0027] Figure 9 This is a flowchart illustrating a method for obtaining the first inference relation structure data in one embodiment;

[0028] Figure 10 This is a structural block diagram of a power document data generation device based on entity relationship extraction in one embodiment;

[0029] Figure 11 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0031] This application provides a method for generating power document data based on entity relationship extraction, which can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104, or it can be located in the cloud or on other network servers. Server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0032] In one exemplary embodiment, such as Figure 2 As shown, a method for generating power document data based on entity relationship extraction is provided, and this method is applied to... Figure 1 Taking the server in the example, the explanation includes the following steps 202 to 208. Wherein:

[0033] Step 202: Obtain the target power grid's power grid operation text information, power grid construction progress information, power grid operation regulations information, and power grid preset relationship structure data.

[0034] The target power grid can be a specific power grid system that requires data analysis, optimization, and operation and maintenance management, including its transmission, substation, distribution, and other aspects.

[0035] Among them, power grid operation text information can be unstructured text data describing the daily operation of the power grid, load dispatching, equipment status, and operational faults.

[0036] Among them, the power grid construction progress information can be text records related to power grid infrastructure construction, expansion and renovation, equipment upgrades, etc., mainly including project plans, construction progress, equipment installation status, and phased acceptance reports.

[0037] Among them, power grid operation regulations information can include relevant policies, technical standards, dispatching principles, safety procedures and other normative documents related to power grid operation.

[0038] Among them, the pre-set relationship structure data of the power grid can be a knowledge base of power grid entity relationships established in advance based on historical operating experience, power grid topology, industry standards, etc. This data defines the connection methods, standard load levels, typical dispatching rules, etc. between power grid equipment.

[0039] Specifically, the data acquisition system collects power grid operation text information, power grid construction progress information, and power grid operation regulation information from various power grid management systems, historical operation and maintenance records, power grid construction and usage regulatory databases, and power grid construction progress archives. Simultaneously, it retrieves pre-defined power grid relational structure data, including power grid equipment topology, historical operating modes, operation and maintenance standards, and optimization objectives, providing a foundation for subsequent data analysis and processing.

[0040] Step 204: Extract power grid operation text information, power grid construction progress information, and power grid operation regulations information to establish entity relationships and obtain power grid entity relationship structure data.

[0041] Among them, the power grid entity relationship structure data can be structured data formed by extracting key entities (such as equipment, faults, and maintenance measures) and their relationships from power grid operation, construction progress, and operation regulations information.

[0042] Specifically, Natural Language Processing (NLP) technology is used to segment the text, identify power grid-related technical terms, and Named Entity Recognition (NER) technology is applied to extract key entities, including equipment (such as transformers, transmission lines, and switching stations), fault types (such as overload, short circuit, and insulation aging), maintenance measures (such as component replacement, load adjustment, and regular inspections), and attribute information (such as voltage level, load capacity, and operating status). Then, a Relation Extraction (RE) model is used to analyze the relationships between these entities, constructing a "equipment-fault-maintenance measure-attribute" structure, for example, "A transformer in a substation experiences an overload fault, and load adjustment and temperature monitoring are taken as maintenance measures." The extracted entity relationship structure data is cleaned, standardized, and formatted before being stored in a relational database or graph database (such as Neo4j), resulting in power grid entity relationship structure data.

[0043] Step 206: Based on the first difference value between the power grid entity relationship structure data and the power grid preset relationship structure data, determine the initial data processing method from the argumentative argumentation and counterfactual reasoning of the target power grid.

[0044] The first difference value can be the deviation between the power grid entity relationship structure data and the power grid preset relationship structure data, reflecting the difference between the current power grid actual operation knowledge base and the preset knowledge base.

[0045] In this context, argumentative argumentation can involve multiple rounds of logical reasoning. The system questions, refutes, and defends conflicts between different data sources, operating standards, or expert opinions, ultimately selecting the most reasonable conclusion. For example, if different documents contradict each other regarding the overheating threshold of equipment, argumentative argumentation can introduce evidence such as historical cases and monitoring data to verify different viewpoints and arrive at an authoritative conclusion.

[0046] Counterfactual reasoning can be an analytical method based on hypothetical scenarios, where the system assesses the impact of a decision or adjustment by simulating possible outcomes under different conditions. For example, the system can assume "if a transformer is repaired 2 hours earlier," and calculate its impact on failure rate, maintenance costs, etc., thereby providing a basis for operation and maintenance decisions.

[0047] The initial data processing method can be the first step data processing strategy determined by the system after analyzing the differences between the power grid entity relationship data and the preset relationship data.

[0048] Specifically, the extracted power grid entity relationship structure data is globally scanned and compared with the preset power grid relationship structure data to identify potential conflicts or differences, obtaining the first difference value. During the comparison process, the system focuses on detecting contradictory situations such as "identical entities with different thresholds" or "mutually exclusive assertions." For example, there may be discrepancies in the overheating criteria for high-voltage switches in different documents (e.g., one document considers 70℃ as overheating, while another considers 80℃). Once such a potential conflict is detected, the system will mark it and extract the assertions from both sides of the conflict, abstracting them into "arguments" that can be used for debate. Next, the system determines the handling method based on the number and severity of conflicts: If there are many conflicts or they involve core power grid operation and maintenance rules (such as load dispatching, equipment safety standards, etc.), that is, if the first difference value is greater than the core operation and maintenance difference threshold, the system will use the argumentative argument of the target power grid as the initial data processing method and directly enter the "argumentative verification" process. Through intelligent argumentation, the rationality of different viewpoints is analyzed and the best standard is selected; if there are few conflicts and the priority is low (such as slight differences in maintenance strategies for non-critical equipment), that is, if the first difference value is less than or equal to the core operation and maintenance difference threshold, the system will use the counterfactual reasoning of the target power grid as the initial data processing method. In this case, the system will postpone entering the argumentation process and instead execute the subsequent counterfactual reasoning first, and trigger the argumentative argument only when necessary.

[0049] Step 208: Using the initial data processing method as a guiding condition, retrieve argumentative arguments and counterfactual reasoning, and perform spiral cyclic iterative processing on the power grid entity relationship structure data and the power grid preset relationship structure data to obtain the target power document data.

[0050] The guiding conditions can be the adjustment direction or rules set by the system based on the current data analysis results during the spiral iterative optimization process. For example, if the system detects a large data deviation, the guiding conditions may favor the execution of argumentative reasoning; if the deviation is small, the system may be guided to first perform counterfactual reasoning to optimize the adjustment strategy.

[0051] Among them, spiral cyclic iterative processing can be a method to gradually optimize power grid data. The system first makes preliminary adjustments to the data through argumentative demonstration or counterfactual reasoning, and then continuously tests and optimizes it to form more accurate entity relationship structure data.

[0052] The target power document data can be the final high-quality power grid operation and maintenance data file formed after spiral cyclic iterative optimization, including dispatch instructions, equipment maintenance plans, optimization schemes, etc.

[0053] Among them, power grid operation instructions can be specific operational instructions or decision suggestions automatically generated by the system based on optimized target power document data, used to guide the real-time operation and maintenance management of the power grid. These instructions cover multiple aspects, including load dispatching adjustments (such as optimizing substation load allocation to reduce overload risk), equipment maintenance arrangements (such as pre-emptive maintenance of a line to reduce the probability of failure), emergency response measures (such as rapid switching strategies in the event of sudden failures), and energy dispatching optimization (such as adjusting generation and transmission and distribution strategies to improve energy utilization). Power grid operation instructions can be directly issued to the power grid control system, dispatch center, or maintenance personnel terminals.

[0054] Specifically, if a significant conflict is detected in the previous stage, the system first enters a debate-based verification process. Through a cycle of "questioning-refutation-defense," it conducts multiple rounds of reasoning analysis on the conflicting assertions (arguments). During these multiple rounds of reasoning analysis, the system retrieves more reference documents, monitoring data, and historical cases to verify the rationality of each party's viewpoint and outputs a conclusion: if one party's argument is completely rejected, the knowledge base is updated; if both parties have valid points, a "conditionalization" strategy is adopted (e.g., using threshold A under high load conditions and threshold B under low load conditions), or the dispute is retained for subsequent optimization. After the debate results reach a consensus or partial consensus, the system moves the updated knowledge base into the counterfactual reasoning stage. In this stage, for optimization questions raised by the operations and maintenance department (e.g., "Can early maintenance reduce the risk of transformer aging?"), relevant variables are identified, and multiple hypothetical scenarios are generated (e.g., "earlier maintenance by 2 hours" and "reducing the load to 85%"). Subsequently, the system evaluates the impact of each scenario based on existing rules in the updated knowledge base, including key indicators such as failure rate, downtime, and maintenance costs, and summarizes the results to form initial power documentation data. If the conclusion of counterfactual reasoning (i.e., the initial power document data) completely contradicts existing knowledge base assertions or historical cases (i.e., the power grid's preset relational structure data), or if an inexplicable conflict arises between different counterfactual scenarios, the system will treat the initial power document data as the power grid entity relational structure data, triggering argumentative verification and counterfactual reasoning again. It will analyze and assess the credibility of newly generated conflicts, and ultimately correct or update the knowledge base (i.e., update the initial power document data). After multiple rounds of iterative optimization, the system ensures the stability and consistency of the knowledge base, meaning the first difference between the initial power document data and the power grid's preset relational structure data is zero, and the final generated initial power document data is used as the target power document data.

[0055] In cases where conflicts are not severe, the system prioritizes counterfactual reasoning to optimize the power grid entity relationship structure data. Following counterfactual reasoning, it enters a debate-based verification process, gradually improving data quality and consistency through a spiral iterative approach. Specifically, the system addresses optimization questions raised by the operations and maintenance department (e.g., "If load allocation is adjusted, can equipment aging be delayed?"), identifies relevant variables (e.g., load level, operating time, ambient temperature), and generates multiple hypothetical scenarios based on existing rules in the power grid entity relationship structure data (e.g., "reducing load to 90%" or "maintaining 1 hour earlier"). The system then evaluates key indicators for each scenario (e.g., equipment failure rate, maintenance cost, downtime) and generates initial counterfactual reasoning data. However, after generating the initial counterfactual reasoning data, the system compares its conclusions with the existing knowledge base (i.e., the power grid's pre-defined relationship structure data). If potential conflicts are found (e.g., the knowledge base states "load reduction has little impact on equipment lifespan," but counterfactual analysis shows "load reduction can significantly extend equipment lifespan"), the conflict area is marked, and the system enters the debate-based verification process. The debate engine retrieves more documents, monitoring data, and historical cases to conduct multiple rounds of questioning, rebuttal, and verification of each party's arguments, ensuring the rationality of the conclusions. If one party's argument is completely refuted, the knowledge base is modified (i.e., the initial counterfactual reasoning data is revised) to obtain the initial power document data. If both sides have valid points, a conditional strategy (e.g., "load reduction has a smaller impact in low-temperature environments and a larger impact in high-temperature environments") or the dispute is retained, and the conditional strategy or retained dispute is added to the knowledge base (i.e., the initial counterfactual reasoning data is modified based on the conditional strategy or retained dispute). The modified counterfactual reasoning data is then used as the power grid entity relationship structure data, and the optimization process is repeated. In each iteration, the system adjusts the input data based on the latest optimization results and repeatedly executes counterfactual reasoning and debate-style arguments, gradually improving the accuracy and consistency of the knowledge base until the first difference between the initial power document data and the preset power grid relationship structure data is zero. Finally, the final output initial power document data is used as the target power document data to generate precise operational instructions, guiding power grid operation and maintenance, and improving the intelligence and reliability of power grid management.

[0056] The aforementioned method for generating power document data based on entity relationship extraction systematically acquires power grid operation text information, power grid construction progress information, and power grid operation regulations information. Combined with pre-set power grid relationship structure data, it constructs comprehensive power grid entity relationship structure data, achieving an accurate depiction of the power grid's operational status. Furthermore, a method combining argumentative argumentation and counterfactual reasoning dynamically analyzes the first difference value between the power grid entity relationship structure data and the pre-set relationship data, intelligently determining the initial data processing method to ensure the scientific rigor and adaptability of data processing. Subsequently, in a spiral iterative process, the interactive optimization mechanism of argumentative argumentation and counterfactual reasoning continuously improves the power grid data processing method, gradually enhancing data quality and the accuracy of information reasoning. Ultimately, it generates high-precision target power document data, which is directly used to generate power grid operation instructions. This meets the power grid operation and maintenance requirements for efficient, accurate, and intelligent data processing, realizing intelligent and automated power grid operation and maintenance, and significantly improving the accuracy of power grid dispatching, the real-time response, and the overall stability and security of operation. Meanwhile, this invention can reduce manual intervention, improve the efficiency and accuracy of power grid data processing, optimize power grid resource allocation, reduce operation and maintenance costs, enhance the power grid's adaptability to abnormal operating conditions and emergencies, and provide strong support for the intelligent upgrading of the power system.

[0057] In one exemplary embodiment, such as Figure 3 As shown, using the initial data processing method as a guiding condition, argumentative reasoning and counterfactual reasoning are invoked to perform spiral iterative processing on the power grid entity relationship structure data and the power grid preset relationship structure data to obtain the target power document data, including steps 302 to 306. Wherein:

[0058] Step 302: Conduct a debate-style argumentation between the power grid entity relationship structure data and the power grid preset relationship structure data to obtain the first argumentation relationship structure data.

[0059] Among them, the argumentation relationship structure data can be the output data obtained through the debate-style argumentation process. It describes the logical relationships between different viewpoints or assertions and their mutual influence. Debate-style argumentation usually involves questioning, refuting, and defending multiple assertions on power grid operation rules, equipment status, operation and maintenance strategies, etc., and finally forming a comprehensive set of argumentation data.

[0060] Specifically, the "first difference value" is calculated by comparing the power grid entity relationship structure data (such as power grid equipment status, load allocation, and dispatch information) with the power grid's preset relationship structure data (i.e., based on standard operating rules and preset information). The first difference value reflects the deviation between the actual power grid operation information and the preset operation information. If the first difference value is greater than the core operation and maintenance difference threshold of the target power grid (this threshold is set based on the key parameters and risk tolerance of power grid operation), it indicates that there is a significant anomaly in the current operation status of the power grid, requiring in-depth processing. Therefore, in this case, the system decides to initiate a debate-style argumentation as a preliminary data processing method to conduct detailed analysis and argumentation of the anomalies found in the power grid operation, and further confirm the nature of the problem.

[0061] After initiating the argumentative process, similar to step 208, when discrepancies arise between the power grid entity relationship structure data and the pre-defined power grid relationship structure data, the system transforms these differences into "arguments" requiring further verification, such as whether the maximum load of power transmission lines should be modified. The system will reason using evidence such as historical fault data, equipment performance test results, and expert opinions through multiple rounds of questioning and argumentation. For example, if existing rules stipulate a maximum load of 300MW for a power transmission line, but in actual operation, certain environmental factors (such as excessively high or low temperatures) may affect the line's load-bearing capacity, the system will retrieve past operating data for that line at different temperatures and evaluate it in conjunction with the performance of similar lines under extreme weather conditions. If historical data shows that despite higher temperatures, the line did not overheat or fail at a load of 350MW, the system may recommend appropriately increasing the line's load limit. However, if data shows that the line's insulation performance deteriorates under high load conditions, potentially leading to accidents, the system will recommend maintaining the original load limit or limiting the load under certain conditions. The final result is the first argument relationship structure data, such as "when the temperature is below 10℃, the line load is allowed to be increased to 350MW".

[0062] In a specific embodiment, suppose a transmission line in the target power grid frequently experiences overload. The grid's preset rules require automatic load regulation when the transmission line's load exceeds 80%. However, in practice, some maintenance personnel believe that regulation is only necessary when the load exceeds 90% to avoid over-dispatch. Yet, power grid standards specify an 80% threshold, considered the safest load limit. After identifying these two differing viewpoints, the system initiates a debate-style argumentation. The system first reviews relevant documents, historical accident cases, and equipment maintenance records, finding that most transmission lines do indeed experience a higher failure rate when the load exceeds 80%, thus 80% should be a reasonable warning line. Ultimately, the system proposes that "80% is the standard threshold, but the load can be flexibly adjusted during abnormally high load peaks to ensure system stability." Thus, the system derives the first argumentation relationship structure data.

[0063] Step 304: Perform counterfactual reasoning on the first argument relation structure data to obtain the first reasoning relation structure data.

[0064] The reasoning relation structure data can be output data obtained through counterfactual reasoning, reflecting how various parts of the power grid system respond and change under different assumptions or scenarios. Counterfactual reasoning infers the operating performance and changes in various indicators of the power grid under different assumptions.

[0065] Specifically, after obtaining the "first-level argument relationship structure data," the system further utilizes counterfactual reasoning to conduct multi-scenario hypothetical analysis of the power grid. Counterfactual reasoning mainly generates a series of "hypothetical scenarios" by adjusting variable factors in power grid operation (such as load allocation, equipment maintenance time, and environmental conditions). For example, the system might simulate "if the load is reduced by 5%" or "if transformer maintenance is carried out 2 hours earlier," and evaluate the impact of these adjustments on indicators such as power grid failure rate, equipment lifespan, and operation and maintenance costs. Based on the analysis results of these hypothetical scenarios, the system derives new "first-level argument relationship structure data." This data structure contains optimized and adjusted power grid operation rules and parameters, reflecting the potential operation and maintenance strategies of the power grid under different scenarios.

[0066] In one specific embodiment, after determining the load regulation rules for transmission lines, the system begins counterfactual reasoning to analyze the transmission line failure rate and operation and maintenance costs under different scheduling schemes. For example, scenario A assumes that regulation begins when the load reaches 75%, while scenario B assumes that 80% is still the threshold for triggering regulation. Based on historical data and simulation reasoning, the system calculates that scenario A, while mitigating the risk of failure, reduces the overall failure rate by 15% despite increasing the frequency of load regulation. Scenario B, while reducing the number of regulation cycles, has a higher failure rate and may lead to higher equipment wear and tear. Through these counterfactual reasoning results, the system generates first-order reasoning relation structure data, providing the operation and maintenance department with a feasibility analysis of different scheduling schemes and helping them select the optimal load regulation strategy.

[0067] Step 306: If the second difference value between the first inference relation structure data and the preset relation structure data of the power grid is not zero, perform spiral cyclic iterative processing on the first inference relation structure data and the preset relation structure data of the power grid to obtain the target power document data.

[0068] Specifically, after obtaining the "first inference relation structure data," the system compares it again with the preset relation structure data of the power grid, calculating a second difference value. This second difference value reflects the degree of difference between the counterfactual reasoning result and the original preset data. If the second difference value is not zero, it indicates that the conclusion generated by the counterfactual reasoning still differs from the preset model. At this point, the system enters a spiral iterative processing stage. Specifically, based on the difference between the new inference data and the original preset data, the system further refines and corrects the analysis through argumentative reasoning and counterfactual reasoning. In each iteration, the system updates the inference model and further examines the results under different scenarios until the difference value approaches zero or reaches an acceptable threshold range. Through multiple rounds of spiral iterative processing, the system ultimately generates the target power document data.

[0069] In one specific embodiment, after selecting scenario A as the load regulation strategy, the system found that the scheme still differed from the grid's preset load dispatching model, particularly in the selection of regulation timing. To further optimize, the system calculated a second difference value and found it was still non-zero, indicating a deviation between the current scheme and the grid's preset model. After entering a spiral iterative cycle, the system readjusted the regulation timing and simulated the regulation effects under different load conditions. After several rounds of adjustments, the system finally derived an accurate dispatching model: load regulation is performed in advance during high load periods, and a dynamic regulation threshold is set. This effectively reduces the failure rate and minimizes the costs associated with frequent regulation. Ultimately, this optimized data was integrated into the target power document data, providing clear and detailed operational instructions for the actual operation of the power grid.

[0070] In this embodiment, by engaging in a debating argumentation process between the power grid entity relationship structure data and the power grid's preset relationship structure data, and then obtaining the first reasoning relationship structure data through counterfactual reasoning, followed by spiral iterative processing, the accuracy and reliability of power grid operation and maintenance decisions can be effectively improved. This process, through multiple rounds of questioning and reasoning, combined with actual operating data and hypothetical scenarios, helps identify potential operational risks and optimization space, ensuring that the target power grid's operation and maintenance strategy is more in line with actual needs. The iterative optimization process continuously adjusts and improves the power grid's operating model, enabling it to more accurately adapt to different operating conditions and emergencies, thereby improving the power grid's stability, efficiency, and security. Ultimately, it generates high-quality power documentation data, supporting more scientific and efficient operation and maintenance decisions.

[0071] In one exemplary embodiment, such as Figure 4 As shown, the target power document data is obtained by spiral iterative processing of the first inference relation structure data and the power grid preset relation structure data, including steps 402 to 406. Wherein:

[0072] Step 402: If the second difference value is greater than the core operation and maintenance difference threshold, the first inference relationship structure data is used as the power grid entity relationship structure data.

[0073] Specifically, after performing counterfactual reasoning on the power grid, the second difference value between the first inference relational structure data calculated by the system and the preset relational structure data of the power grid is not zero. At this point, the system will first assess whether this difference value exceeds the set core operation and maintenance difference threshold. If it exceeds the threshold, it indicates that there is a significant difference between the current inference result and the preset relational structure data, which may affect the normal operation of the power grid. Therefore, the system regards the first inference relational structure data as the new power grid entity relational structure data, replacing the original preset data, and using it as the new data basis for further processing.

[0074] Step 404: Return to the step of performing a debate-style argumentation on the power grid entity relationship structure data and the power grid preset relationship structure data to obtain the first argumentation relationship structure data, until the second difference value is zero, and obtain the target power document data.

[0075] Specifically, once the system uses the first inference relation structure data as the new power grid entity relation structure data, the next step returns to the argumentative argumentation step. At this point, the system, based on the new entity relation structure data and the power grid's preset relation structure data, again and sequentially performs argumentative argumentation and counterfactual reasoning. First, the argumentative argumentation compares the existing knowledge base and rules, conducting multiple rounds of questioning, rebuttal, and verification of the differences between the two to ensure that the new reasoning result is consistent with the power grid's existing standards and to correct any conflicts or contradictions. The argumentative argumentation generates new first-order argument relation structure data. Then, using the new first-order argument relation structure data as input, the system uses counterfactual reasoning—that is, generating a series of hypothetical scenarios based on the existing power grid entity relation data—to explore possible changes in power grid operation under different operating conditions. For example, the system might simulate how the current carrying capacity of a transmission line changes under different loads or weather conditions.

[0076] After completing counterfactual reasoning, the system recalculates the second difference value between the updated first reasoning relational structure data and the preset relational structure data of the power grid. If the second difference value is still not zero, it means that there is a difference between the new data structure and the preset structure, requiring further optimization. At this point, the system will continue to enter the spiral iterative processing stage. Through iteration, the system continuously retrieves new data, corrects and optimizes the difference between the entity relational structure data and the preset relational structure data, and re-performs argumentative and counterfactual reasoning (or performs counterfactual reasoning and argumentative reasoning sequentially). This cyclical process continues until the second difference value drops to zero, indicating that the current output result is completely consistent with the preset data of the power grid, or has reached an acceptable optimization state. Finally, after multiple rounds of argumentative and counterfactual reasoning iterative optimization, the system obtains the target power document data.

[0077] In this embodiment, when the second difference value exceeds the core operation and maintenance difference threshold, the first inference relationship structure data is used as the power grid entity relationship structure data, and the argumentative argumentation is re-executed until the difference value approaches zero, thus achieving precise optimization of the power grid operation model. This process, through continuous iterative argumentation and reasoning, ensures that the data at each stage conforms to the preset relationship structure of the power grid, gradually eliminating potential deviations and inconsistencies, thereby refining power grid operation and maintenance decisions. It can identify and correct strategies that do not meet the actual power grid operation requirements, ensuring that the target power document data, while guaranteeing the safe, stable, and efficient operation of the power grid, fully considers different operating environments and emergencies, ultimately generating the most scientific and effective power grid operation and maintenance guidance.

[0078] In one exemplary embodiment, such as Figure 5As shown, counterfactual reasoning is the initial data processing method. Using this initial data processing method as a guiding condition, argumentative reasoning and counterfactual reasoning are invoked to perform spiral iterative processing on the power grid entity relationship structure data and the power grid preset relationship structure data, resulting in the target power document data, including steps 502 to 506. Wherein:

[0079] Step 502: Perform counterfactual reasoning on the power grid entity relationship structure data to obtain the second reasoning relationship structure data.

[0080] Specifically, the system first calculates a first difference value based on the current state of the power grid and the operational data of the target power grid. This difference value reflects the gap between the actual operation of the power grid and the preset standard information. If the first difference value is less than or equal to the set core operation and maintenance difference threshold, it indicates that the current actual operation of the power grid is consistent with the preset rules or the difference is minor, and will not have a significant impact on the operational safety and stability of the power grid. At this time, the system believes that there is no need to take an overly complex argumentation process and directly adopts counterfactual reasoning as the initial data processing method. The core of counterfactual reasoning is to simulate the performance of the power grid under different operating conditions. For example, assuming that a certain load threshold is set at 85% instead of 90%, the system will evaluate the possible consequences of this assumption on the power grid based on historical data or operating rules, including whether it will be more stable and whether the equipment will age prematurely.

[0081] In the counterfactual reasoning phase, the system generates multiple hypothetical scenarios based on initial data or the existing power grid status. These scenarios typically modify certain variables in the power grid, such as adjusting equipment operating parameters, changing load distribution, or altering maintenance cycles. The system first proposes several "what if" hypotheses by analyzing historical data, rules, and existing operating conditions. For example, assuming the load limit of a transformer is reduced from 80% to 70%, counterfactual reasoning will simulate the transformer's operation under this scenario, including changes in transformer temperature, aging, and failure probability. The system uses existing rules and data to reason and evaluate these hypothetical scenarios, calculating the potential impact of each scenario and providing possible outcomes for each. For example, one counterfactual scenario might lead to a lower failure rate, while another might increase equipment maintenance costs, ultimately resulting in the second inference relation structure data.

[0082] Step 504: Conduct a debate-style argumentation between the second reasoning relation structure data and the power grid preset relation structure data to obtain the second argumentation relation structure data.

[0083] Specifically, in the argumentative argumentation phase, the system compares newly generated data (e.g., second inference relation structure data obtained through counterfactual reasoning) with the existing preset rules or historical knowledge of the power grid, checking for conflicts or inconsistencies. If a conflict exists, the system will proceed to the argumentative argumentation process. Similar to steps 208 and 304, when the system finds differences between the newly generated power grid entity relation structure data and the existing preset power grid relation structure data, these differences are transformed into "arguments" for discussion, such as whether the temperature thresholds of certain equipment should be adjusted. The system will conduct argumentation through multiple rounds of questioning, rebuttal, and defense, retrieving relevant historical data, monitoring records, equipment manuals, expert opinions, and other evidence to verify each argument. For example, if the system proposes that the temperature of a certain substation equipment should be increased from 80°C to 85°C, the argumentative argumentation will review the historical performance and fault records of similar equipment to assess whether this temperature range is safe. If it is found that the equipment did not experience problems under similar conditions in the past and operates more efficiently when the temperature is increased, the system may modify the rules to allow adjustments to the upper limit of the temperature under specific conditions. Finally, after multiple rounds of argumentation, the second argumentation relationship structure data was obtained.

[0084] Step 506: If the third difference value between the second argument relationship structure data and the power grid preset relationship structure data is not zero, perform spiral cyclic iterative processing on the second argument relationship structure data and the power grid preset relationship structure data to obtain the target power document data.

[0085] Specifically, after the second argument relationship structure data is generated, the system calculates the difference between it and the preset relationship structure data of the power grid to obtain the third difference value. If this difference value is not zero, it indicates that there is still inconsistency or conflict between the existing data and the preset rules, requiring further optimization. In this case, the system will enter a spiral iterative process. This process is similar to multiple rounds of verification and correction. First, the system will use counterfactual reasoning to generate new hypothetical scenarios and combine them with new equipment data and operating conditions to infer the impact of possible changes on power grid operation. For example, assuming that the transformer load limit found in the second argument is still controversial, the system will try to generate more scenarios about different load thresholds through counterfactual reasoning and compare the operating results under different scenarios. The system will continuously generate new inference data, compare it with the preset data, and update the power grid entity relationship structure data in each round of verification until the third difference value approaches zero, indicating that the data has been consistent. After multiple rounds of spiral iterative processing of counterfactual reasoning and argumentative verification, the system will finally obtain the target power document data, which contains all verified power grid operation and maintenance strategies, equipment management measures, load regulation schemes, maintenance cycles, threshold settings, and other detailed information.

[0086] In this embodiment, counterfactual reasoning is performed on the power grid entity relationship structure data to generate a second reasoning relationship structure data. This data is then further verified using argumentative reasoning, and a spiral iterative process is performed when the difference value is not zero. This effectively promotes the optimization and precision of power grid operation and maintenance strategies. It can systematically consider the behavior of the power grid system under different hypothetical scenarios and continuously eliminate deviations between the model and the actual power grid operating state through multiple rounds of argumentation and adjustment. By continuously optimizing the power grid relationship structure data, power document data that meets actual needs can be generated, providing reliable decision support for the safe, stable, and efficient operation of the power grid. This ensures that the power grid operation and maintenance strategies can dynamically adapt to different operating environments and conditions, maximizing operation and maintenance efficiency and minimizing potential risks.

[0087] In one exemplary embodiment, such as Figure 6 As shown, when the second difference value between the first inference relation structure data and the preset relation structure data of the power grid is not zero, a spiral iterative process is performed on the first inference relation structure data and the preset relation structure data of the power grid to obtain the target power document data, including steps 602 to 606. Wherein:

[0088] Step 602: If the second difference value is less than or equal to the core operation and maintenance difference threshold, the first inference relationship structure data is used as the power grid entity relationship structure data.

[0089] Specifically, similar to step 402, after performing counterfactual reasoning on the power grid, the second difference value between the first reasoning relational structure data calculated by the system and the preset relational structure data of the power grid is not zero. At this time, the system will first evaluate whether the difference value exceeds the set core operation and maintenance difference threshold. If it does not exceed the threshold, it means that the actual operation of the power grid is consistent with the preset rules or the difference is small, and it will not have a significant impact on the operation safety and stability of the power grid. At this time, the system performs a processing method similar to step 502. The system believes that there is no need to take an overly complicated argumentation process and directly adopts counterfactual reasoning as the initial data processing method. That is, the system regards the first reasoning relational structure data as the new power grid entity relational structure data, replacing the original preset data, and uses it as the new data basis to further perform subsequent processing.

[0090] Step 604: Return to the step of performing counterfactual reasoning on the power grid entity relationship structure data to obtain the second reasoning relationship structure data, until the second difference value is zero, and obtain the target power document data.

[0091] Specifically, once the system uses the first inference relation structure data as the new power grid entity relation structure data, the next step returns to the counterfactual reasoning step. At this point, the system again performs counterfactual reasoning and argumentative argumentation based on the new entity relation structure data and the power grid's preset relation structure data, sequentially. This time, the order is reversed from step 404. First, using the first inference relation structure data as input, the system again uses counterfactual reasoning—that is, generating a series of hypothetical scenarios based on the existing power grid entity relation data—to explore possible changes in power grid operation under different operating conditions. For example, the system will simulate how the current carrying capacity of a transmission line changes under different loads or weather conditions, generating new second inference relation structure data through counterfactual reasoning. Then, using the new second inference relation structure data as input, argumentative argumentation compares the existing knowledge base and rules, conducting multiple rounds of questioning, rebuttal, and verification of the differences between the two to ensure that the new reasoning results are consistent with the existing standards of the power grid and to correct any conflicts or contradictions.

[0092] After completing the argumentative argumentation, a process similar to step 404 is executed, where the system recalculates the second difference value between the updated first argumentative relational structure data and the preset relational structure data of the power grid. If the second difference value is still not zero, it means that there is a difference between the new data structure and the preset structure, requiring further optimization. At this point, the system will continue into a spiral iterative processing phase. Through iteration, the system continuously retrieves new data, corrects and optimizes the difference between the entity relational structure data and the preset relational structure data, and re-performs argumentative argumentation and counterfactual reasoning (or performs counterfactual reasoning and argumentative argumentation sequentially). This cyclical process continues until the second difference value drops to zero, indicating that the current output result is completely consistent with the preset data of the power grid, or has reached an acceptable optimization state. Finally, after multiple rounds of argumentation and iterative optimization, the system obtains the target power document data.

[0093] In this embodiment, when the second difference value is less than or equal to the core operation and maintenance difference threshold, the first inference relationship structure data is used as the power grid entity relationship structure data. Through iterative optimization using counterfactual reasoning until the difference value approaches zero, the accuracy and consistency of the power grid operation and maintenance strategy can be effectively ensured. This process gradually eliminates deviations by repeatedly verifying the differences between the power grid entity relationship data and the preset relationship data, ensuring that the power grid operation model meets actual needs and operation and maintenance objectives. This optimization cycle not only improves the reliability of power grid operation and maintenance decisions but also dynamically adapts to different operating conditions, ensuring the efficient, safe, and stable operation of the power grid. Ultimately, it generates accurate target power document data, providing the operation and maintenance team with strong decision-making support and reducing potential operational risks.

[0094] In one exemplary embodiment, such as Figure 7As shown, when the third difference value between the second argument relationship structure data and the power grid preset relationship structure data is not zero, a spiral iterative process is performed on the second argument relationship structure data and the power grid preset relationship structure data to obtain the target power document data, including steps 702 to 706. Wherein:

[0095] Step 702: If the third difference value is greater than the core operation and maintenance difference threshold, the second argumentation relationship structure data shall be used as the power grid entity relationship structure data.

[0096] Specifically, similar to step 402, after performing counterfactual reasoning on the power grid, the third difference value between the second argument relationship structure data calculated by the system and the preset relationship structure data of the power grid is not zero. At this point, the system will first assess whether this difference value exceeds the set core operation and maintenance difference threshold. If it exceeds the threshold, it indicates that there is a significant difference between the current argument result and the preset relationship structure data, which may affect the normal operation of the power grid. Therefore, the system regards the second argument relationship structure data as new power grid entity relationship structure data, replacing the original preset data, and using it as a new data basis for further processing.

[0097] Step 704: Return to the step of performing counterfactual reasoning on the power grid entity relationship structure data to obtain the second reasoning relationship structure data, until the third difference value is zero, to obtain the target power document data.

[0098] Specifically, similar to step 604, once the system uses the second argument relationship structure data as the new power grid entity relationship structure data, the next step will return to the counterfactual reasoning step. At this point, the system again performs counterfactual reasoning and argumentative argumentation based on the new entity relationship structure data and the power grid's preset relationship structure data, sequentially. First, using the second argument relationship structure data as input, the system uses counterfactual reasoning—that is, generating a series of hypothetical scenarios based on existing power grid entity relationship data—to explore possible changes in power grid operation under different operating conditions. For example, the system will simulate how the current carrying capacity of a transmission line changes under different loads or weather conditions; counterfactual reasoning will generate new second argument relationship structure data. Then, using the new second argument relationship structure data as input, argumentative argumentation compares the existing knowledge base and rules, conducting multiple rounds of questioning, rebuttal, and verification of the differences between the two to ensure that the new reasoning results are consistent with the existing standards of the power grid and to correct any conflicts or contradictions.

[0099] After completing the argumentative argumentation, the system recalculates the third difference value between the updated second argumentative relationship structure data and the preset relationship structure data of the power grid. If the third difference value is still not zero, it means that there is a difference between the new data structure and the preset structure, requiring further optimization. At this point, the system will continue to enter the spiral iterative processing stage. Through iteration, the system continuously retrieves new data, corrects and optimizes the difference between the entity relationship structure data and the preset relationship structure data, and re-performs argumentative argumentation and counterfactual reasoning (or performs counterfactual reasoning and argumentative argumentation sequentially). This cyclical process continues until the third difference value drops to zero, indicating that the current output result is completely consistent with the preset data of the power grid, or has reached an acceptable optimization state. Finally, after multiple rounds of argumentation and iterative optimization, the system obtains the target power document data.

[0100] In this embodiment, when the third difference value exceeds the core operation and maintenance difference threshold, the second argumentation relationship structure data is used as the power grid entity relationship structure data. This is then iterated through counterfactual reasoning until the difference value reaches zero, ensuring accurate optimization of power grid operation and maintenance decisions in complex situations. This process effectively eliminates potential errors and inconsistencies by continuously verifying and adjusting the differences between the power grid entity data and the preset relationship structure. Through repeated reasoning and argumentation, the power grid operation model is gradually improved, ultimately forming target power document data that meets actual operational needs. This optimization iteration process can cope with changing power grid operating conditions, improve the scientific nature and adaptability of operation and maintenance decisions, thereby ensuring the efficient and safe operation of the power grid, providing more precise operational guidance for the operation and maintenance team, and reducing system failures and risks.

[0101] In one exemplary embodiment, such as Figure 8 As shown, when the third difference value between the second argument relationship structure data and the power grid preset relationship structure data is not zero, a spiral iterative process is performed on the second argument relationship structure data and the power grid preset relationship structure data to obtain the target power document data, including steps 802 to 806. Wherein:

[0102] Step 802: If the third difference value is less than or equal to the core operation and maintenance difference threshold, the second argumentation relationship structure data shall be used as the power grid entity relationship structure data.

[0103] Specifically, similar to step 602, after performing counterfactual reasoning on the power grid, the third difference value between the second argument relationship structure data calculated by the system and the preset relationship structure data of the power grid is not zero. At this time, the system will first evaluate whether the difference value exceeds the set core operation and maintenance difference threshold. If it does not exceed the threshold, it means that the actual operation of the power grid is consistent with the preset rules or the difference is small, and it will not have a significant impact on the operation safety and stability of the power grid. At this time, the system performs a processing method similar to step 502. The system believes that there is no need to take an overly complicated argumentation process and directly adopts counterfactual reasoning as the initial data processing method. That is, the system regards the second argument relationship structure data as the new power grid entity relationship structure data, replacing the original preset data, and uses it as the new data basis to further perform subsequent processing.

[0104] Step 804: Return to the step of conducting a debate-style argumentation between the power grid entity relationship structure data and the power grid preset relationship structure data to obtain the first argumentation relationship structure data, until the third difference value is zero, and obtain the target power document data.

[0105] Specifically, similar to step 404, once the system uses the second argument relationship structure data as the new power grid entity relationship structure data, the next step will return to the argumentative argumentation step. At this point, the system, based on the new entity relationship structure data and the power grid's preset relationship structure data, will again conduct argumentative argumentation and counterfactual reasoning in sequence. First, the argumentative argumentation compares the existing knowledge base and rules, conducting multiple rounds of questioning, rebuttal, and verification of the differences between the two to ensure that the new reasoning result is consistent with the existing standards of the power grid and to correct any conflicts or contradictions. The argumentative argumentation will generate new first argument relationship structure data. Then, using the new first argument relationship structure data as input data, the system will use counterfactual reasoning—that is, generating a series of hypothetical scenarios based on the existing power grid entity relationship data—to explore possible changes in power grid operation under different operating conditions. For example, the system will simulate how the current carrying capacity of a transmission line changes under different loads or different weather conditions.

[0106] After completing counterfactual reasoning, the system recalculates the third difference value between the updated first reasoning relational structure data and the preset relational structure data of the power grid. If the third difference value is still not zero, it means that there is a difference between the new data structure and the preset structure, requiring further optimization. At this point, the system will continue to enter the spiral iterative processing stage. Through iteration, the system continuously retrieves new data, corrects and optimizes the difference between the entity relational structure data and the preset relational structure data, and re-performs argumentative and counterfactual reasoning (or performs counterfactual reasoning and argumentative reasoning sequentially). This cyclical process continues until the third difference value drops to zero, indicating that the current output result is completely consistent with the preset data of the power grid, or has reached an acceptable optimization state. Finally, after multiple rounds of argumentative and counterfactual reasoning iterative optimization, the system obtains the target power document data.

[0107] In this embodiment, when the third difference value is less than or equal to the core operation and maintenance difference threshold, the second argumentation relationship structure data is used as the power grid entity relationship structure data. Through iterative optimization using argumentation, until the difference value is zero, the accuracy and consistency of power grid operation and maintenance decisions can be effectively improved. This process, through repeated argumentation and verification, gradually eliminates the deviation between the power grid entity relationship data and the preset relationship structure, thereby ensuring that the final generated power document data accurately reflects the actual operation of the power grid. Iterative optimization not only helps to make flexible adjustments in a dynamic and complex power grid environment but also allows for updating operation and maintenance strategies based on actual conditions, maximizing the stability and efficiency of the power grid, reducing potential risks and failures, and providing reliable decision support for the operation and maintenance team.

[0108] In one exemplary embodiment, such as Figure 9 As shown, counterfactual reasoning is performed on the first argument relation structure data to obtain the first reasoning relation structure data, including steps 902 to 906. Wherein:

[0109] Step 902: Based on the first argumentation relationship structure data, determine the operation and maintenance target information of the target power grid, and the operation and maintenance element information corresponding to any operation and maintenance target information.

[0110] The operation and maintenance target information can be the key objectives set by the target power grid in its daily operation and maintenance, which typically include aspects such as equipment stability, safety, economy, and failure rate. For example, the operation and maintenance target could be "reducing the failure rate of transformers" or "optimizing grid load distribution and reducing energy loss".

[0111] Among these, operation and maintenance element information can be the key variables and conditions of the target power grid operation and maintenance objectives. These elements are specific parameters or factors related to power grid operation, such as equipment temperature, current load, equipment operating time, environmental factors (such as weather and temperature), voltage, load balance, etc.

[0112] Specifically, the system first analyzes the first argument relationship structure data to identify the core operation and maintenance (O&M) objectives of the power grid. These objectives may include stable equipment operation, minimizing failure rates, minimizing power losses, and optimizing load balancing. Each O&M objective requires specific O&M elements to support it; these elements are concrete variables that affect the achievement of the O&M objective, such as equipment operating temperature, current load, maximum transmission line carrying capacity, and climatic conditions. For example, if an O&M objective is to extend the service life of equipment, then the O&M elements might include equipment operating temperature, operating time, and load level. Through the analysis of the first argument relationship structure data, the system identifies the relationships between these objectives and elements, constructing an O&M objective information system for the target power grid.

[0113] Step 904: For any operation and maintenance target information, perform counterfactual reasoning on the first argument relationship structure data based on the operation and maintenance element information to obtain the initial reasoning relationship structure data.

[0114] The initial inference relation structure data can be derived from counterfactual reasoning, reflecting the possible operational outcomes of the power grid system under certain assumptions. This data is generated by modifying existing power grid argument relation data or through hypothetical deduction. For example, assuming how transformer load changes will affect power grid stability under certain temperature conditions, the initial inference relation structure data reflects the predicted outcome of this change.

[0115] Specifically, by setting different hypothetical scenarios, the system infers how changes in certain variables will affect the operation and maintenance (O&M) objectives of the power grid. For any given O&M objective, the system performs counterfactual reasoning based on its related O&M elements. For example, if the objective is to reduce equipment failure rates, the system might counterfactually infer how much the failure rate would decrease if the equipment's operating temperature were reduced by 5°C. In this process, the system simulates changes in variables (such as temperature, load, and operating conditions) to calculate the performance of the power grid under different conditions and generates initial inference relational structure data. This data describes how O&M objectives change with variations in O&M elements under different scenarios.

[0116] Step 906: Based on the multi-objective optimization information of the target power grid, optimize each initial inference relation structure data to obtain the first inference relation structure data.

[0117] Specifically, once the initial inference relation structure data corresponding to different scenarios is obtained, the next step is multi-objective optimization. The core of this step is to weigh and optimize multiple operational and maintenance objectives (such as stability, economy, and failure rate). The system will adjust the initial inference relation data based on the multi-objective optimization information determined by the power grid's operational and maintenance objectives, such as the relative importance and weight allocation of each objective. In practice, contradictions may arise between different operational and maintenance objectives. For example, increasing the operating temperature of equipment may help improve equipment efficiency but may increase the risk of failure. The system will adjust the inference results according to the priority, constraints, and optimization directions of different objectives. Specifically, the system may use mathematical tools such as weighted average methods and Pareto optimization methods to optimize other secondary objectives while ensuring that key objectives are met. Finally, the system will output an optimized first inference relation structure data, ensuring that all operational and maintenance objectives achieve the best balance.

[0118] In this embodiment, by determining the power grid's operation and maintenance (O&M) target information based on the first argument relationship structure data, and combining it with counterfactual reasoning of the element information corresponding to each O&M target, the behavior and O&M effects of the power grid under different conditions can be accurately simulated. This helps identify and analyze key factors that may affect power grid operation, thereby providing optimized O&M strategies for the power grid. By combining the power grid's multi-objective optimization information to optimize the initial reasoning relationship data, not only can the overall O&M efficiency of the power grid be improved, but the balance and coordination between different O&M targets can also be ensured, ultimately resulting in a more accurate and efficient O&M solution. This effectively improves the adaptability and stability of the power grid under different operating scenarios, provides a scientific basis for O&M decisions, and effectively reduces potential fault risks.

[0119] Based on the same inventive concept, this application also provides an entity-relationship-based power document data generation apparatus for implementing the aforementioned power document data generation method based on entity relationship extraction. For example... Figure 10 As shown, a power document data generation device based on entity relationship extraction is provided, including: a power grid data acquisition module 1002, an entity relationship extraction module 1004, a processing order determination module 1006, and a document data generation module 1008. The solution provided by this device is similar to the solution described in the above method.

[0120] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 11 As shown. This computer device includes a processor, memory, input / output interfaces (I / O), and communication interfaces.

[0121] Those skilled in the art will understand that Figure 11 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0122] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0123] In one embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0124] In one embodiment, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and executes the computer instructions, causing the computer device to perform the steps in the above method embodiments.

[0125] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0126] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods.

[0127] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0128] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for generating power document data based on entity relation extraction, characterized in that, The method includes: Acquire the target power grid's power grid operation text information, power grid construction progress information, power grid operation regulations information, and power grid preset relationship structure data; The power grid operation text information, the power grid construction progress information, and the power grid operation regulations information are extracted to establish entity relationships, thereby obtaining power grid entity relationship structure data. Based on the first difference value between the power grid entity relationship structure data and the power grid preset relationship structure data, the initial data processing method is determined from the argumentative arguments and counterfactual reasoning of the target power grid; If the first difference value is greater than the core operation and maintenance difference threshold of the target power grid, the argumentative argument is the initial data processing method; The argumentative argumentation is performed on the power grid entity relationship structure data and the power grid preset relationship structure data to obtain the first argumentation relationship structure data; Counterfactual reasoning is performed on the first argument relationship structure data to obtain the first reasoning relationship structure data; If the second difference value between the first inference relation structure data and the preset relation structure data of the power grid is not zero, the first inference relation structure data and the preset relation structure data of the power grid are subjected to spiral cyclic iterative processing to obtain the target power document data. When the first difference value is less than or equal to the core operation and maintenance difference threshold of the target power grid, the counterfactual reasoning is the initial data processing method; Counterfactual reasoning is performed on the power grid entity relationship structure data to obtain the second reasoning relationship structure data; The second reasoning relation structure data and the preset relation structure data of the power grid are subjected to the argumentative demonstration to obtain the second argumentation relation structure data. If the third difference value between the second argument relationship structure data and the preset relationship structure data of the power grid is not zero, the second argument relationship structure data and the preset relationship structure data of the power grid are subjected to spiral cyclic iterative processing to obtain the target power document data. The target power document data is used to generate power grid operation instructions and perform operation and maintenance operations on the target power grid.

2. The method according to claim 1, characterized in that, When the second difference value between the first inference relation structure data and the preset power grid relation structure data is not zero, a spiral iterative process is performed on the first inference relation structure data and the preset power grid relation structure data to obtain target power document data, including: If the second difference value is greater than the core operation and maintenance difference threshold, the first reasoning relationship structure data will be used as the power grid entity relationship structure data. Return to the step of performing the argumentative argumentation on the power grid entity relationship structure data and the power grid preset relationship structure data to obtain the first argumentation relationship structure data, until the second difference value is zero, and obtain the target power document data.

3. The method according to claim 1, characterized in that, When the second difference value between the first inference relation structure data and the preset power grid relation structure data is not zero, a spiral iterative process is performed on the first inference relation structure data and the preset power grid relation structure data to obtain target power document data, including: If the second difference value is less than or equal to the core operation and maintenance difference threshold, the first reasoning relationship structure data will be used as the power grid entity relationship structure data. Return to the step of performing counterfactual reasoning on the power grid entity relationship structure data to obtain the second reasoning relationship structure data, until the second difference value is zero, and obtain the target power document data.

4. The method according to claim 1, characterized in that, When the third difference value between the second argument relationship structure data and the preset power grid relationship structure data is not zero, a spiral iterative process is performed on the second argument relationship structure data and the preset power grid relationship structure data to obtain the target power document data, including: If the third difference value is greater than the core operation and maintenance difference threshold, the second argumentation relationship structure data will be used as the power grid entity relationship structure data. Return to the step of performing counterfactual reasoning on the power grid entity relationship structure data to obtain the second reasoning relationship structure data, until the third difference value is zero, to obtain the target power document data.

5. The method according to claim 1, characterized in that, When the third difference value between the second argument relationship structure data and the preset power grid relationship structure data is not zero, a spiral iterative process is performed on the second argument relationship structure data and the preset power grid relationship structure data to obtain the target power document data, including: If the third difference value is less than or equal to the core operation and maintenance difference threshold, the second argumentation relationship structure data shall be used as the power grid entity relationship structure data. Return to the step of performing the argumentative argumentation on the power grid entity relationship structure data and the power grid preset relationship structure data to obtain the first argumentation relationship structure data, until the third difference value is zero, and obtain the target power document data.

6. The method according to any one of claims 2 to 5, characterized in that, The step of performing counterfactual reasoning on the first argument relationship structure data to obtain the first reasoning relationship structure data includes: Based on the first argumentation relationship structure data, determine the operation and maintenance target information of the target power grid, and the operation and maintenance element information corresponding to any one of the operation and maintenance target information; For any of the operation and maintenance target information, counterfactual reasoning is performed on the first argument relationship structure data based on each of the operation and maintenance element information to obtain initial reasoning relationship structure data; Based on the multi-objective optimization information of the target power grid, the initial inference relation structure data are optimized to obtain the first inference relation structure data.

7. A power document data generation device based on entity relationship extraction, characterized in that, The device includes: The power grid data acquisition module is used to acquire power grid operation text information, power grid construction progress information, power grid operation regulations information, and power grid preset relationship structure data of the target power grid; The entity relationship extraction module is used to extract the power grid operation text information, the power grid construction progress information, and the power grid operation regulations information to obtain power grid entity relationship structure data. The processing order determination module is used to determine the initial data processing method from the argumentative arguments and counterfactual reasoning of the target power grid based on the first difference value between the power grid entity relationship structure data and the power grid preset relationship structure data; If the first difference value is greater than the core operation and maintenance difference threshold of the target power grid, the argumentative argument is the initial data processing method; The argumentative argumentation is performed on the power grid entity relationship structure data and the power grid preset relationship structure data to obtain the first argumentation relationship structure data; Counterfactual reasoning is performed on the first argument relationship structure data to obtain the first reasoning relationship structure data; If the second difference value between the first inference relation structure data and the preset relation structure data of the power grid is not zero, the first inference relation structure data and the preset relation structure data of the power grid are subjected to spiral cyclic iterative processing to obtain the target power document data. When the first difference value is less than or equal to the core operation and maintenance difference threshold of the target power grid, the counterfactual reasoning is the initial data processing method; Counterfactual reasoning is performed on the power grid entity relationship structure data to obtain the second reasoning relationship structure data; The second reasoning relation structure data and the preset relation structure data of the power grid are subjected to the argumentative demonstration to obtain the second argumentation relation structure data. If the third difference value between the second argument relationship structure data and the preset relationship structure data of the power grid is not zero, the second argument relationship structure data and the preset relationship structure data of the power grid are subjected to spiral cyclic iterative processing to obtain the target power document data. The target power document data is used to generate power grid operation instructions and perform operation and maintenance operations on the target power grid.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

Images