Power distribution network operation optimization method and device based on setting value generation and storage medium

Abstract

The invention discloses a power distribution network operation optimization method and device based on setting value generation and a storage medium, and belongs to the field of power distribution networks. The method comprises the following steps: when the similarity between current operation data and historical operation data is greater than a preset similarity threshold value, taking the corresponding historical operation data as first operation data; the corresponding historical topological structure serves as a first topological structure, and trajectory similarity is obtained through calculation; performing weighted summation according to the trajectory similarity and the historical response characteristics of the corresponding protection equipment to obtain a setting value; constructing a difference measure according to the setting value and a historical constant value in the historical constant value approval list; under the constraint of the constructed penalty function, generating a constant value deviation value according to the difference measure and a preset difference weight coefficient; and determining a final constant value list according to the constant value deviation value and a preset deviation threshold value, and optimizing the current operation data of the to-be-optimized power distribution network. By implementing the invention, the problems of low protection efficiency and accuracy of the power distribution network in the prior art can be solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power distribution network, and particularly relates to a power distribution network operation optimization method and device based on setting value generation and a storage medium. BACKGROUND

[0002] In the field of power distribution network operation, protection setting calculation is a key link to ensure the safe and stable operation of the power grid. Some existing technical solutions have preliminarily realized the automatic generation and arrangement of the setting value sheet, and the related system solutions mainly include the integration of the setting calculation tool and the data platform. Some power distribution enterprises have built professional software with setting calculation function, which can quickly output setting value suggestions according to the operation data in combination with the power grid operation management data platform. Some regions configure an automatic reading module of the setting value sheet based on fixed rule logic in the setting value management system to assist in the preliminary verification of the setting value parameters.

[0003] In the existing setting value sheet approval process, the setting value sheet parameters are manually checked and verified by manually consulting historical experience documents, regulations and specifications. Some systems support the embedding of a part of the standard comparison library and use single threshold judgment logic to assist in the approval. When the submitted setting value sheet meets the set threshold range, it can be automatically approved, otherwise manual intervention is required. For example, whether the current value is within the safety boundary, whether the voltage trip threshold is over-standard, etc. In some actual projects, the setting value sheet still needs to be exported in a document format, and then signed and confirmed by engineers offline or after meeting and auditing. However, in the face of complex fault scenarios and rapid demand response, most single threshold judgment methods cannot find contradictions in the overall configuration and cannot effectively evaluate the interaction between parameters, thereby restricting the efficiency and accuracy of the power distribution network operation protection. SUMMARY

[0004] The embodiments of the present application provide a power distribution network operation optimization method and device based on setting value generation and a storage medium, which can effectively solve the problem that the existing single threshold judgment method cannot find contradictions in the overall configuration and cannot effectively evaluate the interaction between parameters, thereby restricting the efficiency and accuracy of the power distribution network operation protection.

[0005] An embodiment of the present application provides a power distribution network operation optimization method based on setting value generation, comprising: obtaining current operation data, a current topology structure, historical operation data, a historical topology structure, a historical setting value approval sheet of a to-be-optimized power distribution network, and historical response characteristics of corresponding protection equipment; in a case where the similarity between the current operation data and the historical operation data is greater than a preset similarity threshold, taking the corresponding historical operation data as first operation data, and taking the historical topology structure corresponding to the corresponding historical operation data as a first topology structure; Based on the current running data, current topology, current timestamp, first running data, first topology, and corresponding historical timestamps, the trajectory similarity is calculated using a preset weighted time-series matching function. The setting value is obtained by weighted summation based on trajectory similarity and the historical response characteristics of the corresponding protection equipment; and a setting value sheet is generated based on the setting value. A difference metric is constructed based on the set value and the historical set values ​​in the historical set value approval form, and a penalty function is constructed based on the preset target reference set value, the preset maximum offset tolerance, and the parameters corresponding to the set value. Under the constraint of the penalty function, a fixed deviation value is generated based on the difference metric and the preset difference weight coefficient. Based on the deviation value and the preset deviation threshold, the final set value sheet is determined; and the current operating data of the distribution network to be optimized is optimized based on the final set value sheet.

[0006] Furthermore, based on the current running data, current topology, current timestamp, first running data, first topology, and corresponding historical timestamps, trajectory similarity is calculated using a preset weighted temporal matching function, including: Generate a triplet for the current running trajectory based on the current running data, the current topology, and the current timestamp; Based on the first running data, the first topology, and the corresponding historical timestamps, a historical running trajectory triplet is generated; Based on the current trajectory triplet and the historical trajectory triplet, the trajectory similarity is calculated using a preset weighted temporal matching function.

[0007] Furthermore, a weighted summation calculation is performed based on the trajectory similarity and the corresponding historical response characteristics of the protection equipment to obtain the setting value, including: Based on the historical response characteristics of the corresponding protection equipment, determine the total number of factors influencing the setting value of the protection equipment and the variable parameters; Feature transformation is performed on the protection device response records in the historical response characteristics to obtain protection action parameter values; wherein, the protection action parameter values ​​include historical response accuracy, action time deviation rate, and fault type matching degree; Extract the protection action parameter values ​​corresponding to the variable parameters from the historical response characteristics to obtain the historical response characteristic parameter set; The setpoint is obtained by fusing the historical response feature parameter set, the total number of influencing factors, and the trajectory similarity, and then by weighted summation.

[0008] Furthermore, a difference metric is constructed based on the set value and historical set values ​​in the historical set value approval forms, including: The parameters of the setpoint are standardized to obtain the setpoint feature vector; Based on the historical set values ​​in the historical set value approval form, extract the historical set value parameters corresponding to the set value to form a set of historical set value parameters; Calculate the mean vector and covariance matrix of historical constant values ​​based on the historical constant value parameter set; A difference measure is constructed based on the eigenvector of the setpoint, the mean vector of the historical setpoints, and the covariance matrix.

[0009] Furthermore, the penalty function is constructed using the following formula: in, Represents the penalty function; Indicates the kth i The parameters corresponding to each setpoint; This indicates the preset target reference value; This indicates the preset maximum offset tolerance.

[0010] Furthermore, under the constraint of the penalty function, a fixed deviation value is generated based on the difference metric and the preset difference weight coefficient, including: Under the constraint of the penalty function, the parameter corresponding to the tuning value with a penalty function value of 0 is selected as the first tuning parameter; Assign a preset difference weight coefficient based on the first selected tuning parameter; The baseline deviation value is calculated based on the difference measure and the assigned difference weight coefficients. The parameter corresponding to the tuning value of the penalty function that is greater than 0 is used as the second tuning parameter; The penalty function value of the second tuning parameter and the preset penalty weight coefficient are weighted and calculated to obtain the over-limit compensation value; The baseline deviation value and the excess compensation value are summed to obtain the fixed deviation value.

[0011] Furthermore, based on the deviation from the set value and the preset deviation threshold, the final set value sheet is determined, including: If the deviation value is less than or equal to the preset deviation threshold, the generated value sheet is deemed qualified and is used as the final value sheet. If the deviation of the set value is greater than the preset deviation threshold, the second setting parameter is marked so that the operator can review the marked set value sheet and regenerate the set value sheet.

[0012] Furthermore, based on the final setpoint sheet, the current operating data of the distribution network to be optimized is used for optimization, including: The setting parameters in the final setting sheet are categorized according to the type of protection equipment to generate an equipment parameter configuration table; wherein, the equipment parameter configuration table includes parameter name, target value and allowable fluctuation range; The device parameter configuration table is sent to the corresponding protection device, triggering the device parameter update. The current operating data is compared with the parameter thresholds in the final set value sheet to monitor whether there are any operating conditions where the operating data exceeds the allowable fluctuation range of the set value. If present, adjust the load of non-core users to bring the operating data back to the set operating fluctuation range; If it does not exist, continuously monitor the current operating parameters.

[0013] As an improvement to the above solution, another embodiment of the present invention provides a distribution network operation optimization device based on setting value generation, comprising: The data acquisition module is used to acquire the current operating data, current topology, historical operating data, historical topology, historical setting approval forms, and the historical response characteristics of the corresponding protection devices of the distribution network to be optimized. The data similarity matching module is used to select the corresponding historical running data as the first running data when the similarity between the current running data and the historical running data is greater than a preset similarity threshold; and to select the historical topology structure corresponding to the corresponding historical running data as the first topology structure. The trajectory similarity calculation module is used to calculate the trajectory similarity based on the current running data, the current topology, the current timestamp, the first running data, the first topology, and the corresponding historical timestamps, using a preset weighted time-series matching function. The setting value generation module is used to perform weighted summation calculation based on trajectory similarity and the historical response characteristics of the corresponding protection equipment to obtain the setting value; and to generate a setting value sheet based on the setting value. The difference measurement construction module is used to construct difference measurement based on the set value and the historical set value in the historical set value approval form, and to construct a penalty function based on the preset target reference set value, the preset maximum offset tolerance and the parameters corresponding to the set value; The constant value deviation determination module is used to generate a constant value deviation value under the constraint of the penalty function, based on the difference metric and the preset difference weight coefficient. The distribution network data optimization module is used to determine the final setpoint list based on the setpoint deviation value and the preset deviation threshold; and to optimize the current operating data of the distribution network to be optimized based on the final setpoint list.

[0014] Another embodiment of the present invention provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the distribution network operation optimization method based on setting value generation described in the above embodiment.

[0015] By implementing this invention, at least the following beneficial effects are achieved: This invention provides a method, device, and storage medium for optimizing distribution network operation based on setpoint generation. The method automatically acquires historical and current operating data and topology, calculates trajectory similarity using a weighted time-series matching function, and derives setpoints. Most of the process is automated, reducing manual review and verification, and enabling rapid generation of setpoint lists. Setpoints are calculated based on historical response characteristics and trajectory similarity, allowing for timely determination of setpoint lists in rapid demand response scenarios, ensuring timely adjustments to the distribution network and preventing fault escalation. The method comprehensively considers operating data, topology, and time-series information to calculate trajectory similarity, and obtains setpoints by weighting similarity with historical response characteristics. This considers multiple parameters and the overall system operating status, avoiding misjudgments caused by isolated parameter evaluations. A difference metric is constructed based on the setpoints and historical setpoints, and setpoint deviations are calculated under penalty function constraints. The penalty function, combined with target reference setpoints and maximum offset tolerances, effectively identifies setpoints that contradict historical experience or safety requirements, overcoming the limitation of single-threshold logic in detecting overall configuration contradictions. By utilizing historical data and combining it with current timestamps to dynamically calculate setting values, the system can adapt to changes in the operating status of the distribution network. Compared to fixed single threshold judgments, this approach better reflects the complex and ever-changing actual operating conditions, ensuring reliable operation of protection equipment. A comprehensive evaluation of whether each setting parameter conforms to reasonable logical relationships based on historical experience enables panoramic risk control and collaborative optimization of distribution network protection configurations. This effectively improves the overall coordination of the protection system, significantly enhancing the deep logical reasoning and anomaly detection capabilities during the approval process, and effectively improving the intelligence level of the approval process and the rationality of the protection system configuration. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a distribution network operation optimization method based on setting value generation according to an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the working principle of an optimization method implemented in software according to an embodiment of the present invention. Figure 3 This is another working principle diagram of the optimization method provided in one embodiment of the present invention, implemented by software; Figure 4 This is a schematic diagram of the structure of a power distribution network operation optimization device based on setting value generation, provided in an embodiment of the present invention. Detailed Implementation

[0017] 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.

[0018] See Figure 1 To address the problem that existing single-threshold judgment methods struggle to detect overall configuration inconsistencies and effectively assess parameter interactions, thus hindering the efficiency and accuracy of distribution network operation protection, an embodiment of this invention provides a flowchart illustrating a distribution network operation optimization method based on setting value generation, comprising: S1. Obtain the current operating data, current topology, historical operating data, historical topology, historical setting approval forms, and historical response characteristics of the corresponding protection devices of the distribution network to be optimized; Specifically, current operating data includes current, voltage, frequency, and load data; the current topology is the physical structure information of the distribution network, such as the current line connection methods and equipment layout, including line connection relationships, switch status, and transformer locations. Historical operating data consists of records of operating parameters of the distribution network over a past period, consistent with the current operating data type. The historical topology is the physical structure information of the distribution network, including past line connection methods and equipment layout. Historical setting approval forms are documents generated for the approval of setting values ​​for distribution network protection equipment in the past, containing information such as approved setting parameters, approval time, and approval basis. The historical response characteristics of protection equipment are the response characteristics of protection equipment to various faults or operating conditions during past operation, such as action time, action accuracy, false tripping, and failure to trip.

[0019] S2. If the similarity between the current running data and the historical running data is greater than a preset similarity threshold, the corresponding historical running data shall be used as the first running data; and the historical topology corresponding to the corresponding historical running data shall be used as the first topology. Specifically, the preset similarity threshold is a manually set critical value used to judge the degree of similarity between the current running data and historical running data. The first running data refers to the historical running data selected from the historical running data whose similarity to the current running data is greater than the preset similarity threshold. The first topology refers to the historical topology corresponding to the first running data.

[0020] S3. Based on the current running data, current topology, current timestamp, first running data, first topology, and corresponding historical timestamps, calculate the trajectory similarity using a preset weighted time-series matching function; Specifically, the weighted time-series matching function represents a function that considers different parameter weights and time-series features, used to calculate the similarity between the current trajectory and historical trajectories. The trajectory similarity is an index of the similarity between the current trajectory and historical trajectories calculated using the weighted time-series matching function.

[0021] Preferably, the trajectory similarity is calculated using a preset weighted temporal matching function based on the current running data, the current topology, the current timestamp, the first running data, the first topology, and the corresponding historical timestamps, including: Based on the current running data, the current topology, and the current timestamp, generate a current running trajectory triplet; based on the first running data, the first topology, and the corresponding historical timestamp, generate a historical running trajectory triplet; based on the current running trajectory triplet and the historical running trajectory triplet, calculate the trajectory similarity using a preset weighted temporal matching function.

[0022] Specifically, the current operating trajectory triplet represents a triplet consisting of current operating data, current topology, and current timestamp, used to characterize the current operating trajectory of the distribution network. The historical operating trajectory triplet represents a triplet consisting of first operating data, first topology, and the corresponding historical timestamp, used to characterize the historical operating trajectory of the distribution network.

[0023] Schematic, current operating data, current topology, and current timestamp are combined to generate a current operating trajectory triplet, which fully describes the current operating status and time information of the distribution network. First operating data, first topology, and corresponding historical timestamps are combined to generate a historical operating trajectory triplet, which fully describes the operating status and time information of the distribution network under similar historical scenarios. A preset weighted time-series matching function is used to calculate the trajectory similarity between the current and historical operating trajectory triplets, making the trajectory similarity calculation more targeted and accurate. By integrating multiple aspects of information in the form of triplets, and considering the influence of time and topology factors on the operating trajectory, the trajectory similarity calculation results are more closely aligned with actual conditions, providing a more reliable basis for determining subsequent setting values.

[0024] In a preferred embodiment of the present invention, the triplet forms of the current trajectory triplet and the historical trajectory triplet are as follows: ,in, For a moment Vectors of operational data (such as bus voltage, current, power supply feedback, etc.). This refers to the actual response behavior of the protection devices in the topology (such as whether they operate and how long the delay is). The timestamp indicating the occurrence of the event is used to identify the periodicity or seasonality of external disturbances.

[0025] Schematic representation: The current trajectory triples are represented as vectors. The trajectory similarity is calculated using the following formula: Among them, Sim t For trajectory similarity, The preset feature difference penalty coefficient, This is a preset time decay factor used to suppress the effects of events that are too old. This is the current timestamp. This corresponds to the historical timestamp. ||P now -P t || represents the squared Euclidean distance between the current trajectory triplet and the historical trajectory triplet, which measures the degree of difference between them in the feature space.

[0026] S4. Perform a weighted summation calculation based on trajectory similarity and the historical response characteristics of the corresponding protection equipment to obtain the setting value; and generate a setting value sheet based on the setting value. Specifically, setting values ​​refer to parameter values ​​set to ensure the correct operation of protection equipment, such as operating current, operating voltage, and delay time. Setting sheets are documents recording the setting values ​​of protection equipment and are an important basis for the operation of the protection equipment.

[0027] Preferably, the setting value is obtained by weighted summation based on trajectory similarity and the historical response characteristics of the corresponding protection equipment, including: determining the total number of influencing factors and variable parameters of the protection equipment setting value based on the historical response characteristics of the corresponding protection equipment; performing feature transformation on the protection equipment response records in the historical response characteristics to obtain protection action parameter values; wherein, the protection action parameter values ​​include historical response accuracy, action time deviation rate, and fault type matching degree; extracting the protection action parameter values ​​corresponding to the variable parameters from the historical response characteristics to obtain a historical response feature parameter set; and fusing the historical response feature parameter set, the total number of influencing factors, and trajectory similarity, and obtaining the setting value through weighted summation.

[0028] Specifically, the total number of influencing factors refers to the number of factors affecting the setting values ​​of protection equipment, such as fault type and load size. Variable parameters represent parameters in the protection equipment setting values ​​that may change with operating conditions, such as operating current and delay time. Protection action parameter values ​​represent parameter values ​​obtained by feature transformation of the protection equipment's historical response records, used to reflect the protection equipment's operating characteristics. Historical response accuracy represents the probability that the protection equipment correctly responds to faults or operating conditions in historical operation. Action time deviation rate represents the degree of deviation between the actual action time and the preset action time of the protection equipment. Fault type matching degree represents the degree of matching between the protection equipment's action and the fault type. The historical response feature parameter set represents the set of protection action parameter values ​​corresponding to the variable parameters, collectively reflecting the historical response characteristics of the protection equipment.

[0029] Schematic, the tuning value is generated using the following inference function: in, To protect the equipment The set value, Represents the triplet of historical trajectory Medium protection equipment The historical response characteristics, r represents the total number of factors affecting the setting value in the equipment, and t is the variable parameter in the equipment, which can represent different parameters. The action feature extraction function can be a transformation function for trip current, delay time, etc., and is based on the historical response features R of the protection device in each historical operation trajectory triplet. t,i The system extracts representative parameter values ​​that reflect the characteristics of protection actions for subsequent setting inference. The output setting values ​​are adaptable to the current operating conditions and take into account historical environmental conditions and behavioral response differences, which can reduce the setting error rate and is suitable for distribution network environments with frequent load fluctuations, significant seasonal differences, or unstable structures.

[0030] Schematic representation of the action feature extraction function To illustrate, we will extract features from the trip current. Assume a protection device tripped during a historical fault, and the trip current record shows 680A. The record indicates a tripping action, with a current value of 680A during the trip. Therefore, the circuit breaker will trip. Directly extract the operating current value =680. When deriving new overcurrent protection settings, the system can integrate several historical operating current data points to calculate a tripping threshold value suitable for the current operating conditions. Similarly, in the task of extracting operating features from multi-feature historical data, The design aims to extract combined values ​​of multiple action features. For example, it simultaneously extracts action current, action delay, and action type (e.g., trip / alarm) to form a vector feature, which is then used for subsequent, more complex setting derivation logic. For example, defining... : ,in , , These represent operating current, operating delay, and operating type, respectively. Figure 2 As shown, the values, units, and physical meanings of these three parameters are completely different, so they need to be mapped to a corresponding vector for mathematical calculation. Here, the parameter values ​​of the three items are extracted and combined to form an action feature vector.

[0031] In a preferred embodiment of the present invention, the number of factors and variable parameters affecting the setting value are determined based on the historical response characteristics of the protection device, thus clarifying the relevant factors and variables for the setting value calculation. The historical response records of the protection device are processed and transformed to obtain protection action parameter values, including historical response accuracy, action time deviation rate, and fault type matching degree, quantifying the historical response characteristics. Protection action parameter values ​​corresponding to the variable parameters are extracted from the historical response characteristics to form a historical response characteristic parameter set, thus associating the historical response characteristics with the variable parameters. The historical response characteristic parameter set, the total number of influencing factors, and trajectory similarity are fused, and the setting value is obtained through weighted summation. The setting value is determined by comprehensively considering multiple factors, improving its rationality. By fully utilizing the historical response characteristics of the protection device, the calculated setting value better reflects the actual operating conditions of the protection device, improving the accuracy of the setting value. By comprehensively considering multiple factors, the influence of a single factor on the setting value is avoided, making the setting value more reasonable and reliable.

[0032] Building upon traditional disturbance response analysis, this method integrates event sequences and historical equipment behavior trajectories over time to dynamically adjust the setting logic. It addresses three key issues: first, the varying response characteristics of similar topologies in distribution networks at different historical moments; second, the problems of "over-level protection" and "protection gaps" during the setting process; and third, the time delay and multi-stage response issues between environmental disturbances and equipment behavior. This optimized method learns judgment logic from historical approval samples, significantly enhancing the intelligence and reliability of the approval process.

[0033] Existing technologies, based on single-threshold logic judgments, struggle to detect inconsistencies in overall configuration, protection coordination failures, or operational timing conflicts. This embodiment's optimized method, however, introduces trajectory similarity analysis between historical and current operating trajectories. It systematically learns from a large number of approved setting parameter configurations, extracting the collaborative relationships and statistical characteristics between key setting parameters. By fusing similarity with historical response features and using dynamically adaptive weighted summation, it combines similarity with historical response features. During setting value approval, it performs individual compliance judgments on each setting parameter, comprehensively evaluating whether each parameter conforms to reasonable logical relationships from historical experience. This accurately identifies problems that traditional single-threshold judgment methods struggle to detect, such as individual parameter compliance but overall configuration inconsistencies, protection coordination failures, or operational timing conflicts. This achieves panoramic risk control and collaborative optimization of distribution network protection configurations, effectively improving the overall coordination of the protection system. It significantly enhances deep logical reasoning and anomaly detection capabilities during the approval process, effectively improving the intelligence level of the approval process and the rationality of the protection system's configuration.

[0034] On the other hand, traditional setting value calculations are mostly based on current snapshots or static rules, which cannot accurately reflect the differences in historical operating conditions of the distribution network and the behavior and response patterns of equipment. This leads to poor setting accuracy under topology changes or seasonal disturbances, resulting in protection gaps or the risk of cascading trips. This embodiment introduces a setting inference algorithm based on event time series. By constructing a tripartite structure of operating data, response, and time, it extracts historical operating trajectories similar to the current network state and performs weighted inference based on response characteristics to generate setting values. It comprehensively considers the evolution of the power grid structure, load changes, and equipment response timing to achieve dynamic modeling of setting value generation. This significantly improves the adaptability of setting parameters to complex operating conditions, enabling automatic matching of the optimal setting scheme for seasonal changes, high load fluctuations, and multi-source mixed scenarios, reducing the false trip rate and enhancing the accuracy and stability of system protection.

[0035] S5. Construct a difference measure based on the set value and the historical set value in the historical set value approval form, and construct a penalty function based on the preset target reference set value, the preset maximum offset tolerance and the parameters corresponding to the set value; Specifically, the difference metric represents an indicator used to measure the degree of difference between the setting value and the historical setting value in the historical setting approval form. The preset target reference setting value represents the ideal setting value of the protection equipment preset according to the distribution network operation specifications, equipment characteristics, etc. The preset maximum offset tolerance represents the maximum range of deviation of the setting value from the target reference setting value. The penalty function represents a function used to penalize the setting value that exceeds the maximum offset tolerance, so as to constrain the setting value within a reasonable range.

[0036] Preferably, a difference metric is constructed based on the set value and historical set values ​​in historical set value approval forms, including: The parameters of the setpoint are standardized to obtain the setpoint feature vector; Based on the historical set values ​​in the historical set value approval form, extract the historical set value parameters corresponding to the set value to form a set of historical set value parameters; Calculate the mean vector and covariance matrix of historical constant values ​​based on the historical constant value parameter set; A difference measure is constructed based on the eigenvector of the setpoint, the mean vector of the historical setpoints, and the covariance matrix.

[0037] Specifically, standardization refers to processing the parameters of the setpoint to eliminate the influence of different dimensions, making them comparable. The setpoint eigenvector represents the vector formed by combining the standardized parameters of the setpoint, used to characterize the features of the setpoint. The historical setpoint parameter set represents the set of historical setpoint parameters corresponding to the setpoint extracted from historical setpoint approval forms. The mean vector represents the vector composed of the average values ​​of the parameters in the historical setpoint parameter set. The covariance matrix is ​​a matrix used to describe the correlation between parameters in the historical setpoint parameter set.

[0038] Schematic, the parameters of the tuning value are standardized to obtain the tuning value feature vector. , of which each Parameters representing a setting value, such as overcurrent setting, instantaneous overcurrent coefficient, operating delay, and trip current limit. Historical setting parameter sets are denoted as... ,in, This represents the i-th complete historical fixed-value approval form, where... This refers to the first compliant historical setting approval form in the sample set. Essentially, it's a high-dimensional feature vector, where each element represents a specific protection setting parameter, such as "overcurrent action setting" or "trip current upper limit." To characterize the distribution center and fluctuation range of the historical setting parameter set, the mean vector of the historical setting parameters across all historical setting parameter sets is calculated. With covariance matrix They are respectively denoted as: ; Where m represents the number of historical value approval orders in the historical value parameter set, that is, the total number of historical value approval orders, the parentheses are a vector, and T represents transpose, which has no physical meaning.

[0039] The Mahalanobis distance is used to construct a measure of the difference between the distribution of new fixed-value parameters and the historical fixed-value parameter sets. The specific form is as follows: The difference metric is used to measure the degree to which a new parameter point deviates from the center of the historical sample distribution. This represents the eigenvector of the new fixed value, i.e., the eigenvector of the fixed value.

[0040] In an optional implementation, a classification model based on support vector machines (SVMs) is used for approval judgment. SVMs construct a boundary hyperplane in a high-dimensional space to effectively separate compliant and non-compliant samples, demonstrating good discriminative ability for scenarios with small sample sizes but high feature discrimination. Although its interpretability is slightly weaker than the Mahalanobis distance model, it has advantages in model training stability and handling nonlinear relationships, achieving the same technical effect as in this embodiment: quickly identifying and adjusting abnormal parameters and making judgments.

[0041] In a preferred embodiment of the present invention, the parameters of the setpoint are standardized to obtain a feature vector of the setpoint, eliminating the influence of dimensions and facilitating subsequent comparison with historical setpoints. Historical setpoint parameters corresponding to the setpoint are extracted from historical setpoint approval forms to form a historical setpoint parameter set, providing reference data for calculating the difference measure. The mean vector and covariance matrix of the historical setpoints are calculated based on the historical setpoint parameter set, reflecting the overall level of the historical setpoints and the correlation between parameters. Combining the feature vector of the setpoint, the mean vector of the historical setpoints, and the covariance matrix, a difference measure is constructed to quantify the difference between the setpoint and the historical setpoints. This allows the difference measure to more accurately reflect the difference between the setpoint and the historical setpoints, providing a reliable basis for the verification of the setpoint.

[0042] Preferably, the penalty function is constructed using the following formula: ;in, Represents the penalty function; Indicates the kth i The parameters corresponding to each setpoint; This indicates the preset target reference value; This indicates the preset maximum offset tolerance.

[0043] The calculation method of the penalty function is determined by the formula. The penalty is imposed according to the degree of deviation between the setpoint parameter and the target reference setpoint, which constrains the setpoint parameter to a reasonable range. This makes the construction of the penalty function more deterministic and operable, and can effectively constrain the setpoint parameter to prevent it from deviating too much from the target reference setpoint, thus ensuring the rationality and safety of the setpoint.

[0044] S6. Under the constraint of the penalty function, a fixed deviation value is generated based on the difference metric and the preset difference weight coefficient. Specifically, the preset difference weighting coefficient represents the weighting coefficient set according to the importance of different parameters to the operation of the distribution network. The setpoint deviation value represents the index obtained by combining the difference measurement and the difference weighting coefficient under the constraint of the penalty function, which measures the degree to which the setpoint deviates from the reasonable range.

[0045] Preferably, under the constraint of the penalty function, a fixed deviation value is generated based on the difference metric and a preset difference weight coefficient, including: Under the constraint of the penalty function, the parameter corresponding to the tuning value with a penalty function value of 0 is selected as the first tuning parameter; Assign preset difference weight coefficients based on the first selected tuning parameter; The baseline deviation value is calculated based on the difference measure and the assigned difference weight coefficients; The parameter corresponding to the tuning value of the penalty function that is greater than 0 is used as the second tuning parameter; The penalty function value of the second tuning parameter and the preset penalty weight coefficient are weighted and calculated to obtain the over-limit compensation value; The baseline deviation value and the excess compensation value are summed to obtain the fixed deviation value.

[0046] Specifically, the first setting parameter represents the setting value parameter with a penalty function value of 0, meaning these parameters are within the preset maximum offset tolerance range. The basic deviation value represents the deviation value calculated based on the difference measure of the first setting parameter and the corresponding difference weight coefficient. The second setting parameter represents the setting value parameter with a penalty function value greater than 0, meaning these parameters exceed the preset maximum offset tolerance. The over-limit compensation value represents the compensation value calculated based on the penalty function value and penalty weight coefficient of the second setting parameter, reflecting the impact of the over-limit parameter on the setting value deviation. The setting value deviation value is the sum of the basic deviation value and the over-limit compensation value, comprehensively reflecting the degree of deviation of the setting value.

[0047] Indicatively, the deviation from the set value R ,in The differential weighting coefficients are determined primarily based on the proportion of the overall characteristic distribution and individual parameter deviations in the impact on system risk during actual fault handling, through empirical setting, data-driven optimization, and scenario tuning.

[0048] By distinguishing between the first and second setting parameters, calculating the basic deviation value and the over-limit compensation value separately, and then adding the two together to obtain the setting deviation value, the deviation of different parameters is fully considered, making the calculation of the setting deviation value more accurate. This allows the setting deviation value to more comprehensively and accurately reflect the degree of deviation of the setting value, providing a reliable basis for the determination of the final setting sheet.

[0049] S7. Determine the final set value sheet based on the set value deviation and the preset deviation threshold; and optimize the current operating data of the distribution network to be optimized based on the final set value sheet.

[0050] Specifically, the preset deviation threshold represents the critical value used to determine whether the deviation from the setpoint is within an acceptable range. The final setpoint sheet represents the setpoint sheet determined after a series of verifications and adjustments, which can be used to guide the optimization of distribution network operation.

[0051] Preferably, the final setpoint sheet is determined based on the setpoint deviation and a preset deviation threshold, including: If the deviation value is less than or equal to the preset deviation threshold, the generated value sheet is deemed qualified and is used as the final value sheet. If the deviation of the set value is greater than the preset deviation threshold, the second setting parameter is marked so that the operator can review the marked set value sheet and regenerate the set value sheet.

[0052] Specifically, when the deviation of the setpoint is less than or equal to the preset deviation threshold, it indicates that the deviation of the setpoint is within an acceptable range, and the generated setpoint sheet is deemed qualified and used as the final setpoint sheet. When the deviation of the setpoint is greater than the preset deviation threshold, it indicates that the deviation of the setpoint exceeds the acceptable range, the second setpoint parameter is marked, and the operator is prompted to review the marked setpoint sheet and regenerate the setpoint sheet to ensure the accuracy of the final setpoint sheet.

[0053] Indicatively, the preset deviation threshold Determined according to needs, the general reference range is [0.5, 2.0]. For example... Figure 3 As shown, when When the time comes, the system automatically determines that the setting sheet is qualified and performs a secondary knowledge rule matching verification; if If the value is not found, the setting sheet will be marked as an item pending review for further expert confirmation. All calculations are completed in real time, and the sample set and reference features are continuously updated in the background, forming a closed-loop learning capability.

[0054] Existing approval mechanisms largely rely on static rules and manual judgment, making it difficult to identify coupling relationships between parameters and statistical anomalies. This results in low approval efficiency, high misjudgment rates, and a lack of adaptability. This embodiment standardizes the setpoint parameters into high-dimensional feature points, uses Mahalanobis distance to assess their deviation from the distribution of historical approval samples, and introduces a parameter offset penalty mechanism based on physical constraints to construct setpoint deviation values, quantitatively judging the compliance of setpoint sheets. Preliminary approval decisions are automatically completed based on the comparison of the setpoint deviation value with a threshold. This improves the intelligence and interpretability of the approval process, reduces manual intervention, and increases approval efficiency and accuracy, especially in power grid scenarios with multi-source data fusion and significant regional differences. Different processing methods are adopted based on the comparison results of the setpoint deviation value and the preset deviation threshold. Qualified sets are directly determined as final setpoint sheets, while unqualified sets are regenerated through manual review, ensuring the reliability of the final setpoint sheets. The combination of automatic judgment and manual review improves the accuracy and reliability of the final setpoint sheets. Marking and reviewing non-compliant value sheets allows for the timely detection and correction of problems, preventing the use of unreasonable value sheets in distribution network operation and ensuring the safe operation of the distribution network.

[0055] Preferably, optimization is performed based on the current operating data of the distribution network to be optimized according to the final setpoint sheet, including: The setting parameters in the final setting sheet are categorized according to the type of protection equipment to generate an equipment parameter configuration table; wherein, the equipment parameter configuration table includes parameter name, target value and allowable fluctuation range; The device parameter configuration table is sent to the corresponding protection device, triggering the device parameter update. The current operating data is compared with the parameter thresholds in the final set value sheet to monitor whether there are any operating conditions where the operating data exceeds the allowable fluctuation range of the set value. If present, adjust the load of non-core users to bring the operating data back to the set operating fluctuation range; If it does not exist, continuously monitor the current operating parameters.

[0056] Specifically, the equipment parameter configuration table is a list of setting parameters in the final setting sheet, categorized by protection equipment type, facilitating parameter updates and management of the protection equipment. The parameter name is the name of the setting parameter, such as operating current or delay time. The target value is the target set value of the setting parameter. The allowable fluctuation range is the range of fluctuations allowed for the setting parameter during operation.

[0057] In a preferred embodiment of the present invention, the equipment parameter configuration table is sent to the corresponding protection equipment, triggering equipment parameter updates so that the protection equipment operates according to the new setting values. The current operating data is compared with the parameter thresholds in the final setting sheet to monitor whether there are any operating conditions where the operating data exceeds the allowable fluctuation range of the setting, thus promptly grasping the operating status of the distribution network. If such conditions exist, the load of non-core users is adjusted to bring the operating data back to the allowable fluctuation range of the setting, ensuring the stable operation of the distribution network. If no such conditions exist, the current operating parameters are continuously monitored to ensure the normal operation of the distribution network. Through parameter updates, operation monitoring, and load regulation, the operating status of the distribution network can be adjusted in a timely manner to ensure its operation within a reasonable range, improving the operating efficiency and safety of the distribution network. Measures to regulate the load of non-core users in response to abnormal situations balance the grid load and ensure the stable power supply to core users.

[0058] Indicatively, it also includes: S10. Collect parameters from the final setpoint sheet, distribution network operating status, equipment configuration, topology, historical fault trajectories, and protection action responses, and standardize the data to unify the units and formats.

[0059] The key to the secondary review of setting parameters that have already passed intelligent approval lies in integrating expert knowledge and actual operating conditions to conduct in-depth analysis and optimization of the setting sheet's logical rationality, protection coordination, and consistency with historical experience. The collected data, including all parameters in the setting sheet and related information such as distribution network operating status, equipment configuration, topology, historical fault trajectories, and protection action responses, is used as input. All information is uniformly converted into operating condition feature tensors and then enters the expert rule matching process.

[0060] The feature tensor obtained during standardization is a mathematical structure used to represent multi-dimensional distribution network operating condition data. By encoding multi-dimensional information, the operating status of the entire distribution network is represented, ensuring accurate reflection of various parameters under different operating conditions. Specific input data collection and processing can be represented as follows: single-parameter settings, such as overcurrent setting, instantaneous overcurrent coefficient, and action delay; distribution network operating status, such as electrical parameters like voltage, current, and power; equipment configuration, such as protection equipment model, wiring method, and capacity; topology, including structural information such as nodes, branches, and buses of the distribution network; historical fault trajectories, including the occurrence time, type, and fault section of past fault events; and protection action responses, such as protection action time, trip current, and delay.

[0061] S11. Construct a multi-dimensional operating condition feature tensor, corresponding to the following dimensions: setting single parameter, distribution network operating status, equipment configuration, topology, historical fault trajectory, and protection action response, and encode the data of each dimension into a vector or matrix.

[0062] This information is acquired through different data collection systems and standardized (e.g., Z-score standardization) to ensure consistency across different data sources and units. The dimensions of the feature tensor typically correspond to different categories of the input data. Each dimension represents a different feature. Here are some examples of possible dimensions for a feature tensor: Dimension 1: Setting single parameters (e.g., overcurrent setting, action delay, etc.), each setting parameter is an independent dimension; Dimension 2: Distribution network operating status (e.g., bus voltage, current, etc.), each electrical parameter is an independent dimension; Dimension 3: Equipment configuration (e.g., equipment model, capacity, etc.), each type of equipment corresponds to one dimension; Dimension 4: Topology (e.g., nodes, branches, etc.), each key topological element is a dimension; Dimension 5: Historical fault trajectories (e.g., fault time, fault type, etc.), each historical fault record is a dimension; Dimension 6: Protection action response (e.g., trip current, delay, etc.), each response parameter is an independent dimension.

[0063] The dimensions of the feature tensor can be flexibly expanded according to the actual application. The feature value Xi of each dimension is represented by a vector through its corresponding dataset. The feature vectors of each dimension will have a uniform length so that information from different dimensions can be combined into the same tensor. For example, the overcurrent setting (X1) can be a scalar value (e.g., 500A); the bus voltage (X2) may be a vector (e.g., voltage values ​​at different times); the historical fault trajectory (X5) may be a time series vector; therefore, the final feature tensor structure will be A=(X1,X2,X3,X4,X5,X6).

[0064] S12. Extract rules from historical engineering cases, expert experience, and power grid regulations, and store the rules using knowledge graphs or databases.

[0065] An expert database is constructed, consisting of numerous historical engineering cases, expert experience summaries, and rules extracted from power grid operation procedures. Each rule defines a set of applicable operating conditions and corresponding setting and adjustment suggestions, and adopts a structural modeling that separates the condition header from the suggestion body.

[0066] S13. Calculate the matching degree between the current working condition feature tensor and the rules, set a threshold to filter rules that meet the matching degree, and form a rule set.

[0067] By constructing a matching score function, a multidimensional similarity analysis is performed on the input working conditions and expert rules. Given the current input feature tensor... Expert Rules The center of the corresponding conditional domain is 'n' represents the number of features involved in rule matching, i.e., the number of features participating in the matching degree calculation in the expert rules and input conditions. Each feature has a tolerance parameter. The system calculates the degree of matching between the rule and the current operating conditions. , expressed as: in, The importance coefficient for each feature dimension is used to adjust the influence of different parameters in the matching process. This matching function not only considers the similarity of parameter values ​​but also retains a tolerance mechanism to adapt to disturbances and uncertainties in the field environment. The tolerance parameter represents the j-th feature, which is usually related to the standard deviation of the feature and describes the range of deviations that the feature can tolerate under operating conditions.

[0068] In an optional implementation, a knowledge reasoning method based on fuzzy logic rule trees can also be used to determine the matching degree. This method establishes fuzzy membership functions by fuzzifying rule conditions (such as "excessive current" or "complex topology"), and completes the recommendation judgment of tuning suggestions in a hierarchical rule manner. The fuzzy logic method has advantages in rule readability and ease of human maintenance, and is especially suitable for systems with small rule sets and frequent expert participation in rule maintenance. It can also achieve the goal of "automatic matching and suggestion output of expert knowledge under multiple operating conditions".

[0069] S14. For each rule in the rule set, calculate its confidence score for the tuning parameters, and select the working conditions that correspond to the expected confidence scores.

[0070] After matching is complete, the system performs a weighted aggregation of the suggestion strength for each rule in the matched rule set to obtain a confidence score for the tuning parameter adjustment suggestion. Let the rule set be... , No. Rule number 1 The recommended value for each adjustment suggestion is The historical validity weight of the rule is Then the confidence score for this suggestion is Among them, the historical validity weight is By evaluating the success rate, applicability, and expert experience of rules in historical applications, and dynamically adjusting their weights based on statistical data, the effectiveness and credibility of the rules under specific working conditions can be reflected; M r This indicates the degree of matching between rule r and the current operating condition. This represents the recommended value of rule r for the i-th tuning suggestion.

[0071] Determine whether to recommend modifying the value based on the confidence score results. If the value exceeds the set threshold, the item is marked as "needs adjustment," and the optimal suggested value and its rationale are provided. If multiple possible solutions exist, they are output in order of score for manual confirmation or automatic processing. The final verification result includes the recommended adjustment value, a summary of the hit rule, the matching score, and historical case references, and is uniformly fed back for approval. This is used to update the model's experience samples, forming a dynamic collaborative evolution loop between knowledge and algorithms.

[0072] Traditional expert experience exists in the form of unstructured documents, which cannot be recognized, retrieved, or invoked by the system. This results in low experience reuse rates, the inability to automate the knowledge verification process, and heavy reliance on individual experts. This embodiment constructs a structured expert rule base, abstracting experiential knowledge into condition-suggestion pairs of rules. Through feature tensor matching and similarity calculation functions, it achieves dynamic rule matching and multi-rule weighted reasoning, outputting adjustment suggestions and confidence scores. This enables automatic matching and reasoning judgment for knowledge invocation, improving the comprehensiveness and efficiency of verification. It achieves the digitization, computation, and continuous accumulation of expert knowledge, giving the system the ability to make secondary judgments covering complex working conditions, compensating for the blind spots of intelligent algorithms in boundary scenarios, and enhancing overall robustness and decision-making depth.

[0073] Indicatively, it also includes: S15. Encrypt network communication between various distribution network terminal equipment, clients, and control interfaces; synchronize setting sheets to authorized terminals, including dispatch center clients, portable field maintenance devices, and remote control interfaces, through encrypted and authenticated network communication. The encrypted and authenticated network communication method uses secure communication protocols (such as SSL / TLS) to ensure that data is not stolen or tampered with by unauthorized third parties during transmission. This method encrypts the data using encryption algorithms and employs mechanisms such as digital certificates and identity tokens to ensure that only authorized users can access and operate the data, guaranteeing the security, integrity, and reliability of data transmission. In this application, all setting sheet transmissions are performed in this manner, ensuring the confidentiality of information and the legality of user operations.

[0074] S16. Define the permissions for various types of distribution network personnel and visualize the process status. Each setting order is bound to a unique identifier and approval process status. The system provides users with an operating interface and functional scope that matches their permission level. When a user accesses a specific setting order, the operating system will simultaneously display its original generation source, approval path, expert verification results, version evolution records, and related equipment information, ensuring that the user can fully understand the setting background and basis for changes. If the user has editing permissions, they can modify the setting parameters. The system will automatically record the differences before and after the modification and generate a change log, while triggering a re-approval process to ensure process closure and content traceability.

[0075] S17. This module supports cached editing of distribution network operations in offline mode and synchronous submission of operations in online mode. Considering the variability of the distribution network environment, this module supports cached editing in offline mode and synchronous submission in online mode. After editing in a network-free environment, on-site personnel can reliably submit the changes upon reconnection using breakpoint resume and data integrity verification, avoiding process failures or data loss due to communication interruptions. Simultaneously, the distribution network module has terminal adaptive capabilities, dynamically adjusting the interface display logic according to different device resolutions and performance, enabling users to operate smoothly in on-site, office, or remote environments, lowering the technical barrier to entry and improving response efficiency.

[0076] In an optional implementation, the network outage caching and resynchronization mechanism can also be implemented by deploying a lightweight local approval proxy service based on embedded edge computing devices. Edge processing nodes are deployed at the power distribution station, integrating the setting modification interface and approval logic. This allows for local setting querying, modification, and temporary storage of approval results without relying on the central network. When the network recovers, the system can bidirectionally synchronize approval information with the central system through the edge nodes, achieving processing capabilities similar to existing solutions—"offline editable, online uploadable"—and improving the system's adaptability to deployment in remote or network-unstable areas.

[0077] This embodiment constructs a secure communication mechanism and editing interface supporting multiple terminals, allowing users to remotely access, edit, and trigger re-approval processes via authorized devices. It also supports offline caching and automatic online synchronization. This closed-loop mechanism of remote terminal access and approval re-entry enhances the ease of use and flexibility of the entire process, adapting to emergency on-site needs. The optimized methods for the aforementioned steps ensure that this embodiment possesses flexible remote access and online operation capabilities during actual distribution network operation and maintenance, serving as a crucial support for upgrading the setting approval process from "automation" to "digital collaboration." This step primarily receives the setting sheet results output by the expert database verification module and serves multiple user roles, including maintenance personnel, control personnel, and approval personnel, enabling them to access, view, edit, and re-approve setting sheets without physical location restrictions, thus building an efficient collaborative mechanism across regions, roles, and terminals. This significantly improves the response speed and on-site adaptability of setting management, particularly demonstrating strong practical value in typical scenarios such as emergency repairs, on-site commissioning, and holiday duty shifts, optimizing the closed-loop management process of distribution network operation and maintenance.

[0078] Indicatively, it also includes: S18. A permission model is designed based on roles, operations, and objects to determine the boundaries and operations of each distribution network operation role. The permission control mechanism plays a crucial role in ensuring system security, compliant operation, and preventing abuse. The permission control model is designed using a "role-operation-object" model to ensure that each operation and access object is within the role's permission scope. The permission design based on roles, operations, and objects ensures that each operation conforms to the prescribed permissions and can be tracked and audited. A specific implementation example of the role-operation-object modeling is as follows: Roles: Each user's identity within the system (e.g., administrator, approver, expert, maintenance personnel). Operations: The operations a role can perform, such as generating, approving, modifying, viewing, and reverting. Objects: The objects affected by the operations, such as setting sheets, approval processes, and historical records. Roles are a permission management system based on the division of labor among different roles. Each role has different responsibilities and access permissions within the system, including: Administrators, responsible for the overall management of the system, including permission configuration, approval path settings, and user management, possessing the highest system privileges. Approver, responsible for approving setting sheets. Based on the system's permission settings, approvers can approve or reject setting sheets, but cannot change parameters or modify approval paths. Experts, responsible for expert database verification, checking the rationality of setting parameters, and providing modification suggestions. Experts can provide setting suggestions based on rules, but cannot directly modify setting sheets. Maintenance personnel are primarily responsible for operations related to setting sheets, such as modifying parameters (if permitted) and performing equipment-related operations. Maintenance personnel's permissions are limited by the status of the setting sheets. Users can view fixed-value order information and historical records, but cannot perform any modifications or approvals. Operation permissions are restricted based on each role and the object being operated on.

[0079] Common operations include: **Generate:** Create a value sheet; only administrators or the automatic generation module can perform this. **Approve:** Approve each stage of the value sheet (intelligent approval, expert verification, etc.). Approver performs this operation according to their permissions. **Modify:** Modify the parameters of the value sheet. Operations personnel or users with modification permissions can perform this operation, but this operation will trigger a re-approval process. **View:** View the status, history, version, and other information of the value sheet. Most roles, such as operations personnel, experts, and approvers, can perform this operation, but their permissions differ depending on the role. **Rewind:** Track the historical operations of the value sheet approval process, recording all approval steps and status changes. Administrators and approvers typically have this operation permission. **Object:** Elements in the system that require access control.

[0080] In this embodiment, the main permission objects include: Setting value sheets, which are the core object and are involved in approval, modification, and verification operations; approval processes, where the approval process for setting value sheets is a crucial object of permission control, allowing different roles to participate at different stages; expert rules, where rules provided by experts are important for verifying setting value sheets, and experts provide modification suggestions based on these rules; and historical records, including system operation records, change logs, and approval paths. The permission control logic includes: role allocation, where users are assigned permissions based on their roles. For example, approvers can only approve setting value sheets but cannot modify setting value parameters or change process status; operation restrictions, where the system determines whether a user can perform a specific operation based on their role permissions. For example, only administrators can modify system settings, experts can verify setting value sheets but cannot modify setting value parameters; and object access, where different roles have different access permissions to different objects. For example, experts can view detailed information about setting value sheets but cannot view modification history; and maintenance personnel can modify setting values ​​but cannot approve them.

[0081] S19. Perform permission verification and intercept and record abnormal operations. During each operation, the system performs permission verification based on the user's role, operation type, and target object. If the user does not have permission to perform an operation, the system will reject the operation and return the corresponding error message. In addition to permission control, this module also has status tracking and permission verification functions to ensure the compliance and traceability of operations.

[0082] During status tracking, each quotation form's status (e.g., pending approval, approved, pending modification) is monitored throughout the approval process. Information related to the current status is displayed based on the user's role and permissions. For example, only approvers can operate on quotation forms awaiting approval, while modified quotation forms require re-entry into the approval process. During permission verification, user permissions are dynamically verified to ensure that every action performed by a user conforms to the corresponding permission configuration. If a user attempts to perform an operation without permission, the attempt is rejected and recorded.

[0083] S20. Perform full-process status change and flow path scheduling management on the set value.

[0084] Using the value set sheet as the core of the process entity, a process state machine manages its status changes and flow paths between each step. From initial generation, algorithm approval, expert verification, user viewing and remote editing, re-approval, to archiving, all process links are uniformly scheduled and managed by the system. Each status has clearly defined trigger conditions and operation permissions. For example, after the value set sheet is approved by intelligent approval, it automatically changes to the "awaiting expert verification" status and can only be processed by users with expert verification permissions. After verification is completed, the status changes to "awaiting issuance" or "returned for modification," and the system schedules the flow to the remote interaction module or re-enters the approval process accordingly.

[0085] In actual operation, the system supports customized workflows for different power distribution network units or operation and maintenance scenarios through configurable processes. For example, it supports setting approval levels, specifying approval roles, defining automatic redirection conditions, and adding multi-level countersigning processes to meet the needs of scenarios involving multiple organizational levels, multiple roles, and concurrent tasks. Internally, a complete approval process log is maintained. Every status change, every user operation, and every modification is accurately recorded and timestamped by the system, ensuring that the formation process of any later version can be traced, meeting technical compliance and auditing requirements.

[0086] S21. Perform full-process data synchronization to connect data behavior with process operations.

[0087] This optimization method synchronizes data with steps S10-S14 and S15-S17. When expert review suggestions are adopted or users submit modifications via the terminal, the approval workflow management module immediately identifies the change in process status and restarts the relevant steps, automatically generating new approval nodes and updating task allocation information, achieving automatic connection from data behavior to process behavior. After approval, the setpoint sheet enters the archive state, is marked as "approved" by the system, and synchronized to the historical database for subsequent power grid operation and scheduling reference and expansion of the approval model learning sample. Archived data is stored in a structured manner, supporting multi-dimensional retrieval by equipment, region, time, or fault characteristics, providing accurate and rich data support for technical management personnel.

[0088] When the distribution network topology changes or new equipment models are added, existing systems often require manual adjustments to approval rules or algorithm parameters, resulting in poor flexibility. This embodiment features adaptive learning and algorithm parameter adjustment capabilities, dynamically optimizing the setting logic based on actual operating conditions to ensure setting accuracy. In this application, the aforementioned optimization method serves as the central control point for the operational order and state flow of each stage. Its core function is to ensure a closed-loop operation of the setting calculation setpoint sheet, ensuring consistency in process, clarity of status, controlled permissions, and traceability of data across multiple stages, including automatic generation, intelligent approval, expert verification, and remote modification. This module runs throughout the entire setting management process, undertaking functions such as process orchestration, approval path configuration, status tracking, permission verification, and record archiving. It is a crucial foundation for ensuring the system is manageable, controllable, and accountable.

[0089] By implementing this embodiment, historical and current operating data and topology are automatically acquired. A weighted time-series matching function is used to calculate trajectory similarity and derive setting values. Most of the process is automated, reducing manual review and verification, and enabling rapid generation of setting value sheets. Setting values ​​are calculated based on historical response characteristics and trajectory similarity. In scenarios requiring rapid demand response, setting value sheets can be determined promptly, ensuring timely adjustments to the distribution network and preventing fault escalation. Trajectory similarity is calculated by comprehensively considering operating data, topology, and time-series information. Setting values ​​are obtained by weighting similarity with historical response characteristics, taking into account multiple parameters and the overall system operating status, avoiding misjudgments caused by isolated parameter evaluations. A difference metric is constructed based on the setting values ​​and historical setting values, and setting deviation values ​​are calculated under the constraint of a penalty function. The penalty function, combined with the target reference setting value and maximum offset tolerance, effectively identifies setting values ​​that contradict historical experience or safety requirements, overcoming the limitation of single-threshold logic in detecting overall configuration contradictions. By utilizing historical data and combining it with current timestamps to dynamically calculate setting values, the system can adapt to changes in the operating status of the distribution network. Compared to fixed single threshold judgments, this approach better reflects the complex and ever-changing actual operating conditions, ensuring reliable operation of protection equipment. A comprehensive evaluation of whether each setting parameter conforms to reasonable logical relationships based on historical experience enables panoramic risk control and collaborative optimization of distribution network protection configurations. This effectively improves the overall coordination of the protection system, significantly enhancing the deep logical reasoning and anomaly detection capabilities during the approval process, and effectively improving the intelligence level of the approval process and the rationality of the protection system configuration.

[0090] See Figure 4 This is a schematic diagram of a distribution network operation optimization device based on setting value generation according to an embodiment of the present invention, comprising: The data acquisition module is used to acquire the current operating data, current topology, historical operating data, historical topology, historical setting approval forms, and the historical response characteristics of the corresponding protection devices of the distribution network to be optimized. The data similarity matching module is used to select the corresponding historical running data as the first running data when the similarity between the current running data and the historical running data is greater than a preset similarity threshold; and to select the historical topology structure corresponding to the corresponding historical running data as the first topology structure. The trajectory similarity calculation module is used to calculate the trajectory similarity based on the current running data, the current topology, the current timestamp, the first running data, the first topology, and the corresponding historical timestamps, using a preset weighted time-series matching function. The setting value generation module is used to perform weighted summation calculation based on trajectory similarity and the historical response characteristics of the corresponding protection equipment to obtain the setting value; and to generate a setting value sheet based on the setting value. The difference measurement construction module is used to construct difference measurement based on the set value and the historical set value in the historical set value approval form, and to construct a penalty function based on the preset target reference set value, the preset maximum offset tolerance and the parameters corresponding to the set value; The constant value deviation determination module is used to generate a constant value deviation value under the constraint of the penalty function, based on the difference metric and the preset difference weight coefficient. The distribution network data optimization module is used to determine the final setpoint list based on the setpoint deviation value and the preset deviation threshold; and to optimize the current operating data of the distribution network to be optimized based on the final setpoint list.

[0091] This invention provides a distribution network operation optimization device based on setpoint generation. It automatically acquires historical and current operating data and topology, calculates trajectory similarity using a weighted time-series matching function, and derives setpoints. Most of the process is automated, reducing manual review and verification, and enabling rapid generation of setpoint sheets. Setpoints are calculated based on historical response characteristics and trajectory similarity. In scenarios requiring rapid demand response, setpoint sheets can be determined promptly, ensuring timely adjustments to the distribution network and preventing fault escalation. The device comprehensively considers operating data, topology, and time-series information to calculate trajectory similarity, and obtains setpoints by weighting similarity with historical response characteristics. This considers multiple parameters and the overall system operating status, avoiding misjudgments caused by isolated parameter evaluations. A difference metric is constructed based on the setpoints and historical setpoints, and setpoint deviations are calculated under penalty function constraints. The penalty function, combined with target reference setpoints and maximum offset tolerances, effectively identifies setpoints that contradict historical experience or safety requirements, overcoming the limitation of single-threshold logic in detecting overall configuration contradictions. By utilizing historical data and combining it with current timestamps to dynamically calculate setting values, the system can adapt to changes in the operating status of the distribution network. Compared to fixed single threshold judgments, this approach better reflects the complex and ever-changing actual operating conditions, ensuring reliable operation of protection equipment. A comprehensive evaluation of whether each setting parameter conforms to reasonable logical relationships based on historical experience enables panoramic risk control and collaborative optimization of distribution network protection configurations. This effectively improves the overall coordination of the protection system, significantly enhancing the deep logical reasoning and anomaly detection capabilities during the approval process, and effectively improving the intelligence level of the approval process and the rationality of the protection system configuration.

[0092] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles 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 optimizing the operation of a distribution network based on setting values, characterized in that, include: Obtain the current operating data, current topology, historical operating data, historical topology, historical setting approval forms, and historical response characteristics of the corresponding protection devices of the distribution network to be optimized; If the similarity between the current running data and the historical running data is greater than a preset similarity threshold, the corresponding historical running data will be used as the first running data. And the historical topology corresponding to the historical operation data is used as the first topology; Based on the current running data, current topology, current timestamp, first running data, first topology, and corresponding historical timestamps, the trajectory similarity is calculated using a preset weighted time-series matching function. The setting value is obtained by weighted summation based on trajectory similarity and the historical response characteristics of the corresponding protection devices; And generate a setting sheet based on the setting value; A difference metric is constructed based on the set value and the historical set values ​​in the historical set value approval form, and a penalty function is constructed based on the preset target reference set value, the preset maximum offset tolerance, and the parameters corresponding to the set value. Under the constraint of the penalty function, a fixed deviation value is generated based on the difference metric and the preset difference weight coefficient. The final set value is determined based on the deviation value and the preset deviation threshold. And optimize the current operating data of the distribution network to be optimized based on the final set value sheet.

2. The distribution network operation optimization method based on setting value generation as described in claim 1, characterized in that, Based on the current running data, current topology, current timestamp, first running data, first topology, and corresponding historical timestamps, trajectory similarity is calculated using a preset weighted time-series matching function, including: Generate a triplet for the current running trajectory based on the current running data, the current topology, and the current timestamp; Based on the first running data, the first topology, and the corresponding historical timestamps, a historical running trajectory triplet is generated; Based on the current trajectory triplet and the historical trajectory triplet, the trajectory similarity is calculated using a preset weighted temporal matching function.

3. The distribution network operation optimization method based on setting value generation as described in claim 1, characterized in that, The setting value is obtained by weighted summation based on trajectory similarity and the historical response characteristics of the corresponding protection devices, including: Based on the historical response characteristics of the corresponding protection equipment, determine the total number of factors influencing the setting value of the protection equipment and the variable parameters; Feature transformation is performed on the protection device response records in the historical response characteristics to obtain protection action parameter values; wherein, the protection action parameter values ​​include historical response accuracy, action time deviation rate, and fault type matching degree; Extract the protection action parameter values ​​corresponding to the variable parameters from the historical response characteristics to obtain the historical response characteristic parameter set; The setpoint is obtained by fusing the historical response feature parameter set, the total number of influencing factors, and the trajectory similarity, and then by weighted summation.

4. The distribution network operation optimization method based on setting value generation as described in claim 1, characterized in that, A difference metric is constructed based on the set value and historical set values ​​in the historical set value approval forms, including: The parameters of the setpoint are standardized to obtain the setpoint feature vector; Based on the historical set values ​​in the historical set value approval form, extract the historical set value parameters corresponding to the set value to form a set of historical set value parameters; Calculate the mean vector and covariance matrix of historical constant values ​​based on the historical constant value parameter set; A difference measure is constructed based on the eigenvector of the setpoint, the mean vector of the historical setpoints, and the covariance matrix.

5. The distribution network operation optimization method based on setting value generation as described in claim 1, characterized in that, The penalty function is constructed using the following formula: in, Represents the penalty function; Indicates the kth i The parameters corresponding to each setpoint; This indicates the preset target reference value; This indicates the preset maximum offset tolerance.

6. The distribution network operation optimization method based on setting value generation as described in claim 1, characterized in that, Under the constraint of the penalty function, a fixed deviation value is generated based on the difference metric and the preset difference weight coefficient, including: Under the constraint of the penalty function, the parameter corresponding to the tuning value with a penalty function value of 0 is selected as the first tuning parameter; Assign a preset difference weight coefficient based on the first selected tuning parameter; The baseline deviation value is calculated based on the difference measure and the assigned difference weight coefficients. The parameter corresponding to the tuning value of the penalty function that is greater than 0 is used as the second tuning parameter; The penalty function value of the second tuning parameter and the preset penalty weight coefficient are weighted and calculated to obtain the over-limit compensation value; The baseline deviation value and the excess compensation value are summed to obtain the fixed deviation value.

7. The distribution network operation optimization method based on setting value generation as described in claim 6, characterized in that, Based on the deviation from the set value and the preset deviation threshold, the final set value sheet is determined, including: If the deviation value is less than or equal to the preset deviation threshold, the generated value sheet is deemed qualified and is used as the final value sheet. If the deviation of the set value is greater than the preset deviation threshold, the second setting parameter is marked so that the operator can review the marked set value sheet and regenerate the set value sheet.

8. The distribution network operation optimization method based on setting value generation as described in claim 1, characterized in that, The current operating data of the distribution network to be optimized is based on the final setpoint sheet, including: The setting parameters in the final setting sheet are categorized according to the type of protection equipment to generate an equipment parameter configuration table; wherein, the equipment parameter configuration table includes parameter name, target value and allowable fluctuation range; The device parameter configuration table is sent to the corresponding protection device, triggering the device parameter update. The current operating data is compared with the parameter thresholds in the final set value sheet to monitor whether there are any operating conditions where the operating data exceeds the allowable fluctuation range of the set value. If present, adjust the load of non-core users to bring the operating data back to the set operating fluctuation range; If it does not exist, continuously monitor the current operating parameters.

9. A distribution network operation optimization device based on setpoint generation, characterized in that, include: The data acquisition module is used to acquire the current operating data, current topology, historical operating data, historical topology, historical setting approval forms, and the historical response characteristics of the corresponding protection devices of the distribution network to be optimized. The data similarity matching module is used to select the corresponding historical running data as the first running data when the similarity between the current running data and the historical running data is greater than a preset similarity threshold. And the historical topology corresponding to the historical operation data is used as the first topology; The trajectory similarity calculation module is used to calculate the trajectory similarity based on the current running data, the current topology, the current timestamp, the first running data, the first topology, and the corresponding historical timestamps, using a preset weighted time-series matching function. The setting value generation module is used to perform weighted summation calculation based on trajectory similarity and the historical response characteristics of the corresponding protection equipment to obtain the setting value; and to generate a setting value sheet based on the setting value. The difference measurement construction module is used to construct difference measurement based on the set value and the historical set value in the historical set value approval form, and to construct a penalty function based on the preset target reference set value, the preset maximum offset tolerance and the parameters corresponding to the set value; The constant value deviation determination module is used to generate a constant value deviation value under the constraint of the penalty function, based on the difference metric and the preset difference weight coefficient. The distribution network data optimization module is used to determine the final set value sheet based on the set value deviation value and the preset deviation threshold. And optimize the current operating data of the distribution network to be optimized based on the final set value sheet.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform a distribution network operation optimization method based on setting value generation as described in any one of claims 1 to 8.

Images