Text self-recognition spot market intervention reason checking method and device and medium

Abstract

The invention discloses a text self-recognition spot market intervention reason checking method and device and a medium, and the method comprises the steps: recognizing a power grid section and calculating features, and generating a section processing wide table in combination with power grid load information; and constructing a standard section dictionary containing a section name and a power limit value based on the wide table, retrieving a first record containing stable section constraint in the power grid manual intervention wide table, extracting a section description field of an intervention reason, matching the field with the standard section dictionary, and generating a check result wide table containing an intervention record compliance judgment result. According to the text self-recognition spot market intervention reason checking method and device and the medium, the section processing wide table is generated by automatically integrating power grid data, the unified standard section dictionary is constructed, section description in the manual intervention wide table is automatically matched, manual tedious operation can be replaced, efficiency can be improved, and the labor intensity of workers is reduced. The problems of low efficiency and low accuracy of manual checking of intervention reasons of the electric power spot market can be solved.

Description

Technical Field

[0001] This invention relates to the field of power system spot market operation and management, and in particular to a text-based self-identification method, device and medium for verifying the reasons for intervention in the spot market. Background Technology

[0002] The current power system spot market has entered a long period of stable operation. Unit output and electricity prices are cleared uniformly by the system to ensure the principles of "fairness, impartiality, and openness." Under this background, dispatchers should minimize market intervention. However, the widespread integration of power generation, grid, load, and storage leads to large load fluctuations. Coupled with immature load forecasting technology, power plant output imbalances can easily cause equipment overload and capacity exceeding limits. In such cases, dispatchers need to intervene promptly and disclose the relevant reasons to power plants, trading centers, and energy bureaus. Excessive / inappropriate intervention will directly result in losses for power plants, while missed / untimely intervention will affect the safety and stability of the power grid. Therefore, it is urgent to automatically review the reasons for intervention, promptly correct inappropriate interventions, and self-check for missed interventions. Currently, dispatching work still mainly relies on dispatching professionals to manually verify operational data, compare transaction records, and review intervention ledgers, combining professional experience with visual verification and comprehensive judgment.

[0003] However, the existing scheduling method, which relies on manual visual operation for identification, has obvious shortcomings. It requires dispatchers to independently complete the search for potential overloaded equipment and sections, the retrieval of overload time periods, and the word-by-word comparison of intervention reasons to draw conclusions. The whole process is cumbersome and time-consuming. Moreover, during peak load periods, manual processing is not only inefficient, but also prone to omissions or misjudgments due to factors such as staff fatigue and differences in experience, making it difficult to guarantee the efficiency and accuracy of the verification work. Summary of the Invention

[0004] This invention provides a text-based self-identification method, device, and medium for verifying the causes of intervention in the electricity spot market, in order to solve the problems of low efficiency and low accuracy in manual verification of the causes of intervention in the electricity spot market.

[0005] To achieve the above objectives, this invention provides a text-based self-identification method for verifying the reasons for intervention in the spot market, comprising: Obtain basic data about the power grid; Based on the aforementioned basic data, the power grid cross sections are identified and their features are calculated. Combined with the load information of the power grid, a wide table for cross section processing, including power grid clustering result fields, is generated. Based on the cross-section processing wide table, a standard cross-section dictionary containing cross-section names and power limits is constructed. The first record containing stable cross-section constraints in the power grid's manual intervention wide table is retrieved, and the cross-section description field of the intervention reason in the first record is extracted. By matching the cross-section description field with the standard cross-section dictionary, a verification result wide table containing the compliance judgment result of the intervention record is generated.

[0006] This invention integrates basic power grid data, identifies cross-sections, calculates features, and generates a wide table for cross-section processing containing clustering result fields by combining load information. This replaces the tedious process of manually organizing fragmented multi-source data, reduces repetitive work, and directly improves verification efficiency. Furthermore, the standardized cross-section identification, feature calculation, and clustering classification avoid information distortion caused by data chaos and subjective judgment bias in manual processing, providing standardized and consistent basic data support for subsequent intervention cause verification, thereby reducing the error rate and improving accuracy. By constructing a standard cross-section dictionary containing cross-section names and power limits as a unified judgment benchmark, the process of manually searching for scattered standards and interpreting them is replaced. At the same time, the system automatically retrieves cross-section descriptions from manual intervention records and performs accurate matching. This not only saves the tedious operation of manual comparison to improve efficiency, but also completely avoids misjudgment problems caused by inconsistent understanding of stable cross-section constraint standards, differences in individual subjective judgment, and human oversights in manual verification through fixed and unified matching standards and automated execution logic. It ensures the consistency and accuracy of verification results from multiple aspects such as judgment standards and execution process, thereby effectively solving the core pain point of low accuracy of manual verification due to intervention in the electricity spot market.

[0007] Compared to existing technologies, this invention automatically integrates power grid data to generate a cross-section processing wide table, constructs a unified standard cross-section dictionary, and automatically matches the cross-section descriptions in the manual intervention wide table. This not only improves efficiency by replacing tedious manual operations with automation, but also avoids subjective bias and differences in standard understanding through standardized processing and unified benchmarks. Therefore, it can solve the problems of low efficiency and low accuracy of manual verification for intervention reasons in the electricity spot market.

[0008] As a preferred embodiment, based on the aforementioned basic data, power grid sections are identified and their features are calculated. Combined with the power grid load information, a wide table for section processing, including power grid clustering result fields, is generated. Specifically: Based on the aforementioned basic data, the power grid cross sections are identified and their features are calculated using the secondary simulated annealing optimization spectrum clustering algorithm, resulting in an initial cross section calculation wide table. Based on the preset standard naming rules for cross-sections, the cross-sections in the initial cross-section calculation wide table are given standard names, and combined with the load information of the power grid, the cross-section processing wide table including the power grid clustering result field is generated.

[0009] This preferred solution achieves accurate identification and feature calculation of power grid sections through a secondary simulated annealing optimization spectral clustering algorithm, and generates a wide table for section processing by combining preset standard naming rules and load information. Furthermore, the spectral clustering algorithm improves the accuracy and robustness of section identification, better reflecting the actual operating characteristics of the power grid; in addition, standardized naming solves the problem of inconsistent section names, facilitating subsequent data association and retrieval; simultaneously, integrating load information ensures that the wide table for section processing includes both the characteristics of the sections themselves and the associated operating conditions, providing a structured and highly available data foundation for subsequent standard section dictionary construction and intervention cause verification, reducing the tediousness of manual processing and improving the quality of pre-processing data in the verification process.

[0010] As a preferred approach, based on the aforementioned basic data, the power grid cross-sections are identified and their features are calculated using a secondary simulated annealing optimization spectral clustering algorithm to obtain an initial cross-section calculation wide table, specifically: Based on the aforementioned basic data, a weighted adjacency matrix with power flow values ​​as weights is constructed. Combined with preset cluster centers, the first round of spectral clustering is performed using a simulated annealing optimized spectral clustering algorithm to obtain the first round of cross-sectional information. Within the primary category corresponding to the first round of spectral clustering, a second round of spectral clustering is performed with the substation busbar directly connected to the target main transformer node as the center to obtain the second round of cross-sectional information. Within the secondary category corresponding to the second round of spectrum clustering, branch breakage test is performed. If the breakage of a branch causes other lines to overload, the corresponding branch and the overloaded line constitute a local section, and the final round section information is obtained. The first-round cross-section information, the second-round cross-section information, and the last-round cross-section information are integrated to form the initial cross-section calculation wide table containing all heavy overload cross-section information.

[0011] In this preferred scheme, the first round of clustering captures global cross-sectional features based on a weighted adjacency matrix, the second round of clustering focuses on the directly connected area of ​​the main transformer node to achieve local refinement, and the final round of branch circuit test accurately identifies local cross-sections experiencing cascading overloads caused by branch circuit disconnection. This hierarchical identification method avoids the omission of complex power grid cross-sections by a single clustering algorithm and ensures that all types of heavily overloaded cross-sections are included in the calculation table, providing comprehensive and complete basic data for subsequent cross-section processing and intervention cause verification, thus enhancing the reliability of the verification results.

[0012] As a preferred embodiment, the cross-sections in the initial cross-section calculation wide table are given standard names based on preset cross-section standard naming rules, and combined with the load information of the power grid, the cross-section processing wide table including the power grid clustering result field is generated, specifically: According to the preset cross-section standard naming rules, add the cross-section name to the initial cross-section calculation width table to obtain the final cross-section calculation width table. The final cross-section calculation wide table, the load information of the cross-sections to be monitored in the power grid, and the unit sensitivity data are integrated to generate the cross-section processing wide table, which includes the power grid clustering result field.

[0013] This preferred solution adds standardized names to the initial cross-section calculation table using standard cross-section naming rules, and integrates load information of the monitored cross-sections and unit sensitivity data to generate a cross-section processing table, offering multiple advantages. Standardized naming eliminates the ambiguity and diversity of cross-section names, ensuring direct comparison of cross-section information under different scenarios. Furthermore, integrating load information and unit sensitivity data ensures that the cross-section processing table not only includes basic cross-section characteristics but also correlates with operating load status and the unit's impact on the cross-section, enriching the data dimensions. This table provides multi-dimensional support for subsequent construction of a standard cross-section dictionary and precise verification of intervention reasons, reducing matching errors caused by data fragmentation and improving the consistency and efficiency of the verification process.

[0014] As a preferred option, the cross-section standard naming rule is formulated based on the type of equipment that makes up the cross-section; The types of equipment that make up the cross section include combinations of lines, combinations of main transformers, and combinations of main transformers and lines.

[0015] This preferred solution explicitly bases the standard cross-section naming rules on the type of constituent equipment. Its core value lies in establishing a unified and traceable cross-section naming standard. This equipment-type-based naming method allows the cross-section name to intuitively reflect its constituent structure, avoiding issues such as different names for the same cross-section or different cross-sections with the same name due to arbitrary naming. Simultaneously, the standardized naming rules provide a clear comparison benchmark for subsequent matching of cross-section description fields with the standard cross-section dictionary, reducing the complexity and error rate of text matching, ensuring the accurate identification of cross-section descriptions in manual intervention records, and providing a reliable foundation for determining the compliance of intervention reasons.

[0016] As a preferred embodiment, a standard cross-section dictionary containing cross-section names and power limits is constructed based on the cross-section processing wide table. The first record containing stable cross-section constraints is retrieved from the power grid's manual intervention wide table, and the cross-section description field of the intervention reason is extracted from the first record. By matching the cross-section description field with the standard cross-section dictionary, a verification result wide table containing the compliance judgment results of the intervention record is generated, specifically: The cross-section names are summarized from the cross-section processing wide table, and the standard cross-section dictionary containing the cross-section names and power limits is constructed based on the existing power limits in the cross-section processing wide table. The intervention reasons for each record in the wide table of human intervention in the power grid are retrieved, and the first record containing the classification keyword "stable section constraint" is selected. Based on the first record, fields involving cross-section description are extracted from the intervention reasons to obtain the cross-section description field; Data matching is performed based on the cross-section description field and the standard cross-section dictionary. The compliance of the intervention record is determined by combining the reasoning rules in the verification reasoning rule base, and the verification result wide table containing the compliance determination results of the intervention record is generated.

[0017] This preferred solution streamlines and standardizes the verification process for intervention reasons by constructing a standard cross-section dictionary, filtering stable cross-section constraint records, extracting cross-section description fields, and matching and determining the results. The construction of the standard cross-section dictionary provides a unified benchmark for cross-section names and power limits, resolving inconsistencies in verification criteria. Filtering records containing stable cross-section constraints precisely focuses on the core scenarios requiring verification, reducing interference from irrelevant data. The matching of cross-section description fields with the dictionary and the application of inference rules transform the textual information of manual intervention reasons into quantifiable compliance judgment results, replacing the subjectivity and inefficiency of traditional manual verification. This significantly improves verification efficiency and accuracy, providing an operational technical means for the standardized supervision of intervention behavior in the spot market.

[0018] As a preferred embodiment, data matching is performed based on the cross-section description field and the standard cross-section dictionary, and the compliance of the intervention records is determined by combining the inference rules in the verification inference rule base, generating the verification result wide table containing the compliance determination results of the intervention records, specifically as follows: By matching the overloaded section records with load rates exceeding preset values ​​in the section processing wide table with the section description fields involving intervention reasons in the manual intervention wide table, it is determined whether there is an uncontrolled overload; if it has been controlled, by checking the correlation between the time of the section occurrence in the section processing wide table and the intervention period in the manual intervention wide table, it is determined whether the control was not timely, and a first determination result is obtained. For the cross-section description field in the wide table of manual intervention that contains the keyword "stable cross-section constraint", the second judgment result is obtained by matching the standard cross-section dictionary, checking the corresponding cross-section load rate and verifying the time interval between intervention and cross-section. The judgment is then made to determine whether the record is incorrect, the intervention is non-compliant, or the intervention is excessive. By integrating the first and second judgment results, a wide table of verification results is obtained, which includes the judgment results of the compliance of intervention records.

[0019] This preferred solution achieves comprehensive and accurate assessment of intervention record compliance by subdividing judgment scenarios and combining multi-dimensional matching and verification rules. For matching overload sections with intervention records, it can promptly identify uncontrolled or lagging control issues. For stability section constraint records, dictionary matching, load factor verification, and time interval verification can accurately identify record errors, unfounded interventions, and excessive interventions. This multi-scenario coverage judgment logic ensures the completeness of compliance checks while improving the granularity of judgment results, making the verification results more instructive and helping to regulate intervention behavior in the spot market, ensuring market fairness and stable grid operation.

[0020] The present invention also provides a text self-recognition spot market intervention reason verification device, including a data module, a cross-section module and a verification module; The data module is used to acquire basic data of the power grid. The cross-section module is used to identify and calculate the features of the power grid cross-section based on the basic data, and generate a cross-section processing wide table including the power grid clustering result field by combining the load information of the power grid. The verification module is used to construct a standard cross-section dictionary containing cross-section names and power limits based on the cross-section processing wide table, retrieve the first record containing stable cross-section constraints in the power grid's manual intervention wide table, extract the cross-section description field of the intervention reason in the first record, and generate a verification result wide table containing the compliance judgment result of the intervention record by matching the cross-section description field and the standard cross-section dictionary.

[0021] The present invention also provides a storage medium storing a computer program, which is called and executed by a computer to realize the text self-recognition spot market intervention reason verification method as described above.

[0022] The present invention also provides a computer program product, including a computer program or instructions, which, when executed by a communication device, implements the text self-recognition spot market intervention reason verification method as described above. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating a text-based self-identification method for verifying the reasons for intervention in the spot market, provided in an embodiment of the present invention. Figure 2 This is a cross-section calculation logic diagram provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the classification reasoning rules of the verification reasoning rule base provided in the embodiments of the present invention; Figure 4 This is a flowchart of the automatic verification method for the reasons for intervention in the spot market based on text self-recognition provided in this embodiment of the invention; Figure 5 This is a framework diagram of the automatic verification system for the reasons for intervention in the spot market based on text self-recognition provided in this embodiment of the invention; Figure 6 This is a schematic diagram of the structure of a text self-recognition spot market intervention reason verification device provided in an embodiment of the present invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] In the description of this invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, features defined with "first" and "second," except for "second minor feature value," may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a number" means two or more.

[0026] The text-based self-identification method for verifying the reasons for intervention in the electricity spot market provided in this embodiment aims to solve the problems of manual verification of intervention reasons in the electricity spot market, which requires manually searching for overloaded equipment and corresponding time periods and comparing intervention reasons word by word. This is cumbersome when the number of intervention records is large during peak load periods. Furthermore, it is prone to omissions or misjudgments due to personnel fatigue and differences in experience, which may lead to damage to the grid revenue due to excessive / inappropriate intervention, or affect the stability of the grid due to missed or untimely intervention.

[0027] Example 1: Please see Figure 1 The present invention provides a text-based self-identification method for verifying the reasons for intervention in the spot market, including S1 to S3, and the specific implementation steps are as follows: S1. Obtain basic data of the power grid.

[0028] Step S1 of this embodiment of the invention is specifically as follows: Based on the SCADA cross-section interaction module, a stable data interaction is established with the SCADA system through the IEC 104 protocol. Three types of core power grid data are accurately collected every 30 minutes to obtain basic data. All data is verified, decoded, and then stored in a real-time database for unified management. The SCADA system is a power data acquisition and monitoring control system. The three core grid data categories include topology model data, power flow data, and generator sensitivity data. Topology model data is described in CIM / E format, clearly defining the connection logic and topological relationships of various grid devices. Power flow data reflects the real-time operating status of the grid, covering relevant parameters of lines and transformers at voltage levels of 220kV and above, specifically including key indicators such as voltage, phase angle, and active power. Generator sensitivity data is calculated by simulating the N-1 trip scenario of generators, focusing on analyzing whether a sudden reduction in generator output will aggravate the power flow at the cross-section. Finally, the names of generators that have a significant impact on the cross-section are selected and output, providing targeted references for verifying the cause of intervention.

[0029] Furthermore, the output rules for unit sensitivity data are as follows: For a certain section, traverse all units in the network. For a certain unit, XX Power Plant #XX Unit, simulate its tripping and sudden reduction in output, and count the load increase of the section. The sensitivity of the unit = the load increase of the section / the sudden reduction in output of the unit. If the sensitivity > 0, the unit name is output in the wide table record of that section: XX Power Plant #XX Unit.

[0030] S2. Based on the basic data, identify and calculate the features of the power grid sections, and combine the power grid load information to generate a wide table for section processing that includes power grid clustering result fields.

[0031] In this embodiment of the invention, step S2 includes S2.1 to S2.2; wherein, S2.1 is the process of establishing an initial cross-section calculation wide table based on the simulated annealing optimized spectral clustering algorithm, and S2.2 is the process of adjusting and integrating the initial cross-section calculation wide table and other data to generate a cross-section processing wide table, specifically as follows: S2.1 The dynamic calculation module for cross-sections based on secondary annealing optimized spectral clustering constructs a weighted adjacency matrix with power flow values ​​as weights based on the basic data. Combined with preset cluster centers, the module performs the first round of spectral clustering based on the dynamic calculation method for cross-sections based on secondary simulated annealing optimized spectral clustering algorithm (simulated annealing optimized spectral clustering algorithm) which takes into account prior knowledge and is oriented towards 220kV power grids, and obtains the first round of cross-section information.

[0032] Within the primary category corresponding to the first round of spectral clustering, the substation busbar directly connected to the target main transformer node is used as the new cluster center. The spectral clustering process is repeated to carry out the second round of spectral clustering, further refining the cross-sectional division and obtaining the second round of cross-sectional information. Within the secondary category corresponding to the second round of spectral clustering, branch-by-branch test is performed. If the opening of a branch causes overload in other lines, the corresponding branch and the overloaded line are determined to constitute a local section, and the relevant information is recorded as the final section information. By integrating the first, second, and final cross-section information, an initial wide cross-section calculation table containing all heavy overload cross-section information is formed. The core fields of this wide table cover "equipment that constitutes the cross-section" and "reference limit (MW)".

[0033] Among them, the method of spectral clustering based on the cross-section dynamic calculation method of the second-order simulated annealing optimized spectral clustering algorithm for 220kV power grid, which takes into account prior knowledge, is as follows: Step 1: First clustering based on simulated annealing optimized spectral clustering algorithm: ① First, construct a weighted adjacency matrix. W : Based on fundamental data, for a power grid structure containing two 500kV main transformer nodes (A and B), all other nodes being 220kV bus nodes, and edges being 220kV lines, a weighted adjacency matrix is ​​constructed using line power flow values ​​as weights. W ; W Depend on composition, i , j These represent the node numbers respectively. It is the weight of the association between nodes, which is related to The function is represented using the Gaussian kernel function: in, , For node power, These are kernel function parameters; the smaller the node power difference, the better. The larger the number, the less connectivity. , It can accurately reflect the tightness of the connection between nodes.

[0034] ② Subsequently, the core computational process of simulated annealing optimization spectral clustering is executed: The first round of spectral clustering uses nodes A and B as cluster centers and selects the normalized tangent function. The objective function is expressed as shown in Equation 1: in, The sum of edge weights between classes A and B. ; This represents the sum of the edge weights associated with class A nodes. ; The computational logic is the same Similarly, it can be seen In fact Therefore, this embodiment of the invention uses a simulated annealing optimization algorithm to find the function. The optimal solution makes Minimum, then calculate Core intermediate variables ( The formula for () is shown in Equation 2: Equation 1 is used for the first round of coarse clustering, with the input of the initial branch weight matrix (basic topology, power flow correlation), the purpose of which is to quickly eliminate redundant branches, preliminarily divide potential overload clusters, and directly obtain some heavy overload sections; Equation 2 is used for the second round of fine clustering, with the input of the optimized core branch weight matrix (updating real-time power flow coupling strength, etc.), the goal of which is to obtain more refined heavy overload sections and avoid omissions.

[0035] The specific steps for simulated annealing optimization are as follows: First, set the initial temperature. T_high Termination temperature T_low The cooling coefficient α and the number of iterations L; random generation of initial solutions. σ_current (A random positive number), and calculate its corresponding energy value. E_current = NCut(σ_current) .by σ_new = σ_current + η * randn The method involves perturbing the vicinity of the current solution to generate a new solution. σ_new ;in, η For the perturbation step size, randn These are random numbers that follow a standard normal distribution.

[0036] Next, let I be an all-one vector matrix, calculate the normalized Laplacian matrix, extract the Fiedler vector corresponding to its second smallest eigenvalue, and normalize it using L2 to obtain the matrix. V ,node i The coordinates in the normalized feature space are: v i Determine the node A and B Their coordinates in the normalized feature space are respectively C A and C B Furthermore, these coordinates are calculated in advance by placing nodes A and B in the matrix. V The corresponding row vector is obtained from this.

[0037] For each node i Calculate its to C A and C B The Euclidean distance is used to divide the nodes into partitions with smaller distances, and then the Euclidean distance is calculated separately based on the partitioning results. σ_current and σ_new Energy value: ; The normalized Laplace matrix is: ③ Finally, complete the new solution judgment and the first round of cross-section determination: Determine whether the new solution is acceptable: If Directly accept new solutions and update , .like Then, based on probability Accept the new solution. If accepted, update the current solution and energy; otherwise, reject. After completion, initiate the next perturbation generation, and repeat the perturbation, calculation, and judgment process until... If it remains unchanged for 5 consecutive times, confirm. To obtain the optimal solution, we obtain the first round of clustering partitioning results, i.e., the first clustering result.

[0038] Among them, probability The calculation formula is: Step 2: Perform the first round of cross-section calculation based on the first clustering: Based on the results of the first clustering, the first round of cross-sectional calculation is performed. The tie lines directly connected between different categories (i.e., tie lines between primary categories) are retrieved. When an N-1 fault occurs between different 500kV main transformers, between a 500kV main transformer and a tie line of the primary category, or between tie lines of the primary category, it is calculated whether there are overloaded devices. If there are overloaded devices, the names of the two corresponding devices, the power calculation element, the power data, and the calculation method (specifically, the summation of the power data) are obtained. These are then summarized to form a cross-sectional information, thus obtaining the first round of cross-sectional information.

[0039] Step 3, Secondary Clustering and Second Round Cross-Section Calculation: Within each primary category, based on prior knowledge, spectral clustering is performed again using the 220kV substation busbar directly connected to the 500kV main transformer node as the new cluster center, resulting in a secondary clustering result. Within each primary category, it is checked whether an N-1 fault occurs between the following equipment, constituting a cross-section: between secondary category tie lines, and between the connection line between the 500kV main transformer and the directly connected 220kV substation and the secondary category tie line. Based on this, the second round of cross-section calculation is completed, obtaining the second round of cross-section information. Here, "each primary category" refers to the independent category divided from the primary clustering result.

[0040] Step 4, Calculation of the final wheel cross-section: Branch line failure tests are conducted within each secondary category. If the failure of a branch line causes overload on other lines within the category, then that branch line and the overloaded line together constitute a local cross-section, and relevant information is recorded as the final cross-section information. This completes the comprehensive calculation and identification of all heavily overloaded cross-sections from the main power grid to its branches. Here, "each secondary category" refers to the independent category defined in the secondary clustering results.

[0041] For an explanation of the embodiments of the present invention, please refer to [link / reference]. Figure 2 , Figure 2The cross-section calculation logic diagram provided for the embodiments of the present invention illustrates the specific method of spectral clustering based on the dynamic calculation method of cross-section based on the second-order simulated annealing optimized spectral clustering algorithm for 220kV power grid, which takes into account prior knowledge.

[0042] In this embodiment S2.1, the first round of clustering captures global cross-sectional features based on a weighted adjacency matrix, the second round of clustering focuses on the directly connected area of ​​the main transformer node to achieve local refinement, and the final round of branch circuit test accurately identifies local cross-sections experiencing cascading overloads caused by branch circuit disconnection. This hierarchical identification method avoids the omission of complex power grid cross-sections by a single clustering algorithm and ensures that all types of heavily overloaded cross-sections are included in the calculation wide table, providing comprehensive and complete basic data for subsequent cross-section processing and intervention cause verification, thus enhancing the reliability of the verification results.

[0043] S2.2 Based on the scanning section naming module, add a "section name" column to the initial section calculation width table, and perform standard naming of the section according to the preset section standard naming rules to obtain the final section calculation width table and store it in the historical section library. Based on the historical section database, the final section calculation wide table, the load information of the sections to be monitored in the power grid, and the unit sensitivity data are integrated to generate a section processing wide table including power grid clustering result fields, and this table is output to the intervention cause analysis module. Specifically, the historical section database serves two functions: firstly, it performs data aggregation, integrating historical scan section data accumulated from past power grid operations with the currently calculated section data to form a comprehensive section data aggregation wide table covering both historical and real-time dimensions; secondly, it feeds back key operational information of the sections to be monitored to the SCADA system in real time, including maximum load value (MW), load rate (%), and peak load occurrence time and duration (minutes). After receiving the section information to be monitored, the SCADA system's section interaction module calculates and generates unit sensitivity data by simulating the N-1 unit trip scenario, and filters out the names of units that have a significant impact on the section; then, it sends this unit sensitivity data, along with the maximum load value, load rate, occurrence time, and duration data of the section to be monitored, back to the historical section database to complete the data supplementation and update of the historical section database. After integrating all data, the historical cross-section database outputs a complete wide table for cross-section processing. This wide table explicitly includes six core fields: "Cross-section Name," "Reference Limit (MW)," "Maximum Load (MW)," "Load Rate (%)," "Occurrence Time," and "Duration (minutes)." The scanned cross-section naming module and the historical cross-section database belong to the cross-section storage module. This module is used for the standardized storage and management of the cross-section calculation wide table and for providing monitoring cross-section information to the SCADA system. Furthermore, the SCADA cross-section interaction module, the secondary annealing optimized spectrum clustering cross-section dynamic calculation module, and the cross-section storage module belong to the cross-section scanning module. This module is used to scan, calculate, and uniformly store power grid cross-sections, providing cross-section information support for intervention cause analysis.

[0044] The cross-section standard naming rules are based on the types of equipment that make up the cross-section. These types include combinations of lines, combinations of main transformers, and combinations of main transformers and lines. Specifically: Rule 1: The "section composition equipment" is "ab-line A" and "ab-line B", and is named "ab-line A-B"; Rule 2: The "section composition equipment" is "ab line A" and "cd line B", and is named "ab line A + cd line B".

[0045] Rule 3: The "section-composition equipment" consists of "Station A #1 main transformer", "Station A #2 main transformer"... "Station A #x main transformer (#x main transformer is the last main transformer in Station A)", and is named "Station A load". Rule 4: The "section-composition equipment" consists of "station a#y main transformer (#y main transformer becomes any main transformer in station a)" and "station b#z main transformer (#z main transformer becomes any main transformer in station b)", and is named "station a#y transformer + station b#z transformer". Rule 5: The “section-composition equipment” is “station a#y main transformer (#y main transformer is any main transformer of station a)” and “cd A (B) line”, named “station a#y transformer + cd A (B) line”.

[0046] In this embodiment, S2.2, a standardized name is added to the initial cross-section calculation wide table using standard cross-section naming rules. This, combined with load information from the monitored cross-sections and unit sensitivity data, generates a cross-section processing wide table, offering multiple advantages. Standardized naming eliminates the ambiguity and diversity of cross-section names, ensuring direct comparison of cross-section information across different scenarios. Furthermore, integrating load information and unit sensitivity data ensures that the cross-section processing wide table not only includes basic cross-section characteristics but also correlates with operating load status and the unit's influence on the cross-section, enriching the data dimensions. This wide table provides multi-dimensional support for subsequent construction of a standard cross-section dictionary and precise verification of intervention reasons, reducing matching errors caused by data fragmentation and improving the consistency and efficiency of the verification process. Furthermore, the standard naming rules for cross-sections are clearly based on the type of constituent equipment. Their core value lies in establishing a unified and traceable naming standard for cross-sections. This equipment-type-based naming method allows cross-section names to intuitively reflect their constituent structures, avoiding issues such as different names for the same cross-section or different cross-sections with the same name due to arbitrary naming. Simultaneously, the standardized naming rules provide a clear comparison benchmark for subsequent matching of cross-section description fields with the standard cross-section dictionary, reducing the complexity and error rate of text matching, ensuring accurate identification of cross-section descriptions in manual intervention records, and providing a reliable foundation for determining the compliance of intervention reasons.

[0047] In this embodiment, S2 achieves accurate identification and feature calculation of power grid sections through a secondary simulated annealing optimization spectral clustering algorithm, and generates a wide table for section processing by combining preset standard naming rules and load information. Furthermore, the spectral clustering algorithm improves the accuracy and robustness of section identification, better reflecting the actual operating characteristics of the power grid. In addition, standardized naming solves the problem of inconsistent section names, facilitating subsequent data association and retrieval. Simultaneously, integrating load information ensures that the wide table for section processing includes both the characteristics of the sections themselves and the associated operating conditions, providing a structured and highly available data foundation for subsequent standard section dictionary construction and intervention cause verification, reducing the tediousness of manual processing and improving the quality of pre-processing data in the verification process.

[0048] S3. Based on the cross-section processing wide table, construct a standard cross-section dictionary containing cross-section names and power limits. Retrieve the first record containing stable cross-section constraints in the power grid's manual intervention wide table, and extract the cross-section description field of the intervention reason from the first record. By matching the cross-section description field with the standard cross-section dictionary, generate a verification result wide table containing the compliance judgment results of the intervention record.

[0049] Step S3 in this embodiment of the invention includes S3.1 to S3.2; wherein, S3.1 is the process of establishing a standard cross-section dictionary based on the cross-section processing wide table and obtaining the cross-section description field based on the manual intervention wide table, and S3.2 is the process of matching data based on the cross-section description field and the standard cross-section dictionary to obtain a verification result wide table, specifically as follows: S3.1 Based on the SCADA intervention record module, obtain the unit manual intervention wide table from the SCADA system of the power grid, retrieve the intervention reasons for each record in the manual intervention wide table, and filter out the first record containing the target keywords such as "stable section constraint" for classification. The manual intervention wide table includes "power plant name", "unit name", "intervention period", "planned output (MW) before intervention", "planned output (MW) after intervention", "maximum output of cogeneration (MW)", "minimum output of cogeneration (MW)", "effective", "intervention reason", "intervener", and "intervention time".

[0050] The control text intelligent verification module receives the manual intervention wide table and the cross-section processing wide table, summarizes the cross-section names from the cross-section processing wide table, and constructs a standard cross-section dictionary containing cross-section names and power limits based on the existing power limits in the cross-section processing wide table. Based on the first record, fields related to cross-section description are extracted from "Intervention Reasons" using regular expressions to obtain the cross-section description field. The rule for obtaining the cross-section description based on regular expressions is: control(.*?)(?:cross-section|load), meaning the cross-section control specification record is "control" + ×× + "cross-section" / "load", ultimately obtaining the ×× field as the cross-section description.

[0051] It should be noted that when the dispatcher manually intervenes in the unit output curve, the spot system will first require the dispatcher to forcibly select the intervention type. These type names will be simultaneously written into the "Intervention Reason" field to form the target keywords for subsequent retrieval, namely the "target keywords" mentioned in the above scheme, which specifically include five categories: "stability section constraint", "cooperating with power plant work", "cooperating with the upper-level dispatch system for peak shaving", "other", and "accident handling". From the perspective of practical application scenarios, there are significant differences in the intervention records corresponding to these five types of keywords: among them, the intervention of "stable section constraint" accounts for as high as 90%, which not only has a fixed record format, but the system can also automatically trace the correctness of the record; the other four types account for only 0~10%, and their interventions are mostly initiated based on the daily work plan, temporary notices or sudden accident situations. The "intervention reason" has no fixed text format, and compliance needs to be confirmed by manual review later. Furthermore, interventions involving "stable cross-section constraints" have clear recording standards: the "intervention reason" must clearly reflect the logic of "control + specific object + cross-section / load," that is, clearly record "to constrain the stability of the ×× cross-section (if it is a cross-section composed of a single 220kV substation main transformer, it is often described as the ×× substation load), and manually intervene in the output curve of the ×× unit," ensuring that the relationship between the intervention target and the cross-section is traceable. Therefore, when operating, dispatchers must not only accurately select the "stable cross-section constraint" intervention type, but also strictly follow the above fixed format to fill in the "intervention reason," ensuring the compliance and traceability of the intervention record.

[0052] S3.2 Based on the text-based intelligent verification module, according to the classification and reasoning rules of the verification reasoning rule base, the overloaded section records with load rates exceeding preset values ​​in the section processing wide table are matched with the section description fields involving intervention reasons in the manual intervention wide table to determine whether there is an uncontrolled overload; if it has been controlled, the correlation between the occurrence time of the section in the section processing wide table and the intervention period in the manual intervention wide table is checked to determine whether the control is not timely, and the first judgment result is obtained. For the cross-section description field in the wide table of manual intervention that contains the keyword "stable cross-section constraint", the second judgment result is obtained by matching the standard cross-section dictionary, checking the corresponding cross-section load rate and verifying the time interval between intervention and cross-section. The judgment is then made to determine whether the record is incorrect, the intervention is non-compliant, or the intervention is excessive. The system integrates the first and second judgment results, combining the verification classification results such as "overload not controlled, overload control not timely, incorrect section name / intervention reason record, non-compliant intervention section, excessive intervention at the section, and compliant intervention record." These results are then output according to preset priorities and sorting, ultimately yielding a wide table of verification results containing the compliance judgment results of the intervention records. The SCADA intervention record module and the text intelligent verification module belong to the intervention reason analysis module. The intervention reason analysis module is used to automatically identify and verify the compliance of the intervention reason text based on the data provided by the section scanning module.

[0053] The specific reasoning rules categorized in the verification reasoning rule base are as follows: Rule 1: Whether it belongs to "overload section not controlled": For a record in Table 1 whose "load rate (%)" exceeds 100%, does its "section name" not match the "section description field" extracted from the "intervention reason" of any record in Table 2? If yes, the classification is completed; otherwise, after locating the corresponding record in Table 2, proceed to Rule 2. Rule 2: Does it fall under "untimely control of overload section"? For the record in Table ①, is its "occurrence time" not within the "intervention period" + 10 minutes of the corresponding record in Table ②? If yes, the classification is completed; otherwise, the record is not output. After all records in Table ① have completed Rule 1 to Rule 2, proceed to Rule 3. Rule 3: Whether it belongs to "Records pending manual review": Does Table ② "Reason for intervention" contain the classification keyword "stable cross-section constraint"? If not, complete the classification; if yes, proceed to Rule 4. Rule 4: Is it a case of “Incorrect record of section name / intervention reason”? For a record in Table ② whose “intervention reason” contains “stable section control”, is the “section description field” extracted from the “intervention reason” not matched in the standard section dictionary? If yes, the classification is completed; otherwise, after locating the corresponding record in Table ①, proceed to Rule 5. Rule 5: Whether it belongs to "non-compliant intervention section": Does the corresponding record in Table 1 for the record in Table 2 not meet the "load rate (%)" ≥ 90? If yes, the classification is completed; otherwise, proceed to Rule 6. Rule 6: Determine if it belongs to "excessive intervention at the cross section": Does the "intervention time" of the record in Table ② not meet the requirement of being within the range of 1 hour before to 2 hours after the "occurrence time" of the corresponding record in Table ①? If yes, the classification is completed; otherwise, proceed to Rule 7. Rule 7: Determine if the intervention record is "compliant"; after all records in Table ② have completed Rules 3 to 7, proceed to Rule 8; Rule 8: Output a wide table of verification results from top to bottom based on the following verification categories: "Overload section not controlled", "Overload section not controlled in a timely manner", "Intervention section non-compliant", "Section name / intervention reason recorded incorrectly", "Excessive intervention in section", "Records pending manual review".

[0054] Table ① is the cross-section treatment table, and Table ② is the manual intervention table.

[0055] For an explanation of the embodiments of the present invention, please refer to [link / reference]. Figure 3-5 , Figure 3 The schematic diagram of the classification reasoning rules of the verification reasoning rule base provided in the embodiments of the present invention illustrates the specific logic of the classification reasoning rules of the verification reasoning rule base.

[0056] Figure 4The flowchart of the automatic verification method for the cause of intervention in the spot market based on text self-recognition provided in the embodiments of the present invention illustrates the complete execution process of the automatic verification of the cause of intervention in the spot market based on text self-recognition in the embodiments of the present invention. The method obtains the topology map, power flow data and manual intervention wide table from the SCADA system, realizes the dynamic scanning of the cross section, the standard naming of the cross section and the calculation of the unit sensitivity through the second simulated annealing optimization spectral clustering algorithm, and then classifies the data through the text intelligent verification module, and finally outputs the verification result wide table.

[0057] Figure 5 The framework diagram of the automatic verification system for the cause of intervention in the spot market based on text self-identification provided in the embodiments of the present invention illustrates the data interaction relationship between the modules in the embodiments of the present invention.

[0058] This embodiment, S3.2, achieves comprehensive and accurate judgment of the compliance of intervention records by subdividing judgment scenarios and combining multi-dimensional matching and verification rules. For matching overload sections with intervention records, uncontrolled or delayed control issues can be detected in a timely manner. For stability section constraint records, dictionary matching, load factor verification, and time interval verification can accurately identify record errors, unfounded interventions, and excessive interventions. This multi-scenario coverage judgment logic ensures the completeness of compliance checks while improving the granularity of judgment results, making the verification results more instructive and helping to regulate intervention behavior in the spot market, ensuring market fairness and stable grid operation.

[0059] This embodiment S3 achieves a streamlined and standardized process for verifying intervention causes by constructing a standard cross-section dictionary, filtering stable cross-section constraint records, extracting cross-section description fields, and matching and determining the results. The construction of the standard cross-section dictionary provides a unified benchmark for cross-section names and power limits, resolving the issue of inconsistent verification criteria. Filtering records containing stable cross-section constraints precisely focuses on the core scenarios requiring verification, reducing interference from irrelevant data. The matching of cross-section description fields with the dictionary and the application of inference rules transform the textual information of manual intervention causes into quantifiable compliance judgment results, replacing the subjectivity and inefficiency of traditional manual verification. This significantly improves verification efficiency and accuracy, providing an operable technical means for the standardized supervision of intervention behavior in the spot market.

[0060] As an optional implementation, in one embodiment of the invention, during the data interaction process based on the SCADA cross-section interaction module, its four core units can provide stable guarantees for data acquisition and processing: Data acquisition unit: Establishes a communication connection with the SCADA system via the IEC 104 protocol, with an acquisition cycle of 5 minutes / time, ensuring an acquisition success rate of no less than 99.9%; Data parsing unit: Employs a CIM / E format parser, supports the IEEE CIM 17 standard, and controls the parsing error rate to be less than 0.1%; Data verification unit: A 32-bit check code is generated using the CRC-32 check algorithm, and the integrity of the collected data is ensured by comparing the check codes. Data caching unit: Equipped with 8GB DDR4 memory as temporary storage support, it can support a maximum data caching capacity of 10,000 records per second to avoid data processing congestion.

[0061] Overall, the embodiments of the present invention have the following beneficial effects: This invention integrates basic power grid data, identifies cross-sections, calculates features, and generates a wide table for cross-section processing containing clustering result fields by combining load information. This replaces the tedious process of manually organizing fragmented multi-source data, reduces repetitive work, and directly improves verification efficiency. Furthermore, the standardized cross-section identification, feature calculation, and clustering classification avoid information distortion caused by data chaos and subjective judgment bias in manual processing, providing standardized and consistent basic data support for subsequent intervention cause verification, thereby reducing the error rate and improving accuracy. By constructing a standard cross-section dictionary containing cross-section names and power limits as a unified judgment benchmark, the process of manually searching for scattered standards and interpreting them is replaced. At the same time, the cross-section descriptions in the manual intervention records are automatically retrieved and accurately matched. This not only saves the tedious operation of manual comparison to improve efficiency, but also completely avoids the misjudgment problems caused by inconsistent understanding of stable cross-section constraint standards, differences in individual subjective judgment, and human comparison negligence through fixed and unified matching standards and automated execution logic. It ensures the consistency and accuracy of the verification results from multiple aspects such as judgment standards and execution process, thereby effectively solving the core pain point of low accuracy of manual verification due to intervention in the electricity spot market. In summary, this invention, through the design of a secondary simulated annealing optimized spectral clustering algorithm, comprehensively and intelligently extracts potentially overloaded sections, avoiding omissions or misidentifications of heavily overloaded sections due to insufficient dispatcher experience. This provides a complete reasoning basis for the verification of intervention texts. Subsequently, by designing a verification reasoning rule base, the entire process of automatic verification and output of intervention records is realized. This not only significantly reduces the reliance on manual intervention but also improves the processing efficiency of related businesses in the spot market, while ensuring the accuracy of market intervention information disclosure.

[0062] Example 2: Please see Figure 6 The present invention provides a text self-identification spot market intervention reason verification device, including a data module 10, a cross-section module 20 and a verification module 30; Among them, data module 10 is used to acquire basic data of the power grid; The cross-section module 20 is used to identify and calculate the features of power grid cross-sections based on basic data, and generate a cross-section processing wide table including power grid clustering result fields by combining the power grid load information. The verification module 30 is used to construct a standard cross-section dictionary containing cross-section names and power limits based on the cross-section processing wide table, retrieve the first record containing stable cross-section constraints in the power grid's manual intervention wide table, extract the cross-section description field of the intervention reason in the first record, and generate a verification result wide table containing the compliance judgment results of the intervention record by matching the cross-section description field and the standard cross-section dictionary.

[0063] It should be noted that the technical concept of this second embodiment is completely consistent with that of the first embodiment. The two maintain a high degree of synergy at the technical logic level. The specific technical details can be referred to the relevant description of the first embodiment, which will not be repeated here.

[0064] Example 3: This invention provides a computer-readable storage medium including a stored computer program, wherein the computer program, when running, controls the device where the computer-readable storage medium is located to execute the aforementioned text self-recognition spot market intervention reason verification method; The text-based self-identification method for verifying the reasons for intervention in the spot market, when implemented as a software functional unit and used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0065] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer programs or instructions. When the computer program or instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present invention are performed entirely or partially. The computer can be a general-purpose computer, a special-purpose computer, a computer network, a network device, a user equipment, or other programmable device. The computer program or instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another. For example, the computer program or instructions can be transferred from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium, such as a floppy disk, hard disk, or magnetic tape; it can also be an optical medium, such as a digital video optical disc; or it can be a semiconductor medium, such as a solid-state drive. The computer-readable storage medium may be a volatile or non-volatile storage medium, or may include both types of storage media.

[0066] The above are preferred embodiments of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for self-identifying intervention causes in a spot market, characterized by, The method comprises the following steps: acquiring basic data of a power grid; based on the basic data, identifying and calculating the characteristics of the power grid section, and combining the load information of the power grid to generate a section processing wide table including the power grid clustering result field; based on the section processing wide table, constructing a standard section dictionary containing section names and power limits, retrieving a first record containing stable section constraints in the manual intervention wide table of the power grid, extracting the section description field of the intervention reason in the first record, and generating a check result wide table containing the compliance judgment result of the intervention record by matching the section description field and the standard section dictionary.

2. The method of claim 1, wherein the intervention reason is one of a plurality of intervention reasons, and the intervention reason is identified by comparing the text with a plurality of intervention reason templates, each intervention reason template corresponding to one of the plurality of intervention reasons. based on the basic data, identifying and calculating the characteristics of the power grid section, and combining the load information of the power grid to generate a section processing wide table including the power grid clustering result field, specifically: based on the basic data, identifying and calculating the characteristics of the power grid section according to the secondary simulated annealing optimization spectral clustering algorithm, to obtain an initial section calculation wide table; based on the preset section standard naming rule, standard naming the sections in the initial section calculation wide table, and combining the load information of the power grid to generate the section processing wide table including the power grid clustering result field.

3. A method for self-identifying intervention causes in a spot market according to claim 2, characterized in that, based on the basic data, identifying and calculating the characteristics of the power grid section according to the secondary simulated annealing optimization spectral clustering algorithm, to obtain an initial section calculation wide table, specifically: based on the basic data, constructing a weighted adjacency matrix with power flow values as weights, and combining the preset clustering center to perform first-round spectral clustering by the simulated annealing optimization spectral clustering algorithm to obtain first-round section information; within the primary category corresponding to the first-round spectral clustering, performing secondary spectral clustering with the substation bus directly connected to the target main transformer node as the center to obtain secondary section information; within the secondary category corresponding to the secondary spectral clustering, performing branch test, if branch opening causes other lines to be overloaded, then the corresponding branch and the overloaded line form a local section, to obtain final section information; integrating the first-round section information, the secondary section information and the final section information to form the initial section calculation wide table containing all overload section information.

4. The text-based self-identification method for verifying the reasons for intervention in the spot market as described in claim 2, characterized in that, based on the preset section standard naming rule, standard naming the sections in the initial section calculation wide table, and combining the load information of the power grid to generate the section processing wide table including the power grid clustering result field, specifically: according to the preset section standard naming rule, adding section names in the initial section calculation wide table to obtain a final section calculation wide table; integrating the final section calculation wide table, the load information of the monitored section in the power grid and the unit sensitivity data to generate the section processing wide table including the power grid clustering result field.

5. A method of self-identifying intervention in a spot market according to any one of claims 1 to 4, wherein, The section standard naming rule is formulated based on the type of section component equipment; wherein, the type of section component equipment includes line and line combination, main transformer and main transformer combination, and main transformer and line combination.

6. The method of claim 1, wherein the method further comprises: identifying a reason for intervention in the spot market based on the text. Based on the section processing wide table, a standard section dictionary containing section names and power limits is constructed, a first record containing a stable section constraint in the manual intervention wide table of the power grid is retrieved, a section description field of an intervention reason in the first record is extracted, a checking result wide table containing intervention record compliance judgment results is generated by matching the section description field and the standard section dictionary, and specifically: ​ The section names are summarized from the section processing wide table, and the standard section dictionary containing section names and power limits is constructed based on the existing power limits in the section processing wide table; The intervention reasons of each record in the manual intervention wide table of the power grid are retrieved, and the first record containing a classified keyword of a stable section constraint is screened out; Based on the first record, a field related to section description is extracted from the intervention reason to obtain a section description field; Data matching is performed based on the section description field and the standard section dictionary, the compliance of the intervention record is judged in combination with the reasoning rules in the checking reasoning rule library, and the checking result wide table containing the intervention record compliance judgment results is generated.

7. A method for self-identifying intervention causes in a spot market according to claim 6, characterized in that, Data matching is performed based on the section description field and the standard section dictionary, the compliance of the intervention record is judged in combination with the reasoning rules in the checking reasoning rule library, and the checking result wide table containing the intervention record compliance judgment results is generated, and specifically: By matching the overload section record with a load rate exceeding a preset value in the section processing wide table with the section description field related to the intervention reason in the manual intervention wide table, it is judged whether there is overload uncontrolled; if it is controlled, the relevance of the section occurrence time in the section processing wide table and the intervention time period in the manual intervention wide table is checked to judge whether the control is not timely, and a first judgment result is obtained; For the section description field containing the keyword of the stable section constraint in the manual intervention wide table, whether the record is wrong, the intervention is not compliant, and the intervention is excessive is judged in turn by matching the standard section dictionary, checking the corresponding section load rate, and verifying the intervention and section time interval, and a second judgment result is obtained; The first judgment result and the second judgment result are integrated to obtain the checking result wide table containing the intervention record compliance judgment results.

8. A text self-identifying spot market intervention reason checking device, characterized by, It comprises a data module, a section module and a checking module; The data module is used to acquire the basic data of the power grid. The section module is used to identify and calculate the characteristics of the power grid section based on the basic data, and generate a section processing wide table containing the power grid clustering result field in combination with the load information of the power grid. The checking module is used to construct a standard section dictionary containing section names and power limits based on the section processing wide table, retrieve a first record containing a stable section constraint in the manual intervention wide table of the power grid, extract a section description field of an intervention reason in the first record, generate a checking result wide table containing intervention record compliance judgment results by matching the section description field and the standard section dictionary.

9. A storage medium, characterized by The storage medium stores a computer program, which is invoked and executed by a computer to implement the text self-recognition spot market intervention reason checking method according to any one of claims 1 to 7.

10. A computer program product comprising computer programs or instructions, characterized in that, The computer program or instruction is executed by the communication device to implement the text self-recognition spot market intervention reason checking method according to any one of claims 1 to 7.

Images