A power distribution network self-healing control system and method based on a smart grid

Abstract

This invention relates to the field of power distribution control technology, and more particularly to a self-healing control system and method for power distribution networks based on smart grids. Addressing the problems of slow fault detection, inaccurate isolation, and low power restoration efficiency in traditional power distribution networks, this invention uses a data sensing module to monitor grid over-limit conditions in real time, a data processing module to perform time alignment and quality optimization of multi-source data, a fault location module to accurately locate fault areas, a fault isolation module to quickly isolate fault areas, and finally a power restoration module to construct the grid topology and achieve tiered power restoration. The system employs a load grading correction and tripping verification mechanism, forming a complete "sensing-location-isolation-restoration" closed-loop control system. This achieves a leap from hours to seconds in fault location time, improves isolation operation efficiency by 80%, reduces power restoration time by 70%, and significantly improves the power supply reliability and self-healing capability of the power distribution network.

Description

Technical Field

[0001] This invention relates to the field of power distribution control technology, and in particular to a power distribution self-healing control system and method based on smart grids. Background Technology

[0002] In the current operation of power distribution networks, traditional fault handling methods face significant technical bottlenecks. Existing systems primarily rely on manual inspections and reactive repairs, resulting in long fault detection response times, low location accuracy, expanded outage areas, and low restoration efficiency. With the increasing penetration of distributed power sources and the growing complexity of power distribution network structures, traditional overcurrent protection coordination methods are ill-suited to multi-source power supply scenarios, easily leading to maloperation or failure to operate. Existing self-healing control systems generally suffer from incomplete data acquisition, slow fault isolation speed, and simplistic power restoration strategies, lacking a complete "sensing-decision-execution" closed-loop control capability. At the data level, the lack of effective fusion of multi-source heterogeneous data affects the accuracy of fault diagnosis; at the control level, the disconnect between fault isolation and power restoration hinders rapid power restoration.

[0003] For example, Chinese patent CN104701831A discloses a self-healing control method for power distribution networks, which mainly includes fault location, fault recovery, and self-healing evaluation. It utilizes advanced monitoring methods to continuously diagnose and evaluate the operating status of the power grid online, promptly identifying and eliminating potential faults. When a fault occurs, it can quickly isolate the fault and self-recover, minimizing the impact of the fault and ensuring the safe and stable operation of the power grid and high-quality power supply. However, this patent still suffers from problems such as slow fault detection speed, inaccurate isolation, reliance on manual decision-making for power restoration, resulting in large power outage areas, long recovery times, and low power supply reliability. Summary of the Invention

[0004] To address this, the present invention provides a distribution network self-healing control system and method based on smart grids, which overcomes the problems of large power outage range, long recovery time, and low power supply reliability caused by slow fault detection speed, inaccurate isolation, and reliance on manual decision-making for power restoration in the prior art.

[0005] To achieve the above objectives, in one aspect, the present invention provides a distribution network self-healing control system based on a smart grid, comprising:

[0006] The data sensing module is used to collect power grid data based on a preset sensor array, and to judge the local over-limit situation based on the power grid data to obtain the local over-limit situation.

[0007] The data processing module is used to perform time alignment processing on the power grid data to obtain the processed power grid data. It is also used to acquire the processing status of the time alignment processing and push the perception optimization scheme to the management terminal based on the processing status. It is also used to supplement the push process of the perception optimization scheme based on the local over-limit situation.

[0008] The fault location module is used to locate the fault area based on the processed power grid data, and obtain the fault area and its coordinates.

[0009] The fault isolation module is used to isolate the fault area based on the coordinates of the fault area.

[0010] The power restoration module is used to acquire equipment status information, construct a power grid topology map based on the equipment status information, segment the power grid topology map according to the coordinates of the fault area to obtain non-fault areas, classify the load of the non-fault areas, generate power supply commands based on the load classification, restore power supply to the non-fault areas according to the power supply commands, perform graded correction on the generation process of the power supply commands, perform trip verification on the graded correction process, and perform trip adjustment on the over-limit compensation process.

[0011] Furthermore, the data sensing module determines the local limit violation situation based on power grid data, specifically as follows:

[0012] The phase current I in the power grid data is compared with the preset phase current I0, and the local over-limit situation is output based on the comparison result, wherein:

[0013] When I≤I0, the data sensing module outputs the non-local limit violation as a local limit violation case;

[0014] When I > I0, the data sensing module will output the local limit violation as a local limit violation situation;

[0015] The zero-sequence voltage U in the power grid data is compared with the preset zero-sequence voltage U0, and the local over-limit situation is output according to the comparison result, wherein:

[0016] When U≤U0, the data sensing module outputs the non-local limit violation as a local limit violation case;

[0017] When U > U0, the data sensing module will output the local limit violation as a local limit violation situation;

[0018] The frequency change rate Z in the power grid data is compared with the preset frequency change rate Z0, and the local limit violation situation is output according to the comparison result, wherein:

[0019] When Z≤Z0, the data sensing module outputs the non-local limit violation as a local limit violation case;

[0020] When Z > Z0, the data sensing module outputs the local limit violation as a local limit violation situation.

[0021] Furthermore, the data processing module performs time alignment processing on the power grid data to obtain processed power grid data, specifically as follows:

[0022] Step A1: Perform timestamp parsing and standardization on the power grid data to obtain timestamp data in a unified format;

[0023] Step A2: Based on the unified format timestamp data obtained in step S1, construct a reference timeline to obtain the baseline time series;

[0024] Step A3: Resample the power grid data using a linear interpolation algorithm and the reference time series to obtain preliminary aligned data;

[0025] Step A4: Perform anomaly detection and consistency verification on the preliminary aligned data to obtain the processed power grid data.

[0026] Furthermore, the data processing module acquires the processing status of the time alignment process and pushes a perception optimization scheme to the management terminal based on the processing status, specifically as follows:

[0027] The missing percentage C of missing points filled by the linear interpolation algorithm in the time alignment process is obtained. This missing percentage C is compared with a preset missing percentage C0. Based on the comparison result, the data missing situation is judged, and a perception optimization plan is pushed to the management terminal according to the judgment result. Specifically:

[0028] When C≤C0, the data processing module determines that the data loss is not serious and does not push the perception optimization scheme to the management terminal.

[0029] When C > C0, the data processing module determines that the data loss is severe and pushes a perception optimization plan to the management terminal. The perception optimization plan includes:

[0030] The sensors in the preset sensor array are inspected and the number of sensors in the preset sensor array is increased.

[0031] Furthermore, the data processing module supplements the push process of the perception optimization scheme based on the local limit violation situation, specifically as follows:

[0032] Obtain the number of local violations R in the local violation cases, compare the number of local violations R with the preset number of local violations R0, judge the violation situation in the area based on the comparison result, and supplement the violation process of the perception optimization plan according to the judgment result. Among them:

[0033] When R < R0, the data processing module determines that the area exceeding the limit is a non-frequent area exceeding the limit and does not supplement the limit exceeding process in the process of pushing the perception optimization scheme.

[0034] When R≥R0, the data processing module determines the region exceeding the limit as a frequently exceeding the limit region, and performs limit compensation during the push process of the perception optimization scheme. The limit compensation includes:

[0035] If the data loss is not severe, a perception optimization plan will also be pushed to the management terminal.

[0036] Furthermore, the fault location module locates the fault area based on the processed power grid data, obtaining the fault area and its coordinates, specifically as follows:

[0037] The fault judgment data in the power grid data is input into the preset fault judgment model to obtain the regional fault status output by the fault judgment model. The fault judgment data includes the three phase voltages, three phase currents and zero-sequence voltage of each region. The fault status includes whether the region has a fault or not. Based on the regional fault status, the region is divided into fault regions and the fault region coordinates are obtained.

[0038] Furthermore, the fault isolation module isolates the fault area based on the coordinates of the fault area, specifically as follows:

[0039] Input the coordinates of the fault area into the preset isolation scheme generation model, obtain the isolation scheme output by the isolation scheme generation model, and send the isolation scheme to the substation automation system. Control the circuit breaker and disconnector to complete the tripping operation according to the isolation scheme.

[0040] Furthermore, the power restoration module acquires equipment status information and constructs a power grid topology map based on the equipment status information. It then segments the power grid topology map according to the coordinates of the fault area to obtain non-fault areas. Specifically:

[0041] Equipment status information is acquired, and based on the equipment ledger data in the equipment status information, combined with the power grid wiring diagram and equipment connection relationships in the equipment status information, an initial power grid topology diagram with equipment as nodes and electrical connections as edges is constructed using graph theory methods. The connection relationships in the topology diagram are dynamically adjusted according to the real-time status information of switching equipment and circuit breakers in the equipment status information to ensure that the topology diagram is consistent with the actual operation mode, and a power grid topology diagram reflecting the current power grid structure is generated. All faulty nodes and faulty equipment located in the faulty area in the power grid topology diagram are identified according to the coordinates of the faulty area, and all faulty nodes and faulty equipment are removed from the power grid topology diagram to obtain the non-faulty area.

[0042] Furthermore, the power restoration module classifies the load of non-faulty areas, generates power supply commands based on the load classification, and restores power to the non-faulty areas based on the power supply commands, specifically as follows:

[0043] The region attributes of non-faulty areas are obtained, and the load levels of non-faulty areas are classified according to the region attributes to obtain the load levels, where:

[0044] When the area attribute is important, the first-level load is used as the load level of the non-faulty area;

[0045] When the area attribute is a secondary area, the secondary load is treated as the load level of the non-faulty area;

[0046] When the area attribute is a general area, the level 3 load is used as the load level of the non-faulty area.

[0047] Input the load level and coordinates of non-faulty areas into the preset power supply command generation model, obtain the power supply command output by the preset power supply command generation model, and restore power supply to non-faulty areas according to the power supply command.

[0048] The power restoration module performs hierarchical correction on the power supply command generation process, specifically as follows:

[0049] When the load level is level three, the power supply command is subject to graded correction, which includes:

[0050] Add three delayed batch power supply operations as supplementary instructions to the power supply instructions;

[0051] The power restoration module performs trip verification during the graded correction process and trip adjustment during the over-limit replenishment process, specifically as follows:

[0052] The secondary trip rate P of the non-faulty area is acquired and compared with the preset secondary trip rate P0. Based on the comparison result, the secondary tripping situation is judged, and the tripping calibration process for the graded correction is performed based on the judgment result. Furthermore, the tripping adjustment process for over-limit compensation is also performed based on the judgment result. Specifically:

[0053] When P≤P0, the power supply restoration module determines that the secondary tripping situation is a low probability tripping, and does not perform tripping correction during the graded correction process, nor does it perform tripping adjustment during the over-limit supplementation process;

[0054] When P > P0, the power restoration module determines that the secondary tripping situation is a high-probability adjustment and performs tripping verification during the graded correction process. The tripping verification includes:

[0055] Replace the content of the graded correction with "Add the five delayed batch power supply as a supplementary instruction to the power supply instruction";

[0056] The process of replenishing beyond the limit also involves tripping adjustments, which include:

[0057] If the area exceeding the limit is a non-frequent area, the process of pushing the perception optimization plan will also be supplemented with limit exceeding information.

[0058] On the other hand, the present invention also provides a method for a distribution network self-healing control system based on a smart grid, comprising:

[0059] Step S1: Collect power grid data according to the preset sensor array, and judge the local over-limit situation according to the power grid data to obtain the local over-limit situation;

[0060] Step S2: Time alignment processing is performed on the power grid data to obtain the processed power grid data. The processing status of the time alignment processing is also acquired, and a perception optimization scheme is pushed to the management terminal based on the processing status. Furthermore, the process of pushing the perception optimization scheme is supplemented based on the local over-limit situation.

[0061] Step S3: Locate the fault area based on the processed power grid data to obtain the fault area and its coordinates;

[0062] Step S4: Isolate the fault area based on the coordinates of the fault area;

[0063] Step S5: Acquire equipment status information, construct a power grid topology map based on the equipment status information, segment the power grid topology map according to the coordinates of the fault area to obtain non-fault areas, classify the load of the non-fault areas, generate power supply commands according to the load classification, restore power supply to the non-fault areas according to the power supply commands, perform graded correction on the generation process of the power supply commands, perform trip verification on the graded correction process, and perform trip adjustment on the over-limit compensation process.

[0064] Compared with existing technologies, the beneficial effects of this invention are as follows: the system establishes a real-time over-limit monitoring system through a data perception module, providing accurate data support for fault early warning; the system also achieves multi-source data time alignment and intelligent optimization through a data processing module, significantly improving data quality and reliability; the system also quickly and accurately identifies the coordinates of the fault area through a fault location module, transforming traditional manual investigation into second-level automatic location; the system also achieves rapid and accurate isolation of the fault area through a fault isolation module, effectively preventing the fault range from expanding and thus reducing the power outage area caused by the fault; the system also constructs an intelligent power supply restoration system through a power supply restoration module, realizing power supply restoration according to the corresponding load level, and combining trip verification to ensure the safety and reliability of the restoration process. This system upgrades the traditional distribution network's "passive repair" mode to an "active self-healing" mode, shortening the fault location time from hours to seconds, improving isolation operation efficiency by 80%, reducing power supply restoration time by 70%, and significantly improving power supply reliability. Attached Figure Description

[0065] Figure 1 This is a schematic diagram of the structure of the distribution network self-healing control system based on the smart grid in this embodiment;

[0066] Figure 2 This is a flowchart illustrating the method of the distribution network self-healing control system based on the smart grid in this embodiment. Detailed Implementation

[0067] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0068] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0069] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0070] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0071] Please see Figure 1 As shown, this is a schematic diagram of the structure of the distribution network self-healing control system based on the smart grid in this embodiment. The system includes:

[0072] The data sensing module is used to collect power grid data based on a preset sensor array, and to judge the local over-limit situation based on the power grid data to obtain the local over-limit situation.

[0073] The data processing module is used to perform time alignment processing on the power grid data to obtain the processed power grid data. It is also used to acquire the processing status of the time alignment processing and push the perception optimization scheme to the management terminal based on the processing status. It is also used to supplement the push process of the perception optimization scheme based on the local over-limit situation.

[0074] The fault location module is used to locate the fault area based on the processed power grid data, and obtain the fault area and its coordinates.

[0075] The fault isolation module is used to isolate the fault area based on the coordinates of the fault area.

[0076] The power restoration module is used to acquire equipment status information, construct a power grid topology map based on the equipment status information, segment the power grid topology map according to the coordinates of the fault area to obtain non-fault areas, classify the load of the non-fault areas, generate power supply commands based on the load classification, restore power supply to the non-fault areas according to the power supply commands, perform graded correction on the generation process of the power supply commands, perform trip verification on the graded correction process, and perform trip adjustment on the over-limit compensation process.

[0077] Specifically, the smart grid-based distribution network self-healing control system is applied to active distribution network terminals. The system establishes a real-time over-limit monitoring system through a data sensing module, providing accurate data support for fault early warning. The system also achieves multi-source data time alignment and intelligent optimization through a data processing module, significantly improving data quality and reliability. Furthermore, the system quickly and accurately identifies fault area coordinates through a fault location module, transforming traditional manual troubleshooting into second-level automatic location. The system also achieves rapid and accurate isolation of fault areas through a fault isolation module, effectively preventing the fault range from expanding and thus reducing the power outage area caused by the fault. Finally, the system constructs an intelligent power restoration system through a power restoration module, restoring power to the corresponding load level based on the load level, and combining trip verification to ensure a safe and reliable restoration process. This system upgrades the traditional distribution network's "passive repair" mode to an "active self-healing" mode, reducing fault location time from hours to seconds, improving isolation operation efficiency by 80%, and reducing power restoration time by 70%, significantly improving power supply reliability.

[0078] Specifically, the preset sensor array refers to a pre-set sensor array for collecting power grid data, including current transformers and voltage transformers. The power grid data includes phase current, zero-sequence voltage, frequency change rate, and fault judgment data. The phase current refers to the current value flowing through the three-phase conductors (phase A, phase B, and phase C) in the power system. The frequency change rate refers to the rate of change of the power grid frequency over time. The data sensing module collects the three-phase voltage signals of the power grid through the voltage transformer and uses frequency measurement technology based on phase-locked loops to obtain the frequency change rate. The zero-sequence voltage refers to a characteristic quantity reflecting the degree of three-phase voltage imbalance in the power system. The fault judgment data refers to data used for fault judgment, including the three-phase voltage, three-phase current, and zero-sequence voltage of each region. The three-phase voltage of each region refers to the three-phase voltage measured in each monitoring area of ​​the distribution network. The three-phase current of each region refers to the three-phase current measured in each monitoring area of ​​the distribution network. The zero-sequence voltage of each region refers to the zero-sequence voltage measured in each monitoring area of ​​the distribution network.

[0079] Specifically, the data sensing module determines the local limit violation situation based on power grid data, and obtains the local limit violation situation as follows:

[0080] The phase current I in the power grid data is compared with the preset phase current I0, and the local over-limit situation is output based on the comparison result, wherein:

[0081] When I≤I0, the data sensing module outputs the non-local limit violation as a local limit violation case;

[0082] When I > I0, the data sensing module will output the local limit violation as a local limit violation situation;

[0083] The zero-sequence voltage U in the power grid data is compared with the preset zero-sequence voltage U0, and the local over-limit situation is output according to the comparison result, wherein:

[0084] When U≤U0, the data sensing module outputs the non-local limit violation as a local limit violation case;

[0085] When U > U0, the data sensing module will output the local limit violation as a local limit violation situation;

[0086] The frequency change rate Z in the power grid data is compared with the preset frequency change rate Z0, and the local limit violation situation is output according to the comparison result, wherein:

[0087] When Z≤Z0, the data sensing module outputs the non-local limit violation as a local limit violation case;

[0088] When Z > Z0, the data sensing module outputs the local limit violation as a local limit violation situation.

[0089] Specifically, the preset phase current refers to the preset value for outputting in response to local over-limit situations. In this embodiment, I0 is set to 600A. The preset zero-sequence voltage refers to the preset value for outputting in response to local over-limit situations. In this embodiment, U0 is set to 690V. The preset frequency change rate refers to the preset value for outputting in response to local over-limit situations. In this embodiment, Z0 is set to 0.5Hz / s.

[0090] Specifically, the data processing module performs time alignment processing on the power grid data to obtain processed power grid data, as follows:

[0091] Step A1: Perform timestamp parsing and standardization on the power grid data to obtain timestamp data in a unified format;

[0092] Step A2: Based on the unified format timestamp data obtained in step S1, construct a reference timeline to obtain the baseline time series;

[0093] Step A3: Resample the power grid data using a linear interpolation algorithm and the reference time series to obtain preliminary aligned data;

[0094] Step A4: Perform anomaly detection and consistency verification on the preliminary aligned data to obtain the processed power grid data.

[0095] Specifically, this embodiment extracts the timestamp information of each data point in the power grid data and converts it into a unified ISO 8601 time format. Simultaneously, it corrects for time zone and daylight saving time deviations to generate standardized timestamp data with a consistent time reference. This results in unified format timestamp data, thus achieving the step of parsing and standardizing the power grid data to obtain unified format timestamp data. In this embodiment, a continuous, equally spaced time point sequence (e.g., 100 points per second) is generated with millisecond-level precision using the BeiDou synchronous clock signal, serving as the reference time axis for data alignment. This results in a reference time sequence, which is then used to construct a reference time axis based on the unified format timestamp data obtained in step S1, thus obtaining the reference time sequence. This embodiment further achieves this step by analyzing the power grid data at the reference time points. Linear interpolation is performed to fill missing data points, and data with different sampling rates are unified onto a reference time axis to form a preliminary time-synchronized aligned dataset. This allows for the resampling of power grid data using a linear interpolation algorithm and the reference time series to obtain preliminary aligned data. In this embodiment, a sliding window algorithm is applied to detect jumps and outliers in the preliminary aligned data. Combined with physical constraints, such as power balance, the rationality of the data is verified, inconsistent data points are removed, and complete and reliable time-aligned power grid data is output, resulting in processed power grid data. This process achieves anomaly detection and consistency verification of the preliminary aligned data to obtain processed power grid data.

[0096] Specifically, the data processing module acquires the processing status of the time alignment process and pushes a perception optimization scheme to the management terminal based on the processing status, specifically:

[0097] The missing percentage C of missing points filled by the linear interpolation algorithm in the time alignment process is obtained. This missing percentage C is compared with a preset missing percentage C0. Based on the comparison result, the data missing situation is judged, and a perception optimization plan is pushed to the management terminal according to the judgment result. Specifically:

[0098] When C≤C0, the data processing module determines that the data loss is not serious and does not push the perception optimization scheme to the management terminal.

[0099] When C > C0, the data processing module determines that the data loss is severe and pushes a perception optimization plan to the management terminal. The perception optimization plan includes:

[0100] The sensors in the preset sensor array are inspected and the number of sensors in the preset sensor array is increased.

[0101] Specifically, this embodiment counts the number of missing points filled by the linear interpolation algorithm and divides the number of missing points by the total number of data points to obtain the missing percentage. The preset missing percentage refers to a preset value for judging the data missing situation. In this embodiment, C0 is set to 10%. The data missing situation refers to whether the data missing is severe, as judged by the missing percentage and the preset missing percentage. The data missing situation includes mild missing and severe missing. This embodiment does not limit the specific implementation method of pushing the perception optimization scheme to the management terminal. Those skilled in the art can limit it according to the actual situation, such as pushing the perception optimization scheme to the management terminal through wireless signal transmission. The management terminal refers to the terminal that manages the power grid system.

[0102] Specifically, the data processing module supplements the push process of the perception optimization scheme based on the local limit-crossing situation, specifically as follows:

[0103] Obtain the number of local violations R in the local violation cases, compare the number of local violations R with the preset number of local violations R0, judge the violation situation in the area based on the comparison result, and supplement the violation process of the perception optimization plan according to the judgment result. Among them:

[0104] When R < R0, the data processing module determines that the area exceeding the limit is a non-frequent area exceeding the limit and does not supplement the limit exceeding process in the process of pushing the perception optimization scheme.

[0105] When R≥R0, the data processing module determines the region exceeding the limit as a frequently exceeding the limit region, and performs limit compensation during the push process of the perception optimization scheme. The limit compensation includes:

[0106] If the data loss is not severe, a perception optimization plan will also be pushed to the management terminal.

[0107] Specifically, this embodiment obtains the number of local violations by statistically analyzing the local violation cases. The preset number of local violations refers to a preset value for judging the regional violation cases. In this embodiment, R0 is set to 3 times. The regional violation cases refer to whether the regional violation is frequent, as judged by the number of local violations and the preset number of local violations. The regional violation cases include infrequent violations and frequent violations.

[0108] Specifically, the fault location module locates the fault area based on the processed power grid data, obtaining the fault area and its coordinates, as follows:

[0109] The fault judgment data in the power grid data is input into the preset fault judgment model to obtain the regional fault status output by the fault judgment model. The fault judgment data includes the three phase voltages, three phase currents and zero-sequence voltage of each region. The fault status includes whether the region has a fault or not. Based on the regional fault status, the region is divided into fault regions and the fault region coordinates are obtained.

[0110] Specifically, the pre-set fault judgment model refers to a pre-set fault judgment model, which is a recurrent neural network model that takes fault judgment data as input and outputs regional fault conditions. This embodiment does not limit the setting method of the fault judgment model. Those skilled in the art can limit it according to the actual situation, such as training the recurrent neural network model with a fault judgment training dataset to obtain the fault judgment model. The fault judgment training dataset refers to the training dataset obtained by training the recurrent neural network model and obtaining the fault judgment model. The fault judgment training dataset includes historically acquired fault judgment data and the regional fault conditions corresponding to the historically acquired fault judgment data. This embodiment does not limit the method of obtaining the fault area coordinates of the fault area. Those skilled in the art can set it according to the actual situation, such as obtaining the fault area coordinates of the fault area according to the power grid topology map.

[0111] Specifically, the fault isolation module isolates the fault area based on the coordinates of the fault area, as follows:

[0112] Input the coordinates of the fault area into the preset isolation scheme generation model, obtain the isolation scheme output by the isolation scheme generation model, and send the isolation scheme to the substation automation system. Control the circuit breaker and disconnector to complete the tripping operation according to the isolation scheme.

[0113] Specifically, the pre-set isolation scheme generation model refers to a pre-set isolation scheme generation model, which is a recurrent neural network model that takes the fault area coordinates as input and the isolation scheme as output. This embodiment does not limit the setting method of the isolation scheme generation model. Those skilled in the art can set it according to the actual situation, such as training the recurrent neural network model with the isolation scheme training dataset to obtain the isolation scheme generation model. The isolation scheme training dataset refers to the training dataset for training the recurrent neural network model to obtain the isolation scheme production model. The isolation scheme training dataset includes historically acquired fault area coordinates and the isolation schemes corresponding to the historically acquired fault area coordinates. This embodiment does not limit the specific implementation method of sending the isolation scheme to the substation automation system. Those skilled in the art can set it according to the actual situation, such as sending the isolation scheme to the substation automation system via wireless signal.

[0114] Specifically, the power restoration module acquires equipment status information, constructs a power grid topology map based on the equipment status information, and segments the power grid topology map according to the coordinates of the fault area to obtain the non-fault area, specifically:

[0115] Equipment status information is acquired, and based on the equipment ledger data in the equipment status information, combined with the power grid wiring diagram and equipment connection relationships in the equipment status information, an initial power grid topology diagram with equipment as nodes and electrical connections as edges is constructed using graph theory methods. The connection relationships in the topology diagram are dynamically adjusted according to the real-time status information of switching equipment and circuit breakers in the equipment status information to ensure that the topology diagram is consistent with the actual operation mode, and a power grid topology diagram reflecting the current power grid structure is generated. All faulty nodes and faulty equipment located in the faulty area in the power grid topology diagram are identified according to the coordinates of the faulty area, and all faulty nodes and faulty equipment are removed from the power grid topology diagram to obtain the non-faulty area.

[0116] Specifically, this embodiment does not limit the method of obtaining device status information. Those skilled in the art can set it according to the actual situation, such as obtaining it through various sensors, including current transformers and voltage transformers. The device status information includes device ledger data, power grid wiring diagrams, and device connection relationships. The device ledger data refers to the static attribute files of power equipment. The power grid wiring diagram refers to the topological graph describing the electrical connection relationships of the power network. The device connection relationships refer to the logical description of the electrical connectivity between devices. The graph theory method refers to the mathematical tool that applies graph theory algorithms to solve power grid topology problems.

[0117] Specifically, the power restoration module classifies the load of non-faulty areas, generates power supply commands based on the load classification, and restores power to the non-faulty areas based on the power supply commands. Specifically:

[0118] The region attributes of non-faulty areas are obtained, and the load levels of non-faulty areas are classified according to the region attributes to obtain the load levels, where:

[0119] When the area attribute is important, the first-level load is used as the load level of the non-faulty area;

[0120] When the area attribute is a secondary area, the secondary load is treated as the load level of the non-faulty area;

[0121] When the area attribute is a general area, the level 3 load is used as the load level of the non-faulty area.

[0122] Input the load level and coordinates of non-faulty areas into the preset power supply command generation model, obtain the power supply command output by the preset power supply command generation model, and restore power supply to non-faulty areas according to the power supply command.

[0123] Specifically, the important areas refer to important public power supply locations such as hospitals and subway stations; the secondary areas refer to large power supply locations such as factories; and the general areas refer to ordinary power supply locations such as residential areas. The pre-set power supply command generation model is a decision tree model that takes load level and non-faulty area coordinates as input and power supply commands as output. The coordinates of the non-faulty areas are obtained through the power grid topology map. This embodiment does not limit the specific setting method of the power supply command generation model. Those skilled in the art can set it according to the actual situation, such as training the decision tree model with a command training dataset to obtain the power supply command generation model. The command training dataset includes historically acquired load levels and non-faulty areas, as well as power supply commands corresponding to the historically acquired load levels and non-faulty areas.

[0124] Specifically, the power restoration module performs graded correction on the power supply command generation process, specifically as follows:

[0125] When the load level is level three, the power supply command is subject to graded correction, which includes:

[0126] Add three delayed, batch power supply operations as supplementary instructions to the power supply instructions.

[0127] Specifically, the delayed batch power supply refers to the control strategy of closing the circuit and supplying power in batches according to a preset time interval during the power restoration process after a power distribution network failure.

[0128] Specifically, the power restoration module performs trip verification during the graded correction process and trip adjustment during the over-limit replenishment process, as follows:

[0129] The secondary trip rate P of the non-faulty area is acquired and compared with the preset secondary trip rate P0. Based on the comparison result, the secondary tripping situation is judged, and the tripping calibration process for the graded correction is performed based on the judgment result. Furthermore, the tripping adjustment process for over-limit compensation is also performed based on the judgment result. Specifically:

[0130] When P≤P0, the power supply restoration module determines that the secondary tripping situation is a low probability tripping, and does not perform tripping correction during the graded correction process, nor does it perform tripping adjustment during the over-limit supplementation process;

[0131] When P > P0, the power restoration module determines that the secondary tripping situation is a high-probability adjustment and performs tripping verification during the graded correction process. The tripping verification includes:

[0132] Replace the content of the graded correction with "Add the five delayed batch power supply as a supplementary instruction to the power supply instruction";

[0133] The process of replenishing beyond the limit also involves tripping adjustments, which include:

[0134] If the area exceeding the limit is a non-frequent area, the process of pushing the perception optimization plan will also be supplemented with limit exceeding information.

[0135] Specifically, in this embodiment, the number of secondary trips during power restoration in the area within one year is recorded, and the number of secondary trips is divided by the number of power restorations to obtain the secondary trip rate. The preset secondary trip rate is a preset value for judging the secondary trip situation. In this embodiment, P0 is set to 0.15. The secondary trip situation refers to the level of trip probability judged based on the secondary trip rate and the preset secondary trip rate. The secondary trip situation includes high probability trip and low probability trip.

[0136] Please see Figure 2 As shown, this is a flowchart illustrating the method of the distribution network self-healing control system based on a smart grid in this embodiment. The method includes:

[0137] Step S1: Collect power grid data according to the preset sensor array, and judge the local over-limit situation according to the power grid data to obtain the local over-limit situation;

[0138] Step S2: Time alignment processing is performed on the power grid data to obtain the processed power grid data. The processing status of the time alignment processing is also acquired, and a perception optimization scheme is pushed to the management terminal based on the processing status. Furthermore, the process of pushing the perception optimization scheme is supplemented based on the local over-limit situation.

[0139] Step S3: Locate the fault area based on the processed power grid data to obtain the fault area and its coordinates;

[0140] Step S4: Isolate the fault area based on the coordinates of the fault area;

[0141] Step S5: Acquire equipment status information, construct a power grid topology map based on the equipment status information, segment the power grid topology map according to the coordinates of the fault area to obtain non-fault areas, classify the load of the non-fault areas, generate power supply commands according to the load classification, restore power supply to the non-fault areas according to the power supply commands, perform graded correction on the generation process of the power supply commands, perform trip verification on the graded correction process, and perform trip adjustment on the over-limit compensation process.

[0142] 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 self-healing control system for distribution networks based on smart grids, characterized in that, include: The data sensing module is used to collect power grid data based on a preset sensor array, and to judge the local over-limit situation based on the power grid data to obtain the local over-limit situation. The data processing module is used to perform time alignment processing on the power grid data to obtain the processed power grid data. It is also used to acquire the processing status of the time alignment processing and push the perception optimization scheme to the management terminal based on the processing status. It is also used to supplement the push process of the perception optimization scheme based on the local over-limit situation. The fault location module is used to locate the fault area based on the processed power grid data, and obtain the fault area and its coordinates. The fault isolation module is used to isolate the fault area based on the coordinates of the fault area. The power restoration module is used to acquire equipment status information, construct a power grid topology map based on the equipment status information, segment the power grid topology map according to the coordinates of the fault area to obtain non-fault areas, classify the load of the non-fault areas, generate power supply commands based on the load classification, restore power supply to the non-fault areas according to the power supply commands, perform hierarchical correction on the generation process of power supply commands, perform trip verification on the hierarchical correction process, and perform trip adjustment on the over-limit compensation process. The data sensing module determines the local limit violation situation based on power grid data, and obtains the local limit violation situation as follows: The phase current I in the power grid data is compared with the preset phase current I0, and the local over-limit situation is output based on the comparison result, wherein: When I≤I0, the data sensing module outputs the non-local limit violation as a local limit violation case; When I > I0, the data sensing module will output the local limit violation as a local limit violation situation; The zero-sequence voltage U in the power grid data is compared with the preset zero-sequence voltage U0, and the local over-limit situation is output according to the comparison result, wherein: When U≤U0, the data sensing module outputs the non-local limit violation as a local limit violation case; When U > U0, the data sensing module will output the local limit violation as a local limit violation situation; The frequency change rate Z in the power grid data is compared with the preset frequency change rate Z0, and the local limit violation situation is output according to the comparison result, wherein: When Z≤Z0, the data sensing module outputs the non-local limit violation as a local limit violation case; When Z > Z0, the data sensing module will output the local limit violation as a local limit violation situation; The data processing module performs time alignment processing on the power grid data to obtain processed power grid data, specifically: Step A1: Perform timestamp parsing and standardization on the power grid data to obtain timestamp data in a unified format; Step A2: Based on the unified format timestamp data obtained in step S1, construct a reference timeline to obtain the baseline time series; Step A3: Resample the power grid data using a linear interpolation algorithm and the reference time series to obtain preliminary aligned data; Step A4: Perform anomaly detection and consistency verification on the preliminary aligned data to obtain the processed power grid data; The data processing module acquires the processing status of the time alignment process and pushes a perception optimization plan to the management terminal based on the processing status, specifically: The missing percentage C of missing points filled by the linear interpolation algorithm in the time alignment process is obtained. This missing percentage C is compared with a preset missing percentage C0. Based on the comparison result, the data missing situation is judged, and a perception optimization plan is pushed to the management terminal according to the judgment result. Specifically: When C≤C0, the data processing module determines that the data loss is not serious and does not push the perception optimization scheme to the management terminal. When C > C0, the data processing module determines that the data loss is severe and pushes a perception optimization plan to the management terminal. The perception optimization plan includes: Inspect the sensors in the preset sensor array and increase the number of sensors in the preset sensor array; The data processing module supplements the push process of the perception optimization scheme according to the local over-limit situation, specifically as follows: Obtain the number of local violations R in the local violation cases, compare the number of local violations R with the preset number of local violations R0, judge the violation situation in the area based on the comparison result, and supplement the violation process of the perception optimization plan according to the judgment result. Among them: When R < R0, the data processing module determines that the area exceeding the limit is a non-frequent area exceeding the limit and does not supplement the limit exceeding process in the process of pushing the perception optimization scheme. When R≥R0, the data processing module determines the region exceeding the limit as a frequently exceeding the limit region, and performs limit compensation during the push process of the perception optimization scheme. The limit compensation includes: If the data loss is not severe, a perception optimization plan will also be pushed to the management terminal.

2. The self-healing control system for distribution networks based on smart grids according to claim 1, characterized in that, The fault location module locates the fault area based on the processed power grid data, obtaining the fault area and its coordinates, specifically: The fault judgment data in the power grid data is input into the preset fault judgment model to obtain the regional fault status output by the fault judgment model. The fault judgment data includes the three-phase voltage, three-phase current and zero-sequence voltage of each region. The fault status includes whether the region has a fault or not. Based on the regional fault status, the region is divided into fault regions and the fault region coordinates are obtained.

3. The self-healing control system for distribution networks based on smart grids according to claim 2, characterized in that, The fault isolation module isolates the fault area based on the coordinates of the fault area, specifically as follows: Input the coordinates of the fault area into the preset isolation scheme generation model, obtain the isolation scheme output by the isolation scheme generation model, and send the isolation scheme to the substation automation system. Control the circuit breaker and disconnector to complete the tripping operation according to the isolation scheme.

4. The self-healing control system for distribution networks based on smart grids according to claim 3, characterized in that, The power restoration module acquires equipment status information and constructs a power grid topology map based on this information. It then segments the power grid topology map according to the coordinates of the fault area to obtain non-fault areas. Specifically: Equipment status information is acquired, and based on the equipment ledger data in the equipment status information, combined with the power grid wiring diagram and equipment connection relationships in the equipment status information, an initial power grid topology diagram with equipment as nodes and electrical connections as edges is constructed using graph theory methods. The connection relationships in the topology diagram are dynamically adjusted according to the real-time status information of switching equipment and circuit breakers in the equipment status information to ensure that the topology diagram is consistent with the actual operation mode, and a power grid topology diagram reflecting the current power grid structure is generated. All faulty nodes and faulty equipment located in the faulty area in the power grid topology diagram are identified according to the coordinates of the faulty area, and all faulty nodes and faulty equipment are removed from the power grid topology diagram to obtain the non-faulty area.

5. The self-healing control system for distribution networks based on smart grids according to claim 4, characterized in that, The power restoration module performs load classification on non-faulty areas, generates power supply commands based on the load classification, and restores power supply to non-faulty areas based on the power supply commands, specifically as follows: The region attributes of non-faulty areas are obtained, and the load levels of non-faulty areas are classified according to the region attributes to obtain the load levels, where: When the area attribute is important, the first-level load is used as the load level of the non-faulty area; When the area attribute is a secondary area, the secondary load is treated as the load level of the non-faulty area; When the area attribute is a general area, the level 3 load is used as the load level of the non-faulty area. Input the load level and coordinates of non-faulty areas into the preset power supply command generation model, obtain the power supply command output by the preset power supply command generation model, and restore power supply to non-faulty areas according to the power supply command. The power restoration module performs hierarchical correction on the power supply command generation process, specifically as follows: When the load level is level three, the power supply command is subject to graded correction, which includes: Add three delayed batch power supply operations as supplementary instructions to the power supply instructions; The power restoration module performs trip verification during the graded correction process and trip adjustment during the over-limit replenishment process, specifically as follows: The secondary trip rate P of the non-faulty area is acquired and compared with the preset secondary trip rate P0. Based on the comparison result, the secondary tripping situation is judged, and the tripping calibration process for the graded correction is performed based on the judgment result. Furthermore, the tripping adjustment process for over-limit compensation is also performed based on the judgment result. Specifically: When P≤P0, the power supply restoration module determines that the secondary tripping situation is a low probability tripping, and does not perform tripping correction during the graded correction process, nor does it perform tripping adjustment during the over-limit replenishment process; When P > P0, the power restoration module determines that the secondary tripping situation is a high-probability adjustment and performs tripping verification during the graded correction process. The tripping verification includes: Replace the content of the graded correction with "Add the 5 delayed batch power supply as a supplementary instruction to the power supply instruction"; The process of replenishing beyond the limit also involves tripping adjustments, which include: If the area exceeding the limit is a non-frequent area, the process of pushing the perception optimization plan will also be supplemented with limit exceeding information.

6. A method applied to a smart grid-based distribution network self-healing control system as described in any one of claims 1 to 5, characterized in that, include: Step S1: Collect power grid data according to the preset sensor array, and judge the local over-limit situation according to the power grid data to obtain the local over-limit situation; Step S2: Time alignment processing is performed on the power grid data to obtain the processed power grid data. The processing status of the time alignment processing is also acquired, and a perception optimization scheme is pushed to the management terminal based on the processing status. Furthermore, the process of pushing the perception optimization scheme is supplemented based on the local over-limit situation. Step S3: Locate the fault area based on the processed power grid data to obtain the fault area and its coordinates; Step S4: Isolate the fault area based on the coordinates of the fault area; Step S5: Acquire equipment status information, construct a power grid topology map based on the equipment status information, segment the power grid topology map according to the coordinates of the fault area to obtain non-fault areas, classify the load of the non-fault areas, generate power supply commands according to the load classification, restore power supply to the non-fault areas according to the power supply commands, perform graded correction on the generation process of the power supply commands, perform trip verification on the graded correction process, and perform trip adjustment on the over-limit compensation process.

Images