Distributed fault self-healing method suitable for active power distribution network

Abstract

The invention relates to the technical field of power grid regulation, in particular to a distributed fault self-healing method suitable for an active power distribution network, and the method comprises the steps: obtaining a target protection list through the step S1, calibrating the current time of an STU through the step S2, and carrying out the high-frequency collection of target high-frequency patrol data through the step S3, the method further comprises the steps of performing signal smoothing processing on target high-frequency patrol data through the step S4, generating electrical fault fingerprint characteristic values through the step S5, calculating fingerprint similarity through the step S6, and adopting a reasonable power supply mode for the sufficient degree of island available power through the step S7. And in the step S8, the precision requirement of time synchronization is calibrated in real time when the target protection list is frequently updated, and in the step S9, an extreme scene is set and a fault self-healing mode is correspondingly updated.

Description

Technical Field

[0001] This invention relates to the field of power grid regulation technology, specifically to a distributed fault self-healing method suitable for active distribution networks. Background Technology Fault location and protection in active distribution networks are crucial for ensuring power supply reliability. Traditional methods primarily rely on overcurrent protection and time-delay coordination, which presents a trade-off between speed and selectivity, making them unsuitable for complex network structures. With the increasing penetration of distributed power sources, fault characteristics are becoming more complex, placing higher demands on protection performance. To further improve location accuracy and speed, more advanced signal processing and intelligent algorithms are needed to achieve millisecond-level precise fault isolation, meeting the self-healing operation requirements of smart grids.

[0002] Chinese patent application CN115395489B discloses a method and system for distributed fault self-healing in active distribution networks based on hierarchical partitioning. The method includes, during normal operation of the active distribution network system, intelligent terminal devices using their configured local topology information to hierarchically divide the system into regions and distribute the hierarchical region information; when the operating state of the active distribution network system changes, the intelligent terminal devices adaptively update the stored hierarchical region information; when a fault occurs in a feeder section of the active distribution network system, the fault section is located and isolated based on the adaptive hierarchical region information and the differential protection principle; after the fault section is isolated, the relevant intelligent terminal devices use their stored network hierarchical region information to perform distributed power supply restoration. However, this approach still suffers from problems such as the lack of generated electrical fault fingerprint features leading to complex fault identification, and the failure to plan and update fault self-healing based on the target protection list, resulting in significant power grid fault losses. Summary of the Invention

[0003] The purpose of this invention is to provide a distributed fault self-healing method suitable for active distribution networks, so as to overcome the problems in the prior art that the failure to generate electrical fault fingerprint feature values ​​leads to complex fault identification, and the failure to perform fault self-healing and updating according to the target protection list leads to large power grid fault losses.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: A distributed fault self-healing method applicable to active distribution networks includes the following steps: Step S1: Obtain the target protection list; Step S2: Configure the STU and monitor the time difference of the STU to obtain the device time difference. Then, perform time synchronization processing on the STU based on the device time difference to obtain the synchronized STU. Step S3: After synchronization, the STU performs high-frequency acquisition of high-frequency patrol data of the target. Step S4: Perform signal smoothing processing on the target high-frequency patrol data to obtain processed high-frequency patrol data; Step S5: Obtain the fingerprint feature value of electrical fault based on the processed high-frequency patrol data; Step S6: Perform fault isolation processing on the distributed power grid based on the electrical fault fingerprint feature value to obtain the distributed power grid after isolation processing; Step S7: Based on the target protection list, islanding is established for the isolated distributed power grid to obtain the fault self-healing mode; Step S8: When executing the fault self-healing mode, the update time interval is obtained, and the time synchronization process is updated with difference based on the update time interval. The protection list update deviation is also obtained, and the difference update process is optimized based on the protection list update deviation. Step S9: Acquire extreme scenarios and update the fault self-healing mode according to the extreme scenarios.

[0005] Furthermore, in step S1, when obtaining the target protection list, the target protection list is obtained through the government platform. The target protection list includes the unit name, load location, electricity consumption and allowable power outage time. In step S2, when setting up the STU and monitoring the time difference of the STU, the time difference of the STU is monitored by the time difference monitoring method. The time difference monitoring method includes: Step A01: Set the master clock and slave clock; Step A02: Send a Sync message to the slave clock via the master clock and record the time point t1 at which the Sync message is sent; Step A03: Record the time point t2 when the Sync message is received; Step A04: Send a Delay message from the clock to the master clock and record the time point t3 when the Delay message is sent; Step A05: Record the time point t4 when the Delay message is received; Step A06: Calculate the device time difference Δt based on the time point t1 when the Sync message is sent, the time point t2 when the Sync message is received, the time point t3 when the Delay message is sent, and the time point t4 when the Delay message is received. Set Δt=[(t2-t1)-(t4-t3)] / 2 to obtain the device time difference Δt.

[0006] Further, in step S2, when performing time synchronization processing on the STU based on the device time difference, the device time difference Δt is compared with a preset device time difference Δt0, where Δt0 = 1 microsecond. The state of the device time difference is determined based on the comparison result, and the STU is then time synchronized based on the determination result, wherein: When Δt≤Δt0, the time difference of the equipment is considered acceptable, and no time synchronization processing is performed on the STU. When Δt > Δt0, the device time difference is deemed unacceptable. Time synchronization is then performed on the STU. The current time t of the STU is obtained, and the STU update time tp is calculated based on the current time t and the device time difference Δt. The STU update time tp is set to t - Δt. The current time t of the STU is then calibrated to the STU update time tp, resulting in the synchronized STU.

[0007] Furthermore, in step S3, the target high-frequency patrol data is obtained by the synchronized STU acquiring the target high-frequency patrol data at a rate of 1 million acquisitions per second. In step S4, when performing signal smoothing processing on the target high-frequency patrol data, a window is pre-set for the target high-frequency patrol data to obtain a preset filtering window, and the target high-frequency patrol data is windowed according to the preset filtering window to obtain processed high-frequency patrol data.

[0008] Further, in step S5, when obtaining the electrical fault fingerprint feature value based on the processed high-frequency patrol data, the current mutation value La is calculated based on the real-time current Lb in the processed high-frequency patrol data and the real-time current Lc in the processed high-frequency patrol data at the previous moment. La is set to |Lb-Lc|, and the current mutation value La is obtained. In step S5, the current mutation value La is compared with the preset current mutation value La0. Based on the comparison result, the real-time current mutation situation is judged, and the electrical fault fingerprint feature value is obtained based on the judgment result, wherein: When La≤La0, the sudden change in real-time current is determined to be non-existent, and no electrical fault fingerprint feature value is acquired. When La > La0, the sudden change in real-time current is determined to be a sudden change. The fingerprint feature value of electrical fault is obtained, the initial fault waveform is truncated, and the fingerprint of the initial fault waveform is compressed by wavelet transform energy distribution to obtain the fingerprint feature value of electrical fault.

[0009] Further, in step S6, when performing fault isolation processing on the distributed power grid based on the electrical fault fingerprint feature value, the synchronized STU prioritizes sending the electrical fault fingerprint feature value to the neighboring STU via a dual-channel transmission method. The neighboring STU acquires the fingerprint similarity Df and compares the fingerprint similarity Df with a preset fingerprint similarity Df0, setting Df0=0.35. Based on the comparison result, it judges the fingerprint similarity state and outputs the fault location based on the judgment result, wherein: When Df≤Df0, the fingerprint similarity status is determined to be similar, the non-protected section is output as the fault location, and the distributed power grid is not isolated in parallel. When Df > Df0, the fingerprint similarity state is determined to be dissimilar, the protected section is output as the fault location, and the distributed power grid is isolated in parallel to obtain the isolated distributed power grid.

[0010] Further, in step S7, when establishing islands for the distributed power grid after isolation processing according to the target protection list, non-faulty areas are used as power grid islands. The available power Pt and required power Pm of the island are obtained through the SCADA system. The available power Pt and required power Pm of the island are compared, and the sufficiency of the available power of the island is judged based on the comparison result. Based on the judgment result, the fault self-healing mode is output, wherein: When Pt≥Pm, the available power of the island is deemed sufficient, and the fault self-healing mode output is: zero power outage processing for all power-consuming units in the target protection list; When Pt < Pm, the available power of the island is determined to be insufficient, and the fault self-healing mode output is: orderly load reduction of the power consumption units in the target protection list. When the main power grid resumes power supply, the isolated distributed power grid will be matched and added to the fault self-healing mode.

[0011] Further, in step S8, when executing the fault self-healing mode, the update time interval is acquired, and when performing differential updates on the time synchronization process based on the update time interval, the update time interval Δta is acquired through the STU system log, the update time interval Δta is compared with the preset update time interval Δta0, the status of the update time interval is judged based on the comparison result, and the time synchronization process is differentially updated based on the judgment result, wherein: When Δta≥Δta0, the update time interval is determined to be long, and the difference update is not performed during the time synchronization process. When Δta < Δta0, the update time interval is determined to be short. The time synchronization process is then updated by difference. The preset device time difference Δt0 is updated by difference according to the update coefficient gx. The updated preset device time difference is set to Δt01, and Δt01 = Δt0 × gx. The updated preset device time difference Δt01 is output as the preset device time difference Δt0, and the device time difference Δt is compared with the preset device time difference Δt0 again.

[0012] Further, in step S8, when the protection list update deviation is obtained and the deviation optimization is performed on the difference update process based on the protection list update deviation, the government platform updates the target protection list to obtain the updated target protection list, and sends the updated target protection list to the synchronized STU. The DMS master station sends a version query command to the synchronized STU according to a preset frequency. After receiving the version query command, the synchronized STU returns the current target protection list to the DMS master station. The current target protection list is compared with the updated target protection list to obtain the protection list update deviation. The protection list update deviation includes list consistency and list inconsistency. When the protection list update deviation is that the list is consistent, no deviation optimization is performed on the difference update process; When the protection list update deviation is that the list is inconsistent, the deviation update process is optimized, the updated target protection list is resent to the synchronized STU, and the version query command is resent to the synchronized STU through the DMS master station, and an alarm is activated.

[0013] Furthermore, in step S9, extreme scenarios are acquired, and the fault self-healing mode is updated according to the extreme scenarios. The extreme scenarios include complete communication failure, switch failure, and complete blackout of the main network. When the extreme scenario is a complete communication failure, the fault self-healing mode is updated and the switch to the backup channel is output as the fault self-healing mode. When the extreme scenario is that the switch fails to operate, the fault self-healing mode is updated, and the adjacent STU is used as the fault self-healing mode for output. When the mainnet goes completely black in an extreme scenario, the fault self-healing mode will be updated, and the mobile energy storage vehicle fleet will be activated as the fault self-healing mode for output.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: the method obtains the target protection list in step S1 to prioritize the protection of units with high importance; the method also calibrates the current time of the STU in step S2 to ensure time consistency of all STUs; the method further performs high-frequency acquisition of target high-frequency patrol data in step S3 to improve the acquisition accuracy of the target high-frequency patrol data; the method further performs signal smoothing processing on the target high-frequency patrol data in step S4 to improve the clarity and data quality of the target high-frequency patrol data; the method further generates electrical fault fingerprint feature values ​​in step S5 to reduce the complexity of power grid fault identification; and the method further performs fingerprint similarity analysis in step S6. The method performs calculations to quickly and accurately locate and isolate faults in parallel, adapting to the characteristics of distributed power grids. Step S7 further optimizes the power supply mode for islanded areas based on available power, reducing the impact of grid faults. Step S8 calibrates the time synchronization accuracy in real time when frequently updating the target protection list and monitors the distribution of the updated list to synchronized STUs to avoid inconsistencies and untimely updates, thus minimizing fault losses. Step S9 sets and updates the fault self-healing mode for extreme scenarios to minimize fault losses, reduce power restoration time, and improve the efficiency of distributed power grid fault self-healing. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating the distributed fault self-healing method applicable to active distribution networks in this embodiment. Figure 2 This is a flowchart illustrating the time difference monitoring method in this embodiment. Detailed Implementation

[0016] The present invention will now be described in further detail: This invention provides a distributed fault self-healing method suitable for active distribution networks. The method is applied to the control terminals of distributed distribution networks, such as distribution master station rooms. The method compares the electrical fault fingerprint feature values ​​between system units (STUs) to facilitate timely isolation of the distributed network when a grid fault occurs, and to acquire fault self-healing modes to adapt to the characteristics of distributed networks and reduce fault losses. The specific method is as follows: A distributed fault self-healing method suitable for active distribution networks includes the following steps: Step S1: Obtain the target protection list; Step S2: Configure the STU and monitor the time difference of the STU to obtain the device time difference. Then, perform time synchronization processing on the STU based on the device time difference to obtain the synchronized STU. Step S3: After synchronization, the STU performs high-frequency acquisition of high-frequency patrol data of the target. Step S4: Perform signal smoothing processing on the target high-frequency patrol data to obtain processed high-frequency patrol data; Step S5: Obtain the fingerprint feature value of electrical fault based on the processed high-frequency patrol data; Step S6: Perform fault isolation processing on the distributed power grid based on the electrical fault fingerprint feature value to obtain the distributed power grid after isolation processing; Step S7: Based on the target protection list, islanding is established for the isolated distributed power grid to obtain the fault self-healing mode; Step S8: When executing the fault self-healing mode, the update time interval is obtained, and the time synchronization process is updated with difference based on the update time interval. The protection list update deviation is also obtained, and the difference update process is optimized based on the protection list update deviation. Step S9: Acquire extreme scenarios and update the fault self-healing mode according to the extreme scenarios.

[0017] Specifically, in step S1, when obtaining the target protection list, the target protection list is obtained through the government platform. The target protection list includes the unit name, load location, electricity consumption and allowable power outage time.

[0018] Specifically, the target protection list refers to the priority order of electricity-consuming units that need power protection, as provided by the government platform. The unit name refers to the specific name of the electricity-consuming unit that needs power protection, such as a hospital. The load location refers to the specific protection location within the electricity-consuming unit that needs power protection, such as the ICU ward of a hospital. ICU stands for Intensive Care Unit. The power consumption refers to the total amount of electrical energy required for the load location to maintain normal operation. The allowable power outage time refers to the length of power outage that the load location can accept, such as an allowable power outage time of 0 seconds.

[0019] Specifically, in step S1, a target protection list is formulated to prioritize the protection of units with high importance, thereby adapting to the characteristics of the distributed power grid and reducing the failure losses of the distribution network.

[0020] Specifically, in step S2, when setting the STU and monitoring the time difference of the STU, the time difference of the STU is monitored by the time difference monitoring method to obtain the equipment time difference Δt.

[0021] Specifically, the STU refers to an intelligent terminal device that collects and processes high-frequency patrol data of the target. Its full name is Smart Terminal Unit. This embodiment does not limit the specific setting method of the STU in the distributed power grid. Those skilled in the art can freely choose according to actual needs, such as setting up the STU in the distribution network switch station and distribution room.

[0022] Specifically, in step S2, when performing time synchronization processing on the STU based on the device time difference, the device time difference Δt is compared with a preset device time difference Δt0, where Δt0 = 1 microsecond. The state of the device time difference is determined based on the comparison result, and the STU is then synchronized based on the determination result. When Δt≤Δt0, the time difference of the equipment is considered acceptable, and no time synchronization processing is performed on the STU. When Δt > Δt0, the device time difference is deemed unacceptable. Time synchronization is then performed on the STU. The current time t of the STU is obtained, and the STU update time tp is calculated based on the current time t and the device time difference Δt. The STU update time tp is set to t - Δt. The current time t of the STU is then calibrated to the STU update time tp, resulting in the synchronized STU.

[0023] Specifically, the preset device time difference refers to a preset value for judging the state of the device time difference. This embodiment does not limit the specific value of the preset device time difference Δt0. Those skilled in the art can freely choose according to actual needs. For example, in this embodiment, Δt0 is set to 1 microsecond according to the time synchronization accuracy of the IEEE 1588 standard. The state of the device time difference refers to the acceptability of the device time difference judged based on the device time difference and the preset device time difference. The state of the device time difference includes acceptable and unacceptable. The current STU time refers to the time currently used by the STU when time synchronization is not performed. The derivation process of the formula tp=t-Δt is as follows: the current STU time t has a time delay. The STU update time tp is obtained by subtracting the device time difference Δt from the unified STU time.

[0024] Specifically, in step S2, the state of the device time difference is judged so that when the state of the device time difference is unacceptable, the current time of the STU is calibrated in a timely manner, thereby ensuring the time consistency of all STUs.

[0025] Specifically, in step S3, the target high-frequency patrol data is obtained by the synchronized STU acquiring the target high-frequency patrol data at a rate of 1 million acquisitions per second.

[0026] Specifically, the target high-frequency patrol data includes real-time current and real-time voltage. In this embodiment, the real-time current and real-time voltage are collected by installing current sensors and voltage sensors at the STU setting node. The real-time current refers to the grid current value collected in real time by the current sensor at the STU setting node, and the real-time voltage refers to the grid voltage value collected in real time by the voltage sensor at the STU setting node. The STU setting node refers to the location node in the distributed power grid where the synchronized STU is set.

[0027] Specifically, in step S3, the target high-frequency patrol data is collected at high frequency by the synchronized STU in order to improve the accuracy of the target high-frequency patrol data collection.

[0028] Specifically, in step S4, when performing signal smoothing processing on the target high-frequency patrol data, a window is pre-set for the target high-frequency patrol data to obtain a preset filtering window, and the target high-frequency patrol data is windowed according to the preset filtering window to obtain processed high-frequency patrol data.

[0029] Specifically, the window presetting refers to the process of extracting data from the target high-frequency patrol data according to a preset data length. This embodiment does not limit the preset data length, and those skilled in the art can freely choose it according to actual needs, as long as the preset data length is a power of 2. For example, in this embodiment, the preset data length is set to 256 to ensure the response speed of signal smoothing processing. The windowing process refers to the process of multiplying the target high-frequency patrol data within the preset filtering window with the data at the corresponding position in the Hann window function. The Hann window function refers to the existing standard mathematical tool for windowing the target high-frequency patrol data.

[0030] Specifically, in step S4, signal smoothing processing is performed on the high-frequency patrol data of the target to improve the clarity and data quality of the high-frequency patrol data of the target, thereby improving the accuracy of fault identification in the distributed power grid.

[0031] Specifically, in step S5, when obtaining the electrical fault fingerprint feature value based on the processed high-frequency patrol data, the current mutation value La is calculated based on the real-time current Lb in the processed high-frequency patrol data and the real-time current Lc in the processed high-frequency patrol data at the previous moment. La is set to |Lb-Lc|, and the current mutation value La is obtained. In step S5, the current mutation value La is compared with the preset current mutation value La0. Based on the comparison result, the real-time current mutation situation is judged, and the electrical fault fingerprint feature value is obtained based on the judgment result, wherein: When La≤La0, the sudden change in real-time current is determined to be non-existent, and no electrical fault fingerprint feature value is acquired. When La > La0, the sudden change in real-time current is determined to be a sudden change. The fingerprint feature value of electrical fault is obtained, the initial fault waveform is truncated, and the fingerprint of the initial fault waveform is compressed by wavelet transform energy distribution to obtain the fingerprint feature value of electrical fault.

[0032] Specifically, the real-time current in the high-frequency patrol data after processing at the previous moment refers to the real-time current in the processed high-frequency patrol data obtained after the STU collected the target high-frequency patrol data after the previous synchronization and after signal smoothing processing. The preset current mutation value refers to a preset value for judging the mutation of the real-time current. This embodiment does not limit the specific value of the preset current mutation value La0. Those skilled in the art can freely choose according to actual needs. For example, in this embodiment, based on actual operating experience, La0 is set to 10kV, and for a power distribution line with a rated current of 100A, La0 is set to 300A. The mutation of the real-time current refers to whether there is a mutation in the real-time current, judged by the current mutation value and the preset current mutation value. This includes both the presence and absence of sudden changes. The initial fault waveform refers to the current sudden change waveform in the first 20 milliseconds after synchronization, where the sudden change in real-time current intercepted by the STU is determined to be a sudden change. The wavelet transform energy distribution refers to the existing technology of quantifying the energy distribution disorder of the initial fault waveform to generate a unique and quantifiable fault feature value. The fingerprint compression refers to the process of converting the initial fault waveform into an electrical fault fingerprint feature value through wavelet packet energy entropy. The electrical fault fingerprint feature value refers to the value obtained after fingerprint compression to describe the characteristics of the current sudden change fault. The derivation of the formula La=|Lb-Lc| is as follows: take the absolute value of the difference between the real-time current Lb and the real-time current Lc in the high-frequency patrol data processed at the previous moment to measure the sudden change in current.

[0033] Specifically, in step S5, by judging the sudden changes in real-time current, the sudden changes in current are used as a sign of a fault, and the fingerprint feature value of electrical fault is generated in a timely manner, so as to compress the initial fault waveform into feature values ​​that characterize the nature of the fault, thereby realizing the rapid and accurate identification of the fault type of distributed power grid.

[0034] Specifically, in step S6, when performing fault isolation processing on the distributed power grid based on electrical fault fingerprint feature values, the synchronized STU prioritizes sending electrical fault fingerprint feature values ​​to neighboring STUs via dual-channel transmission. The neighboring STUs acquire the fingerprint similarity Df and compare it with a preset fingerprint similarity Df0, setting Df0=0.35. Based on the comparison result, they determine the fingerprint similarity status and output the fault location based on the determination result, wherein: When Df≤Df0, the fingerprint similarity status is determined to be similar, the non-protected section is output as the fault location, and the distributed power grid is not isolated in parallel. When Df > Df0, the fingerprint similarity state is determined to be dissimilar, the protected section is output as the fault location, and the distributed power grid is isolated in parallel to obtain the isolated distributed power grid.

[0035] Specifically, the dual-channel transmission method refers to setting up two independent communication network ports to transmit electrical fault fingerprint feature values. For example, if one network port is interrupted, the electrical fault fingerprint feature value will be automatically transmitted from the other network port to the neighboring STU within 50 microseconds. The neighboring STU refers to the synchronized STU that is adjacent to the feature generation STU in the electrical connection of the distributed power grid. This embodiment does not limit the specific method of prioritizing the transmission of electrical fault fingerprint feature values ​​to the neighboring STU; those skilled in the art can freely choose according to actual needs, such as setting TSN gating. TSN stands for Time-Sensitive Network. In Networking, the fingerprint similarity refers to a numerical value that measures the similarity between the electrical fault fingerprint feature values ​​in the synchronized STU and the electrical fault fingerprint feature values ​​in the neighboring STU. This embodiment does not limit the specific method of obtaining fingerprint similarity. Those skilled in the art can freely choose according to actual needs, such as calculating the cosine similarity between the electrical fault fingerprint feature values ​​in the synchronized STU and the electrical fault fingerprint feature values ​​in the neighboring STU, and using the cosine similarity as the fingerprint similarity. The preset fingerprint similarity refers to a preset value for judging the fingerprint similarity state. In this embodiment, Df0=0.35 is set based on historical experimental verification. The fingerprint similarity state refers to the similarity between the electrical fault fingerprint feature values ​​in the synchronized STU and the electrical fault fingerprint feature values ​​in the neighboring STU. The fingerprint similarity state includes similarity and dissimilarity. The non-protected section refers to other grid sections in the distributed grid other than the protected section. The protected section refers to the grid section between the synchronized STU and the neighboring STU. The parallel isolation means that when the protected section is output as the fault location, the synchronized STU and the neighboring STU isolate the fault location by simultaneously tripping the circuit breaker on one side of the protected section.

[0036] Specifically, in step S6, by transmitting electrical fault fingerprint feature values ​​to neighboring STUs, and by having adjacent STUs exchange electrical fault fingerprint feature values ​​and calculate fingerprint similarity, the fault location can be quickly and accurately located and isolated in parallel to adapt to the characteristics of distributed power grids, thereby reducing the complexity of power grid fault identification.

[0037] Specifically, in step S7, when establishing islands for the isolated distributed power grid according to the target protection list, non-faulty areas are designated as power grid islands. The available power Pt and required power Pm of the island are acquired through the SCADA system. The available power Pt and required power Pm are compared, and the sufficiency of the available power is determined based on the comparison result. Based on the determination result, the fault self-healing mode is output, wherein: When Pt≥Pm, the available power of the island is deemed sufficient, and the fault self-healing mode output is: zero power outage processing for all power-consuming units in the target protection list; When Pt < Pm, the available power of the island is determined to be insufficient, and the fault self-healing mode output is: orderly load reduction of the power consumption units in the target protection list. When the main power grid resumes power supply, the isolated distributed power grid will be matched and added to the fault self-healing mode.

[0038] Specifically, the available power of the island refers to the power that can be allocated and used in the current moment of the power grid island, as obtained from the SCADA system. The required power of the island refers to the power required to maintain normal power supply in the current moment of the power grid island, as obtained from the SCADA system. SCADA refers to Supervisory Control And Data Acquisition System. The sufficiency of the available power of the island refers to the sufficiency of the available power of the island to maintain normal power supply. The sufficiency of the available power of the island includes sufficiency and insufficiency. The zero-outage processing refers to the mode of continuous power supply to the power users in the target protection list in the power grid island. The orderly load shedding refers to the mode of power outage starting from the power users with lower priority according to the target protection list in the power grid island. The matching power supply refers to the process of unifying the voltage, frequency and phase of the main grid side and the power grid island side through the STU at the tie switch and controlling the circuit breaker to close to restore power supply.

[0039] Specifically, in step S7, by adopting a reasonable power supply mode based on the sufficiency of available power in the island, the impact of grid faults is reduced, and the power demand of the grid island is guaranteed.

[0040] Specifically, in step S8, when executing the fault self-healing mode, the update time interval is obtained, and when performing differential updates on the time synchronization process based on the update time interval, the update time interval Δta is obtained through the STU system log, the update time interval Δta is compared with the preset update time interval Δta0, the status of the update time interval is judged based on the comparison result, and the time synchronization process is differentially updated based on the judgment result, wherein: When Δta≥Δta0, the update time interval is determined to be long, and the difference update is not performed during the time synchronization process. When Δta < Δta0, the update time interval is determined to be short. The time synchronization process is then updated by difference. The preset device time difference Δt0 is updated by difference according to the update coefficient gx. The updated preset device time difference is set to Δt01, and Δt01 = Δt0 × gx. The updated preset device time difference Δt01 is output as the preset device time difference Δt0, and the device time difference Δt is compared with the preset device time difference Δt0 again.

[0041] Specifically, the STU system log refers to a log file that records all critical operations, status changes, internal diagnostics, and communication events within the STU. The update interval refers to the time interval at which the government platform updates the target protection list. The preset update interval refers to a preset value for judging the status of the update interval. This embodiment does not limit the specific value of the preset update interval Δta0; those skilled in the art can freely choose it according to actual needs. For example, in this embodiment, Δta0 is set to 6 months based on the difference update verification. The difference update verification refers to the verification of the status corresponding to the historical update interval. The process of updating the difference to obtain the optimal preset update time interval, wherein the state of the update time interval refers to the length of the update time interval, including long time and short time. This embodiment does not limit the specific value of the update coefficient. Those skilled in the art can freely choose according to actual needs, as long as the requirement of 0 < gx < 1 is met. For example, in this embodiment, gx = 0.66 is set according to historical experience. The derivation process of the formula Δt01 = Δt0 × gx is as follows: multiply the preset device time difference Δt0 by the update coefficient gx, and 0 < gx < 1, so as to reduce the value of the preset device time difference Δt0.

[0042] Specifically, in step S8, by judging the state of the update time interval, when the state of the update time interval is short, the preset device time difference is reduced by the update coefficient, so as to calibrate the accuracy requirements of time synchronization in real time when frequently updating the target protection list, thereby improving the sensitivity of distributed power grid fault self-healing.

[0043] Specifically, in step S8, when the protection list update deviation is obtained and the deviation optimization is performed on the difference update process based on the protection list update deviation, the government platform updates the target protection list to obtain the updated target protection list, and sends the updated target protection list to the synchronized STU. The DMS master station sends a version query command to the synchronized STU according to a preset frequency. After receiving the version query command, the synchronized STU returns the current target protection list to the DMS master station. The current target protection list is compared with the updated target protection list to obtain the protection list update deviation. The protection list update deviation includes list consistency and list inconsistency. When the protection list update deviation is that the list is consistent, no deviation optimization is performed on the difference update process; When the protection list update deviation is that the list is inconsistent, the deviation update process is optimized, the updated target protection list is resent to the synchronized STU, and the version query command is resent to the synchronized STU through the DMS master station, and an alarm is activated.

[0044] Specifically, the preset frequency refers to the pre-set frequency at which the version query command is sent to the synchronized STU. This embodiment does not limit the preset frequency, and those skilled in the art can freely choose it according to actual needs. For example, in this embodiment, the preset frequency is set to 1 time / hour. The version query command refers to a specific communication instruction actively sent by the DMS master station to the synchronized STU to request and verify the protection list update deviation. The DMS refers to the Distribution Management System. The current target protection list refers to the target protection list that the synchronized STU is currently executing. This embodiment does not limit the specific method of comparing the current target protection list with the updated target protection list. Those skilled in the art can freely choose it according to actual needs. For example, the hash values ​​of the updated target protection list and the current target protection list can be calculated and compared separately. This embodiment does not limit the alarm method when the protection list update deviation is that the lists are inconsistent. Those skilled in the art can freely choose it according to actual needs. For example, an alarm can be triggered by flashing an audible and visual alarm.

[0045] Specifically, in step S8, the deviation of the protection list update is checked, and the updated target protection list is promptly sent to the synchronized STU to avoid problems such as inconsistent lists and untimely updates, thereby reducing fault losses and improving the security of the distributed power grid.

[0046] Specifically, in step S9, extreme scenarios are acquired, and the fault self-healing mode is updated according to the extreme scenarios. The extreme scenarios include complete communication failure, switch failure, and complete blackout of the main network. When the extreme scenario is a complete communication failure, the fault self-healing mode is updated and the switch to the backup channel is output as the fault self-healing mode. When the extreme scenario is that the switch fails to operate, the fault self-healing mode is updated, and the adjacent STU is used as the fault self-healing mode for output. When the mainnet goes completely black in an extreme scenario, the fault self-healing mode will be updated, and the mobile energy storage vehicle fleet will be activated as the fault self-healing mode for output.

[0047] Specifically, "complete communication failure" refers to the complete interruption of the high-speed communication network in the distributed power grid; "switching to backup channel" refers to the distributed power grid switching to a backup communication mode of a 230MHz wireless private network; "switch failure" refers to the situation where the circuit breaker cannot execute the trip command for fault isolation due to a fault; "adjacent STU supplementary trip" refers to the method where a neighboring STU detects the circuit breaker failure within 150 microseconds and performs parallel isolation on its behalf; "main grid blackout" refers to the situation where the power grid is completely de-energized and cannot start on its own, requiring external power support; and "activating the mobile energy storage vehicle group" refers to the method of calling the mobile energy storage vehicle group to restore power supply one by one according to the target protection list.

[0048] Specifically, in step S9, by setting extreme scenarios and updating the fault self-healing mode accordingly, the fault loss can be minimized in extreme scenarios, the power restoration time can be reduced, and the efficiency of distributed power grid fault self-healing can be improved.

[0049] Please see Figure 2 As shown, this is a flowchart illustrating the time difference monitoring method of this embodiment. The time difference monitoring method includes: Step A01: Set the master clock and slave clock; Step A02: Send a Sync message to the slave clock via the master clock and record the time point t1 at which the Sync message is sent; Step A03: Record the time point t2 when the Sync message is received; Step A04: Send a Delay message from the clock to the master clock and record the time point t3 when the Delay message is sent; Step A05: Record the time point t4 when the Delay message is received; Step A06: Calculate the device time difference Δt based on the time point t1 when the Sync message is sent, the time point t2 when the Sync message is received, the time point t3 when the Delay message is sent, and the time point t4 when the Delay message is received. Set Δt=[(t2-t1)-(t4-t3)] / 2 to obtain the device time difference Δt.

[0050] Specifically, in this embodiment, the Grandmaster Clock is used as the master clock. The Grandmaster Clock refers to a high-precision clock server deployed at the power distribution station, primarily based on a satellite timing system. The slave clock refers to the clock system of the synchronized STU. The Sync message is information sent from the master clock to the slave clock, carrying the timestamp of the Sync message transmission. The Delay message is information sent from the slave clock to the master clock, carrying the timestamp of the Delay message transmission. The derivation of the formula Δt=[(t2-t1)-(t4-t3)] / 2 is as follows: The transmission time of the Sync message is t2-t1, which includes the actual network delay D and the device time difference Δt. The transmission time of the Delay message is t4-t3, which is equal to the actual network delay D minus the device time difference Δt. t2-t1=D+Δt, t4-t3=D-Δt, and based on t2-t1=D+Δt and t4-t3=D-Δt, we can derive (t2) t1) (t4) t3) = (D + Δt) (D) Δt), which simplifies to Δt=[(t2-t1)-(t4-t3)] / 2.

[0051] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A distributed fault self-healing method suitable for active power distribution networks, characterized by, The method comprises the following steps: Step S1, obtaining a target protection list; Step S2, setting an STU, monitoring a time difference of the STU, obtaining a device time difference, and performing time synchronization processing on the STU according to the device time difference to obtain a synchronized STU; Step S3, collecting high-frequency patrol data of a target through the synchronized STU; Step S4, performing signal smoothing processing on the target high-frequency patrol data to obtain processed high-frequency patrol data; Step S5, obtaining an electrical fault fingerprint characteristic value according to the processed high-frequency patrol data; Step S6, performing fault isolation processing on a distributed power grid according to the electrical fault fingerprint characteristic value to obtain an isolated distributed power grid; Step S7, establishing an island according to the target protection list to obtain a fault self-healing mode; Step S8, obtaining an update time interval when the fault self-healing mode is executed, performing difference updating on the time synchronization processing according to the update time interval, obtaining a protection list update deviation, and performing deviation optimization on the difference updating according to the protection list update deviation; Step S9, obtaining an extreme scenario and performing mode updating on the fault self-healing mode according to the extreme scenario; When the extreme scenario is a complete communication interruption, a standby channel is switched as the fault self-healing mode for output; When the extreme scenario is a switch refusal, an adjacent STU is supplemented as the fault self-healing mode for output; When the extreme scenario is a main network complete black, a mobile energy storage vehicle group is started as the fault self-healing mode for output.

2. The method for distributed fault self-healing suitable for active distribution network as claimed in claim 1 wherein, In the step S1, the target protection list is obtained through a government platform, and the target protection list comprises a unit name, a load position, a power consumption, and an allowed power outage time; In the step S2, the time difference of the STU is monitored through a time difference monitoring method; The time difference monitoring method comprises: Step A01, setting a master clock and a slave clock; Step A02, sending a Sync message from the master clock to the slave clock, and recording a time point t1 of sending the Sync message; Step A03, recording a time point t2 of receiving the Sync message; Step A04, sending a Delay message from the slave clock to the master clock, and recording a time point t3 of sending the Delay message; Step A05, recording a time point t4 of receiving the Delay message; Step A06, calculating a device time difference Δt according to the time point t1 of sending the Sync message, the time point t2 of receiving the Sync message, the time point t3 of sending the Delay message, and the time point t4 of receiving the Delay message, setting Δt = [(t2-t1)-(t4-t3)] / 2, and obtaining the device time difference Δt.

3. The method for distributed fault self-healing suitable for active distribution network according to claim 2, characterized in that, In the step S2, when the device time difference is synchronized with the STU, the device time difference Δt is compared with a preset device time difference Δt0, and Δt0=1 microsecond. The state of the device time difference is judged according to the comparison result, and the STU is time-synchronized according to the judgment result, wherein: When Δt≤Δt0, it is determined that the state of the device time difference is acceptable, and the STU is not time-synchronized. When Δt>Δt0, it is determined that the state of the device time difference is unacceptable, and the STU is time-synchronized. The current time t of the STU is obtained, and the updated time tp of the STU is calculated according to the current time t of the STU and the device time difference Δt, which is set as tp=t-Δt. The updated time tp of the STU is obtained, and the current time t of the STU is calibrated as the updated time tp of the STU to obtain the synchronized STU.

4. The method for distributed fault self-healing suitable for active distribution network according to claim 3, characterized in that, In the step S3, the target high-frequency patrol data is collected at a speed of 100 million times per second by the synchronized STU to obtain the target high-frequency patrol data. In the step S4, when the target high-frequency patrol data is signal-smoothed, the target high-frequency patrol data is window-pre-set to obtain a preset filter window, and the target high-frequency patrol data is windowed according to the preset filter window to obtain processed high-frequency patrol data.

5. The method for distributed fault self-healing suitable for active distribution network as claimed in claim 4 wherein, In the step S5, when the electrical fault fingerprint characteristic value is obtained according to the processed high-frequency patrol data, the current mutation value La is calculated according to the real-time current Lb in the processed high-frequency patrol data and the real-time current Lc in the previous time processed high-frequency patrol data, which is set as La=|Lb-Lc| to obtain the current mutation value La. In the step S5, the current mutation value La is compared with a preset current mutation value La0. The mutation of the real-time current is judged according to the comparison result, and the electrical fault fingerprint characteristic value is obtained according to the judgment result, wherein: When La≤La0, it is determined that the mutation of the real-time current does not exist, and the electrical fault fingerprint characteristic value is not obtained. When La>La0, it is determined that the mutation of the real-time current exists, and the electrical fault fingerprint characteristic value is obtained. The initial segment fault waveform is intercepted, and the initial segment fault waveform is fingerprint-compressed by wavelet transform energy distribution to obtain the electrical fault fingerprint characteristic value.

6. The method for distributed fault self-healing suitable for active distribution network according to claim 5, characterized in that, In the step S6, when the distributed power grid is fault-isolated according to the electrical fault fingerprint characteristic value, the synchronized STU preferentially sends the electrical fault fingerprint characteristic value to the neighbor STU through a double-channel transmission mode. The neighbor STU obtains the fingerprint similarity Df, and compares the fingerprint similarity Df with a preset fingerprint similarity Df0, which is set as Df0=0.

35. The fingerprint similarity state is judged according to the comparison result, and the fault position is output according to the judgment result, wherein: When Df≤Df0, it is determined that the fingerprint similarity state is similar, and the non-protected segment is output as the fault position, and the distributed power grid is not parallel-isolated. When Df>Df0, it is determined that the fingerprint similarity state is dissimilar, the protection section is output as a fault position, and the distributed power grid is parallelly isolated to obtain an isolated distributed power grid.

7. The method for distributed fault self-healing suitable for active distribution network according to claim 6, characterized in that, In the step S7, when the isolated distributed power grid is established according to the target protection list, a non-fault area is taken as a power grid island, the island available power Pt and the island demand power Pm are obtained through the SCADA system, the island available power Pt and the island demand power Pm are compared, the sufficiency of the island available power is judged according to the comparison result, and the fault self-healing mode is output according to the judgment result, wherein: When Pt≥Pm, it is determined that the sufficiency of the island available power is sufficient, and the fault self-healing mode is output as: zero power-off processing is performed on all power consumption units in the target protection list; When Pt<Pm, it is determined that the sufficiency of the island available power is insufficient, and the fault self-healing mode is output as: ordered load shedding is performed on the power consumption units in the target protection list; When the main power grid restores power supply, the matched power supply of the isolated distributed power grid is added to the fault self-healing mode.

8. The method for distributed fault self-healing suitable for active distribution network as claimed in claim 3 wherein, In the step S8, when the fault self-healing mode is executed, the update time interval is obtained, and when the time synchronization process is differentially updated according to the update time interval, the update time interval Δta is obtained through the STU system log, the update time interval Δta and the preset update time interval Δta0 are compared, the state of the update time interval is judged according to the comparison result, and the time synchronization process is differentially updated according to the judgment result, wherein: When Δta≥Δta0, it is determined that the state of the update time interval is long, and the time synchronization process is not differentially updated; When Δta<Δta0, it is determined that the state of the update time interval is short, the time synchronization process is differentially updated, the preset device time difference Δt0 is differentially updated according to the update coefficient gx, the updated preset device time difference is set as Δt01, and Δt01=Δt0×gx, the updated preset device time difference Δt01 is output as the preset device time difference Δt0, and the device time difference Δt and the preset device time difference Δt0 are compared again.

9. The method for distributed fault self-healing suitable for active distribution network according to claim 8, characterized in that, In the step S8, when the protection list update deviation is obtained and the deviation optimization process of differential update is performed according to the protection list update deviation, the target protection list is updated by the government platform to obtain an updated target protection list, and the updated target protection list is sent to the synchronized STU. The DMS master station sends a version query command to the synchronized STU according to a preset frequency, the synchronized STU returns the current target protection list to the DMS master station after receiving the version query command, the current target protection list and the updated target protection list are compared to obtain the protection list update deviation, and the protection list update deviation includes list consistency and list inconsistency. When the protection list update deviation condition is list consistent, the difference value update process is not deviatedly optimized; When the protection list update deviation condition is list inconsistent, the difference value update process is deviatedly optimized, the updated target protection list is re-sent to the synchronized STU, a version query command is re-sent to the synchronized STU by the DMS master station, and an alarm is started.

10. The method for distributed fault self-healing suitable for active distribution network as claimed in claim 7 wherein, In the step S9, an extreme scenario is acquired, and a fault self-healing mode is updated according to the extreme scenario, the extreme scenario including communication full interruption, switch refusal and main network full black; When the extreme scenario is communication full interruption, the fault self-healing mode is updated, and a switching standby channel is output as the fault self-healing mode; When the extreme scenario is switch refusal, the fault self-healing mode is updated, and an adjacent STU jump is output as the fault self-healing mode; When the extreme scenario is main network full black, the fault self-healing mode is updated, and a mobile energy storage vehicle group is started as the fault self-healing mode.

Images