A power distribution network self-healing function test method and system

Abstract

This invention relates to the field of distribution network self-healing testing technology, and in particular to a method and system for testing the self-healing function of distribution networks. Based on timestamps and confidence levels, this invention constructs fault evolution paths, enabling dynamic simulation of the concurrent occurrence and evolution of multiple fault types under complex fault scenarios in distribution networks. This allows for the reproduction of the dynamic characteristics of faults over time and with environmental changes. Secondly, this invention eliminates random interference by using confidence thresholds and constructs fault evolution paths through correlation condition judgments, ensuring the effectiveness and reliability of test cases. Thirdly, by distinguishing between evolution paths and isolated event paths, this invention addresses the operational testing needs of both fault evolution scenarios and independent fault scenarios. Finally, this invention achieves quantitative evaluation of the self-healing function of distribution networks through indicators such as equipment state change control rate, self-healing response timing matching rate, and initial state recovery rate, thereby significantly improving the effectiveness and accuracy of test results.

Description

Technical Field

[0001] This invention relates to the field of distribution network self-healing testing technology, and in particular to a method and system for testing the self-healing function of distribution networks. Background Technology

[0002] With the rapid development of society and economy and the continuous progress of science and technology, electricity demand continues to grow, and users' requirements for power supply quality and reliability are also increasing. Currently, distribution networks utilize self-healing technology to achieve a high degree of autonomy in grid operation and control, enabling them to restore normal power supply when encountering faults, reducing reliance on human intervention. During normal grid operation, self-healing functions can be used to employ preventative control strategies, proactively monitoring for potential fault signs. Once an anomaly is detected, immediate measures are taken to prevent small problems from escalating into major faults.

[0003] Self-healing functionality is a key capability for maintaining the continuity and reliability of power supply in distribution networks. Its operation, maintenance, and testing are of great significance for ensuring power supply continuity and reliability. Due to the multi-causative and dynamic nature of distribution network faults, the testing of distribution network self-healing functionality has undergone multiple stages of evolution and is beginning to shift towards system-level verification. On the one hand, self-healing analysis models are constructed by collecting real-time operational data of the distribution network, attempting to improve the fit between test scenarios and actual power grid operation scenarios. On the other hand, online testing systems, by deploying dedicated testing instruments, achieve real-time acquisition and comparison of distribution terminal response signals, realizing online testing of the distribution network's self-healing functionality.

[0004] Although current distribution network self-healing function testing has initially achieved data-driven and real-time verification capabilities, current testing methods all use isolated terminal commands for testing. The test environment and test case design fail to consider the concurrent and evolutionary processes of multiple types of faults in distribution networks under complex fault scenarios. It is difficult to verify the performance of distribution network self-healing function in the face of dynamically evolving fault characteristics due to environmental uncertainty and randomness. It is also difficult to support the comprehensive verification requirements of self-healing function under complex fault scenarios, resulting in low comprehensiveness and effectiveness of current distribution network self-healing function testing. Summary of the Invention

[0005] This invention aims to provide a method and system for testing the self-healing function of a distribution network. By simulating the fault evolution characteristics in the real operation scenario of the distribution network, it solves the technical problem that current self-healing function testing fails to consider the concurrent and evolutionary process of multiple types of faults in the distribution network under complex fault scenarios, making it difficult to verify the performance of the distribution network's self-healing function when facing fault characteristics that dynamically evolve due to environmental uncertainties and randomness. This invention improves the comprehensiveness and effectiveness of the self-healing function testing of the distribution network.

[0006] To achieve the above objectives, the first aspect of the present invention provides a method for testing the self-healing function of a power distribution network, comprising the following steps: Obtain historical events of the distribution network and the occurrence time of the historical events; combine any two historical events with the same occurrence time as a historical event combination to obtain several historical event combinations. Several combinations of failure events are selected from several combinations of historical events; Obtain the confidence level of each of the fault event combinations, and then construct a fault event set based on several fault event combinations whose corresponding confidence levels are higher than a preset confidence threshold; Determine whether any two combinations of fault events in the set of fault events satisfy a preset association condition, and connect any two combinations of fault events that satisfy the preset association condition into a local evolution path to obtain several local evolution paths. Obtain the timestamp of each of the local evolution paths, and concatenate several of the local evolution paths into a first fault evolution path based on the timestamp of each of the local evolution paths, and then generate a first test case based on the first fault evolution path; The first test case is input into a pre-built digital twin model so that the digital twin model responds to the first test case to simulate a self-healing process; The test data of the digital twin model during the self-healing process is obtained, and then the self-healing function of the digital twin model is evaluated through the test data to obtain the test evaluation results.

[0007] The aforementioned method for testing the self-healing function of distribution networks provides a virtual operating environment consistent with the actual structure and operation scenarios of distribution networks by constructing a digital twin model of the distribution network. By building combinations of historical events occurring at the same time based on historical events and their occurrence times, it can recreate the real scenario of multiple events occurring concurrently in distribution network operation. Based on this, fault event combinations are selected and a fault event set is constructed based on confidence levels, which can eliminate random interference from massive historical data and extract fault correlations and evolutionary relationships with high reliability. Then, by judging the correlation conditions between fault event combinations, each combination is connected into a local evolution path, and finally, a first fault evolution path is formed based on timestamps. This allows for the discovery of fault concurrency and evolution patterns hidden in historical data, enabling the generated first test case to simulate the dynamic evolution process of multiple types of faults in the distribution network under complex fault scenarios. Finally, the first test case is input into the digital twin model, and by acquiring test data during the self-healing process, the self-healing function is evaluated, verifying the true performance of the distribution network's self-healing function in the face of dynamically evolving fault characteristics. Compared with the existing technology that uses isolated terminal commands for testing, the above-mentioned technical means effectively solves the technical problem that current self-healing function testing is difficult to simulate the concurrent and evolution process of faults under complex fault scenarios. This makes the testing and verification of the self-healing function of the distribution network more in line with actual operation, and significantly improves the comprehensiveness and effectiveness of self-healing function testing.

[0008] Furthermore, the step of selecting several combinations of fault events from several combinations of historical events includes: The historical event combinations that include fault characteristic fields and device status fields are selected as candidate combinations. The probability of each candidate combination appearing in all the historical event combinations is obtained, and then the candidate combinations whose occurrence probability is higher than a preset probability threshold are taken as the fault event combinations, thereby obtaining a number of fault event combinations.

[0009] In this implementation, by detecting historical event combinations with fault characteristic fields and device status fields as candidate combinations, key data segments containing fault information can be accurately extracted from a large number of historical event combinations, avoiding interference from irrelevant data. Then, by obtaining the occurrence probability of each candidate combination in all historical event combinations, and selecting candidate combinations with occurrence probabilities higher than a preset probability threshold as fault event combinations, occasional low-probability abnormal events can be effectively eliminated. This ensures that the selected fault event combinations have statistical universality and representativeness, avoiding test case realism deviations caused by fault evolution paths formed by occasional event combinations, thereby improving the effectiveness of subsequent fault evolution path and self-healing function test results.

[0010] Further, the step of obtaining the confidence level of each of the fault event combinations, and then constructing a fault event set based on a number of fault event combinations whose corresponding confidence levels are higher than a preset confidence threshold, includes: For any combination of fault events: Identify the preceding and following events in the fault event combination; Based on several historical events and combinations of historical events, the conditional probability of the subsequent event occurring after the preceding event occurs is obtained, and the conditional probability is used as the confidence level of the fault event combination.

[0011] In this implementation, the strength of the causal relationship between two fault events in a fault event combination is quantified by identifying the preceding and subsequent events in the combination and calculating the conditional probability of the subsequent event occurring given the occurrence of the preceding event. Using this confidence level to filter fault event combinations and construct a fault event set retains combinations with strong causal relationships while eliminating those with weaker relationships. This ensures that the fault event combinations in the constructed set conform to the objective laws of fault evolution in the distribution network, improving the effectiveness of subsequent self-healing function testing.

[0012] Further, after determining whether any two combinations of fault events in the fault event set satisfy a preset association condition, and concatenating any two combinations of fault events that satisfy the preset association condition into a local evolution path to obtain several local evolution paths, the method further includes: For any combination of fault events in the set of fault events, if neither the combination of fault events nor the other combinations of fault events in the set of fault events satisfy the preset association condition, then the combination of fault events is regarded as an isolated combination of events. Obtain the timestamp of each isolated event combination, and then concatenate several isolated event combinations into a second fault evolution path based on the timestamp of each isolated event combination, and then generate a second test case according to the second fault evolution path; The second test case is input into the digital twin model so that the digital twin model simulates a self-healing process in response to the second test case; The test data of the digital twin model during the self-healing process is obtained, and then the self-healing function of the digital twin model is evaluated through the test data to obtain the test evaluation results.

[0013] In this implementation, fault event combinations that cannot satisfy preset association conditions with other combinations are defined as isolated event combinations. A second fault evolution path and a second test case input digital twin model are generated based on timestamps to simulate and evaluate the self-healing process. This implementation also considers isolated or independent fault scenarios that may exist in complex fault scenarios without obvious evolutionary characteristics, supplementing the second test cases for these isolated event combinations. These isolated event combinations are concatenated according to temporal relationships, enabling the generated second test cases to simulate test scenarios consisting of multiple independent but chronologically occurring fault events, supplementing the distribution network operation scenarios that the first test cases failed to simulate. Through the combination of the first and second test cases, the aforementioned self-healing function test can cover complex fault scenarios with evolutionary associations and isolated fault scenarios without evolutionary associations in distribution network operation, improving the comprehensiveness of the self-healing function test.

[0014] Further, generating the first test case based on the first fault evolution path includes: Based on the first fault evolution path, fault type evolution information and equipment status change information are generated; Generate timing-triggered instructions based on the device status change information; The first test case is generated based on the fault type evolution information, the equipment state change information, and the timing triggering instruction.

[0015] In this implementation, fault type evolution information and equipment state change information are extracted based on the first fault evolution path, and timing trigger instructions are generated accordingly. Finally, the first test case is generated by integrating these instructions and converting the fault evolution event data into specific instruction programs that can be recognized and executed by the digital twin model. This achieves a precise digital expression of the complex fault evolution process, enabling the first test case to accurately drive the digital twin model to simulate the entire process of fault dynamic evolution in the real power grid, and providing test case input for self-healing function testing.

[0016] Further, the step of acquiring test data of the digital twin model during the self-healing process, and then evaluating the self-healing function of the digital twin model using the test data to obtain test evaluation results, includes: Based on the test data, the digital twin model is used to calculate the device state change control rate, self-healing response timing matching rate, and initial state recovery rate during the self-healing process. A comprehensive self-healing performance score is obtained based on the device state change control rate, the self-healing response timing matching rate, and the initial state recovery rate, and the comprehensive self-healing performance score is used as the test evaluation result.

[0017] In this implementation, the self-healing function is quantitatively evaluated from three dimensions: equipment state change control rate, self-healing response timing matching rate, and initial state recovery rate, based on test data and calculated using a digital twin model during the self-healing process. These dimensions represent equipment control capability, response timing accuracy, and system recovery capability. The equipment state change control rate reflects the self-healing function's effect on equipment state regulation; the self-healing response timing matching rate measures the timing accuracy of self-healing actions; and the initial state recovery rate characterizes the self-healing function's ability to restore the system to a normal state. Obtaining a comprehensive self-healing performance score based on these three indicators as the test evaluation result overcomes the limitations of single-indicator evaluation, achieving a multi-dimensional comprehensive evaluation of the distribution network's self-healing function and improving the comprehensiveness and accuracy of the test evaluation results.

[0018] Further, the calculation of the device state change control rate, self-healing response timing matching rate, and initial state recovery rate of the digital twin model during the self-healing process based on the test data includes: Based on the test data, obtain the final fault status and importance coefficient of each device in the digital twin model; The control rate of equipment state change is calculated based on the final fault state and importance coefficient of each of the aforementioned devices.

[0019] In this implementation, the final fault state and importance coefficient of each device in the digital twin model are obtained based on test data. Then, the device state change control rate is calculated based on the final fault state and importance coefficient of each device. This calculation method introduces the final fault state and importance coefficient of the devices. Considering the different levels of importance of different devices in the distribution network topology, the fault state of each device is weighted using the importance coefficient, giving greater weight to the fault control status of key devices in the evaluation results. This allows for a more accurate reflection of the actual effect of self-healing functions on ensuring the safe operation of the core areas of the power grid.

[0020] Further, the calculation of the device state change control rate, self-healing response timing matching rate, and initial state recovery rate of the digital twin model during the self-healing process based on the test data includes: Obtain the preset trigger time for each self-healing action in the digital twin model; Based on the test data, obtain the actual trigger time of each self-healing action in the digital twin model; The self-healing response timing matching rate is calculated based on the preset trigger time and actual trigger time of each self-healing action in the digital twin model.

[0021] In this implementation, the preset trigger time of each self-healing action in the digital twin model is obtained, and the actual trigger time of each self-healing action is obtained based on test data. Then, the self-healing response timing matching rate is calculated based on the preset trigger time and the actual trigger time. This calculation method, by comparing the preset trigger time and the actual trigger time, quantitatively evaluates the timing accuracy of the self-healing action execution. It can effectively verify whether the self-healing function can respond according to the expected timing logic under complex fault scenarios, providing a quantitative indicator for evaluating the real-time response capability of the distribution network's self-healing function.

[0022] Further, the calculation of the device state change control rate, self-healing response timing matching rate, and initial state recovery rate of the digital twin model during the self-healing process based on the test data includes: Based on the test data, obtain the final recovery status of each device in the digital twin model; Based on the test data, obtain the actual stable runtime and minimum stable runtime of each device in the digital twin model after the self-healing process ends; The recovery integrity sub-item is calculated based on the final recovery status of each device, and the post-recovery stability sub-item is calculated based on the actual stable operating time and minimum stable operating time of each device. Then, the initial state recovery rate is calculated based on the recovery integrity sub-item and the post-recovery stability sub-item.

[0023] In this implementation, the final recovery state of each device in the digital twin model is obtained based on test data, along with the actual stable runtime and minimum stable runtime of each device after the self-healing process ends. A recovery integrity sub-item is calculated based on the final recovery state of each device, and a post-recovery stability sub-item is calculated based on the actual and minimum stable runtimes. Finally, the initial state recovery rate is calculated based on these two sub-items. This calculation method not only focuses on whether the device has recovered to its normal operating state but also on whether the device can maintain continuous and stable operation after recovery. It can effectively identify problems such as the device failing again shortly after recovery, ensuring that the initial state recovery rate reflects the self-healing function's ability to restore the system to a normal state and maintain stable operation.

[0024] A second aspect of the present invention provides a power distribution network self-healing function testing system, comprising: The data acquisition module is used to acquire historical events of the distribution network and the occurrence time of the historical events, and to combine any two historical events with the same occurrence time as a historical event combination to obtain a number of historical event combinations. The data filtering module is used to filter out several combinations of fault events from several combinations of historical events; The fault event filtering module is used to obtain the confidence level of each fault event combination, and then construct a fault event set based on several fault event combinations whose corresponding confidence levels are higher than a preset confidence threshold. The fault evolution path generation module is used to determine whether any two fault event combinations in the fault event set meet a preset association condition, and to connect any two fault event combinations that meet the preset association condition into a local evolution path to obtain a number of local evolution paths. The test case generation module is used to obtain the timestamp of each of the local evolution paths, connect several local evolution paths into a first fault evolution path based on the timestamp of each of the local evolution paths, and then generate a first test case based on the first fault evolution path. The self-healing function testing module is used to input the first test case into a pre-built digital twin model so that the digital twin model responds to the first test case to simulate a self-healing process; The self-healing performance evaluation module is used to acquire test data of the digital twin model during the self-healing process, and then evaluate the self-healing function of the digital twin model through the test data to obtain test evaluation results.

[0025] The method and system for testing the self-healing function of a power distribution network provided by the present invention have at least the following advantages compared with the prior art: This invention breaks through the limitations of existing distribution network self-healing function testing methods that rely on isolated terminal commands. By mining historical event data and constructing fault evolution paths based on timestamps and confidence levels, it achieves dynamic simulation of the concurrent and evolutionary processes of multiple types of faults under complex fault scenarios in distribution networks. It can restore the dynamic characteristics of faults over time and with environmental changes, effectively solving the problem that existing technologies cannot verify self-healing performance under dynamically evolving faults, and significantly improving the fit between test scenarios and actual operating conditions.

[0026] This invention eliminates random interference by using confidence thresholds and constructs fault evolution paths through correlation condition judgments, ensuring the effectiveness and reliability of the first test case. Simultaneously, by distinguishing between evolution paths and isolated event paths, it addresses the operational testing needs of both fault evolution scenarios and independent fault scenarios, improving the comprehensiveness of distribution network self-healing function testing.

[0027] This invention establishes a multi-dimensional testing and evaluation system. Through indicators such as equipment state change control rate, self-healing response timing matching rate, and initial state recovery rate, it achieves quantitative evaluation of self-healing function from multiple perspectives such as fault control, action timing, and recovery quality, thereby significantly improving the effectiveness and accuracy of test results. Attached Figure Description

[0028] Figure 1 This is a flowchart illustrating a method for testing the self-healing function of a power distribution network according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of a power distribution network self-healing function testing system provided in an embodiment of the present invention. Detailed Implementation

[0029] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that the following detailed descriptions are exemplary and intended to provide further detailed explanation of the invention. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used herein in the specification is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings, are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings are used to distinguish different objects, not to describe a particular order.

[0030] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0031] Please refer to Figure 1 To address the technical problem that current self-healing function testing fails to consider the concurrent and evolving processes of multiple types of faults in complex fault scenarios, making it difficult to verify the performance of the self-healing function in the face of dynamically evolving fault characteristics caused by environmental uncertainties and randomness, and to improve the comprehensiveness and effectiveness of distribution network self-healing function testing, the first embodiment of this invention provides a distribution network self-healing function testing method, including the following steps: S1. Obtain historical events of the distribution network and the occurrence time of the historical events, and combine any two historical events with the same occurrence time as a historical event combination to obtain several historical event combinations; S2. Select several fault event combinations from the several combinations of historical events; S3. Obtain the confidence level of each fault event combination, and then construct a fault event set based on several fault event combinations whose corresponding confidence levels are higher than a preset confidence threshold. S4. Determine whether any two combinations of fault events in the fault event set satisfy a preset association condition, and connect any two combinations of fault events that satisfy the preset association condition into a local evolution path to obtain several local evolution paths. S5. Obtain the timestamp of each of the local evolution paths, and connect several local evolution paths into a first fault evolution path based on the timestamp of each of the local evolution paths, and then generate a first test case based on the first fault evolution path. S6. Input the first test case into the pre-built digital twin model so that the digital twin model responds to the first test case to simulate a self-healing process; S7. Obtain test data of the digital twin model during the self-healing process, and then evaluate the self-healing function of the digital twin model through the test data to obtain test evaluation results.

[0032] The aforementioned method for testing the self-healing function of distribution networks provides a virtual operating environment consistent with the actual structure and operation scenarios of distribution networks by constructing a digital twin model of the distribution network. By building combinations of historical events occurring at the same time based on historical events and their occurrence times, it can recreate the real scenario of multiple events occurring concurrently in distribution network operation. Based on this, fault event combinations are selected and a fault event set is constructed based on confidence levels, which can eliminate random interference from massive historical data and extract fault correlations and evolutionary relationships with high reliability. Then, by judging the correlation conditions between fault event combinations, each combination is connected into a local evolution path, and finally, a first fault evolution path is formed based on timestamps. This allows for the discovery of fault concurrency and evolution patterns hidden in historical data, enabling the generated first test case to simulate the dynamic evolution process of multiple types of faults in the distribution network under complex fault scenarios. Finally, the first test case is input into the digital twin model, and by acquiring test data during the self-healing process, the self-healing function is evaluated, verifying the true performance of the distribution network's self-healing function in the face of dynamically evolving fault characteristics. Compared with the existing technology that uses isolated terminal commands for testing, the above-mentioned technical means effectively solves the technical problem that current self-healing function testing is difficult to simulate the concurrent and evolution process of faults under complex fault scenarios. This makes the testing and verification of the self-healing function of the distribution network more in line with actual operation, and significantly improves the comprehensiveness and effectiveness of self-healing function testing.

[0033] In a preferred embodiment, the pre-built digital twin model specifically needs to be constructed based on the distribution network physical topology, equipment parameters, and load parameters.

[0034] The physical topology of the distribution network includes the location of each node device in the distribution network and the line connection relationship between each device; the equipment parameters include transformer capacity, line impedance and switch rated current; the load parameters include the power characteristics and voltage withstand threshold of various loads.

[0035] The digital twin model comprises a perception layer, a data layer, a model layer, a functional layer, and an application layer. The perception layer is used for data acquisition and access, supporting test environment initialization; the data layer is used for data storage and scheduling, ensuring data availability throughout the entire process; the model layer is used to build multiphysics models for each type of equipment, such as lines, inverters, and buses, ensuring that equipment state changes conform to physical laws; the functional layer is used for parsing the first test case, fault simulation, self-healing response, and performance evaluation; and the application layer displays the visual interface for editing the first test case, the visual interface for monitoring the simulation process, and the visual interface for evaluation results, supporting user operation and decision-making.

[0036] After constructing the basic physical topology model of the distribution network, it is also necessary to integrate the self-healing system of the distribution network to be tested into the digital twin model. By integrating the self-healing strategy algorithms of the distribution network, such as load shedding algorithms and low voltage ride-through protection algorithms, the digital twin model can automatically match the response strategy based on the input of the first test case.

[0037] Specifically, the distribution network self-healing system in this embodiment uses an integrated hierarchical distributed control algorithm to simulate operations such as global load transfer path optimization at the regional master station, fault strategy generation, real-time fault detection at the local terminal, and command execution. The digital twin model also includes intelligent hardware terminals and a real-time communication module, which are the core carriers for the digital twin model to simulate the distribution network's self-healing function. These are intelligent collaborative systems capable of automatically completing fault detection, location, isolation, and load restoration. The intelligent hardware terminals include intelligent boundary switches, automatic backup power switching, and photovoltaic inverter protection devices.

[0038] In one possible embodiment, data processing equipment is deployed in the distribution network nodes to establish communication with the digital twin model, thereby synchronizing the digital twin model with the physical distribution network.

[0039] Specifically, distribution network nodes can be transformers, new energy grid connection points, and 10kV line sectionalizing switches, etc.; data processing equipment is used to collect real-time data from physical equipment, such as real-time voltage and current; the data processing equipment is also used to receive instructions from the digital twin model, such as generating voltage sag signals. Communication between the data processing equipment and the digital twin model is established via 5G communication to achieve synchronization between the digital twin model and the physical distribution network. Preferably, the communication latency needs to be less than or equal to 10ms, and the data packet loss rate needs to be less than or equal to 0.1%.

[0040] In a preferred embodiment, the historical event data of the distribution network includes fault characteristics, equipment status, equipment-related topology status, and environmental interference-related data. Fault characteristics include fault type and fault duration; equipment status includes basic equipment information, real-time operating parameters, and status identifiers. Basic equipment information includes equipment ID, equipment type, and installation location; real-time operating parameters include voltage, current, and power factor; status identifiers include switch open / closed status. Equipment-related topology status refers to the connection relationship and load distribution status of the distribution network topology within the fault's influence range; environmental interference-related data includes external environmental factors at the time of the fault, including: environmental interference type, such as thunderstorms or harmonics exceeding limits; interference intensity, such as lightning current amplitude of 150kA or harmonic content equal to 3%; and also includes the interference occurrence timestamp.

[0041] Furthermore, the step of selecting several combinations of fault events from several combinations of historical events includes: The historical event combinations that include fault characteristic fields and device status fields are selected as candidate combinations. The probability of each candidate combination appearing in all the historical event combinations is obtained, and then the candidate combinations whose occurrence probability is higher than a preset probability threshold are taken as the fault event combinations, thereby obtaining a number of fault event combinations.

[0042] As an example, the historical event combination that simultaneously possesses a fault characteristic field and an equipment status field includes, for example: a load change field and a transformer overcurrent field; a fault duration of 350ms field and a switch tripping field; a voltage sag of 30%, a fault duration of 600ms field and a photovoltaic inverter disconnection field, etc.

[0043] In a preferred embodiment, the probability of each candidate combination appearing among all historical event combinations is obtained, and historical event combinations with a probability of not less than 20% are included in the frequent itemset as fault event combinations. The 20% parameter setting is based on the distribution characteristics of historical fault data, ensuring that the selected frequent itemsets have a certain degree of universality and avoiding random patterns due to insufficient sample size.

[0044] In this embodiment, by detecting historical event combinations with fault characteristic fields and device status fields as candidate combinations, key data segments containing fault information can be accurately extracted from a large number of historical event combinations, avoiding interference from irrelevant data. Then, by obtaining the occurrence probability of each candidate combination in all historical event combinations, and selecting candidate combinations with occurrence probabilities higher than a preset probability threshold as fault event combinations, occasional low-probability abnormal events can be effectively eliminated. This ensures that the selected fault event combinations have statistical universality and representativeness, avoiding test case realism deviations caused by fault evolution paths formed by occasional event combinations, thereby improving the effectiveness of subsequent fault evolution path and self-healing function test results.

[0045] Further, the step of obtaining the confidence level of each of the fault event combinations, and then constructing a fault event set based on a number of fault event combinations whose corresponding confidence levels are higher than a preset confidence threshold, includes: For any combination of fault events: Identify the preceding and following events in the fault event combination; Based on several historical events and combinations of historical events, the conditional probability of the subsequent event occurring after the preceding event occurs is obtained, and the conditional probability is used as the confidence level of the fault event combination.

[0046] In a preferred embodiment, the conditional probability of the subsequent event occurring after the preceding event in the fault event combination is calculated, and the conditional probability is used as the confidence level of the fault event combination. Fault event combinations with a confidence level of not less than 80% are included in the confidence rule set to construct a fault event set.

[0047] Here, using conditional probability as confidence level refers to the probability that a subsequent event (such as transformer overcurrent) will occur after a preceding event (such as a sudden load change) has occurred. Confidence level The calculation formula is: In the formula, A is the preceding event and B is the following event; Let A be the probability that A and B occur simultaneously. The probability of A occurring alone; if the confidence level is less than 80%, the rule for this combination of failure events is not reliable enough and should be discarded.

[0048] In this embodiment, by identifying the preceding and subsequent events in a fault event combination and calculating the conditional probability of the subsequent event occurring given the occurrence of the preceding event as a confidence level, the causal correlation strength between two fault events in the fault event combination is quantified. Using the confidence level to filter fault event combinations and construct a fault event set can retain fault event combinations with strong causal correlations and eliminate those with weak causal correlations. This ensures that the fault event combinations in the constructed fault event set conform to the objective laws of fault evolution in the distribution network, improving the effectiveness of subsequent self-healing function testing.

[0049] In a preferred embodiment, determining whether any two combinations of fault events in the fault event set satisfy a preset association condition, and concatenating any two combinations of fault events that satisfy the preset association condition into a local evolution path to obtain several local evolution paths, includes: Determine whether any two fault event combinations in the fault event set satisfy the association condition. If they do, then connect the two fault event combinations into a local evolution path. Sort several local evolution paths according to the timestamp of the equipment state change. If the timestamps of the equipment state change are the same, sort them from high to low confidence to form the first fault evolution path.

[0050] Ultimately, the first fault evolution path contains multiple sets of fault event combinations arranged in sequence, and each set of fault event combinations includes two fault events ordered according to the relationship between the preceding and following events, so that the first fault evolution path actually contains multiple fault events arranged in sequence.

[0051] In one possible embodiment, in the first fault evolution path, the subsequent event of a fault event combination may also be the preceding event of the next fault event combination, causing the same event to appear repeatedly. In this case, two identical and adjacent events in the first fault evolution path can be merged into one event.

[0052] For any combination of fault events in the set of fault events, if neither the combination of fault events nor the other combinations of fault events in the set of fault events satisfies the preset association condition, then the combination of fault events is included as an isolated event combination in a single rule group, and a second fault evolution path is formed according to the single rule group.

[0053] The correlation conditions include: for any two fault event combinations, the fault type of the fault event in one fault event combination is the same as the fault type of the fault event in the other fault event combination; for any two fault event combinations, there is a causal relationship between the equipment state of the fault event in one fault event combination and the equipment state of the fault event in the other fault event combination, that is, whether the equipment state of the former fault event directly leads to the equipment state of the latter fault event; for example, inverter disconnection in event A and line overload in event B. Inverter disconnection will lead to load transfer and cause line overload, so there is a causal relationship between event A and event B.

[0054] Further, after determining whether any two combinations of fault events in the fault event set satisfy a preset association condition, and concatenating any two combinations of fault events that satisfy the preset association condition into a local evolution path to obtain several local evolution paths, the method further includes: For any combination of fault events in the set of fault events, if neither the combination of fault events nor the other combinations of fault events in the set of fault events satisfy the preset association condition, then the combination of fault events is regarded as an isolated combination of events. Obtain the timestamp of each isolated event combination, and then concatenate several isolated event combinations into a second fault evolution path based on the timestamp of each isolated event combination, and then generate a second test case according to the second fault evolution path; The second test case is input into the digital twin model so that the digital twin model simulates a self-healing process in response to the second test case; The test data of the digital twin model during the self-healing process is obtained, and then the self-healing function of the digital twin model is evaluated through the test data to obtain the test evaluation results.

[0055] In this embodiment, fault event combinations that cannot satisfy preset association conditions with other combinations are defined as isolated event combinations. A second fault evolution path and a second test case input digital twin model are generated based on timestamps to simulate and evaluate the self-healing process. This implementation also considers isolated or independent fault scenarios that may exist in complex fault scenarios without obvious evolutionary characteristics, supplementing the test cases for these isolated event combinations. These isolated event combinations are concatenated according to temporal relationships, enabling the generated second test cases to simulate test scenarios consisting of multiple independent but chronologically occurring fault events, supplementing the distribution network operation scenarios that the first test cases failed to simulate. Through the combination of the first and second test cases, the above self-healing function test can cover complex fault scenarios with evolutionary associations and isolated fault scenarios without evolutionary associations in distribution network operation, improving the comprehensiveness of the self-healing function test.

[0056] Further, generating the first test case based on the first fault evolution path includes: Based on the first fault evolution path, fault type evolution information and equipment status change information are generated; Generate timing-triggered instructions based on the device status change information; The first test case is generated based on the fault type evolution information, the equipment state change information, and the timing triggering instruction.

[0057] In a preferred embodiment, generating the first test case specifically includes: All fault event combinations in the first fault evolution path are identified, and all fault events within them are identified. Then, the fault characteristics and equipment status of each fault event are obtained. The fault characteristics include fault type and fault duration; the equipment status includes basic equipment information, real-time operating parameters, and status identifiers. The basic equipment information includes equipment ID, equipment type, and installation location; real-time operating parameters include voltage, current, and power factor; and the status identifiers include switch open / closed status: open or closed.

[0058] It is understandable that the aforementioned steps have already acquired the fault characteristics, equipment status, equipment associated topology status, and environmental interference associated data of each historical event when acquiring historical event data of the distribution network. Therefore, the fault type and equipment status of each fault event in the first fault evolution path can be directly obtained based on the historical event data.

[0059] Based on the order of fault events in the first fault evolution path, fault type evolution information is generated based on the timestamps and fault types of every two fault events.

[0060] Based on the order of fault events in the first fault evolution path, equipment status change information is generated based on the timestamps and equipment status of every two fault events.

[0061] Finally, the first fault evolution path generates multiple fault type evolution information and equipment state change information based on the timestamps of fault event changes. Each fault type evolution information corresponds to the timestamp of the change from the fault type of the previous fault event to the fault type of the next fault event in the path, including the fault type of the next fault event. For example, at timestamp 0ms, the fault type of the next fault event evolves to a voltage sag fault. Each equipment state change information corresponds to the timestamp of the change from the equipment state of the previous fault event to the equipment state of the next fault event in the path, including the initial equipment state and the final equipment state. The equipment state of the previous fault event is used as the initial equipment state, and the equipment state of the next fault event is used as the final equipment state.

[0062] Based on the timestamp of each device status change, a timing trigger command is generated for each timestamp. The timing trigger command consists of the timestamp of the device status change information and the command action, which includes fault injection actions and device status monitoring actions. For example: T=0ms, inject a 10kV voltage sag fault; T=200ms, monitor the inverter status. Here, injecting a 10kV voltage sag fault is the fault injection action, and monitoring the inverter status is the device status monitoring action.

[0063] Then, the fault type evolution information, device status change information and timing trigger instructions corresponding to each timestamp are converted into JSON format files as the first test case.

[0064] Similarly, for generating the second test case based on the second fault evolution path, fault type evolution information and equipment state change information are also obtained based on the second fault evolution path. Then, a timing trigger instruction is generated based on the timestamp of the equipment state change information. The timing trigger instruction consists of the timestamp of the equipment state change information and the instruction action, which includes a fault injection action and an equipment state monitoring action. Then, the fault type evolution information, equipment state change information, and timing trigger instruction are converted into a JSON format file as the second test case.

[0065] Specifically, the fault type can be represented by two fields: VoltageLevel and FaultType, with preset coded values, such as voltage sag = 01 and single-phase grounding = 02. Ultimately, a 10kV voltage sag fault type can be converted into a JSON file in the following format: VoltageLevel=10kV, FaultType=01.

[0066] Device status change information can use DeviceID (device code), InitialState (initial state code), and TargetState (final state code) as fields, with preset code values, such as grid-connected = 1, grid-off = 0; closed = 1, open = 0. Ultimately, device status change information for an inverter going from grid-connected to grid-off can be converted into the following JSON format file: DeviceID=Inv001, InitialState=1, TargetState=0.

[0067] Timing-triggered commands can use the fields Time (time value), ActionType (action code, custom settings, e.g., inject fault = A01, check status = A02), and TargetDeviceID (target device code). Ultimately, a timing-triggered command with T=200ms for checking inverter status can be converted into the following JSON format file: Time=200, ActionType=A02, TargetDeviceID=Inv001.

[0068] Input the first or second test case into the digital twin model to simulate the fault injection, self-healing response, and state recovery process.

[0069] When the system time (the simulation clock time built into the digital twin model of the distribution network) reaches the time value for the fault injection action, the fault simulation interface is triggered, and fault injection is performed according to the first test case or the second test case, such as adjusting the grounding resistance corresponding to a single-phase grounding.

[0070] When the system time reaches the time value for the device status monitoring action, the device status acquisition interface is triggered to obtain the real-time operating status of the corresponding device. If the real-time operating status of the device evolves into the target fault state, a self-healing action matching the fault type and device status is triggered; otherwise, the device status is restored, the status is reset, and the status is verified to verify the recovery effect of the devices within the path. The device status is continuously monitored to confirm that all devices are in normal status.

[0071] For example: if the real-time voltage of Inv001 is lower than the grid disconnection protection threshold, the real-time operating status of the equipment is determined to evolve into the target fault state.

[0072] In this embodiment, fault type evolution information and equipment state change information are extracted based on the first fault evolution path or the second fault evolution path, and timing trigger instructions are generated accordingly. Finally, the first test case or the second test case is generated by integrating these instructions. The fault evolution event data is transformed into specific instruction programs that can be identified and executed by the digital twin model. This achieves a precise digital expression of the complex fault evolution process, enabling the first test case or the second test case to accurately drive the digital twin model to simulate the entire process of fault dynamic evolution in the real power grid, and providing test case input for self-healing function testing.

[0073] In a preferred embodiment, inputting the first test case into a pre-built digital twin model to cause the digital twin model to simulate a self-healing process in response to the first test case includes: The first test case is input into the digital twin model to simulate the fault injection, self-healing response, and state recovery process. When the simulation clock time built into the distribution network digital twin model reaches the time value for the fault injection action, the fault simulation interface is triggered, and fault injection is performed according to the first test case, such as adjusting the grounding resistance corresponding to a single-phase grounding.

[0074] When the simulation clock time built into the distribution network digital twin model reaches the time value of the equipment status monitoring action, the equipment status acquisition interface is triggered to obtain the real-time operating status of the corresponding equipment. If the real-time operating status of the equipment evolves into the target fault state, the digital twin model triggers a self-healing action that matches the fault type and equipment status; otherwise, the digital twin model restores the equipment status, performs a status reset, and verifies the status to verify the recovery effect of the equipment within the path, continuously monitors the equipment status, and confirms that all equipment statuses are normal.

[0075] As an example, if the real-time voltage of device number Inv001 is lower than the grid disconnection protection threshold, the device's real-time operating state is determined to evolve into the target fault state.

[0076] Further, the step of acquiring test data of the digital twin model during the self-healing process, and then evaluating the self-healing function of the digital twin model using the test data to obtain test evaluation results, includes: Based on the test data, the digital twin model is used to calculate the device state change control rate, self-healing response timing matching rate, and initial state recovery rate during the self-healing process. A comprehensive self-healing performance score is obtained based on the device state change control rate, the self-healing response timing matching rate, and the initial state recovery rate, and the comprehensive self-healing performance score is used as the test evaluation result.

[0077] It should be noted that the equipment state change control rate, self-healing response timing matching rate, and initial state recovery rate are calculated based on the simulation results of the test data. The equipment state change control rate is used to evaluate the control effect of the distribution network self-healing system on the state changes of equipment of different importance. The self-healing response timing matching rate is used to evaluate the synchronization between self-healing actions and timing trigger commands. The initial state recovery rate is used to evaluate the degree of equipment recovery.

[0078] In this embodiment, the self-healing function is quantitatively evaluated from three dimensions—equipment control capability, response timing accuracy, and system recovery capability—by calculating the equipment state change control rate, self-healing response timing matching rate, and initial state recovery rate of the digital twin model during the self-healing process based on test data. The equipment state change control rate reflects the effect of the self-healing function on the equipment state; the self-healing response timing matching rate measures the timing accuracy of the self-healing action; and the initial state recovery rate characterizes the ability of the self-healing function to restore the system to a normal state. Obtaining a comprehensive self-healing performance score based on these three indicators as the test evaluation result overcomes the limitations of single-indicator evaluation, achieving a multi-dimensional comprehensive evaluation of the distribution network's self-healing function and improving the comprehensiveness and accuracy of the test evaluation results.

[0079] Further, the calculation of the device state change control rate, self-healing response timing matching rate, and initial state recovery rate of the digital twin model during the self-healing process based on the test data includes: Based on the test data, obtain the final fault status and importance coefficient of each device in the digital twin model; The control rate of equipment state change is calculated based on the final fault state and importance coefficient of each of the aforementioned devices.

[0080] In one specific embodiment, the device state change control rate The calculation is shown in the following formula: In the formula, n represents the total number of devices in the distribution network digital twin model. Let be the importance coefficient of the i-th device. The final fault state value of the i-th device is the value when the target fault state of the first test case or the second test case is reached. =1, when the target fault state of the first or second test case is not reached. =0. The accuracy of evaluating the control effectiveness of core equipment is improved by introducing an equipment importance weighting coefficient.

[0081] It should be noted that when the equipment reaches the target fault state, =1, the more devices that reach the target fault state, the higher the equipment state change control rate. The lower the value, the worse the effect of preventing the fault; when the equipment has not reached the target fault state... The fewer devices that reach the target fault state, the higher the control rate of equipment state changes. The higher the value, the better the effect of preventing the fault.

[0082] In this implementation, the final fault state and importance coefficient of each device in the digital twin model are obtained based on test data. Then, the device state change control rate is calculated based on the final fault state and importance coefficient of each device. This calculation method introduces the final fault state and importance coefficient of the devices. Considering the different levels of importance of different devices in the distribution network topology, the fault state of each device is weighted using the importance coefficient, giving greater weight to the fault control status of key devices in the evaluation results. This allows for a more accurate reflection of the actual effect of self-healing functions on ensuring the safe operation of the core areas of the power grid.

[0083] Further, the calculation of the device state change control rate, self-healing response timing matching rate, and initial state recovery rate of the digital twin model during the self-healing process based on the test data includes: Obtain the preset trigger time for each self-healing action in the digital twin model; Based on the test data, obtain the actual trigger time of each self-healing action in the digital twin model; The self-healing response timing matching rate is calculated based on the preset trigger time and actual trigger time of each self-healing action in the digital twin model.

[0084] In one specific embodiment, the self-healing response timing matching rate The calculation is shown in the following formula: In the formula, m is the total number of status monitoring instructions for the first test case or the second test case. The actual trigger time of the self-healing action corresponding to the j-th monitoring instruction. This is the preset trigger time for the self-healing action corresponding to the j-th status monitoring instruction. This is a fault propagation rate correction term, used to adjust for the timeliness requirements of response under different fault scenarios.

[0085] In this implementation, the preset trigger time of each self-healing action in the digital twin model is obtained, and the actual trigger time of each self-healing action is obtained based on test data. Then, the self-healing response timing matching rate is calculated based on the preset trigger time and the actual trigger time. This calculation method, by comparing the preset trigger time and the actual trigger time, quantitatively evaluates the timing accuracy of the self-healing action execution. It can effectively verify whether the self-healing function can respond according to the expected timing logic under complex fault scenarios, providing a quantitative indicator for evaluating the real-time response capability of the distribution network's self-healing function.

[0086] Further, the calculation of the device state change control rate, self-healing response timing matching rate, and initial state recovery rate of the digital twin model during the self-healing process based on the test data includes: Based on the test data, obtain the final recovery status of each device in the digital twin model; Based on the test data, obtain the actual stable runtime and minimum stable runtime of each device in the digital twin model after the self-healing process ends; The recovery integrity sub-item is calculated based on the final recovery status of each device, and the post-recovery stability sub-item is calculated based on the actual stable operating time and minimum stable operating time of each device. Then, the initial state recovery rate is calculated based on the recovery integrity sub-item and the post-recovery stability sub-item.

[0087] In one specific embodiment, the initial state recovery rate The calculation is shown in the following formula: in: In the formula, The integrity weight is set according to the recovery priority; For the stability weights after recovery, ; To restore the integrity of the sub-item, This is a sub-item for stability after recovery; To restore the identification value of the device, =1 indicates that the system has returned to its initial state. =0 indicates that it has not been restored; Let i be the importance coefficient of the i-th device; The actual stable operating time after the i-th device is restored. This represents the minimum stable operation time required after the i-th device is restored.

[0088] Preferably, by calculating the two sub-items of restored integrity and post-restoration stability separately and then weighting them together, the insufficient accuracy caused by single-dimensional evaluation is avoided, which is more in line with the long-term operation requirements of the distribution network.

[0089] In this implementation, the final recovery state of each device in the digital twin model is obtained based on test data, along with the actual stable runtime and minimum stable runtime of each device after the self-healing process ends. A recovery integrity sub-item is calculated based on the final recovery state of each device, and a post-recovery stability sub-item is calculated based on the actual and minimum stable runtimes. Finally, the initial state recovery rate is calculated based on these two sub-items. This calculation method not only focuses on whether the device has recovered to its normal operating state but also on whether the device can maintain continuous and stable operation after recovery. It can effectively identify problems such as the device failing again shortly after recovery, ensuring that the initial state recovery rate reflects the self-healing function's ability to restore the system to a normal state and maintain stable operation.

[0090] In a preferred embodiment, the step of acquiring test data of the digital twin model during the self-healing process, and then evaluating the self-healing function of the digital twin model using the test data to obtain test evaluation results, includes: The comprehensive self-healing performance score S is calculated based on the equipment state change control rate, self-healing response timing matching rate, and initial state recovery rate. The comprehensive performance evaluation scoring rules are set as follows: In the formula, This refers to the control rate for changes in equipment status. For self-healing response timing matching rate, This represents the initial state recovery rate.

[0091] The overall performance evaluation rules are as follows: When S≥0.9, the evaluation result is excellent. At this time, the distribution network self-healing system fully meets the requirements of the distribution network scenario and can be directly put into operation under complex working conditions. When 0.7≤S<0.9, the evaluation result is qualified. At this time, the distribution network self-healing system cannot fully meet the needs of the distribution network scenario. The fault injection parameters and the triggering time of the self-healing action need to be adjusted. For example, the original 400ms monitoring of Inv001 is adjusted to 380ms to pre-trigger the protection command and reduce the delay.

[0092] When S < 0.7, the evaluation result is unqualified. At this time, the distribution network self-healing system cannot meet the requirements of the distribution network scenario, and the self-healing strategy needs to be rebuilt.

[0093] Please refer to Figure 2 The second embodiment of the present invention provides a power distribution network self-healing function testing system, comprising: The data acquisition module 100 is used to acquire historical events of the power distribution network and the occurrence time of the historical events, and to combine any two historical events with the same occurrence time as a historical event combination to obtain a number of historical event combinations. The data filtering module 200 is used to filter out several combinations of fault events from several combinations of historical events; The fault event filtering module 300 is used to obtain the confidence level of each fault event combination, and then construct a fault event set based on several fault event combinations whose corresponding confidence levels are higher than a preset confidence threshold. The fault evolution path generation module 400 is used to determine whether any two fault event combinations in the fault event set meet a preset association condition, and to connect any two fault event combinations that meet the preset association condition into a local evolution path to obtain a number of local evolution paths. The test case generation module 500 is used to obtain the timestamp of each of the local evolution paths, connect several local evolution paths into a first fault evolution path based on the timestamp of each of the local evolution paths, and then generate a first test case based on the first fault evolution path. The self-healing function test module 600 is used to input the first test case into a pre-built digital twin model so that the digital twin model responds to the first test case to simulate a self-healing process; The self-healing performance evaluation module 700 is used to acquire test data of the digital twin model during the self-healing process, and then evaluate the self-healing function of the digital twin model through the test data to obtain test evaluation results.

[0094] The method and system for testing the self-healing function of a power distribution network provided by the present invention have at least the following advantages compared with the prior art: This invention breaks through the limitations of existing distribution network self-healing function testing methods that rely on isolated terminal commands. By mining historical event data and constructing fault evolution paths based on timestamps and confidence levels, it achieves dynamic simulation of the concurrent and evolutionary processes of multiple types of faults under complex fault scenarios in distribution networks. It can restore the dynamic characteristics of faults over time and with environmental changes, effectively solving the problem that existing technologies cannot verify self-healing performance under dynamically evolving faults, and significantly improving the fit between test scenarios and actual operating conditions.

[0095] This invention eliminates random interference by using confidence thresholds and constructs fault evolution paths through correlation condition judgments, ensuring the effectiveness and reliability of the first test case. Simultaneously, by distinguishing between evolution paths and isolated event paths, it addresses the operational testing needs of both fault evolution scenarios and independent fault scenarios, improving the comprehensiveness of distribution network self-healing function testing.

[0096] This invention establishes a multi-dimensional testing and evaluation system. Through indicators such as equipment state change control rate, self-healing response timing matching rate, and initial state recovery rate, it achieves quantitative evaluation of self-healing function from multiple perspectives such as fault control, action timing, and recovery quality, thereby significantly improving the effectiveness and accuracy of test results.

[0097] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0098] The term "embodiment" as used herein means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described; however, any combination of these technical features that does not contradict each other should be considered within the scope of this specification.

[0099] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various improvements and substitutions without departing from the concept of this application, and these improvements and substitutions should also be considered within the scope of protection of this invention. Therefore, the scope of protection of this application should be determined by the appended claims.

Claims

1. A method for testing the self-healing function of a power distribution network, characterized in that, include: Obtain historical events of the distribution network and the occurrence time of the historical events; combine any two historical events with the same occurrence time as a historical event combination to obtain several historical event combinations. Several combinations of failure events are selected from several combinations of historical events; Obtain the confidence level of each of the fault event combinations, and then construct a fault event set based on several fault event combinations whose corresponding confidence levels are higher than a preset confidence threshold; Determine whether any two combinations of fault events in the set of fault events satisfy a preset association condition, and connect any two combinations of fault events that satisfy the preset association condition into a local evolution path to obtain several local evolution paths. Obtain the timestamp of each of the local evolution paths, and concatenate several of the local evolution paths into a first fault evolution path based on the timestamp of each of the local evolution paths, and then generate a first test case based on the first fault evolution path; The first test case is input into a pre-built digital twin model so that the digital twin model responds to the first test case to simulate a self-healing process; The test data of the digital twin model during the self-healing process is obtained, and then the self-healing function of the digital twin model is evaluated through the test data to obtain the test evaluation results.

2. The power distribution network self-healing function test method of claim 1, wherein, The step of selecting several fault event combinations from several combinations of historical events includes: The historical event combinations that include fault characteristic fields and device status fields are selected as candidate combinations. The probability of each candidate combination appearing in all the historical event combinations is obtained, and then the candidate combinations whose occurrence probability is higher than a preset probability threshold are taken as the fault event combinations, thereby obtaining a number of fault event combinations.

3. The method of claim 1, wherein, The step of obtaining the confidence level of each of the fault event combinations, and then constructing a fault event set based on a number of fault event combinations whose corresponding confidence levels are higher than a preset confidence threshold, includes: For any combination of fault events: Identify the preceding and following events in the fault event combination; Based on several historical events and several combinations of historical events, the conditional probability of the subsequent event occurring after the preceding event occurs is obtained, and the conditional probability is used as the confidence level of the fault event combination.

4. The method of claim 1, wherein, After determining whether any two combinations of fault events in the fault event set satisfy a preset association condition, and concatenating any two combinations of fault events that satisfy the preset association condition into a local evolution path to obtain several local evolution paths, the method further includes: For any combination of fault events in the set of fault events, if neither the combination of fault events nor the other combinations of fault events in the set of fault events satisfy the preset association condition, then the combination of fault events is regarded as an isolated event combination. Obtain the timestamp of each isolated event combination, and then concatenate several isolated event combinations into a second fault evolution path based on the timestamp of each isolated event combination, and then generate a second test case according to the second fault evolution path; The second test case is input into the digital twin model so that the digital twin model simulates a self-healing process in response to the second test case; The test data of the digital twin model during the self-healing process is obtained, and then the self-healing function of the digital twin model is evaluated through the test data to obtain the test evaluation results.

5. The method of claim 1, wherein, The step of generating the first test case based on the first fault evolution path includes: Based on the first fault evolution path, fault type evolution information and equipment status change information are generated; Generate timing-triggered instructions based on the device status change information; The first test case is generated based on the fault type evolution information, the equipment state change information, and the timing triggering instruction.

6. The method of claim 1, wherein, The process of acquiring test data of the digital twin model during the self-healing process, and then evaluating the self-healing function of the digital twin model using the test data to obtain test evaluation results, includes: Based on the test data, the digital twin model is used to calculate the device state change control rate, self-healing response timing matching rate, and initial state recovery rate during the self-healing process. A comprehensive self-healing performance score is obtained based on the device state change control rate, the self-healing response timing matching rate, and the initial state recovery rate, and the comprehensive self-healing performance score is used as the test evaluation result.

7. The method of claim 6, wherein, The calculation of the device state change control rate, self-healing response timing matching rate, and initial state recovery rate of the digital twin model based on the test data during the self-healing process includes: Based on the test data, obtain the final fault status and importance coefficient of each device in the digital twin model; The control rate of equipment state change is calculated based on the final fault state and importance coefficient of each of the aforementioned devices.

8. The method of claim 6, wherein, The calculation of the device state change control rate, self-healing response timing matching rate, and initial state recovery rate of the digital twin model based on the test data during the self-healing process includes: Obtain the preset trigger time for each self-healing action in the digital twin model; Based on the test data, obtain the actual trigger time of each self-healing action in the digital twin model; The self-healing response timing matching rate is calculated based on the preset trigger time and actual trigger time of each self-healing action in the digital twin model.

9. The method of claim 6, wherein, The calculation of the device state change control rate, self-healing response timing matching rate, and initial state recovery rate of the digital twin model based on the test data during the self-healing process includes: Based on the test data, obtain the final recovery status of each device in the digital twin model; Based on the test data, obtain the actual stable runtime and minimum stable runtime of each device in the digital twin model after the self-healing process ends; The recovery integrity sub-item is calculated based on the final recovery status of each device, and the post-recovery stability sub-item is calculated based on the actual stable operating time and minimum stable operating time of each device. Then, the initial state recovery rate is calculated based on the recovery integrity sub-item and the post-recovery stability sub-item.

10. A power distribution grid self-healing function testing system, characterized by, include: The data acquisition module is used to acquire historical events of the distribution network and the occurrence time of the historical events, and to combine any two historical events with the same occurrence time as a historical event combination to obtain a number of historical event combinations. The data filtering module is used to filter out several combinations of fault events from several combinations of historical events; The fault event filtering module is used to obtain the confidence level of each fault event combination, and then construct a fault event set based on several fault event combinations whose corresponding confidence levels are higher than a preset confidence threshold. The fault evolution path generation module is used to determine whether any two fault event combinations in the fault event set meet a preset association condition, and to connect any two fault event combinations that meet the preset association condition into a local evolution path to obtain a number of local evolution paths. The test case generation module is used to obtain the timestamp of each of the local evolution paths, connect several local evolution paths into a first fault evolution path based on the timestamp of each of the local evolution paths, and then generate a first test case based on the first fault evolution path. The self-healing function testing module is used to input the first test case into a pre-built digital twin model so that the digital twin model responds to the first test case to simulate a self-healing process; The self-healing performance evaluation module is used to acquire test data of the digital twin model during the self-healing process, and then evaluate the self-healing function of the digital twin model through the test data to obtain test evaluation results.

Images