Real-time adaptive constant value checking method and system for relay protection device

Abstract

The invention provides a real-time adaptive constant value checking method and system for a relay protection device, and the method comprises the following steps: firstly, carrying out the automatic recognition and analysis of an effective constant value table from the real-time data of the relay protection device through information entropy and geometric topology analysis, and obtaining an equipment end constant value item; and respectively constructing a mixed feature vector containing text semantics, numerical values and contexts for the equipment end constant value item and a preset standard constant value item. By calculating the semantic similarity among the vectors, intelligent and accurate matching among constant value items with different sources and different naming modes is realized. And finally, unit normalization and numerical value comparison are carried out on the successfully matched constant value item pairs, and a final check report is automatically generated. According to the invention, automation of the relay protection setting value checking process is realized, and the problems of low efficiency and error proneness of manual checking are solved.

Description

Technical Field

[0001] This invention belongs to the field of relay protection device testing technology, specifically relating to a real-time adaptive setting value verification method and system for relay protection devices. Background Technology

[0002] Relay protection devices are the nerve center ensuring the safe and stable operation of the power system. The accuracy of their internal protection settings directly affects whether faults can be accurately and quickly isolated. Therefore, before commissioning and during regular maintenance, specialized testing equipment must be used to read the actual internal settings and verify them item by item against the standard settings sheet issued by the design institute to ensure complete consistency. This verification process is a critical link in ensuring power grid safety, and its accuracy requirements are extremely high.

[0003] However, in current practice, this crucial verification work still largely relies on manual operation by field engineers. Engineers need to compare each setting value obtained from the relay protection device with a paper or electronic standard setting sheet. This process faces significant challenges: different manufacturers and models of relay protection devices have significant differences in the naming conventions, display order, and unit representations of their internal settings. For example, the same "overcurrent stage I setting value" might be displayed as "OC-1 Set" on one device and as "I> Set Value" on another. Engineers must rely on their professional knowledge and extensive experience to understand and associate these setting items with varying names and structures, resulting in a huge workload and a high risk of omissions or errors due to negligence or fatigue. Summary of the Invention

[0004] This invention provides a real-time adaptive setting verification method and system for relay protection devices to solve the problems of huge manual workload and easy omissions or errors in setting verification.

[0005] In a first aspect, the present invention provides a method for real-time adaptive setting verification of a relay protection device, the method comprising the following steps: The entropy of data information is calculated on the real-time printed data obtained from the relay protection device, and candidate data streams with significant structured features are identified from the real-time printed data based on a preset entropy threshold. Extract the frame structure feature points in the candidate data stream, and confirm the candidate data stream as a valid value table by constructing a global grid topology model through a geometric consistency test algorithm; The table boundaries of the effective setting table are defined based on the global grid topology model, and the effective setting table is parsed with local abnormal row fault tolerance capability based on the table boundaries to obtain the equipment-side setting items of the relay protection device. The device-side setpoints and the standard setpoints in the preset standard setpoint sheet are respectively fused to obtain a hybrid feature vector containing text semantics, numerical attributes and contextual information; Calculate the semantic similarity between the mixed feature vectors of the device-side fixed value and the standard fixed value, and perform cross-source matching between the device-side fixed value and the standard fixed value based on the semantic similarity; The device-side setpoints that are successfully matched across sources are normalized in units and compared with the standard setpoints. A setpoint verification report is then generated based on the comparison results.

[0006] Optionally, the step of calculating the data information entropy of the real-time printed data obtained from the relay protection device and identifying candidate data streams with significant structured features from the real-time printed data based on a preset entropy threshold includes the following steps: Set the window size and sliding step of the sliding window, move the sliding window along the real-time printed data, and calculate the frequency distribution of data bytes within the window position of each sliding window; Based on the frequency distribution of data bytes and applying the Shannon entropy formula, the information entropy value is calculated for each window position, and an entropy value sequence curve corresponding to the information entropy value and the data stream position is generated. When the entropy sequence curve crosses the preset first entropy threshold from high to low, the current data stream position is marked as the starting point of the candidate data stream. When the entropy sequence curve crosses the preset second entropy threshold from low to high, the current data stream position is marked as the ending point of the candidate data stream. The second entropy threshold is greater than the first entropy threshold.

[0007] Optionally, the step of extracting the frame structure feature points within the candidate data stream and confirming the candidate data stream as a valid value table using the global mesh topology model constructed through the geometric consistency check algorithm includes the following steps: A predefined character set containing various table drawing characters is used. Traverse all data within the candidate data stream, identify all target characters belonging to the character set, and extract the two-dimensional coordinates of the target characters within the candidate data stream; Combine all two-dimensional coordinates into a two-dimensional point cloud; The global grid topology model, constructed based on two-dimensional point clouds and using a geometric consistency check algorithm, confirms the candidate data stream as a valid value table.

[0008] Optionally, the process of confirming the candidate data stream as a valid value table using the global mesh topology model constructed based on two-dimensional point clouds and a geometric consistency check algorithm includes the following steps: A minimum point set is randomly sampled from a 2D point cloud, and a global grid topology model is assumed based on the minimum point set. Traverse all points in the 2D point cloud according to the global grid topology model, and determine the points that match the geometric position of the global grid topology model as interior points; Repeat the iterative process of random sampling, model assumptions, and interior point statistics until the preset number of iterations is reached; In all iterations, the global grid topology model that receives the most interior point support is selected as the optimal model. If the proportion of interior points in the optimal model exceeds the preset threshold for the proportion of interior points, then the candidate data stream is confirmed as a valid setpoint table.

[0009] Optionally, the step of fusing the device-side setting items with the standard setting items in the preset standard setting sheet to obtain a hybrid feature vector containing text semantics, numerical attributes, and contextual information includes the following steps: For the standard setting items in the device-side setting items or the preset standard setting items, extract the setting item information and the relative position information of the setting item in the corresponding customization table or customization order. The setting item information includes the setting item description text, the setting item value and the setting item unit. Text semantic vectors are generated based on the description text of the fixed value terms. Numerical attribute vectors are generated by combining the numerical value and unit of the fixed value terms and encoding. Contextual feature vectors are generated based on relative position information and encoding. The text semantic vector, numerical attribute vector, and context feature vector are weighted and fused to obtain a hybrid feature vector.

[0010] Optionally, the step of generating a text semantic vector based on a fixed-value term to describe the text includes the following steps: The text describing the fixed-value items is segmented into words to obtain a word sequence; Load the pre-trained word vector model and use the word vector model to map each word in the word sequence into a high-dimensional word vector; A weighted average is calculated for all high-dimensional word vectors corresponding to the text describing the fixed-value item, and the calculated weighted average vector is used as the text semantic vector of the fixed-value item.

[0011] Optionally, the semantic similarity between the hybrid feature vectors of the device-side fixed value and the standard fixed value, and the cross-source matching between the device-side fixed value and the standard fixed value based on the semantic similarity, includes the following steps: For any device-side fixed value, calculate the cosine similarity between the first mixed feature vector of the device-side fixed value and the second mixed feature vector of all standard fixed value items; Select the highest cosine similarity among all cosine similarities. If the highest cosine similarity exceeds the preset confidence threshold, it is determined that the device-side fixed value item and the standard fixed value item corresponding to the highest cosine similarity have successfully matched across sources, forming a set of fixed value item pairs. Repeat the above steps until the cosine similarity between the first mixed feature vector and the second mixed feature vector has been calculated for all device-side values.

[0012] Optionally, the step of performing unit normalization processing and numerical comparison on the device-side setpoints that have successfully matched across sources and the standard setpoints, and generating a setpoint verification report based on the numerical comparison results, includes the following steps: Iterate through all pairs of fixed values, perform unit normalization on each pair of fixed values, compare the values, and classify the comparison results into consistent or inconsistent items. Identify all standard value items that exist in the standard value sheet but have not been successfully matched across sources, and classify them as missing items; Identify all device-side settings that exist in the device-side settings but have not been successfully matched across sources, and classify them as unplanned items; Summarize all consistent, inconsistent, missing, and unplanned items according to the preset format to generate the final value verification report.

[0013] In a second aspect, the present invention also provides a real-time adaptive setting verification system for a relay protection device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the real-time adaptive setting verification method for the relay protection device as described in the first aspect.

[0014] Thirdly, the present invention also provides a computer-readable storage medium storing instructions that, when executed by a processor, configure the processor to perform the dual-source magnetic field integrated detection and analysis method based on buried metal pipelines as described in the first aspect.

[0015] The beneficial effects of this invention are: This invention solves the core problems of tedious, inefficient, and error-prone manual comparison in traditional relay protection setting verification work, caused by the diverse sources of equipment and inconsistent naming conventions. By introducing information entropy and geometric topology analysis, this invention achieves automatic identification and parsing of setting tables from complex real-time data streams, ensuring accurate content extraction even when the tables have local flaws. Furthermore, it constructs a hybrid feature vector containing textual semantics, numerical attributes, and contextual information, using semantic similarity calculation to achieve intelligent matching between device-side setting items and standard setting items. This allows the automatic identification process to understand the same technical connotations under different expressions, thus overcoming barriers between manufacturers, models, and versions, ensuring the uniqueness and accuracy of the match. This invention transforms the entire verification process from manual operation relying on human experience to an automated intelligent data processing and comparison process, greatly improving the efficiency and reliability of setting verification. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the real-time adaptive setting verification method for a relay protection device in one embodiment of this application.

[0017] Figure 2 This is a schematic diagram of the device connection for real-time acquisition of device-side setpoints in one embodiment of this application.

[0018] Figure 3 This is a schematic diagram of value comparison in one embodiment of this application.

[0019] Figure 4 This is a schematic diagram of the frame structure feature points in the candidate data stream in one embodiment of this application.

[0020] Figure 5 This is a preview diagram of a value verification report in one embodiment of this application. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0022] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0023] Figure 1 This is a flowchart illustrating a real-time adaptive setting verification method for a relay protection device in one embodiment. It should be understood that, although... Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1At least some steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps. For example Figure 1 As shown, the real-time adaptive setting verification method for a relay protection device disclosed in this invention specifically includes the following steps: S101. Perform data information entropy calculation on the real-time printed data obtained from the relay protection device, and identify candidate data streams with significant structured features from the real-time printed data based on a preset entropy threshold.

[0024] Among them, reference Figure 2 A message acquisition module is installed inside the substation cabinet equipped with relay protection devices and connected to the printer port of the protection device. The acquisition module operates continuously, and a mobile terminal can be temporarily connected to the print data receiver as needed (e.g., using an RJ11 quick-connect interface) to synchronize the real-time print data acquired. When calculating the information entropy of the real-time print data obtained from the relay protection device, the size and step parameters of the sliding window are first set. Then, the window is moved along the data stream, and the frequency distribution of bytes within each window is statistically analyzed. The Shannon entropy formula is used to calculate the information entropy value at each window position. When the data stream contains structured content such as tables, the character distribution is relatively regular, and the information entropy is low; while when it contains random text, the character distribution is chaotic, and the information entropy is high. By setting two entropy thresholds, when the entropy curve crosses the first threshold from high to low, it is marked as the starting point of the candidate data stream; when it crosses the second threshold from low to high, it is marked as the ending point. This method can automatically identify data segments with obvious structured features and effectively filter out irrelevant text information.

[0025] S102. Extract the frame structure feature points in the candidate data stream, and confirm the candidate data stream as a valid value table by constructing a global grid topology model through a geometric consistency test algorithm.

[0026] Among them, reference Figure 3When extracting the frame structure features within the candidate data stream, a character set containing various table drawing characters is predefined, such as horizontal lines, vertical lines, and intersections. All characters in the candidate data stream are traversed, and target characters belonging to this character set are identified and their two-dimensional coordinate positions in the data stream are recorded, forming two-dimensional point cloud data. The RANSAC geometric consistency test algorithm is used to analyze the point cloud. A minimum point set is randomly sampled to assume a global grid topology model, and then the number of interior points conforming to the geometric position of this model is counted. After multiple iterations, the model with the most interior point support is selected as the optimal solution. When the proportion of interior points in the optimal model exceeds a preset threshold, the candidate data stream is confirmed as a valid fixed-value table. This method effectively distinguishes between real and pseudo-table structures through geometric constraints, ensuring that only data with a complete grid topology is identified as a fixed-value table, significantly improving the accuracy and robustness of table recognition.

[0027] S103. Define the table boundary of the effective setting table according to the global grid topology model, and perform content parsing with local abnormal row tolerance capability on the effective setting table based on the table boundary to obtain the equipment end setting items of the relay protection device.

[0028] The algorithm extracts data from the setpoint table by defining the table boundaries using a global grid topology model and then employing a grid-based content parsing algorithm. This algorithm is capable of handling local anomaly rows, addressing potential issues such as formatting irregularities, missing characters, or misalignments within the table. The boundary range of each cell is determined using the grid topology model, and text content is extracted within these boundaries. For detected anomaly rows, the algorithm infers and corrects based on the structural characteristics of adjacent normal rows, ensuring the integrity of the extracted data. During parsing, the algorithm identifies key information such as the name, value, and unit of the setpoint items and retains their relative positions within the table. This fault-tolerance mechanism enables the algorithm to handle situations such as poor print quality and formatting changes in practical applications, ensuring that critical relay protection device setpoint information is accurately extracted.

[0029] S104. Perform feature fusion between the device-side setting items and the standard setting items in the preset standard setting sheet to obtain a hybrid feature vector containing text semantics, numerical attributes and contextual information.

[0030] The process involves extracting complete information for each value item, including its descriptive text, numerical value, unit, and relative position within the table. After word segmentation, a pre-trained word vector model maps words to high-dimensional vectors, which are then weighted to obtain a semantic vector. Combining the numerical and unit information, an encoding technique generates a numerical attribute vector that captures the magnitude and unit type of the value. Based on the row and column positions of the value items within the table, a contextual feature vector is generated, reflecting the structured positional relationships of the value items. Finally, these three types of feature vectors are fused according to preset weights to form a hybrid feature vector containing textual semantics, numerical attributes, and contextual information.

[0031] S105. Calculate the semantic similarity between the mixed feature vectors of the device-side fixed value and the standard fixed value, and perform cross-source matching between the device-side fixed value and the standard fixed value based on the semantic similarity.

[0032] In calculating the semantic similarity between device-side defined items and standard defined items, a cosine similarity algorithm is used to measure the similarity of the mixed feature vectors. For each device-side defined item, its cosine similarity with all standard defined items is calculated, and the highest similarity value is selected for judgment. When the highest similarity exceeds a preset confidence threshold, the device-side defined item is considered to have successfully matched the corresponding standard defined item, forming a defined item pair. This cross-source matching method based on semantic similarity can handle expression differences between different data sources, such as synonyms, abbreviations, and changes in description order, achieving intelligent defined item association. By setting an appropriate confidence threshold, matching accuracy can be ensured while avoiding false matches.

[0033] S106. Perform unit normalization and numerical comparison on the device-side setpoints that have successfully matched across sources and the standard setpoints, and generate a setpoint verification report based on the numerical comparison results.

[0034] Among them, reference Figure 4 First, standardize the numerical representation under different unit systems, such as converting kilovolts to volts and milliamperes to amperes, to ensure the accuracy of numerical comparisons. After unit normalization, directly compare the magnitudes of two setpoints and determine their consistency based on a preset error tolerance. Classify the comparison results into consistent and inconsistent items, and identify missing items in the standard setpoint sheet that have not been matched, as well as unplanned items that exist on the equipment side but have not been matched. Summarize all categories of setpoint information according to a preset format, including detailed information such as setpoint name, standard value, actual value, and degree of deviation, to generate a structured setpoint verification report. The setpoint verification report is as follows: Figure 5As shown, "Uncertain" in the verification results column of the report refers to "missing items" or "unplanned items." This report clearly demonstrates the accuracy of relay protection device setting configurations, helping maintenance personnel quickly identify incorrect, omitted, or redundant setting items, ensuring the reliable operation of power system protection devices.

[0035] In one embodiment, the process of calculating the data information entropy of real-time printed data obtained from the relay protection device and identifying candidate data streams with significant structured features from the real-time printed data based on a preset entropy threshold includes the following steps: Set the window size and sliding step of the sliding window, move the sliding window along the real-time printed data, and calculate the frequency distribution of data bytes within the window position of each sliding window; Based on the frequency distribution of data bytes and applying the Shannon entropy formula, the information entropy value is calculated for each window position, and an entropy value sequence curve corresponding to the information entropy value and the data stream position is generated. When the entropy value sequence curve crosses the preset first entropy value threshold from high to low, the current data stream position is marked as the starting point of the candidate data stream. When the entropy value sequence curve crosses the preset second entropy value threshold from low to high, the current data stream position is marked as the ending point of the candidate data stream. The second entropy value threshold is greater than the first entropy value threshold.

[0036] In this implementation, two key parameters need to be determined first: window size and sliding step size. The window size determines the data range for each analysis, typically set to a length sufficient to contain a adequate number of character samples, such as 64 or 128 bytes, to ensure the reliability of the statistical results. The sliding step size controls the distance the window moves; a smaller step size provides finer analytical granularity but increases computational load, while a larger step size has the opposite effect. The sliding window moves along the real-time printed data stream at the set step size, performing statistical analysis on all data bytes within each window position. During the statistical process, the frequency of each byte value is recorded, and its frequency distribution within the current window is calculated. This frequency distribution reflects the character composition characteristics of the data: structured data such as tables have relatively regular character distributions, while random text exhibits a more uniform or chaotic distribution pattern. Based on the frequency distribution of data bytes at each window position, the Shannon entropy formula is applied to calculate the information entropy value. Then, the information entropy values ​​corresponding to each window position are arranged according to the positional order of the data stream, forming a continuous entropy value sequence curve. This curve visually displays the changing trend of information complexity throughout the data stream, providing a quantitative indicator for identifying structured data segments.

[0037] This method identifies the boundary positions of candidate data streams based on entropy sequence curves and a pre-defined dual-threshold mechanism. Two entropy thresholds are set: a lower first threshold detects the start of structured data, and a higher second threshold detects the end of structured data. When the entropy sequence curve changes from a high-entropy region to a low-entropy region and crosses the first threshold, it indicates that the data stream has shifted from random text to structured content, and the current position is marked as the starting point of the candidate data stream. Conversely, when the curve changes from a low-entropy region to a high-entropy region and crosses the second threshold, it indicates the end of structured content, and the current position is marked as the ending point of the candidate data stream. The dual-threshold design avoids the frequent switching problems that may occur with a single threshold and improves the stability of boundary detection by introducing a hysteresis effect. This method can automatically identify data segments containing significant structured features such as tables and lists, effectively filtering out irrelevant text information.

[0038] In one implementation, calculating the information entropy value for each window position based on the data byte frequency distribution and applying the Shannon entropy formula includes the following steps: Based on the structural role of characters in the fixed value table, all possible data bytes are pre-divided into multiple character categories, which include at least: table borders, numbers and decimal points, and letters and whitespace. For each window position, the frequency distribution of data bytes falling into each preset character category is statistically analyzed; By independently applying the Shannon entropy formula to the frequency distribution of each character category, a set of classification information entropy values ​​corresponding to each character category are calculated for the window position. A preset weight coefficient is set for the classification information entropy value of each character category, with the table border category having the highest weight coefficient. The composite information entropy value that enhances the perception of structured features is calculated by weighting and summing a set of classification information entropy values ​​at the window position with their corresponding weight coefficients.

[0039] In this implementation, characters are classified based on their structural role in the fixed-value table, specifically establishing a classification system based on the functional characteristics of different character types in table construction. The table border class includes characters used to draw table boundaries and separator lines, such as... Figure 4As shown, the characters include, but are not limited to, the top-left border symbol “┏”, top-right border symbol “┓”, bottom-left border symbol “┗”, bottom-right border symbol “┛”, cell separators “┯” and “┠”, which are core elements constituting the visual structure of the table. The number and decimal point category covers all numeric characters and decimal points, primarily carrying specific numerical information. The letter and whitespace category includes various alphabetic characters, spaces, tabs, etc., typically used to represent item names, unit identifiers, and formatting adjustments. This function-based character classification method allows for the orderly categorization of previously mixed byte data according to its actual function within the table. Each character category carries a specific information expression function, and its distribution characteristics reflect the degree of structure in the data content. During the frequency statistics of character categories at each sliding window position, it is necessary to traverse all data bytes within the current window and assign each byte to its corresponding category according to the preset character classification rules.

[0040] The statistical process involves counting the occurrences of each specific character within each character category and then calculating its frequency distribution within that category. For table borders, the distribution of various line characters is statistically analyzed; for numbers and decimal points, the frequency of different numbers and decimal points is recorded; and for letters and spaces, the distribution patterns of letters and spaces are analyzed. This classification statistical method reveals the composition ratio and distribution patterns of different functional characters in the current data segment. When the data contains a table structure, border characters exhibit a relatively concentrated distribution, number characters appear densely in the numerical area, while letter characters are mainly distributed in the title and description areas. When independently applying the Shannon entropy formula to calculate the classification information entropy for each character category, the character frequency distribution within the category is used as the basis for calculation. In the calculation process, the occurrence count of each character within each category is first divided by the total number of characters in that category to obtain the probability distribution within the category, and then substituted into the formula to obtain the classification information entropy of that category. This classification entropy calculation method can independently assess the complexity and randomness within each character category. Low entropy values ​​for table borders indicate a regular distribution pattern of line characters, entropy values ​​for numbers reflect the diversity of numerical values, and entropy values ​​for letters reflect the complexity of text content.

[0041] Next, the weight allocation is determined based on the importance of different character categories to table structure recognition. Characters representing table borders are key elements constituting the visual structure of a table; their distribution directly determines whether the data possesses table characteristics, therefore they are assigned the highest weight coefficient. Numbers and decimal points carry core information and play a crucial role in table recognition, thus their weight coefficient is set to a medium level. Letters and whitespace characters mainly serve auxiliary and formatting functions, having a relatively small impact on table structure judgment, therefore their weight coefficient is set to a low level. When weighted summing the classification information entropy values ​​of each character category with their corresponding weight coefficients, a linear combination method is used to calculate the composite information entropy.

[0042] The formula for calculating the composite information entropy is as follows:

[0043] in, It is a composite information entropy value, where m represents the total number of character categories and k represents the index of the character category. The weight coefficient representing the category of the k-th character. The set representing the k-th character category. The summation range is all elements belonging to the category. bytes , This indicates that the known byte belongs to a category. Under the premise that the byte is exactly The probability is actually calculated as follows: within the data window, the number of bytes... The number of times it appears, divided by the number of all items in the window belonging to that category. The total number of bytes. The part in parentheses calculates the internal entropy of the k-th character category itself, considering the variation within a single character category. For example, if table borders use only one type of character, their classification entropy will be low; if multiple types are used, the classification entropy will be higher. During the calculation, the classification entropy of each category is multiplied by its corresponding weight coefficient, and then all weighted results are summed to obtain the final composite entropy value. This weighted fusion method comprehensively considers the contributions of different character categories, forming a more comprehensive and accurate measure of structured features. Composite entropy inherits the advantages of each classification entropy while highlighting the role of key features through a weighting mechanism. When the data contains a regular table structure, the low entropy value of border characters will significantly reduce the composite entropy value under high weighting; when the data is random text, the entropy values ​​of each category are relatively high, and the composite entropy value also increases accordingly.

[0044] In one implementation, extracting the frame structure feature points within the candidate data stream and confirming the candidate data stream as a valid value table using a global mesh topology model constructed through a geometric consistency check algorithm includes the following steps: A predefined character set containing various table drawing characters is used. Traverse all data within the candidate data stream, identify all target characters belonging to the character set, and extract the two-dimensional coordinates of the target characters within the candidate data stream; Combine all two-dimensional coordinates into a two-dimensional point cloud; The global grid topology model, constructed based on two-dimensional point clouds and using a geometric consistency check algorithm, confirms the candidate data stream as a valid value table.

[0045] In this implementation, when predefining the table drawing character set, it is necessary to collect all possible character types used to construct the table structure and establish a complete character library. This character set includes horizontal line characters for drawing the horizontal boundaries and separators of the table; vertical line characters for constructing the vertical boundaries and column separators; intersection characters located at the intersection of horizontal and vertical lines, forming grid nodes; corner characters for identifying the four corners of the table; and other auxiliary drawing characters such as T-shaped connectors and L-shaped corner symbols. The establishment of the character set needs to consider the differences in different printing devices and character encoding standards to ensure coverage of all possible table drawing symbols. These characters have corresponding numerical representations in ASCII, Unicode, or other encoding systems. By establishing a mapping relationship between character code values ​​and character functions, a standardized character recognition library is formed. When traversing the candidate data stream for target character recognition, each character in the data stream needs to be checked one by one to determine whether it belongs to the predefined table drawing character set. The recognition process uses character code value matching, comparing the current character's encoding value with the standard encoding in the character set to determine if it is the target character. For each identified target character, its two-dimensional coordinate position information in the candidate data stream needs to be accurately extracted. During the coordinate extraction process, a coordinate system is established with the starting position of the data stream as the origin. The horizontal axis represents the column position of the character in the current row, and the vertical axis represents the row number of the character.

[0046] The extracted coordinates of all target characters are aggregated into a two-dimensional point cloud. This point cloud consists of a series of coordinate points, each representing the spatial position of a character in the data stream. The coordinate values ​​of the points reflect the row and column positions of the characters. Point cloud data structures are typically stored in array or list form, with each element containing two values: an x-coordinate and a y-coordinate. During point cloud construction, coordinate standardization is required to eliminate potential coordinate offsets and scaling differences from different data sources. The quality of the point cloud directly affects the accuracy of subsequent geometric analysis; therefore, it is necessary to detect and filter out abnormal coordinate points, removing coordinate values ​​that significantly deviate from the normal range. Next, a global grid topology model is constructed based on the two-dimensional point cloud using a geometric consistency check algorithm, and the RANSAC random sampling consensus algorithm is used to identify regular grid structures in the point cloud. The algorithm first randomly selects a minimum set of points from the point cloud and hypothesizes a candidate grid topology model based on these points. This model defines the row and column spacing, angles, and overall layout parameters of the grid. Then, it iterates through all other points in the point cloud, calculating the geometric deviation of each point from the hypothetical model. Points with deviations less than a preset threshold are identified as interior points supporting the model.

[0047] The iterative process of random sampling, model hypothesis testing, and interior point statistics is repeated, generating a candidate model and a corresponding set of interior points in each iteration. After all iterations, the model with the most interior point support is selected as the optimal global grid topology model. When the proportion of interior points in the optimal model to the total number of points exceeds a preset threshold, the candidate data stream is confirmed to have a complete tabular grid structure and is recognized as a valid value table. This geometric consistency check method effectively distinguishes between a true tabular structure and randomly distributed characters, ensuring that only data with regular grid characteristics is identified as a value table.

[0048] In one implementation, confirming candidate data streams as valid value tables based on a global mesh topology model constructed using a two-dimensional point cloud and a geometric consistency check algorithm includes the following steps: A minimum point set is randomly sampled from a 2D point cloud, and a global grid topology model is assumed based on the minimum point set. Traverse all points in the 2D point cloud according to the global grid topology model, and determine the points that match the geometric position of the global grid topology model as interior points; Repeat the iterative process of random sampling, model assumptions, and interior point statistics until the preset number of iterations is reached; In all iterations, the global grid topology model that receives the most interior point support is selected as the optimal model. If the proportion of interior points in the optimal model exceeds the preset threshold for the proportion of interior points, then the candidate data stream is confirmed as a valid setpoint table.

[0049] In this embodiment, for a regular rectangular grid structure, the minimum point set typically contains four non-collinear points that define the basic geometric parameters of the grid. The random sampling process uses a uniformly distributed random number generator to randomly select a specified number of point indices from the total number of points in the point cloud, ensuring that each point has an equal probability of being selected. Based on the selected minimum point set, key parameters of the global grid topology model are derived through geometric calculations, including the row spacing, column spacing, starting coordinate position, and overall grid tilt angle. The grid model assumes that data points are distributed in a regular rectangular array, with fixed vertical spacing between adjacent rows and fixed horizontal spacing between adjacent columns. The calculation of model parameters needs to consider the geometric relationships of the point set, and the optimal combination of grid parameters is determined using the least squares method or other fitting algorithms. For any point in the point cloud, the theoretical grid node position corresponding to that point is first calculated based on the grid model parameters, and then the Euclidean distance between the actual point coordinates and the theoretical position is calculated. The distance calculation formula is... ,in Represents the actual point coordinates. This represents the coordinates of the corresponding theoretical mesh node. When the calculated distance is less than the preset geometric tolerance threshold, the point is determined to be an interior point supporting the current mesh model; otherwise, it is determined to be an exterior point.

[0050] The iterative process of random sampling, model assumptions, and interior point statistics is repeated. Each iteration independently samples the minimum point set to avoid the influence of previous iterations. The number of iterations is typically determined based on probability theory, considering the probability that, given a certain proportion of outliers, at least one iteration will sample a minimum point set consisting entirely of inliers. More iterations increase the probability of finding the optimal model but increase computational cost; fewer iterations may lead to missing the best solution. In each iteration, the entire process from random sampling to interior point statistics needs to be executed completely, and the grid model parameters and the corresponding number of inliers obtained in the current iteration are recorded. After all iterations are completed, the number of inlier support for the grid models obtained in each iteration is compared, and the model with the most inliers is selected as the global optimum, because more inlier support means that the model can better explain the distribution of the actual data. When the number of inliers is the same, auxiliary indicators such as geometric fitting error and model complexity can be further considered for optimization.

[0051] Next, the proportion of interior points in the optimal model to the total number of points needs to be calculated and compared with a preset threshold. When the proportion of interior points exceeds the preset threshold, it indicates that most of the table characters conform to a regular grid distribution pattern, confirming the candidate data stream as a valid fixed-value table. When the proportion is below the threshold, it indicates that the data lacks obvious table structure features and is not recognized as a fixed-value table. The threshold setting needs to strike a balance between recognition accuracy and error tolerance. An excessively high threshold may lead to missed detection of real tables, while an excessively low threshold may misidentify non-table data. Through this quantitative evaluation method based on geometric consistency, it is possible to objectively and accurately determine whether the data has a complete table structure.

[0052] In one implementation, feature fusion is performed between the device-side setting items and the standard setting items in the preset standard setting sheet to obtain a hybrid feature vector containing text semantics, numerical attributes, and contextual information. This includes the following steps: For the standard setting items in the device-side setting items or the preset standard setting items, extract the setting item information and the relative position information of the setting item in the corresponding customization table or customization order. The setting item information includes the setting item description text, the setting item value and the setting item unit. Text semantic vectors are generated based on the description text of the fixed value terms. Numerical attribute vectors are generated by combining the numerical value and unit of the fixed value terms and encoding. Contextual feature vectors are generated based on relative position information and encoding. The text semantic vector, numerical attribute vector, and context feature vector are weighted and fused to obtain a hybrid feature vector.

[0053] In this embodiment, the extraction of setpoint information comprises three core components: the setpoint description text, typically located in the left column or header row of the table, includes the name, function, and technical parameters of the setpoint; the setpoint value is the specific quantified value of the setpoint, which may include integers, decimals, or values ​​expressed in scientific notation; and the setpoint unit represents the dimension of the value, such as volts for voltage, amperes for current, and seconds for time. Relative position information extraction requires recording the row and column numbers of the setpoint in the table, as well as its spatial relationship with adjacent setpoints. Position information is represented using a standardized coordinate system, establishing row and column coordinates with the top left corner of the table as the origin. Through regular expressions and pattern matching techniques, the three components—text, value, and unit—can be accurately identified and separated.

[0054] The descriptive text is then preprocessed, including punctuation removal, case neutralization, and word segmentation, converting continuous text into a word sequence. A pre-trained word vector model maps each word to a high-dimensional numerical vector, capturing semantic information and contextual relationships. For descriptive text containing multiple words, a weighted average method is used to calculate the overall semantic vector, with weights determined based on word frequency, importance, or positional information. Generating numerical attribute vectors requires jointly encoding the numerical value and unit information of the fixed-value items. The numerical part is standardized to a fixed range, while the unit part is converted to a numerical representation through one-hot encoding or embedding vectors. Contextual feature vectors are generated based on the relative positional information of the fixed-value items, including row numbers, column numbers, and distances to key reference points. Spatial positional information is converted into vector form using positional encoding techniques.

[0055] The hybrid feature vector is constructed using a linear combination method, and the fusion formula is as follows: ,in Represents a mixed feature vector. Represents a text semantic vector. Represents a vector of numerical attributes. Represents the context feature vector. , , These are the corresponding weight coefficients and satisfy... The weighting coefficients need to be adjusted according to the importance of matching the fixed-value terms based on different feature types. Textual semantics usually carry the most important recognition information and are given higher weights; numerical attributes reflect the quantitative characteristics of the fixed value and have moderate importance; positional information provides structured context and has relatively lower weights. Through weighted fusion, the hybrid feature vector can comprehensively express the semantic content, numerical features, and structural position of the fixed-value terms, among other multi-dimensional information.

[0056] In one implementation, generating a text semantic vector based on a fixed-value term describing the text includes the following steps: The text describing the fixed-value items is segmented into words to obtain a word sequence; Load the pre-trained word vector model and use the word vector model to map each word in the word sequence into a high-dimensional word vector; A weighted average is calculated for all high-dimensional word vectors corresponding to the text describing the fixed-value item, and the calculated weighted average vector is used as the text semantic vector of the fixed-value item.

[0057] In this implementation, the word segmentation process first involves text preprocessing, including removing redundant whitespace characters, punctuation marks, and special symbols, and standardizing the capitalization of the text to ensure the standardization of the input text. For Chinese text, a word segmentation algorithm combining dictionary and statistical models is used, such as the jieba word segmenter or other mature Chinese word segmentation tools, which can accurately identify word boundaries and handle ambiguous segmentation problems. For English text, word segmentation is relatively simple, mainly based on spaces and punctuation marks. Special handling is required for technical terms in the relay protection field during word segmentation; a dedicated domain dictionary is established to ensure that technical terms are correctly identified as complete word units and not incorrectly segmented. The word segmentation results generate an ordered sequence of words, maintaining the relative positions of words in the original text. Word vector models are typically trained on massive amounts of text data using algorithms such as Word2Vec, GloVe, or FastText, mapping words to dense vectors of fixed dimensions, where each dimension contains specific semantic information. The model loading process includes steps such as reading the pre-trained model file and initializing the vocabulary and vector matrix. For each word obtained from word segmentation, the corresponding high-dimensional word vector is obtained by querying the vocabulary of the word vector model. The dimension of the word vector is usually set to 100 to 300 dimensions, which can effectively capture the semantic features and contextual relationships of words.

[0058] Next, a weighted average needs to be calculated on all high-dimensional word vectors to generate a vector representation that can represent the semantics of the entire text. The weighted average formula is:

[0059] in This represents the weighted average vector. This represents the word vector of the i-th word. Let L represent the weight of the i-th word, and L represent the total number of words. Weights can be determined based on factors such as word frequency statistics, TF-IDF values, or the word's position in the sentence. For specialized terms in the field of relay protection, higher weights can be assigned to highlight their importance. During the calculation process, it is necessary to ensure that all word vectors have the same dimension; if the dimensions are inconsistent, vector alignment is required.

[0060] In one implementation, calculating the semantic similarity between the mixed feature vectors of the device-side fixed value and the standard fixed value, and performing cross-source matching between the device-side fixed value and the standard fixed value based on the semantic similarity includes the following steps: For any device-side fixed value, calculate the cosine similarity between the first mixed feature vector of the device-side fixed value and the second mixed feature vector of all standard fixed value items; Select the highest cosine similarity among all cosine similarities. If the highest cosine similarity exceeds the preset confidence threshold, it is determined that the device-side fixed value item and the standard fixed value item corresponding to the highest cosine similarity have successfully matched across sources, forming a set of fixed value item pairs. Repeat the above steps until the cosine similarity between the first mixed feature vector and the second mixed feature vector has been calculated for all device-side values.

[0061] In this implementation, the cosine similarity algorithm measures the similarity between two vectors by calculating the cosine of the angle between them. The calculation process first involves calculating the dot product of the two vectors (multiplying corresponding elements and summing the results); then, calculating the magnitudes of the two vectors (square roots of the sum of squares of each element); finally, dividing the dot product by the product of the magnitudes yields the cosine similarity value. The cosine similarity value ranges from -1 to 1, with values ​​closer to 1 indicating greater similarity and values ​​closer to -1 indicating less similarity; a value of 0 indicates orthogonality. For any given device-side value, similarity calculations are performed with each standard value in the standard value sheet, forming a similarity array. All similarity values ​​are then compared, and the maximum value and its corresponding standard value index are recorded. The confidence threshold needs to strike a balance between matching accuracy and coverage, typically determined based on historical data statistics and cross-validation results. When the highest cosine similarity exceeds a preset threshold, the device-side value is confirmed to have successfully matched the corresponding standard value, establishing an association between the value pairs.

[0062] The iterative process of similarity calculation and matching judgment is repeated, using a loop to process each item in the device-side value item list sequentially. For each device-side value item, the complete process of similarity calculation with all standard value items, selection of the highest similarity, and confidence judgment must be executed. During the iteration, a matching status record needs to be maintained to track which device-side value items have been successfully matched and which standard value items have been occupied, avoiding duplicate matching and conflict allocation. After the iteration is completed, complete matching result statistics can be obtained, including the number of successfully matched value item pairs, the number of unmatched device-side value items, and the number of unmatched standard value items.

[0063] In one implementation, the process of normalizing the unit values ​​of the device-side settings that have successfully matched across sources and comparing their values ​​with the standard settings, and generating a setting verification report based on the comparison results, includes the following steps: Iterate through all pairs of fixed values, perform unit normalization on each pair of fixed values, compare the values, and classify the comparison results into consistent or inconsistent items. Identify all standard value items that exist in the standard value sheet but have not been successfully matched across sources, and classify them as missing items; Identify all device-side settings that exist in the device-side settings but have not been successfully matched across sources, and classify them as unplanned items; Summarize all consistent, inconsistent, missing, and unplanned items according to the preset format to generate the final value verification report.

[0064] In this implementation, the unit normalization process establishes a standard unit conversion table, containing unit conversion relationships for various physical quantities such as voltage, current, time, and power, e.g., converting kilovolts to volts, milliamperes to amperes, and milliseconds to seconds. For each setpoint pair, the unit types of the equipment-side setpoint and the standard setpoint are first identified, and then the corresponding conversion coefficients are applied to unify the two values ​​to the same reference unit. Numerical comparison uses a relative error calculation method. When the relative error is less than a preset tolerance threshold, it is determined to be a consistent item; when the error exceeds the threshold, it is determined to be an inconsistent item. The tolerance threshold setting needs to consider factors such as measurement accuracy, equipment error, and engineering practice, and is typically set to a range of 1% to 5%. The missing item identification process iterates through all setpoint items in the standard setpoint list, checking whether each standard setpoint item appears in a successfully matched setpoint pair. Unplanned item identification iterates through the equipment-side setpoint list, checking whether each equipment-side setpoint item participated in a successful cross-source matching. For device-side settings that are not matched, the reasons for their occurrence need to be analyzed. These may include non-standard settings defined by the device manufacturer, settings added after device upgrades, or redundant settings due to configuration errors. The existence of unplanned items may indicate that the device configuration exceeds standard requirements, and their impact on system operation needs to be assessed. For each unplanned item, its complete setting information needs to be recorded, including the setting name, actual value, unit, and its position in the device-side setting table.

[0065] A final setting verification report is generated according to a preset format. The report typically uses a table format, listing detailed information on consistent, inconsistent, missing, and unplanned items. The report also needs to include statistical summary information, such as the total number of setting items, the number of items in each category, and key indicators like the overall consistency rate. The report generation uses a template-based approach to ensure format consistency and readability. The final verification report provides maintenance personnel with a comprehensive and accurate assessment of the setting configuration status, helping to quickly identify configuration problems and develop corresponding solutions, ensuring the reliable operation of relay protection devices.

[0066] The present invention also discloses a real-time adaptive setting verification system for a relay protection device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the real-time adaptive setting verification method for the relay protection device as described in any of the above embodiments.

[0067] The processor can be a central processing unit (CPU). Of course, depending on the actual use, it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc., and this application does not limit it.

[0068] The memory can be an internal storage unit of a computer device, such as a hard disk or RAM, or an external storage device, such as a plug-in hard disk, smart memory card (SMC), secure digital card (SD), or flash memory card (FC) provided on the computer device. Furthermore, the memory can be a combination of internal storage units and external storage devices of a computer device. The memory is used to store computer programs and other programs and data required by the computer device. The memory can also be used to temporarily store data that has been output or will be output. This application does not limit this.

[0069] The present invention also discloses a computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the real-time adaptive setting verification method for relay protection devices described in any of the above embodiments.

[0070] The computer program can be stored in a machine-readable medium. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or certain middleware. The machine-readable medium includes any entity or device capable of carrying computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the machine-readable medium includes, but is not limited to, the above-mentioned components.

[0071] The real-time adaptive setting verification method of the relay protection device in the above embodiments is stored in the computer-readable storage medium and loaded and executed on the processor to facilitate the storage and application of the above method.

[0072] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of protection of this application is limited to these examples; within the framework of this application, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of one or more embodiments of this application as described above, which are not provided in detail for the sake of brevity.

[0073] One or more embodiments in this application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of this application. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of one or more embodiments in this application should be included within the protection scope of this application.

Claims

1. A method for real-time adaptive setting verification of a relay protection device, characterized in that, Includes the following steps: The entropy of data information is calculated on the real-time printed data obtained from the relay protection device, and candidate data streams with significant structured features are identified from the real-time printed data based on a preset entropy threshold. Extract the frame structure feature points in the candidate data stream, and confirm the candidate data stream as a valid value table by constructing a global grid topology model through a geometric consistency test algorithm; The table boundaries of the effective setting table are defined based on the global grid topology model, and the effective setting table is parsed with local abnormal row fault tolerance capability based on the table boundaries to obtain the equipment-side setting items of the relay protection device. The device-side setpoints and the standard setpoints in the preset standard setpoint sheet are respectively fused to obtain a hybrid feature vector containing text semantics, numerical attributes and contextual information; Calculate the semantic similarity between the mixed feature vectors of the device-side fixed value and the standard fixed value, and perform cross-source matching between the device-side fixed value and the standard fixed value based on the semantic similarity; The device-side setpoints that are successfully matched across sources are normalized in units and compared with the standard setpoints. A setpoint verification report is then generated based on the comparison results.

2. The real-time adaptive setting verification method for relay protection devices according to claim 1, characterized in that, The step of calculating the data information entropy of the real-time printed data obtained from the relay protection device and identifying candidate data streams with significant structured features from the real-time printed data based on a preset entropy threshold includes the following steps: Set the window size and sliding step of the sliding window, move the sliding window along the real-time printed data, and calculate the frequency distribution of data bytes within the window position of each sliding window; Based on the frequency distribution of data bytes and applying the Shannon entropy formula, the information entropy value is calculated for each window position, and an entropy value sequence curve corresponding to the information entropy value and the data stream position is generated. When the entropy value sequence curve crosses the preset first entropy value threshold from high to low, the current data stream position is marked as the starting point of the candidate data stream. When the entropy value sequence curve crosses the preset second entropy value threshold from low to high, the current data stream position is marked as the ending point of the candidate data stream. The second entropy value threshold is greater than the first entropy value threshold.

3. The real-time adaptive setting verification method for relay protection devices according to claim 1, characterized in that, The step of extracting the frame structure feature points within the candidate data stream and confirming the candidate data stream as a valid value table using a global mesh topology model constructed through a geometric consistency check algorithm includes the following steps: A predefined character set containing various table drawing characters is used. Traverse all data within the candidate data stream, identify all target characters belonging to the character set, and extract the two-dimensional coordinates of the target characters within the candidate data stream; Combine all two-dimensional coordinates into a two-dimensional point cloud; The global grid topology model, constructed based on two-dimensional point clouds and using a geometric consistency check algorithm, confirms the candidate data stream as a valid value table.

4. The real-time adaptive setting verification method for relay protection devices according to claim 3, characterized in that, The process of confirming candidate data streams as valid value tables using a global grid topology model constructed based on two-dimensional point clouds and through a geometric consistency check algorithm includes the following steps: A minimum point set is randomly sampled from a 2D point cloud, and a global grid topology model is assumed based on the minimum point set. Traverse all points in the 2D point cloud according to the global grid topology model, and determine the points that match the geometric position of the global grid topology model as interior points; Repeat the iterative process of random sampling, model assumptions, and interior point statistics until the preset number of iterations is reached; In all iterations, the global grid topology model that receives the most interior point support is selected as the optimal model. If the proportion of interior points in the optimal model exceeds the preset threshold for the proportion of interior points, then the candidate data stream is confirmed as a valid setpoint table.

5. The real-time adaptive setting verification method for relay protection devices according to claim 1, characterized in that, The step of fusing features between the device-side setpoints and the standard setpoints in the preset standard setpoint sheet to obtain a hybrid feature vector containing text semantics, numerical attributes, and contextual information includes the following steps: For the standard setting items in the device-side setting items or the preset standard setting items, extract the setting item information and the relative position information of the setting item in the corresponding customization table or customization order. The setting item information includes the setting item description text, the setting item value and the setting item unit. Text semantic vectors are generated based on the description text of the fixed value terms. Numerical attribute vectors are generated by combining the numerical value and unit of the fixed value terms and encoding. Contextual feature vectors are generated based on relative position information and encoding. The text semantic vector, numerical attribute vector, and context feature vector are weighted and fused to obtain a hybrid feature vector.

6. The real-time adaptive setting verification method for relay protection devices according to claim 5, characterized in that, The process of generating a text semantic vector based on a fixed-value term description includes the following steps: The text describing the fixed-value items is segmented into words to obtain a word sequence; Load the pre-trained word vector model and use the word vector model to map each word in the word sequence into a high-dimensional word vector; A weighted average is calculated for all high-dimensional word vectors corresponding to the text describing the fixed-value item, and the calculated weighted average vector is used as the text semantic vector of the fixed-value item.

7. The real-time adaptive setting verification method for relay protection devices according to claim 1, characterized in that, The semantic similarity between the hybrid feature vectors of the device-side fixed value and the standard fixed value, and the cross-source matching between the device-side fixed value and the standard fixed value based on the semantic similarity, includes the following steps: For any device-side fixed value, calculate the cosine similarity between the first mixed feature vector of the device-side fixed value and the second mixed feature vector of all standard fixed value items; Select the highest cosine similarity among all cosine similarities. If the highest cosine similarity exceeds the preset confidence threshold, it is determined that the device-side fixed value item and the standard fixed value item corresponding to the highest cosine similarity have successfully matched across sources, forming a set of fixed value item pairs. Repeat the above steps until the cosine similarity between the first mixed feature vector and the second mixed feature vector has been calculated for all device-side values.

8. The real-time adaptive setting verification method for relay protection devices according to claim 7, characterized in that, The process of performing unit normalization and numerical comparison on the device-side setpoints that have successfully matched across sources and the standard setpoints, and generating a setpoint verification report based on the numerical comparison results, includes the following steps: Iterate through all pairs of fixed values, perform unit normalization on each pair of fixed values, compare the values, and classify the comparison results into consistent or inconsistent items. Identify all standard value items that exist in the standard value sheet but have not been successfully matched across sources, and classify them as missing items; Identify all device-side settings that exist in the device-side settings but have not been successfully matched across sources, and classify them as unplanned items; Summarize all consistent, inconsistent, missing, and unplanned items according to the preset format to generate the final value verification report.

9. A real-time adaptive setting verification system for a relay protection device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the real-time adaptive setting verification method for relay protection devices as described in any one of claims 1 to 8.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When executed by a processor, this instruction causes the processor to be configured to perform the real-time adaptive setting verification method for a relay protection device according to any one of claims 1 to 8.

Images