Power distribution network fault self-recovery method and system based on silicon carbide electric energy router

Abstract

The invention discloses a power distribution network fault self-recovery method and system based on a silicon carbide electric energy router, and relates to the technical field of power system automation. The method comprises the following steps: collecting transient electrical quantity signals of each branch, and calculating abnormal characteristic indexes; based on the abnormal characteristic indexes, calculating a correction fault probability in combination with neighborhood state differences, and determining a healthy branch set; calculating an optimized power supply amount according to the corrected fault probability and the bearable power; generating a control instruction of each healthy branch based on the corrected fault probability, the optimized power supply amount and the bearable power; issuing a control instruction to the silicon carbide electric energy router, and verifying execution consistency; and generating a branch control predicted value at a future moment based on the historical data. And the operation safety of the power distribution network is improved by identifying healthy branches and predicating faults.

Description

Technical Field

[0001] This invention relates to the field of power system automation technology, specifically to a method and system for self-healing of distribution network faults based on silicon carbide power routers. Background Technology

[0002] With the development of smart grids and distributed energy resources, the safety, reliability, and power supply optimization of distribution networks have become critical issues. Traditional distribution networks often rely on manual intervention or fixed protection devices for branch isolation and load dispatching when faults occur, resulting in slow response times and a lack of intelligent control methods. Especially in high-voltage or medium-voltage distribution systems using silicon carbide power routers, the large number of branches, complex power distribution, and rapid changes in transient electrical signals make it difficult for traditional methods to accurately identify faulty branches, optimize power supply, and control branch access or isolation. This can lead to increased load losses or higher system operational risks. Furthermore, existing fault handling methods lack full utilization of historical operating data, failing to predict future branch control strategies and achieve intelligent dispatching. Summary of the Invention

[0003] Based on the shortcomings of the prior art described above, the purpose of this invention is to provide a method and system for self-healing of power distribution network faults based on silicon carbide power routers, so as to solve the above-mentioned technical problems.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a self-healing method for distribution network faults based on a silicon carbide power router, comprising:

[0005] S1: Collect transient electrical quantity signals of each branch of the distribution network, perform high-dimensional feature mapping on the transient electrical quantity signals, and calculate the abnormal characteristic indicators of each branch.

[0006] S2: Based on abnormal feature indicators and combined with the state differences of neighboring branches, calculate the corrected fault probability of each branch, and determine the set of healthy branches based on the corrected fault probability.

[0007] S3: Calculate the optimized power supply for each healthy branch based on the corrected fault probability and the power it can carry.

[0008] S4: Based on the corrected fault probability, optimized power supply and load capacity of each healthy branch, control commands for each healthy branch are generated according to preset decision conditions.

[0009] S5: Send control commands to the silicon carbide power router and verify the consistency of execution;

[0010] S6: Generate branch control predictions for future times based on historical data.

[0011] The present invention is further configured such that S1 includes:

[0012] The transient electrical quantity signals of each branch of the distribution network are collected, including transient current signals and transient voltage signals;

[0013] Transient electrical quantity signals are captured based on a preset sampling window;

[0014] A high-dimensional feature mapping function is constructed to perform nonlinear combination and normalization operations on transient electrical signals within the window, and to calculate the abnormal feature indexes of each branch.

[0015] The present invention is further configured such that S2 includes:

[0016] Based on the abnormal characteristic indicators of each branch, the initial fault probability of each branch is calculated.

[0017] The difference in abnormal characteristic indicators of adjacent branches is introduced to correct the initial fault probability by neighborhood correlation, and the corrected fault probability of each branch is generated.

[0018] Branches with a fault correction probability less than a preset health threshold are identified as healthy branches and included in the healthy branch set.

[0019] The present invention is further configured such that S3 includes:

[0020] The power allocation weight of each healthy branch is inversely correlated with the corrected fault probability of each healthy branch.

[0021] The power allocation quota of each healthy branch is positively correlated with the power carrying capacity of each healthy branch.

[0022] By combining the allocation relationships of reverse and forward correlations, the optimal power supply for each healthy branch is determined.

[0023] The present invention is further configured such that S4 includes:

[0024] Based on the comparison between the corrected fault probability of each healthy branch and the preset access threshold, as well as the comparison between its carrying power and the optimized power supply, control instructions for each healthy branch are generated. The control instructions include access instructions and isolation instructions.

[0025] When the probability of correcting a fault is less than the preset access threshold and the power that can be carried is greater than or equal to the optimized power supply, an access command is generated.

[0026] An isolation command is generated when the probability of correcting a fault is greater than or equal to the preset access threshold, or when the power that can be carried is less than the optimized power supply.

[0027] The present invention is further configured such that S5 includes:

[0028] Send control commands to the silicon carbide power router and obtain the actual control status and output power feedback from each branch;

[0029] Based on the control commands, actual control status, optimized power supply and output power of each health branch, the consistency of command execution is verified and verification result indicators are generated.

[0030] When the verification result index is less than the preset success threshold, the control command execution is determined to have failed, and the backup isolation scheme is triggered.

[0031] The present invention is further configured such that S6 includes:

[0032] Based on the historical abnormal characteristic indicators and historical control command data of each healthy branch, a prediction function is constructed to generate the predicted control values ​​of the branch at future times.

[0033] The predicted control values ​​are stored and used as a reference for decision-making on branch access or isolation in subsequent fault handling cycles.

[0034] The present invention is further configured such that the transient electrical quantity signal is acquired through distributed synchronous measurement points, and each measurement unit achieves time synchronization based on a high-precision time synchronization protocol.

[0035] The present invention is further configured such that the method also includes real-time visualization of the fault correction probability, control commands and actual control status of each branch.

[0036] This invention also provides a power distribution network fault self-healing system based on a silicon carbide power router, the system comprising:

[0037] Feature index calculation module: Collects transient electrical quantity signals from each branch of the distribution network, performs high-dimensional feature mapping on the transient electrical quantity signals, and calculates the abnormal feature indexes of each branch;

[0038] Fault probability calculation module: Based on abnormal feature indicators and combined with the state differences of neighboring branches, calculate the corrected fault probability of each branch, and determine the set of healthy branches based on the corrected fault probability.

[0039] Optimized power supply calculation module: Calculates the optimized power supply for each healthy branch based on the corrected fault probability and the power it can carry.

[0040] Command generation module: Based on the corrected fault probability, optimized power supply and load capacity of each healthy branch, the module generates control commands for each healthy branch according to preset decision conditions.

[0041] Control execution verification module: issues control commands to the silicon carbide power router and verifies the consistency of execution;

[0042] Control prediction module: Generates branch control prediction values ​​for future times based on historical data.

[0043] This invention provides a method and system for self-healing faults in distribution networks based on silicon carbide power routers. The method comprises the following steps: S1: Collecting transient electrical quantity signals from each branch of the distribution network, performing high-dimensional feature mapping on the transient electrical quantity signals, and calculating abnormal characteristic indicators for each branch; S2: Based on the abnormal characteristic indicators and the state differences of neighboring branches, calculating the corrected fault probability for each branch, and determining a set of healthy branches based on the corrected fault probability; S3: Calculating the optimized power supply for each healthy branch based on its corrected fault probability and carrying capacity; S4: Generating control commands for each healthy branch based on its corrected fault probability, optimized power supply, and carrying capacity, according to preset decision conditions; S5: Sending the control commands to the silicon carbide power router and verifying execution consistency; S6: Generating predicted branch control values ​​for future times based on historical data. The beneficial effects include:

[0044] 1. By collecting transient electrical quantity signals of each branch and performing high-dimensional feature mapping, calculating abnormal characteristic indicators, and combining the fault probability with the state differences of neighboring branches, the system can accurately identify healthy branches, thereby reducing misjudgments and missed judgments and improving the operational safety of the distribution network.

[0045] 2. Based on historical anomaly characteristic indicators and control commands, a prediction function is constructed to generate predicted values ​​for branch control at future times, providing a reference for subsequent fault handling cycles and realizing the forward-looking and dynamic decision-making capabilities of fault response.

[0046] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0048] Figure 1 A flowchart illustrating a distribution network fault self-healing method based on a silicon carbide power router, as an exemplary embodiment of the present invention;

[0049] Figure 2 This is a schematic diagram illustrating the structure of a power distribution network fault self-healing system based on a silicon carbide power router, as an exemplary embodiment of the present invention. Detailed Implementation

[0050] The embodiments of the present invention will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention and not for limiting the scope of protection of the present invention.

[0051] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0052] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0053] Example 1

[0054] A self-healing method for distribution network faults based on silicon carbide power routers, such as Figure 1 As shown, it includes:

[0055] S1: Collect transient electrical quantity signals of each branch of the distribution network, perform high-dimensional feature mapping on the transient electrical quantity signals, and calculate the abnormal characteristic indicators of each branch.

[0056] S2: Based on abnormal feature indicators and combined with the state differences of neighboring branches, calculate the corrected fault probability of each branch, and determine the set of healthy branches based on the corrected fault probability.

[0057] S3: Calculate the optimized power supply for each healthy branch based on the corrected fault probability and the power it can carry.

[0058] S4: Based on the corrected fault probability, optimized power supply and load capacity of each healthy branch, control commands for each healthy branch are generated according to preset decision conditions.

[0059] S5: Send control commands to the silicon carbide power router and verify the consistency of execution;

[0060] S6: Generate branch control predictions for future times based on historical data.

[0061] The present invention is further configured such that S1 includes:

[0062] The transient electrical quantity signals of each branch of the distribution network are collected, including transient current signals and transient voltage signals;

[0063] Transient electrical quantity signals are captured based on a preset sampling window;

[0064] A high-dimensional feature mapping function is constructed to perform nonlinear combination and normalization operations on transient electrical quantity signals within the window, calculating the abnormal characteristic indicators of each branch. Specifically, this step is used to perform high-dimensional feature mapping on the transient current and voltage signals from the distribution to each branch, calculating the abnormal characteristic indicators of the branches to quantify the abnormal intensity of the branches under short-term transients. The transient current and voltage signals of each branch are collected in real time through a silicon carbide power router, and the collected signals are truncated according to the preset sampling window length to form a continuous time series segment for the calculation of abnormal characteristic indicators. The construction logic of the high-dimensional feature mapping function is as follows: Among them, M i (t) represents the abnormal characteristic index of the i-th branch at time t; N is the sampling window length; I i (t+k) and I i (t+k-1) represent the transient currents of the i-th branch at times t+k and t+1-k, respectively, used to capture the transient current fluctuations of the branch; V i (t+k) and V i (t+k-1) represents the transient voltage at time t+k and time t+1-k, respectively, used to capture the voltage transient fluctuation information of the branch; γ1, γ2, γ3 and γ4 are power exponent coefficients. γ1 and γ2 normalize the current and voltage amplitudes respectively to suppress the influence of background fluctuations on abnormal feature indicators, with a value range of [0.5,2]. γ3 and γ4 are used to amplify the higher-order features of transient current changes and transient voltage changes respectively, with a value range of [1,4].

[0065] The present invention is further configured such that S2 includes:

[0066] Based on the abnormal characteristic indicators of each branch, the initial fault probability of each branch is calculated.

[0067] The difference in abnormal characteristic indicators of adjacent branches is introduced to correct the initial fault probability by neighborhood correlation, and the corrected fault probability of each branch is generated.

[0068] Branches with a corrected fault probability less than a preset health threshold are identified as healthy branches and included in the healthy branch set. Specifically, this step integrates the abnormal characteristic indicators of the branch and the differences in the state of the surrounding branches to calculate the corrected fault probability, thereby accurately identifying healthy branches. The initial fault probability is obtained by transforming the high-dimensional feature mapping value through an exponential function. The exponential amplification mechanism is used to highlight the abnormality of abnormal branches, making them more prominent in the probability space. The initial fault probability only reflects the abnormality of the branch itself. To identify distributed faults and reduce misjudgments caused by measurement noise, the differences in the state of surrounding branches are introduced to correct the initial fault probability, resulting in the corrected fault probability. The calculation logic for the corrected fault probability is as follows: in, α is the corrected fault probability of the i-th branch at time t, used to quantify the likelihood of a branch fault; α is an adjustment coefficient used to exponentially amplify the abnormal characteristic indicators of the branch, highlighting the probability of branches with high abnormal characteristic indicator values ​​to enhance the sensitivity to faulty branches, with a value range of [0.5, 3]; B is the total number of branches within the monitoring range of the distribution network; N i Let M be the set of neighboring branches of the i-th branch; β is the adjustment coefficient, used to control the degree of influence of neighborhood differences on the fault correction probability, with a value range of [0.1, 1]; δ is the exponential coefficient, used to nonlinearly amplify extreme differences between neighborhoods, with a value range of [1, 3]; M i (t), M j (t) and M k (t) represents the abnormal feature indices of the i, j, and k branches, respectively; The initial fault probability is used to highlight the relative weight of the anomaly of the i-th branch in the entire distribution network. This is the average difference between the abnormal characteristic indicators of the i-th branch and its neighboring branches, measuring the degree of abnormal deviation of the branch in the local neighborhood, and used for neighborhood correlation correction; when the corrected fault probability is less than a preset health threshold, i.e.: The branch is classified as healthy, and all branches that meet this condition are grouped together to form a healthy branch set, where ξ healthy Set a preset health threshold.

[0069] The present invention is further configured such that S3 includes:

[0070] The power allocation weight of each healthy branch is inversely correlated with the corrected fault probability of each healthy branch.

[0071] The power allocation quota of each healthy branch is positively correlated with the power carrying capacity of each healthy branch.

[0072] By combining the allocation relationships of reverse and forward correlations, the optimal power supply for each healthy branch is determined. Specifically, the core objective of this step is to calculate the optimal power allocation scheme, i.e., the optimal power supply, for each healthy branch after a fault occurs, based on the corrected fault probability and the power carrying capacity of each branch. This ensures that the remaining healthy power grid can restore power to critical loads to the greatest extent possible while maintaining safety and stability. The calculation logic for the optimal power supply is as follows: Among them, Q i (t) represents the optimized power supply for the i-th branch; This is a safety weighting factor used to adjust the sensitivity of power allocation to branch fault correction probability, expressed as a negative exponent. By performing reverse correlation, the allocation weight of branches with higher fault probability decreases exponentially, thereby reducing the power allocation ratio of high-risk branches and ensuring that power is preferentially allocated to branches with higher safety and reliability; S i (t) θ S is a capacity weighting factor used to adjust the amplification effect of power allocation on the branch's carrying capacity. i (t) represents the power carrying capacity of the i-th branch, expressed as a positive exponent with respect to S. i (t) Perform positive correlation so that the allocation quota of the branch with the larger power carrying capacity increases exponentially, thereby prioritizing the increase of the power supply ratio of high-capacity branches, fully releasing the grid transmission capacity, and achieving efficient and balanced distribution of load among healthy branches, reducing the risk of power supply bottlenecks. η is a normalization factor used to constrain the power allocation results of each health journey, ensuring that the total power supply is equal to the current available capacity limit. This ensures that the optimized power supply satisfies the power balance conditions of the power grid and has the meaning of a command that can be directly issued and executed, avoiding secondary overload or undersupply problems caused by allocation exceeding the limit; η is an adjustment coefficient used to control the power allocation reduction intensity of branches with high fault probability, with a value range of [0.5,3]; θ is an adjustment coefficient used to control the allocation amplification degree of branches with high load-bearing power, with a value range of [1,2].

[0073] The present invention is further configured such that S4 includes:

[0074] Based on the comparison between the corrected fault probability of each healthy branch and the preset access threshold, as well as the comparison between its carrying power and the optimized power supply, control instructions for each healthy branch are generated. The control instructions include access instructions and isolation instructions.

[0075] When the probability of correcting a fault is less than the preset access threshold and the power that can be carried is greater than or equal to the optimized power supply, an access command is generated.

[0076] When the corrected fault probability is greater than or equal to the preset access threshold, or the power carrying capacity is less than the optimized power supply, an isolation command is generated; specifically, the corrected fault probability of each healthy branch is... With the preset access threshold ξ in The comparison was performed, and the power S that this branch can carry was also compared. i (t) and optimized power supply Q i (t) is compared, when And S i (t)≥Q i When (t), an access command is generated, indicating that the branch can be put back into operation; when or S i (t) i When (t), an isolation command is generated, indicating that the branch should remain disconnected or be taken out of operation; through the above dual-condition determination, high-risk branch access is avoided, while ensuring that the power supply capacity is not overloaded. The dual-condition determination can be formalized as a branch state function: R i (t)= Among them, R i (t) = 1 indicates access, R i (t) = 0 indicates isolation; ξ in The preset access threshold; For reliability constraints, the probability of a branch correcting a fault must be less than a preset access threshold. Only branches that achieve sufficient reliability can be reconnected. The preset access threshold is ξ. in The setting is usually higher than the preset health threshold ξ healthy More stringent; S i (t)≥Q i (t) represents the capacity constraint. The power that the branch can carry must be greater than or equal to the optimized power supply to provide a physical safety margin and ensure that the branch's lines, switches and other equipment will not be overloaded when actually carrying the allocated power, thereby effectively preventing secondary faults caused by power restoration. An access command will only be generated when the branch meets both the reliability constraint and the capacity constraint. If either condition is not met, an isolation command will be generated.

[0077] The present invention is further configured such that S5 includes:

[0078] Send control commands to the silicon carbide power router and obtain the actual control status and output power feedback from each branch;

[0079] Based on the control commands, actual control status, optimized power supply and output power of each health branch, the consistency of command execution is verified and verification result indicators are generated.

[0080] ​When the verification result index is less than the preset success threshold, the control command execution is deemed to have failed, and a backup isolation scheme is triggered. Specifically, the control command is sent to each branch silicon carbide power router to perform physical switching operations, and the actual control status and output power of each branch are collected in real time. To quantify the consistency of control command execution, a consistency discrimination function is constructed for each branch. The construction logic of the consistency discrimination function is as follows: Among them, Γ i (t) represents the consistency verification result of instruction execution in the i-th branch; This represents the actual control state of the i-th branch, used to determine the execution status of instructions. The value can be 0 or 1, where 0 indicates that the switch is in the open state and 1 indicates that the switch is in the closed state; Γ is the output power, used to compare with the optimized power supply to determine the power execution deviation; χ is the adjustment coefficient, used to control the sensitivity of the consistency discrimination function to differences, with a value range of [3,5]; ρ and σ are power coefficients, used to nonlinearly amplify the deviation between the control command and the actual control state and the deviation between the optimized power supply and the output power, respectively, with a value range of [2,4]; the calculated verification result index is compared with the preset success threshold to determine whether the command was executed successfully; when the verification result index is less than the preset success threshold, i.e.: Γ i If (t) ≥ τ, the control command is considered successfully executed. The system confirms the status of the branch and continues monitoring. When the verification result index is less than the preset success threshold, i.e.: Γ i If (t) < τ, the control command execution is deemed to have failed, and the backup isolation scheme is immediately triggered. The backup isolation scheme may include: activating redundant backup switches to operate the backup switchgear connected in parallel with the faulty branch to achieve rapid disconnection; expanding the isolation range by tripping the upstream circuit breaker or tie switch of the faulty branch to ensure complete fault isolation and prevent its spread; issuing an abnormal alarm to prompt maintenance personnel to intervene immediately and conduct on-site manual troubleshooting; and simultaneously applying a safety interlock signal to the faulty branch to which the isolation command was executed, prohibiting any automatic or manual reclosing operation until maintenance personnel confirm that the fault has been eliminated and manually release the interlock, thereby preventing accidental reclosing before the fault is eliminated and causing the accident to escalate.

[0081] The present invention is further configured such that S6 includes:

[0082] Based on the historical abnormal characteristic indicators and historical control command data of each healthy branch, a prediction function is constructed to generate the predicted control values ​​of the branch at future times.

[0083] The predicted control values ​​are stored as a reference for branch access or isolation decisions in subsequent fault handling cycles. Specifically, based on the historical abnormal characteristic indicators and historical control command values ​​of each healthy branch, a prediction function is constructed to generate branch control prediction values ​​for future times. This prediction function calculates the similarity between current and historical abnormal characteristic indicators, weights and integrates historical control decisions, thereby predicting the control state of each branch at future times and achieving a forward-looking judgment on the branch operation trend. The construction logic of the prediction function is as follows: in, Here, H represents the predicted control value of the i-th branch at time t+Δt; H is the preset historical observation window length; h is the index; ||M i (t)-M i (th)|| is used to calculate the distance between the anomaly indicator at time th and the current time t. The larger the distance, the less similar the historical state is to the current state; M i (th) is the anomaly characteristic index of the i-th branch at time th; R i (th) represents the control command value of the i-th branch at time th; γ is the adjustment coefficient, used to control the sensitivity of the prediction function to the difference between historical and current abnormal characteristic indicators, with a value range of [0.5, 5]; μ is the power exponent, used to nonlinearly amplify the difference in abnormal characteristic indicators, with a value range of [1, 2]; ν is the adjustment coefficient, used to nonlinearly transform the historical control command value, with a value range of [0.5, 2]. The branch control prediction value is stored in the database and called in subsequent fault handling cycles. When the system executes step S4, the branch control prediction value can be called as an auxiliary decision-making basis. For example, by comparing the branch control prediction value with a preset threshold, it can be used as a third constraint condition in addition to reliability constraints and capacity constraints, thereby generating more accurate control commands and realizing dynamic decision optimization.

[0084] The present invention is further configured such that the transient electrical quantity signals are acquired through distributed synchronous measurement points, and each measurement unit achieves time synchronization based on a high-precision time synchronization protocol. Specifically, in this embodiment, the transient electrical quantity signals are acquired through distributed synchronous measurement units deployed at the beginning of each branch. Each measurement unit maintains microsecond-level time synchronization with the master clock source based on a high-precision time synchronization protocol, ensuring that the acquired data of the entire distribution network has a unified time standard. The high-precision time synchronization protocol preferably adopts the IEEE 1588 precision clock protocol or the IRIG-B code method. This synchronization mechanism ensures that the transient voltage and current signals between different branches are strictly aligned in time, providing an accurate time-domain consistency basis for subsequent collaborative analysis of multi-branch signals and high-dimensional feature mapping.

[0085] The present invention is further configured such that the method also includes real-time visualization of the corrected fault probability, control commands, and actual control status of each branch; specifically, the corrected fault probability is presented in the form of a dynamic curve or heat map to reflect the current risk level of each branch; the control commands and actual control status are displayed with icons or colors to intuitively distinguish between branches that have been connected, isolated, and those that are executing abnormally; users can monitor the power grid operation status in real time through the interface and promptly identify and handle abnormal branches, thereby improving the transparency and operability of the system operation.

[0086] Example 2

[0087] Please see Figure 2 This exemplary distribution network fault self-healing system based on silicon carbide power routers includes:

[0088] Feature index calculation module: Collects transient electrical quantity signals from each branch of the distribution network, performs high-dimensional feature mapping on the transient electrical quantity signals, and calculates the abnormal feature indexes of each branch;

[0089] Fault probability calculation module: Based on abnormal feature indicators and combined with the state differences of neighboring branches, calculate the corrected fault probability of each branch, and determine the set of healthy branches based on the corrected fault probability.

[0090] Optimized power supply calculation module: Calculates the optimized power supply for each healthy branch based on the corrected fault probability and the power it can carry.

[0091] Command generation module: Based on the corrected fault probability, optimized power supply and load capacity of each healthy branch, the module generates control commands for each healthy branch according to preset decision conditions.

[0092] Control execution verification module: issues control commands to the silicon carbide power router and verifies the consistency of execution;

[0093] Control prediction module: Generates branch control prediction values ​​for future times based on historical data.

[0094] It should be noted that the distribution network fault self-healing system based on silicon carbide power routers provided in the above embodiments and the distribution network fault self-healing method based on silicon carbide power routers provided in the above embodiments belong to the same concept. The specific operation methods of each module and unit have been described in detail in the method embodiments and will not be repeated here. In practical applications, the distribution network fault self-healing system based on silicon carbide power routers provided in the above embodiments can be assigned to different functional modules as needed, that is, the internal structure of the system can be divided into different functional modules to complete all or part of the functions described above. This is not a limitation here.

[0095] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for power distribution grid fault self-healing based on silicon carbide power router, characterized in that, The method comprises the following steps: S1: Collecting transient electrical quantity signals of each branch of the power distribution network, performing high-dimensional feature mapping on the transient electrical quantity signals, and calculating abnormal feature indexes of each branch; S2: Based on the abnormal feature indexes, combining the state differences of the adjacent branches, calculating the corrected fault probability of each branch, and determining the healthy branch set according to the corrected fault probability; S3: According to the corrected fault probability and the bearable power of each healthy branch, the optimized power supply of each healthy branch is calculated; S4: Based on the corrected fault probability, the optimized power supply and the bearable power of each healthy branch, the control instruction of each healthy branch is generated according to the preset decision condition; S5: The control instruction is sent to the silicon carbide electric energy router, and the execution consistency is verified; S6: Based on the historical data, the branch control prediction value at the future time is generated.

2. The method for power distribution grid fault self-healing based on silicon carbide power router according to claim 1, characterized in that, The S1 comprises: Collecting transient electrical quantity signals of each branch of the power distribution network, the transient electrical quantity signals comprising transient current signals and transient voltage signals; The transient electrical quantity signals are intercepted based on a preset sampling window; A high-dimensional feature mapping function is constructed to perform nonlinear combination and normalization operation on the transient electrical quantity signals in the window, and the abnormal feature indexes of each branch are calculated.

3. The method for power distribution grid fault self-healing based on silicon carbide power electronic router of claim 1, wherein, The S2 comprises: Based on the abnormal feature indexes of each branch, the initial fault probability of each branch is calculated; The initial fault probability is corrected by introducing the abnormal feature index difference of the adjacent branches, and the corrected fault probability of each branch is generated; The branch with a corrected fault probability less than a preset health threshold is determined as a healthy branch and is included in the healthy branch set.

4. The method for power distribution grid fault self-healing based on silicon carbide power electronic router of claim 1, wherein, The S3 comprises: Based on the corrected fault probability of each healthy branch, the power distribution weight of each healthy branch is reversely associated; Based on the bearable power of each healthy branch, the power distribution quota of each healthy branch is positively associated; The reverse association and positive association are combined to determine the optimized power supply of each healthy branch.

5. The method for power distribution grid fault self-healing based on silicon carbide power electronic router of claim 1, wherein, The S4 comprises: Based on the comparison of the corrected fault probability of each healthy branch with the preset access threshold, and the comparison of the bearable power with the optimized power supply, the control instruction of each healthy branch is generated, including the access instruction and the isolation instruction; When the corrected fault probability is less than the preset access threshold and the bearable power is greater than or equal to the optimized power supply, the access instruction is generated; When the corrected fault probability is greater than or equal to the preset access threshold, or the bearable power is less than the optimized power supply, the isolation instruction is generated.

6. The method for power distribution grid fault self-healing based on silicon carbide power electronic router of claim 1, wherein, The S5 comprises: The control instruction is sent to the silicon carbide electric energy router, and the actual control state and the output power of each branch are fed back; Based on the control instruction, the actual control state, the optimized power supply and the output power of each healthy branch, the instruction execution consistency verification is performed and the verification result index is generated; When the verification result index is less than the preset success threshold, it is determined that the control instruction execution fails, and a backup isolation scheme is triggered.

7. The method for power distribution grid fault self-healing of silicon carbide based power electronic router according to claim 1, wherein, The S6 comprises: Based on the historical abnormal feature indexes and the historical control instruction data of each healthy branch, a prediction function is constructed to generate the branch control prediction value at the future time; The control prediction value is stored as a reference for branch access or isolation decision-making in the subsequent fault handling period.

8. The method for power distribution grid fault self-healing of silicon carbide based power electronic router according to claim 1, wherein, The collection of the transient electrical quantity signals is performed through distributed synchronous measurement points, and each measurement unit realizes time synchronization based on a high-precision time synchronization protocol.

9. The method for power distribution grid fault self-healing of silicon carbide based power electronic router according to claim 1, wherein, The method further includes real-time visual display of the corrected fault probability, the control instruction and the actual control state of each branch.

10. A power distribution grid fault self-healing system based on silicon carbide power routers for implementing the method of power distribution grid fault self-healing based on silicon carbide power routers according to any one of claims 1 to 9, characterized in that, The method comprises the following steps: a feature index calculation module: collecting transient electrical quantity signals of each branch of the power distribution network, performing high-dimensional feature mapping on the transient electrical quantity signals, and calculating abnormal feature indexes of each branch; a fault probability calculation module: based on the abnormal feature indexes, combining the state differences of the neighboring branches, calculating the corrected fault probability of each branch, and determining a set of healthy branches according to the corrected fault probability; an optimized power supply calculation module: calculating the optimized power supply of each healthy branch according to the corrected fault probability and the bearable power of each healthy branch; an instruction generation module: based on the corrected fault probability, the optimized power supply and the bearable power of each healthy branch, generating a control instruction of each healthy branch according to a preset decision condition; a control execution verification module: issuing the control instruction to the silicon carbide electric energy router and verifying the execution consistency; a control prediction module: generating a branch control prediction value at a future time based on historical data.

Images