Power distribution network based on different working conditions of electric energy control method and system

Abstract

The application discloses a power distribution network based on different working conditions of electric energy control method and system, the application relates to the technical field of power distribution network, according to the power supply curve of the power distribution network, the area position and time of the corresponding power consumption area determine a plurality of abnormal power supply nodes, according to a plurality of abnormal power supply nodes and the power consumption peak of the power consumption area determine the power supply abnormal event corresponding to the power consumption area, improve the accuracy of the power supply abnormal event corresponding to the power consumption area. According to each power supply abnormal event, the real-time state of the corresponding abnormal power supply node and the power distribution network determines five electric energy control areas, based on five electric energy control areas, the working condition scene and the past electric energy control event of the virtual power distribution network determine the corresponding electric energy control system, and then realize the overall consideration of the dynamic optimization project of the power distribution network, the peak power consumption event of each power consumption area and the load state of the power distribution network, improve the accuracy of the self-regulation system of the power distribution network.

Description

Technical Field

[0001] This invention relates to the technical field of power distribution networks, and in particular to a power control method and system for power distribution networks based on different operating conditions. Background Technology

[0002] With the development of technology, the distribution network, as a power supply component, supplies power to target areas in multiple directions. The distribution network is equipped with corresponding lines, and electricity is transmitted to the corresponding target areas along these lines. These target areas have multiple power-consuming zones that consume the power output from the power supply network. In existing technologies, multiple power supply data from the distribution network are collected, and corresponding power supply events are determined based on these data and power-consuming zones. However, multiple abnormal power supply nodes are ignored, resulting in lower accuracy of the power control system and affecting the power control events of the distribution network under different operating conditions. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art. This invention provides a power control method and system for power distribution networks based on different operating conditions.

[0004] This invention provides a power control method for a power distribution network based on different operating conditions, comprising:

[0005] Based on the power supply data set of the distribution network to the target area and the multiple power consumption areas of the target area, determine the power consumption conditions of each power consumption area, and determine the digital twin model of the power consumption area based on the power consumption conditions of each power consumption area and the corresponding real-time power consumption data.

[0006] The power supply curve of the distribution network is determined by the detection of the digital twin model of each power consumption area. Multiple abnormal power supply nodes are determined according to the power supply curve of the distribution network, the regional location and time of the corresponding power consumption area. The power supply abnormal event corresponding to the power consumption area is determined according to the multiple abnormal power supply nodes and the power consumption peak of the power consumption area.

[0007] Five power control zones are determined based on each power supply anomaly event, the corresponding abnormal power supply node, and the real-time status of the distribution network. Based on the five power control zones, the virtual operating scenario of the distribution network, and previous power control events, the corresponding power control system is determined, and the power control events of the distribution network under different operating conditions are marked.

[0008] Based on the identification of this power control system, power control events of the distribution network under different operating conditions are determined, and multiple key power control factors are determined based on the identification of each power control event.

[0009] The dynamic optimization projects of the distribution network are determined based on multiple key power control factors and digital twin models of each power consumption area; the autonomous control system of the distribution network is determined based on the dynamic optimization projects of the distribution network, the peak power consumption events of each power consumption area and the load status of the distribution network.

[0010] This invention provides a power control system for a power distribution network based on different operating conditions. This power control system is applied to the aforementioned power control method for a power distribution network based on different operating conditions. The power control system includes:

[0011] The digital twin module is used to determine the power consumption conditions of each power consumption area based on the power supply data set of the distribution network to the target area and multiple power consumption areas in the target area, and to determine the digital twin model of the power consumption area based on the power consumption conditions of each power consumption area and the corresponding real-time power consumption data.

[0012] The power supply anomaly event module is used to determine the power supply curve of the distribution network based on the detection of digital twin models of each power consumption area. Based on the power supply curve of the distribution network, the location and time of the corresponding power consumption area, multiple abnormal power supply nodes are determined. Based on the power consumption peak of the multiple abnormal power supply nodes and the power consumption area, the power supply anomaly event corresponding to the power consumption area is determined.

[0013] The power control module is used to determine five power control zones based on various power supply anomaly events, corresponding abnormal power supply nodes, and the real-time status of the distribution network. Based on the five power control zones, the virtual operating scenario of the distribution network, and previous power control events, the corresponding power control system is determined, and the power control events of the distribution network under different operating conditions are marked.

[0014] The key power control factor module is used to determine the power control events of the distribution network under different operating conditions based on the identification of the power control system, and to determine multiple key power control factors based on the identification of each power control event.

[0015] The autonomous control module is used to determine the dynamic optimization projects of the distribution network based on multiple key power control factors and digital twin models of each power consumption area; and to determine the autonomous control system of the distribution network based on the dynamic optimization projects of the distribution network, the peak power consumption events of each power consumption area and the load status of the distribution network.

[0016] Compared with the prior art, the beneficial effects of the present invention are:

[0017] (1) Based on the power supply data set of the distribution network to the target area and the multiple power consumption areas of the target area, the power consumption conditions of each power consumption area are determined. Based on the power consumption conditions of each power consumption area and the corresponding real-time power consumption data, the digital twin model of the power consumption area is determined. Based on the detection of the digital twin model of each power consumption area, the power supply curve of the distribution network is determined. Based on the power supply curve of the distribution network, the regional location and time of the corresponding power consumption area, multiple abnormal power supply nodes are determined. Based on the power consumption peak of multiple abnormal power supply nodes and power consumption areas, the power supply abnormal event corresponding to the power consumption area is determined. The digital twin model of each power consumption area is introduced to control the power supply curve of the distribution network. It takes into account the consideration of multiple abnormal power supply nodes and power consumption peak of power consumption areas, and improves the accuracy of the power supply abnormal event corresponding to the power consumption area.

[0018] (2) Five power control areas are determined based on each power supply abnormality event, the corresponding abnormal power supply node and the real-time status of the distribution network. Based on the five power control areas, the virtual working conditions of the distribution network and previous power control events, the corresponding power control system is determined. The five power control areas are targeted for control. The virtual working conditions of the distribution network are introduced to improve the accuracy of the power control system and mark the power control events of the distribution network under different working conditions.

[0019] (3) Based on the identification of the power control system, the power control events of the distribution network under different operating conditions are determined, and multiple key power control factors are determined according to the identification of each power control event; the dynamic optimization projects of the distribution network are determined according to the multiple key power control factors and the digital twin models of each power consumption area; the autonomous control system of the distribution network is determined based on the dynamic optimization projects of the distribution network, the peak power consumption events of each power consumption area and the load status of the distribution network, and the control of the dynamic optimization projects of the distribution network is realized, thus realizing the overall consideration of the dynamic optimization projects of the distribution network, the peak power consumption events of each power consumption area and the load status of the distribution network, and improving the accuracy of the autonomous control system of the distribution network. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the power control method for a power distribution network based on different operating conditions in an embodiment of the present invention.

[0021] Figure 2 This is a flowchart illustrating step S11 in the power control method for power distribution networks based on different operating conditions in an embodiment of the present invention.

[0022] Figure 3 This is a flowchart illustrating step S12 in the power control method for power distribution networks based on different operating conditions in an embodiment of the present invention.

[0023] Figure 4This is a flowchart illustrating step S13 in the power control method for power distribution networks based on different operating conditions in an embodiment of the present invention.

[0024] Figure 5 This is a flowchart illustrating step S14 in the power control method for power distribution networks based on different operating conditions in an embodiment of the present invention.

[0025] Figure 6 This is a flowchart illustrating step S15 in the power control method for power distribution networks based on different operating conditions in an embodiment of the present invention.

[0026] Figure 7 This is a schematic diagram of the structural composition of the power control system for the power distribution network based on different operating conditions in an embodiment of the present invention. Detailed Implementation

[0027] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0028] Please see Figures 1 to 7 A power control method for power distribution networks based on different operating conditions is proposed and applied to power distribution network scenarios. The power control method for power distribution networks based on different operating conditions includes:

[0029] Step S11: Determine the power consumption conditions of each power consumption area based on the power supply data set of the distribution network to the target area and multiple power consumption areas in the target area; and determine the digital twin model of the power consumption area based on the power consumption conditions of each power consumption area and the corresponding real-time power consumption data.

[0030] Step S12: Determine the power supply curve of the distribution network based on the detection of the digital twin model of each power consumption area. Determine multiple abnormal power supply nodes based on the power supply curve of the distribution network, the location and time of the corresponding power consumption area. Determine the power supply abnormal event corresponding to the power consumption area based on the multiple abnormal power supply nodes and the power consumption peak of the power consumption area.

[0031] Step S13: Based on each power supply anomaly event, the corresponding abnormal power supply node, and the real-time status of the distribution network, determine 5 power control areas. Based on the 5 power control areas, the virtual operating scenario of the distribution network, and previous power control events, determine the corresponding power control system and mark the power control events of the distribution network under different operating conditions.

[0032] Step S14: Based on the identification of the power control system, determine the power control events of the distribution network under different operating conditions, and determine multiple key power control factors based on the identification of each power control event;

[0033] Step S15: Determine the dynamic optimization projects of the distribution network based on multiple key power control factors and digital twin models of each power consumption area; determine the autonomous control system of the distribution network based on the dynamic optimization projects of the distribution network, the peak power consumption events of each power consumption area and the load status of the distribution network.

[0034] refer to Figure 2 In step S11, the specific steps are as follows:

[0035] S111: Collect power supply tasks of the distribution network, determine the corresponding target area based on the identification of the power supply task, and collect multiple power supply data of the distribution network to the target area; determine the power supply data set based on multiple power supply data, the corresponding time and the power supply duration of the distribution network;

[0036] S112: Collect data on the target area, determine each power consumption location based on the target area and the corresponding power consumption markers, determine the corresponding power consumption area based on the tracing of each power consumption location, mark multiple power consumption areas in the target area, determine the power consumption conditions of each power consumption area based on the power supply data set of the distribution network to the target area, the multiple power consumption areas in the target area and the corresponding power consumption status; at the same time, collect real-time power consumption data of each power consumption area, and construct a digital twin model of the power consumption area based on the power consumption conditions of each power consumption area and the corresponding real-time power consumption data.

[0037] In the embodiments of this application, the power supply task of the distribution network is collected, the corresponding target area is determined based on the identification of the power supply task, and multiple power supply data of the distribution network to the target area are collected; the power supply data set is determined based on the multiple power supply data, the corresponding time, and the power supply duration of the distribution network, which takes into account the overall consideration of multiple power supply data, the corresponding time, and the power supply duration of the distribution network, and ensures the accuracy of the power supply data set.

[0038] At this point, the system obtains macroscopic "power supply task" instructions through the interactive distribution automation master station or dispatch system. Using semantic parsing and topology association algorithms, the abstract instructions containing semantic tags such as "routine power supply", "new energy consumption" or "economic operation" are transformed into specific physical spatial coordinates, thereby accurately determining the corresponding target area. For the locked target area, the system relies on intelligent sensing devices deployed at transformer outlets, key nodes of lines and user sides, and uses communication protocols such as IEC61850 or Modbus to collect multi-source heterogeneous power supply data in real time, covering basic electrical quantities such as three-phase voltage, current, active and reactive power, as well as power quality indicators such as harmonic distortion rate and voltage deviation, and equipment status data.

[0039] The system timestamps and cleans the collected multi-source data, removes outliers and noise through preprocessing, calculates the "power supply duration" characteristics of specific states such as voltage over-limit or heavy load operation, and encapsulates the original numerical data, time tags and duration characteristics in a structured manner to form a power supply data set that can reflect the power grid operation status and evolution characteristics.

[0040] Specifically, the target area is set as a rural mountainous area in Province A, and the distribution network is a 10kV #5 transformer substation with typical "high penetration rate of new energy". The power supply radius of this substation is 150 meters and the transformer capacity is 400kVA. The system obtains the power supply task of "full consumption of new energy and voltage quality management in rural areas" from the dispatch center. The system parses out the keywords "new energy" and "voltage quality" and, combined with the distribution network topology database, accurately locates the 10kV #5 transformer substation in the mountainous area, which contains a large number of household photovoltaic systems, as the control object, excluding purely industrial load substations.

[0041] The system uses smart integrated terminals deployed in the area to collect three-phase voltage and current data from the low-voltage side of the transformer and users at the end of the line at high frequency. It focuses on the direction of active power to identify the photovoltaic back-feeding status and records the specific value of 300kW photovoltaic output when the load rate is only 20% at noon.

[0042] The system uses the period from 11:00 to 14:00 on October 3, 2025 as the time window, integrates the change curve of the line end voltage rising from 225V to 242V during this period, calculates that the power supply duration caused by "photovoltaic reverse feed" is about 3.5 hours per day, and the cumulative voltage over-limit time is as long as 1.2 hours. Finally, it outputs a set of structured power supply data containing timestamps, end voltage of 242V, photovoltaic reverse feed power of 280kW, and over-limit duration of 72min, providing accurate input for subsequent steps.

[0043] Furthermore, data is collected from the target area, and power consumption locations are determined based on the target area and corresponding power consumption markers. Based on the tracing of each power consumption location, corresponding power consumption areas are identified, thus marking multiple power consumption areas within the target area. The power consumption conditions of each power consumption area are determined based on the power supply data set from the distribution network to the target area, the multiple power consumption areas in the target area, and their corresponding power consumption states. Simultaneously, real-time power consumption data for each power consumption area is collected, and a digital twin model of the power consumption area is constructed based on its power consumption conditions and corresponding real-time power consumption data. This approach incorporates a holistic consideration of the power supply data set from the distribution network to the target area, the multiple power consumption areas in the target area, and their corresponding power consumption states, ensuring the accuracy of the power consumption conditions for each power consumption area.

[0044] At this point, the system performs a fine scan of the locked target area based on GIS topology data, locates specific physical nodes using preset "power consumption markers", and aggregates and traces power consumption locations that are geographically adjacent or have similar electrical characteristics through graph computing algorithms. These logical units are marked as different "power consumption areas", thereby dividing the huge power distribution network into several subdomains that can be independently analyzed and differentiated for control.

[0045] The system combines previously generated power supply data sets to analyze the current power consumption status of each power-consuming area and deeply integrates it with static archive features. Based on geographical location features and application scenario features, it defines specific operating condition labels such as "rural mountainous areas + high-penetration new energy sources" or "urban commercial areas + impact loads".

[0046] The system utilizes edge intelligent sensing devices to collect real-time power consumption data from various regions at high frequency. It integrates the determined power consumption conditions as physical constraints and rule bases with the real-time data, and constructs a digital twin model with the ability to perform calculations and inferences based on physical laws in the cloud or edge computing gateway, making it a "virtual test field" for verifying the effectiveness of control strategies.

[0047] Specifically, for the 10kV #5 transformer substation in a mountainous rural area of ​​Province A, the system performed a topology scan and identified a large number of user meters marked with "photovoltaic power generation" on the end branch line. The system aggregated these scattered photovoltaic users and the surrounding affected residential users into an independent logical unit, named "Regional-End Photovoltaic Aggregation Zone".

[0048] The system retrieved power supply data sets to analyze the status of the area and found that the active power was reversed and the voltage was rising during the noon period. Combined with the geographical characteristics of the area being a rural mountainous area, the physical parameters of long power supply radius and high photovoltaic coverage, the power consumption condition of the area was finally determined to be "high penetration rate new energy scenario" superimposed on "long line remote load scenario".

[0049] For this region, the system collects real-time data such as voltage, current and harmonics at the beginning of the line every second through intelligent fusion terminals. Based on the operating parameters, a digital twin model is built. This model can not only reproduce the voltage rise process at the end in real time, but also simulate the projection that the voltage will exceed 250V when the photovoltaic penetration rate increases from 25% to 30%. This reveals the serious overvoltage risk in this region and provides solid digital basis for subsequent activation and suppression of voltage rise strategies.

[0050] refer to Figure 3 In step S12, the specific steps are as follows:

[0051] S121: In the digital twin model of each power consumption area, the digital twin model is dynamically tested, and multiple key power supply data of the distribution network are determined during the testing process. Based on the multiple key power supply data, the corresponding time and the load parameters of the distribution network, the power supply curve of the distribution network is constructed.

[0052] S122: Based on the identification of the power supply curve, multiple power supply waveform regions are determined. Based on the identification of each power supply waveform region, multiple waveform abnormal locations are determined. Based on the multiple waveform abnormal locations and the corresponding power consumption regions, a first-level abnormal node combination is determined. Based on the multiple waveform abnormal locations and time, a second-level abnormal node combination is determined. Based on the matching of the first-level abnormal node combination and the second-level abnormal node combination, multiple abnormal power supply nodes are determined.

[0053] S123: Monitor each power consumption area in real time and mark the power consumption peak of each power consumption area. Based on the power consumption peak and corresponding area location of multiple abnormal power supply nodes and power consumption areas, determine each power supply anomaly factor. Based on each power supply anomaly factor, the area shape of the power consumption area and the corresponding power consumption data, determine the power supply anomaly event corresponding to the power consumption area.

[0054] In the embodiments of this application, the digital twin model of each power consumption area is dynamically detected, and multiple key power supply data of the distribution network are determined during the detection process. The power supply curve of the distribution network is constructed based on the multiple key power supply data, the corresponding time, and the load parameters of the distribution network. This takes into account the overall consideration of multiple key power supply data, the corresponding time, and the load parameters of the distribution network, and ensures the accuracy of the power supply curve of the distribution network.

[0055] At this point, the system activates the constructed digital twin model of the power-consuming area, putting it into real-time operation or high-frequency simulation. The model not only receives real-time data uploaded by the edge gateway as boundary conditions, but also performs dynamic deduction based on internal physical mechanisms (such as power flow calculation and state estimation). The system continuously scans and monitors the operating status of the model, which is equivalent to constructing a "shadow power grid" that operates synchronously with the physical power grid in virtual space. This can capture small and rapid changes in electrical quantities in the power grid, providing a dynamic data source for subsequent analysis.

[0056] The system continuously scans and monitors the model's operating status, which is equivalent to constructing a "shadow power grid" that operates synchronously with the physical power grid in virtual space. It can capture small and rapid changes in electrical quantities in the power grid, providing a dynamic data source for subsequent analysis. Key power supply data refers to the core indicators that can characterize power quality and operating status. Key power supply data includes voltage data, current data, power data, and power quality data.

[0057] Voltage data includes three-phase voltage amplitude, voltage deviation, and phase angle; current data includes three-phase current amplitude and zero-sequence / negative-sequence current; power data includes active power, reactive power, and power factor; power quality data includes harmonic distortion rate, voltage fluctuation, and flicker value. While extracting these data, the system adds a high-precision timestamp to ensure that each data point accurately corresponds to a specific moment in the physical world.

[0058] The system uses the extracted key power supply data as the dependent variable and time as the independent variable to construct a basic time series curve. To give the curve physical meaning, the system maps the load parameters of the distribution network (such as transformer rated capacity, line impedance, and load peak) as reference quantities onto the curve. The final generated "power supply curve" not only shows the trajectory of electrical quantities over time, but also intuitively reflects the position of the current operating state relative to the equipment's carrying capacity (load parameters). For example, the curve can clearly show whether the current voltage exceeds the limit, whether the load rate exceeds the standard, and the reversal process of power flow direction (forward power reception / reverse power transmission) over time.

[0059] Specifically, for the "regional-terminal photovoltaic aggregation zone" of the 10kV#5 transformer substation in a mountainous rural area of ​​Province A, the constructed digital twin model is activated, enabling it to access the measurement data of the substation's intelligent integrated terminal in real time. The power flow calculation algorithm is used to simulate the operation status of the low-voltage distribution network, and the node voltage and branch current are dynamically monitored at the millisecond level, virtually reproducing the real-time response process of the mountainous distribution network under changes in sunlight.

[0060] The system focused on extracting key power supply data, detecting negative active power values ​​indicating reverse power transmission from photovoltaics, and a continuous increase in voltage amplitude at the end of the line. It also recorded the power factor fluctuations caused by the photovoltaic inverter and precisely marked these data points with timestamps from 10:00 to 14:00. Based on the above data and time series, combined with load parameters such as a transformer capacity of 400kVA and a rated voltage of 220V, the system constructed a "voltage-time power supply curve" and an "active power-time power supply curve." The voltage-time curve showed a steep upward trend from 11:00 to 14:00, exceeding the safe operating limit of 235V, and was superimposed with a transformer capacity limit reference line, directly exposing the severe overvoltage risk faced by the distribution network during periods of high photovoltaic power generation.

[0061] Furthermore, multiple power supply waveform regions are determined based on the identification of the power supply curve, and multiple waveform anomaly locations are determined based on the identification of each power supply waveform region. A first-level anomaly node combination is determined based on the multiple waveform anomaly locations and the corresponding power consumption regions. A second-level anomaly node combination is determined based on the multiple waveform anomaly locations and time. Multiple abnormal power supply nodes are determined based on the matching of the first-level and second-level anomaly node combinations. This overall consideration of matching the first-level and second-level anomaly node combinations ensures the accuracy of multiple abnormal power supply nodes.

[0062] At this point, the system uses signal processing algorithms (such as wavelet transform, slope detection, or threshold segmentation) to perform morphological analysis on the power supply curve generated by S121; based on the changing characteristics of electrical quantities, the curve on the time axis is divided into different "power supply waveform regions"; for example, it is divided into "stable operation region" and "transient disturbance region" according to voltage fluctuation rate; into "light load region" and "heavy load region" according to load level; or into "forward power receiving region" and "reverse power supply region" according to power flow direction. This process divides the continuous curve into segments with specific physical meaning.

[0063] Within each defined power supply waveform area, the system compares real-time data with preset power quality standards (such as GB / T standards) or predicted values ​​from digital twin models; it identifies feature points or segments in the curve that deviate from the normal range, namely "abnormal waveform positions". These positions are specifically manifested as voltage over-limit points, harmonic distortion abrupt increase points, or the lowest point of voltage sag.

[0064] The system establishes a mapping relationship between waveform data and physical topology; when an abnormal waveform location is detected, the physical line or user node where it is located is locked by tracing the sensor device ID of the data source; all physical locations where waveform anomalies occur are aggregated to form a "first-level abnormal node combination", which defines the geographical coverage of the anomaly (e.g., whether it is a few specific users, a branch of a line, or the entire substation area).

[0065] The system extracts the precise timestamp and duration of each waveform anomaly location; analyzes the temporal patterns of the anomalies, summarizes them into specific time windows, and forms a "second-level anomaly node combination". This combination defines the temporal distribution characteristics of the anomalies (e.g., only during midday, only during evening peak load, or random occurrence).

[0066] The system performs a logical AND operation on spatial and temporal combinations; only when a specific spatial region continuously or repeatedly exhibits anomalies within a specific time window is it identified as a valid "abnormal power supply node"; through this multi-dimensional cross-validation, the system filters out accidental noise interference and ultimately identifies "abnormal power supply nodes" with clear physical locations and temporal attributes.

[0067] Specifically, by analyzing the characteristics of voltage changes over time, the system identifies two characteristic areas: the "normal nighttime operation zone" and the "midday photovoltaic high-penetration disturbance zone." Within the "midday photovoltaic high-penetration disturbance zone," the system performs threshold detection and discovers multiple abnormal waveform locations where the voltage amplitude exceeds the 235V safe operating limit, reaching a maximum of 242V, and these are concentrated in the period from 11:30 to 13:30.

[0068] The system traces the location of the sensors that generate these abnormal data, aggregates the monitoring points located at the end of the main line and on the branch lines, and determines the first-level abnormal node combination as "distribution network - line end area"; at the same time, the system analyzes the time pattern of the abnormality and determines the second-level abnormal node combination as "daily midday period (11:00-14:00)".

[0069] Through spatiotemporal matching, the system determined that the voltage over-limit in the "line end area" only occurred during the "daily midday period" and returned to normal at night, thus ruling out the possibility of a permanent line fault. Finally, the specific abnormal power supply node was determined to be the "distribution network - line end node [time window: 11:00-14:00]", accurately locating the physical phenomenon of photovoltaic backfeeding causing the voltage over-limit at the end.

[0070] Therefore, real-time monitoring of each power-consuming area and marking the peak power consumption of each area are performed. Based on the peak power consumption of multiple abnormal power supply nodes and power-consuming areas and their corresponding locations, various power supply anomaly factors are determined. Based on each power supply anomaly factor, the regional shape of the power-consuming area, and the corresponding power consumption data, the corresponding power supply anomaly event is determined. This approach takes into account all power supply anomaly factors, the regional shape of the power-consuming area, and the corresponding power consumption data, ensuring the accuracy of the power supply anomaly event corresponding to the power-consuming area. At the same time, a digital twin model of each power-consuming area is introduced to control the power supply curve of the distribution network. This approach takes into account the peak power consumption of multiple abnormal power supply nodes and power-consuming areas, improving the accuracy of the power supply anomaly event corresponding to the power-consuming area.

[0071] At this time, the system continuously monitors each power-consuming area in the distribution network in real time, and uses a sliding window algorithm or extreme value detection algorithm to capture the power extreme value of each area within a specific time window; the system identifies and records the "power consumption peak value", which refers not only to the highest value of the forward load, but also to the peak value of reverse generation (backfeeding) in a source network; at the same time, it records the precise time of the peak occurrence and the duration, providing quantitative indicators for analyzing the intensity of the anomaly.

[0072] The system integrates the "abnormal power supply node" (spatiotemporal coordinates) determined in the previous step (S122) with the current "peak power consumption" (intensity index) and "regional location" (spatial attribute). By analyzing whether the time of occurrence of the abnormal node is synchronized with the peak power consumption, and whether the spatial distribution of the abnormality coincides with a specific load area, the system infers the root cause of the abnormality, namely the "abnormal power supply factor". For example, if the "high voltage abnormal node" coincides with the "peak photovoltaic power generation" in time, the abnormal factor is determined to be "voltage rise caused by photovoltaic backfeeding"; if the "voltage sag node" coincides with the "peak charging pile start-up" in time, the abnormal factor is determined to be "voltage fluctuation caused by impulsive load".

[0073] The system introduces "regional morphology" as a key criterion; regional morphology includes physical constraints such as topology (e.g., long lines, short lines), impedance characteristics, and equipment capacity; the system comprehensively judges "power supply anomaly factors" (causes), "regional morphology" (constraints), and specific "power consumption data" (exceeding limits, THD values, etc.) to ultimately define specific "power supply anomaly events"; this step translates technical parameters into operation and maintenance language; for example, if the factor is "voltage rise," the morphology is "long lines," and the data is "severe exceeding limits," it is defined as an "end-of-line overvoltage event"; if the factor is "load surge," the morphology is "thin wire diameter," and the data is "large voltage drop," it is defined as an "end-of-line low voltage event."

[0074] Specifically, for the "regional-terminal photovoltaic aggregation area" of the 10kV#5 transformer station in a mountainous rural area of ​​Province A, the system conducts high-frequency real-time monitoring. Within the abnormal time window, the system detected a significant negative peak in active power (reverse power generation), with a value of -285kW. This moment was marked as the "peak moment of photovoltaic output" and recorded as lasting for about 2.5 hours.

[0075] The system performs a deep correlation analysis between the abnormal nodes identified in S122 and the monitored power consumption peaks. It finds that the time interval of "abnormal increase in end voltage" highly overlaps with the interval of "peak photovoltaic power generation" and both are located in the end area of ​​the line. Therefore, it is determined that the abnormal power supply factor is "power reverse transmission caused by high penetration photovoltaic backfeed, which leads to the line voltage rise effect".

[0076] The system further retrieved static files to confirm that the area was a "long line with a power supply radius of 150 meters" with a large line impedance parameter. Combined with the power consumption data that the voltage at the end reached a maximum of 242V, which seriously exceeded the qualified upper limit of 235V, the system finally determined that the power supply anomaly event corresponding to this power consumption area was a "high penetration photovoltaic area overvoltage event", providing a precise target for the subsequent development of a targeted power control system.

[0077] refer to Figure 4 In step S13, the specific steps are as follows:

[0078] S131: Collect multiple status data of the distribution network, determine the real-time status of the distribution network based on the multiple status data and the corresponding time, and determine the first power control position based on the real-time status of the distribution network and various power supply anomaly events.

[0079] S132: Determine the second power control position based on the real-time status of the distribution network and the abnormal power supply nodes corresponding to each power supply anomaly event; determine 5 power control areas based on the first power control position, the second power control position and the distribution network distribution map.

[0080] S133: Based on the identification of past operating conditions of the distribution network, multiple virtual operating condition data are determined. Based on the multiple virtual operating condition data, the corresponding operating condition scenarios are determined to mark the virtual operating condition scenarios of the distribution network. Based on the virtual operating condition scenarios of the distribution network and 5 power control areas, a power control framework is determined. Based on the power control framework and past power control events, the corresponding power control system is determined. By tracing the virtual operating condition scenarios of the distribution network, the power control events of the distribution network under different operating conditions are determined. The virtual operating condition scenarios of the distribution network are used as dynamically virtual scenarios.

[0081] In the embodiments of this application, multiple state data of the distribution network are collected, the real-time state of the distribution network is determined based on the multiple state data and the corresponding time, and the first power control position is determined based on the real-time state of the distribution network and various power supply anomaly events. This approach takes into account the overall consideration of multiple state data and corresponding time of the distribution network, ensuring the accuracy of the real-time state of the distribution network.

[0082] At this time, the system comprehensively collects the operating parameters of the distribution network through the SCADA system, the intelligent integrated terminal of the distribution area and the UPQC local controller. The collected "status data" includes not only electrical quantities (voltage, current, switch position), but also equipment status (UPQC operating mode, DC bus voltage, module temperature) and environmental data.

[0083] The system timestamps and fuses the collected multidimensional state data to eliminate differences caused by data transmission delays. Through state estimation algorithms, it generates a "real-time state snapshot" of the current distribution network. This snapshot accurately reflects the current topology of the power grid (whether it is a ring network, whether switches are open or closed), operating level (heavy load / light load), and availability of key equipment (whether it has regulation capabilities).

[0084] The system matches and analyzes the "power supply abnormal event" (such as overvoltage in a certain area) identified in step S12 with the current "real-time status". This locked node is the "first power control position". It usually refers to the grid connection point of power quality management equipment (such as UPQC), the tap position of on-load tap-changing transformer, or the outlet of distributed energy storage.

[0085] Specifically, the system performs panoramic status acquisition, obtaining key status data such as "grid-connected operation, standby, and normal communication" of the UPQC device, 40% load rate and normal temperature of the transformer in the distribution area, closed main line switch, and detected reverse active power. The system integrates these data to determine the current real-time status of the distribution network as "photovoltaic reverse power transmission operation status, UPQC device online and available, and complete grid topology", indicating that the grid has the physical basis for power quality regulation.

[0086] The system combines real-time status with previously identified "overvoltage events" to make decisions. Analysis revealed that the UPQC device installed in the middle of the distribution area is located at a critical node in the electrical connection, and its series-side voltage regulation capability can cover voltage fluctuations at the end. Based on the real-time status of the UPQC being "online available" and its controllability over the end node, the system ultimately determined the first power control location to be "distribution network - middle of distribution area - UPQC device grid connection point", laying the execution foundation for subsequent division of control areas and issuance of specific instructions.

[0087] Furthermore, the second power control position is determined based on the real-time status of the distribution network and the abnormal power supply nodes corresponding to each power supply anomaly event. Five power control areas are determined based on the first power control position, the second power control position, and the distribution network distribution map. This approach takes into account the overall considerations of the first power control position, the second power control position, and the distribution network distribution map, ensuring the accuracy of the five power control areas.

[0088] At this time, the system combines the "real-time status" of the distribution network (such as load distribution and power flow direction) to accurately locate the "abnormal power supply node" corresponding to the "power supply abnormal event" identified in S12 on the physical topology; the physical location or electrical connection point of these abnormal nodes is defined as the "secondary power control location". This is the root cause of power quality problems or the area with the most serious impact. It is the target object that the control strategy needs to focus on adjusting or protecting (such as the end of the line with voltage exceeding the limit, the grid connection point of harmonic source).

[0089] The system overlays first-level locations (control resources) and second-level locations (control objects) onto the distribution network geographic information system (GIS) map. Using graph theory algorithms or electrical distance analysis, it calculates the coupling strength between control resources and control objects. The system aggregates nodes with similar electrical distances, control characteristics, or those affected by the same control strategy. Based on the aggregation results, the entire distribution network is divided into five logical or physical "power control zones." These zones are typically divided based on: sensitivity zones, pollution source zones, power source zones, ordinary load zones, and central node zones. This division ensures that each zone has a clear primary control objective and corresponding control resources.

[0090] Specifically, the system verifies the real-time status and confirms that the voltage in the area at the end of the line remains at around 242V during the peak output period of photovoltaic power generation at noon and that a large number of residential photovoltaic inverters are concentrated there. Based on the correspondence between the real-time status and the "overvoltage event in the high-penetration photovoltaic area", the system locks the control target to the abnormal power supply node, thereby determining the second power control location as "distribution network - photovoltaic aggregation area at the end of the line".

[0091] Based on the distribution map, the system performs topological clustering based on the first-level location (midstream UPQC) and the second-level location (end-of-line photovoltaic area), analyzes the single-radial topology of the distribution area, and divides the distribution network into 5 power control zones according to voltage sensitivity and load characteristics: the "source-end stability control zone" covering the transformer outlet and near-end users, the "mid-section main control zone" located near the UPQC installation point, the "end-of-line sensitive control zone" covering the second-level location and its surroundings, the "photovoltaic source control zone" for high-density photovoltaic access branch lines, and the "conventional load buffer zone" located in the middle of the main line. This refined regional division enables the system to implement precise control of overvoltage issues in the "end-of-line sensitive control zone" using the UPQC in the "mid-section main control zone" without affecting the normal operation of other areas.

[0092] Therefore, multiple virtual operating condition data are determined based on the identification of past operating condition events of the distribution network. Corresponding operating condition scenarios are then determined based on these virtual operating condition data to mark the virtual operating condition scenarios of the distribution network. An energy control framework is determined based on the virtual operating condition scenarios of the distribution network and five energy control areas. A corresponding energy control system is determined based on this energy control framework and past energy control events. Furthermore, by tracing back along the virtual operating condition scenarios of the distribution network, energy control events of the distribution network under different operating conditions are determined. The virtual operating condition scenarios of the distribution network, as dynamically virtual scenarios, are compatible with the overall consideration of the energy control framework and past energy control events, ensuring the accuracy of the corresponding energy control system. Simultaneously, targeted control is implemented for the five energy control areas, and the introduction of the virtual operating condition scenarios of the distribution network improves the accuracy of the energy control system and marks the energy control events of the distribution network under different operating conditions.

[0093] At this point, the system delves into the historical operation database to identify typical "operating condition events" that have occurred in the past (such as seasonal overload, voltage fluctuations under extreme weather, and peak loads during holidays). Based on the characteristic parameters of these events (such as peak load, penetration rate of new energy sources, line impedance, temperature, etc.), the system generates or extracts multiple sets of typical "virtual operating condition data".

[0094] Clustering algorithms are used to classify virtual operating condition data, aggregating data with similar characteristics into one category and defining corresponding "operating condition scenarios" (such as "high penetration rate new energy + long line" and "impact load + voltage fluctuation"). The system marks these scenarios as virtual operating condition scenarios of the distribution network. They are not simple static labels, but dynamic virtual models with parameter boundaries generated based on historical data, used to characterize various extreme or typical operating states that the power grid may encounter.

[0095] The system performs topological association and logical mapping between the virtual operating scenario and the "5 power control areas" divided in step S132; Framework construction: The system analyzes which control areas will become the "master control area" (strategy executor) and which will become the "controlled area" (protected object) under different virtual operating scenarios; for example, in the "harmonic exceedance" virtual scenario, the area where the harmonic source is located is the controlled object, while the area where the treatment equipment is located is the master control area; thus, a "power control framework" is established, which defines the scheduling logic and coordination relationship of control resources (UPQC, etc.) between different control areas under specific scenarios.

[0096] Based on the control framework, the system introduces control strategies and parameter libraries accumulated from "past power control events"; it collects a refined mapping table, namely the "power control system," which establishes multi-dimensional correlation logic: the input is "virtual operating scenario + current operating task (such as renewable energy consumption) + regional real-time operating conditions (such as voltage, harmonics)", and the output is specific "control strategy combinations" and "parameter settings" (such as UPQC series side voltage target value, parallel side harmonic compensation command); its core function is to ensure that no matter what virtual or actual operating conditions the distribution network is in, the system can match the optimal control action, realizing the transformation from "universal solution" to "precise policy implementation".

[0097] The system utilizes virtual operating scenarios as a dynamic virtual test field to simulate or monitor the performance of the power control system in actual operation. The system traces the evolution trajectory of the virtual scenario, monitoring the execution frequency and duration of specific control strategies in that scenario. When a certain strategy (such as reverse voltage regulation) is triggered for a long time and at high intensity in the real-time operating conditions corresponding to a specific virtual scenario, it indicates that the current governance capacity is approaching its limit or there is a persistent risk. The system marks this abnormal state of strategy execution as "power control events of the distribution network under different operating conditions". These events not only reflect the current operating status, but also suggest potential equipment hazards or grid structure problems, providing direct basis for subsequent maintenance and transformation.

[0098] Specifically, based on the operation records of the past year, the system identified voltage over-limit phenomena that often occur at midday in summer. The system extracts characteristic parameters such as photovoltaic penetration rate greater than 25%, power supply radius of 150 meters, ambient temperature of 35℃, and reverse power of 250kW to generate "virtual operating condition data". Then, it constructs and marks the virtual operating condition scenario of "high penetration rate new energy + long line + seasonal high temperature" to dynamically simulate the extreme operating conditions during the summer photovoltaic power generation.

[0099] The system matches the virtual scenario with five power control areas, determining that the "end-sensitive control area" is the most severely affected area for voltage overruns, i.e., the controlled area, while the "mid-section main control area" has adjustment capabilities, i.e., the main control area. Thus, a power control framework is established that uses the UPQC device in the "mid-section main control area" to clamp the voltage in the "end-sensitive control area" through series side voltage regulation.

[0100] Based on the framework, the system constructs a power control system by combining previous control experience. It stipulates that when the real-time operating conditions match the virtual scenario and the current task is renewable energy consumption and the voltage is high, the "suppress voltage rise" mode on the UPQC series side is activated and the target voltage setting value is set below 235V. The system runs the virtual scenario and continuously monitors the actual operation. It was found that the continuous duration of the UPQC executing the "suppress voltage rise" strategy during the period matching the virtual operating conditions exceeded 3 hours per day. Based on this, the system marked "frequent overvoltage regulation events", indicating that the existing regulation resources in this area have been operating under high load for a long time in the high-penetration photovoltaic scenario, suggesting that attention should be paid to the equipment life and the load-bearing capacity of the area.

[0101] refer to Figure 5 In step S14, the specific steps are as follows:

[0102] S141: Monitor the power control system in real time, dynamically identify the power control system, identify multiple power control contents during the identification process, mark the corresponding power control areas, and determine the power control events of the distribution network under different operating conditions based on each power control content, the corresponding power control area, and the virtual operating scenario of the distribution network.

[0103] S142: In this power control event, multiple power control items are identified based on the identification of the power control event. The corresponding power control factors are determined according to the project content and corresponding control scope of each power control item. Multiple key power control factors are determined according to the types, corresponding priorities and current operating conditions of the power distribution network of the multiple power control factors.

[0104] In the embodiments of this application, the power control system is monitored in real time, dynamically identified, and multiple power control contents are determined during the identification process. The corresponding power control areas are marked, and power control events of the distribution network under different operating conditions are determined based on each power control content, the corresponding power control area, and the virtual operating scenario of the distribution network. This approach takes into account the overall consideration of each power control content, the corresponding power control area, and the virtual operating scenario of the distribution network, ensuring the accuracy of power control events of the distribution network under different operating conditions.

[0105] At this point, the system establishes a full-link monitoring mechanism for the "power control system" constructed in step S13. The core of the monitoring includes not only the electrical parameters of the power grid (voltage, current), but also the internal control status of core nodes of intelligent operation and maintenance such as UPQC. The system captures in real time the issuance of control commands, execution feedback, switching of strategy modes (such as switching from standby mode to voltage regulation mode), and the operating parameters of equipment (such as modulation ratio, DC bus voltage, output power). This ensures that the system can grasp whether the control system is working as expected and what the current regulation intensity is.

[0106] The system utilizes state machine recognition algorithms or pattern recognition technology to dynamically analyze the captured monitoring data. By analyzing the UPQC control logic, the system determines the specific physical action that the equipment is currently performing, i.e., the "power control content." For example, the UPQC series side is injecting negative sequence voltage to offset the line voltage drop; the UPQC parallel side is injecting harmonic current of a specific frequency to filter out pollution; the Dynamic Reactive Power Compensation (D-STATCOM) function is generating capacitive reactive power to support the voltage. The system also extracts key characteristic parameters of this control content, such as adjustment amplitude, response delay, and duration of action.

[0107] The system associates the identified "power control content" with the physical topology of the distribution network; the system analyzes the electrical node or load range targeted by the control action; for example, it identifies whether the closed-loop feedback signal of the control strategy is taken from the PT (voltage transformer) at the end of the line or from the transformer outlet; based on the feedback source and the scope of influence, the system marks the control action as acting on one or more "power control areas" defined in step S132; for example, it marks the adjustment action for the terminal voltage as acting on the "terminal sensitive control area".

[0108] The system performs multi-dimensional logical integration and judgment of "control behavior (content)," "object of action (area)," and "operating environment (virtual operating scenario)." When a specific "power control content" is triggered in a specific "virtual operating scenario of the distribution network" for a specific "power control area" and meets specific frequency or threshold conditions (such as excessive duration, excessive adjustment range, or frequent back-and-forth adjustment), the system upgrades it to a "power control event of the distribution network under different operating conditions." This process transforms simple equipment actions into "events" with operational significance. For example, "continuous voltage suppression action performed on the terminal area in a long-line high photovoltaic scenario" is defined as a "frequent overvoltage adjustment event," indicating that the current governance pressure of the system is approaching the threshold.

[0109] Specifically, the system continuously monitors the UPQC device installed in the middle of the transformer area, and detects that its series-side converter is in an "active" state with a real-time output power of -15kVA. The internal logic of the controller shows that it is executing the "voltage closed-loop regulation" algorithm. The system analyzes the operating status and identifies that the specific action currently being performed by the device is "series-side dynamic voltage suppression". That is, the UPQC injects an inductive voltage vector with a phase lead of 90 degrees into the grid to offset the voltage rise caused by the photovoltaic backfeed at the end of the line. The system also identifies that the regulation amplitude is -20V and has been running continuously for 45 minutes.

[0110] The system checks the voltage sampling loop of UPQC and confirms that the feedback signal is taken from the intelligent fusion terminal at the end of the line, thus marking the current "series-side dynamic voltage suppression" action as acting on the "end-sensitive control area". The system confirms that the current distribution network is in the virtual operating condition scenario of "high penetration rate new energy + long line". It determines that the duration of the "voltage suppression" action for the "end-sensitive control area" in this scenario has exceeded the preset threshold. Based on this, the system finally identifies a power control event called "frequent overvoltage regulation event", which indicates that the system has identified that the current operation is under high load by relying solely on equipment regulation.

[0111] Furthermore, in this power control event, multiple power control items are identified based on the identification of the power control event. Corresponding power control factors are determined according to the item content and corresponding control scope of each power control item. Multiple key power control factors are determined according to the types, corresponding priorities and current operating conditions of the power distribution network of the multiple power control factors. This approach takes into account the overall consideration of the types, corresponding priorities and current operating conditions of the power distribution network of the multiple power control factors, ensuring the accuracy of the multiple key power control factors.

[0112] At this point, the system performs structural decoupling on the "power control events" (such as frequent overvoltage regulation events) determined in S141; a macro control event often contains multiple specific execution actions; the system decomposes the event into several specific "power control items" according to the execution logic of the control strategy, and these items usually correspond to specific functional modules of equipment such as UPQC.

[0113] Example: If the event is "frequent overvoltage regulation", the breakdown items may include: "series side voltage amplitude regulation", "active power throughput control", or "reactive power dynamic compensation". If the event is "harmonic exceedance control", the breakdown items may include: "5th harmonic current injection" or "7th harmonic current injection".

[0114] For each decomposed "power control project," the system analyzes its physical nature and scope of action, identifying all variables affecting the project's performance and boundary conditions, i.e., "power control factors." Project content analysis: This analyzes what physical quantities (such as voltage, current, phase angle) the project regulates. Control range analysis: This analyzes the physical area affected by the project (such as the end of the line, transformer outlet) and the network characteristics of that area. Combining the above analysis, specific factors are determined. For example, for voltage regulation projects, factors include: line impedance, active load, reactive load, photovoltaic output, transformer tap position, etc.; for harmonic mitigation projects, factors include: harmonic source current magnitude, background harmonic voltage, filter branch impedance, etc.

[0115] The system selects the factors that have the greatest impact on the current operating state and require priority control from numerous "power control factors," namely, the "key power control factors." The selection dimensions are: Type: Distinguishing between "primary electrical factors" (such as voltage, current, and power) and "auxiliary factors" (such as temperature and humidity), with primary electrical factors typically given priority; Priority: Assessing the weight of factors' impact on system safety and power quality; for example, "voltage exceeding limits" has a higher priority than "voltage deviation," and "short-circuit current" has a higher priority than "load rate"; Current operating conditions: Judging the activity level of factors by combining the real-time operating status of the distribution network (such as high photovoltaic power generation, low load, and fault recovery); In scenarios with high photovoltaic power generation, "photovoltaic reverse power" is the core factor; while in heavy load scenarios, "line voltage drop" may be the core factor. Through comprehensive evaluation, the system selects several of the most representative "key power control factors" to guide subsequent dynamic optimization and autonomous control.

[0116] Specifically, for the "frequent overvoltage regulation event" identified in the 10kV #5 transformer substation in a mountainous rural area of ​​Province A, the system deconstructed the event and identified two specific power control projects actually executed by UPQC: Project A is the "series-side dynamic voltage suppression project" and Project B is the "reactive power absorption project". For these two projects, the system combined their control range to mine physical factors and identified a preliminary set of factors including photovoltaic reverse active power, line impedance parameters, load current, power factor and bus voltage.

[0117] The system performs multi-dimensional screening of the above factors based on the current operating conditions. Through category analysis, it prioritizes and retains the main electrical factors. Through priority analysis, it determines that photovoltaic backfeeding is the root cause of voltage rise and has the highest priority. Combined with the characteristics of drastic fluctuations in photovoltaic output under the "high penetration rate new energy" operating conditions, it finally determines three key power control factors: photovoltaic reverse active power, line impedance voltage drop, and UPQC capacity utilization rate, which are used to guide the subsequent formulation of targeted dynamic optimization strategies.

[0118] refer to Figure 6 In step S15, the specific steps are as follows:

[0119] S151: Based on multiple key power control factors and the current power supply task of the distribution network, determine multiple power supply combinations, and determine the dynamic optimization projects of the distribution network according to the combination content of multiple power supply combinations, the corresponding combination priority and the digital twin model of each power consumption area.

[0120] S152: Based on the tracing of each power consumption area, determine the peak power consumption event of each power consumption area, and determine the first autonomous control parameter according to the peak power consumption event of each power consumption area and the dynamic optimization project;

[0121] S153: Collect multiple load data from the distribution network, determine the load status of the distribution network based on the multiple load data, determine the second autonomous control parameter based on the load status of the distribution network and the dynamic optimization project, and determine the autonomous control system of the distribution network based on the mapping relationship table of the first autonomous control parameter, the second autonomous control parameter and the autonomous control system.

[0122] In the embodiments of this application, multiple power supply combinations are determined based on multiple key power control factors and the current power supply task of the distribution network. The dynamic optimization project of the distribution network is determined according to the combination content of the multiple power supply combinations, the corresponding combination priority and the digital twin model of each power consumption area. This approach takes into account the overall consideration of the combination content of the multiple power supply combinations, the corresponding combination priority and the digital twin model of each power consumption area, thus ensuring the accuracy of the dynamic optimization project of the distribution network.

[0123] At this point, the system uses the "key power control factors" (such as photovoltaic reverse power, line impedance, and harmonic distortion rate) extracted in step S142 as state variables and logically couples them with the current "power supply tasks" of the distribution network (such as new energy consumption, routine power supply, and economic operation); combination construction: the system constructs a multi-dimensional state matrix and generates multiple "power supply combinations"; each combination represents a set of operating characteristics of the power grid under a specific task.

[0124] For example: Combination A = [PV output: high] + [current task: renewable energy consumption] + [load level: low]; Combination B = [line impedance: high] + [current task: routine supply guarantee] + [load level: peak]. These combinations cover various typical operating conditions that the distribution network may encounter, providing input samples for subsequent strategy matching.

[0125] Based on safety and stability principles and power quality requirements, the system conducts a risk assessment on the generated "power supply combinations" and determines the "combination priority." Combinations involving safety risks such as voltage overruns and equipment overloads are assigned the highest priority, while combinations involving energy efficiency optimization have the next lowest priority. The system calls upon the "digital twin models of each power consumption area" and inputs the high-priority power supply combinations into the models for simulation. It evaluates which control strategy can restore the grid state to the optimal range with the least cost (such as minimal energy storage charging and discharging, and minimal equipment actions). Based on the simulation results, the system finally determines one or more "dynamic optimization projects." These projects are not just simple control commands, but complete optimization schemes that include objective functions, constraints, and execution strategies.

[0126] Specifically, for the situation in a rural mountainous area of ​​Province A where the 10kV #5 transformer station is currently at noon and the power supply task is "new energy consumption", the R&D personnel executed step S151 and constructed two power supply combinations based on the identified key factors: Combination I is a high-risk combination of "extremely high photovoltaic output (>250kW) + large line impedance voltage drop (>20V) + new energy consumption", and Combination II is a safe combination of "low photovoltaic output (<100kW) + small line impedance voltage drop (<10V) + new energy consumption".

[0127] The system determined that combination I had an extremely high risk of voltage exceeding limits and assigned it the highest priority. It then input this combination into a digital twin model for simulation. The simulation showed that maintaining a fixed voltage reference value or only adjusting the fixed voltage drop would lead to voltage exceeding limits at the end or photovoltaic shutdown. After optimization, the model calculated that if the UPQC voltage reference value could be dynamically fine-tuned as photovoltaic output increased and dynamic reactive power compensation could be implemented when the voltage was about to exceed limits, the voltage could be kept within acceptable limits. Based on this simulation conclusion, the system finally determined the "UPQC Adaptive Voltage Control and Reactive Power Co-optimization Project Based on Photovoltaic Penetration Rate," which includes a coordinated strategy of dynamically adjusting the target voltage curve according to real-time photovoltaic output and linking parallel-side units for capacitive reactive power absorption, achieving refined dynamic optimization.

[0128] Furthermore, the peak power consumption events of each power consumption area are determined based on the tracing of each power consumption area. The first autonomous control parameter is determined based on the peak power consumption events and dynamic optimization projects of each power consumption area, which takes into account the overall consideration of the peak power consumption events and dynamic optimization projects of each power consumption area, and ensures the accuracy of the first autonomous control parameter.

[0129] At this point, the system initiates a time-series backtracking mechanism for historical operating data of each power-consuming area. In the historical database, the system searches for extreme points that occurred in each area within a specific period in the past (such as a year or a quarter). This includes not only peak active power, but also peak voltage exceeding limits, peak harmonic distortion rate, peak reactive power surge, etc.

[0130] When an extreme point is captured, the system combines the current operating conditions (such as weather, temperature, and power generation) to encapsulate the extreme point into a specific "peak power consumption event"; for example, "photovoltaic full-power overvoltage event during the summer high temperature period" or "heating load low voltage event during the year-end cold wave".

[0131] The system applies the "dynamic optimization project" (i.e., the optimization strategy objective determined in S151) to the virtual simulation of the "peak power consumption event" to test whether the optimization project can effectively cope if the event occurs again. In order to ensure that the system can withstand the impact of historical peak events, the system reverse calculates the required equipment capacity boundary and safety margin to determine the "first autonomous control parameter". These parameters are usually "static thresholds" or "hard constraints", including: maximum adjustment depth limit, action threshold setting, protection setting value, equipment capacity limit, etc. At the same time, the first autonomous control parameter must cover the worst case of historical peak events and reserve an appropriate safety margin (e.g., requiring the adjustment capacity to be greater than 1.2 times the historical maximum deviation) to prevent control failure when similar events occur again.

[0132] Specifically, historical extreme value data is used to set basic safety parameters for the "UPQC Adaptive Voltage Control and Reactive Power Co-optimization Project Based on Photovoltaic Penetration Rate"; the system traces historical data of the "end-point sensitive control area". After scanning the operation logs of the past year, it was found that an extreme voltage rise occurred at 12:45 noon on June 21 of the previous year. At that time, the photovoltaic penetration rate soared to 28%, and the measured voltage at the end of the line reached 242V and lasted for 15 minutes. The system confirmed this record as an "extreme overvoltage event at full photovoltaic power generation in summer".

[0133] Combining the aforementioned peak events and the "adaptive voltage control" optimization project, the system conducted simulation verification in a digital twin model. It was found that if the upper limit of UPQC regulation is set to only -15V, although the voltage can be reduced to the acceptable range, the regulation capability is exhausted. In order to not only solve the peak of 242V but also leave a dynamic margin to cope with instantaneous fluctuations, the system calculated that the required regulation depth should reach 27V. Therefore, the system determined the first autonomous control parameters as follows: the maximum regulation depth on the UPQC series side is set to -30V, the overvoltage protection action threshold is set to 238V, and the device capacity safety margin coefficient is set to 1.2 to ensure that the system can still operate stably when reproducing the worst historical conditions.

[0134] Therefore, by collecting multiple load data from the distribution network, determining the load status of the distribution network based on the multiple load data, determining the second autonomous control parameter based on the load status and dynamic optimization projects of the distribution network, and determining the autonomous control system of the distribution network based on the mapping relationship table of the first autonomous control parameter, the second autonomous control parameter, and the autonomous control system, the autonomous control system of the distribution network is determined. This approach takes into account the overall consideration of the first autonomous control parameter, the second autonomous control parameter, and the mapping relationship table of the autonomous control system, ensuring the accuracy of the autonomous control system of the distribution network. At the same time, by controlling the dynamic optimization projects of the distribution network, the system achieves an overall consideration of the dynamic optimization projects of the distribution network, the peak power consumption events of each power consumption area, and the load status of the distribution network, thereby improving the accuracy of the autonomous control system of the distribution network.

[0135] At this time, the system collects current electrical quantity data at high frequency through intelligent sensing devices deployed in the distribution area and key nodes. Data dimensions: The collected "load data" includes real-time active power, reactive power, current amplitude, power factor, and load change rate, etc. The system uses pattern recognition algorithms or preset rule bases to perform real-time analysis on the collected multi-dimensional data to determine the current "load status" of the distribution network. The status is usually divided into typical types, such as: "heavy load status", "light load status", "severe fluctuation status", "three-phase imbalance status" or "renewable energy backfeed status".

[0136] Unlike the "first autonomous control parameter" (a static safety limit based on historical extreme values) determined in S152, the "second autonomous control parameter" is a dynamic adjustment parameter for the current specific operating state. The system combines the "dynamic optimization project" (i.e. optimization strategy) determined in S151 and the current "load state" to calculate real-time control commands or correction coefficients. These parameters determine the intensity, rate, and specific target value of the control action. For example, if the load state is "severely fluctuating", the second parameter will include "faster PI controller gain" or "larger damping coefficient". If the optimization project is voltage optimization, the second parameter will include "the current target voltage setpoint" (this value will be finely adjusted with load fluctuations, rather than being a fixed value).

[0137] The system simultaneously inputs the "first autonomous control parameter" (safety baseline) and the "second autonomous control parameter" (current command) into a pre-configured "mapping table of autonomous control system"; the mapping table is a multi-dimensional decision matrix used to determine the current control level of the system.

[0138] Level 1 (millisecond level): When the second parameter is within the safe range defined by the first parameter, the system activates Level 1 adaptive optimization, and the device performs real-time optimal control locally; Level 2 (hourly / daily level): When the second parameter approaches the warning threshold of the first parameter, the system triggers Level 2 warning, generates a work order, and reports it; Level 3 (weekly / monthly level): When the second parameter frequently touches the limit of the first parameter, indicating that the optimization project can no longer meet the requirements, the system activates Level 3 proactive intervention, generates technical transformation suggestions, and through the above logic, the system finally establishes a complete "distribution network autonomous control system" including the strategy layer, execution layer, and warning layer, realizing comprehensive coverage from real-time control to operation and maintenance closed loop.

[0139] Specifically, step S153 is executed to construct the final autonomous control system; the system monitors the distribution network in real time, collecting data such as photovoltaic reverse power of 280kW, end load current fluctuation rate of 15% / min, and measured end voltage of 233V, and determines the current load state as "photovoltaic reverse power heavy load accompanied by fluctuation state"; combined with this fluctuation state and the "adaptive voltage control" project, the system calculates the second autonomous control parameters: dynamic target voltage correction value of -13V, dynamic response gain coefficient set to 1.5 (to speed up response speed), and reactive power injection depth set to 80%.

[0140] The system inputs these dynamic parameters into a mapping table with the first autonomous control parameters (maximum depth -30V, threshold 238V) set in stage S152. It determines that the current adjustment demand has not exceeded the safety limit and has not triggered the threshold, thus it is in the "L1 level millisecond-level adaptive optimization" range. The system activates the L1 level strategy and directly sends instructions to UPQC to perform voltage suppression with a gain of 1.5 times. The target voltage is set to 220V. At the same time, it monitors the trend. Once the adjustment demand approaches the limit, it will automatically switch to the L2 level early warning mode, thereby realizing precise, layered, and autonomous control of power quality under different operating conditions.

[0141] Please see Figure 7 , Figure 7 This is a schematic diagram of the structural composition of a power control system for a power distribution network based on different operating conditions in an embodiment of the present invention; the power control system for a power distribution network based on different operating conditions is applied to the above-mentioned power control method for a power distribution network based on different operating conditions; the power control system for a power distribution network based on different operating conditions includes:

[0142] The digital twin module 21 is used to determine the power consumption conditions of each power consumption area based on the power supply data set of the distribution network to the target area and multiple power consumption areas in the target area, and to determine the digital twin model of the power consumption area based on the power consumption conditions of each power consumption area and the corresponding real-time power consumption data.

[0143] The power supply anomaly event module 22 is used to determine the power supply curve of the distribution network based on the detection of the digital twin model of each power consumption area, determine multiple abnormal power supply nodes based on the power supply curve of the distribution network, the regional location and time of the corresponding power consumption area, and determine the power supply anomaly event corresponding to the power consumption area based on the multiple abnormal power supply nodes and the power consumption peak of the power consumption area.

[0144] The power control module 23 is used to determine five power control areas based on each power supply anomaly event, the corresponding abnormal power supply node, and the real-time status of the distribution network. Based on the five power control areas, the virtual operating scenario of the distribution network, and previous power control events, the corresponding power control system is determined, and the power control events of the distribution network under different operating conditions are marked.

[0145] The key power control factor module 24 is used to determine the power control events of the distribution network under different operating conditions based on the identification of the power control system, and to determine multiple key power control factors based on the identification of each power control event.

[0146] The autonomous control module 25 is used to determine the dynamic optimization projects of the distribution network based on multiple key power control factors and digital twin models of each power consumption area; and to determine the autonomous control system of the distribution network based on the dynamic optimization projects of the distribution network, the peak power consumption events of each power consumption area and the load status of the distribution network.

[0147] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A power control method for a power distribution network based on different operating conditions, characterized in that, include: Based on the power supply data set of the distribution network to the target area and the multiple power consumption areas of the target area, determine the power consumption conditions of each power consumption area, and determine the digital twin model of the power consumption area based on the power consumption conditions of each power consumption area and the corresponding real-time power consumption data. The power supply curve of the distribution network is determined by the detection of the digital twin model of each power consumption area. Multiple abnormal power supply nodes are determined according to the power supply curve of the distribution network, the regional location and time of the corresponding power consumption area. The power supply abnormal event corresponding to the power consumption area is determined according to the multiple abnormal power supply nodes and the power consumption peak of the power consumption area. Five power control zones are determined based on each power supply anomaly event, the corresponding abnormal power supply node, and the real-time status of the distribution network. Based on the five power control zones, the virtual operating scenario of the distribution network, and previous power control events, the corresponding power control system is determined, and the power control events of the distribution network under different operating conditions are marked. Based on the identification of this power control system, power control events of the distribution network under different operating conditions are determined, and multiple key power control factors are determined based on the identification of each power control event. Dynamic optimization projects for the distribution network are determined based on multiple key power control factors and digital twin models of various power consumption areas. The autonomous control system of the distribution network is determined based on the dynamic optimization project of the distribution network, the peak power consumption events of each power consumption area, and the load status of the distribution network. The process involves determining the corresponding power control system based on five power control zones, virtual operating scenarios of the distribution network, and past power control events, and marking power control events of the distribution network under different operating conditions. This includes: identifying multiple virtual operating condition data based on the identification of past operating condition events of the distribution network; determining corresponding operating scenarios based on the multiple virtual operating condition data to mark the virtual operating scenarios of the distribution network; determining a power control framework based on the virtual operating scenarios of the distribution network and the five power control zones; determining the corresponding power control system based on the power control framework and past power control events; and determining power control events of the distribution network under different operating conditions by tracing the virtual operating scenarios of the distribution network. The virtual operating scenarios of the distribution network are considered as dynamically virtual scenarios. The five power control zones include a source-end stability control zone covering transformer outlets and near-end users, a mid-section main control zone near the UPQC installation point, a terminal sensitive control zone covering the terminal photovoltaic area and its surroundings, a photovoltaic source control zone for high-density photovoltaic access branch lines, and a conventional load buffer zone located in the middle section of the main line.

2. The power control method for a distribution network based on different operating conditions according to claim 1, characterized in that, The process of determining the power consumption conditions of each power consumption area based on the power supply data set of the distribution network to the target area and multiple power consumption areas in the target area, and determining the digital twin model of each power consumption area based on the power consumption conditions of each power consumption area and the corresponding real-time power consumption data, includes: The system collects power supply tasks from the distribution network, identifies the corresponding target area based on the identification of the power supply task, and collects multiple power supply data from the distribution network to the target area; based on the multiple power supply data, the corresponding time, and the power supply duration of the distribution network, the system determines the power supply data set. Data is collected from the target area, and power consumption locations are determined based on the target area and corresponding power consumption markers. Based on the tracing of each power consumption location, the corresponding power consumption areas are determined to mark multiple power consumption areas in the target area. Based on the power supply data set of the distribution network to the target area, the multiple power consumption areas in the target area, and the corresponding power consumption status, the power consumption conditions of each power consumption area are determined. At the same time, real-time power consumption data of each power consumption area is collected, and a digital twin model of the power consumption area is constructed based on the power consumption conditions of each power consumption area and the corresponding real-time power consumption data.

3. The power control method for a distribution network based on different operating conditions according to claim 1, characterized in that, The power supply curve of the distribution network is determined by detecting digital twin models of each power consumption area. Multiple abnormal power supply nodes are identified based on the power supply curve of the distribution network, the location of the corresponding power consumption area, and the time. Power supply anomaly events corresponding to the power consumption area are determined based on the multiple abnormal power supply nodes and the power consumption peak value of the power consumption area, including: In the digital twin model of each power consumption area, the digital twin model is dynamically tested, and multiple key power supply data of the distribution network are determined during the testing process. Based on the multiple key power supply data, the corresponding time and the load parameters of the distribution network, the power supply curve of the distribution network is constructed. Based on the identification of the power supply curve, multiple power supply waveform regions are determined. Based on the identification of each power supply waveform region, multiple waveform abnormal locations are determined. Based on the multiple waveform abnormal locations and the corresponding power consumption regions, a first-level abnormal node combination is determined. Based on the multiple waveform abnormal locations and time, a second-level abnormal node combination is determined. Based on the matching of the first-level abnormal node combination and the second-level abnormal node combination, multiple abnormal power supply nodes are determined. Real-time monitoring of each power consumption area and marking the peak power consumption of each power consumption area. Based on the peak power consumption of multiple abnormal power supply nodes and power consumption areas and the corresponding area location, determine each power supply anomaly factor. Based on each power supply anomaly factor, the area shape of the power consumption area and the corresponding power consumption data, determine the power supply anomaly event corresponding to the power consumption area.

4. The power control method for a distribution network based on different operating conditions according to claim 1, characterized in that, The process involves determining five power control zones based on various power supply anomaly events, corresponding abnormal power supply nodes, and the real-time status of the distribution network. Based on these five power control zones, the virtual operating scenario of the distribution network, and past power control events, a corresponding power control system is established. Power control events of the distribution network under different operating conditions are then marked, including: Collect multiple status data of the distribution network, determine the real-time status of the distribution network based on the multiple status data and the corresponding time, and determine the first power control position based on the real-time status of the distribution network and various power supply anomaly events. The second power control location is determined based on the real-time status of the distribution network and the abnormal power supply nodes corresponding to each power supply anomaly event. Five power control areas are determined based on the first power control location, the second power control location, and the distribution network distribution map.

5. The power control method for a distribution network based on different operating conditions according to claim 1, characterized in that, The power control system identifies power control events in the distribution network under different operating conditions, and determines multiple key power control factors based on the identification of each power control event, including: The power control system is monitored in real time, dynamically identified, and multiple power control contents are determined during the identification process. The corresponding power control areas are marked, and the power control events of the distribution network under different operating conditions are determined based on each power control content, the corresponding power control area, and the virtual operating scenario of the distribution network.

6. The power control method for a distribution network based on different operating conditions according to claim 5, characterized in that, The method of identifying power control events in the distribution network under different operating conditions based on the power control system, and determining multiple key power control factors based on the identification of each power control event, also includes: In this power control event, multiple power control items are identified based on the identification of the power control event. Corresponding power control factors are determined according to the project content and corresponding control scope of each power control item. Multiple key power control factors are determined according to the types, corresponding priorities and current operating conditions of the power distribution network of the multiple power control factors.

7. The power control method for a distribution network based on different operating conditions according to claim 1, characterized in that, The dynamic optimization project of the distribution network is determined based on multiple key power control factors and digital twin models of various power consumption areas. Based on the dynamic optimization project of the distribution network, the peak power consumption events of each power consumption area, and the load status of the distribution network, an autonomous control system for the distribution network is determined, including: Based on multiple key power control factors and the current power supply tasks of the distribution network, multiple power supply combinations are determined. Based on the combination content of multiple power supply combinations, the corresponding combination priority, and the digital twin model of each power consumption area, the dynamic optimization project of the distribution network is determined.

8. The power control method for a distribution network based on different operating conditions according to claim 7, characterized in that, The dynamic optimization project of the distribution network is determined based on multiple key power control factors and digital twin models of various power consumption areas. Based on the dynamic optimization project of the distribution network, the peak power consumption events of each power consumption area, and the load status of the distribution network, an autonomous control system for the distribution network is determined, which also includes: The peak power consumption events of each power consumption area are determined by tracing the power consumption areas, and the first autonomous control parameter is determined based on the peak power consumption events of each power consumption area and the dynamic optimization project. Collect multiple load data from the distribution network, determine the load status of the distribution network based on the multiple load data, determine the second autonomous control parameter based on the load status of the distribution network and dynamic optimization projects, and determine the autonomous control system of the distribution network based on the mapping relationship table of the first autonomous control parameter, the second autonomous control parameter and the autonomous control system.

9. A power control system for a power distribution network based on different operating conditions, characterized in that, The power control system for the power distribution network based on different operating conditions is applied to the power control method for the power distribution network based on different operating conditions as described in any one of claims 1-8; The power control system for the power distribution network based on different operating conditions includes: The digital twin module is used to determine the power consumption conditions of each power consumption area based on the power supply data set of the distribution network to the target area and multiple power consumption areas in the target area, and to determine the digital twin model of the power consumption area based on the power consumption conditions of each power consumption area and the corresponding real-time power consumption data. The power supply anomaly event module is used to determine the power supply curve of the distribution network based on the detection of digital twin models of each power consumption area. Based on the power supply curve of the distribution network, the location and time of the corresponding power consumption area, multiple abnormal power supply nodes are determined. Based on the power consumption peak of the multiple abnormal power supply nodes and the power consumption area, the power supply anomaly event corresponding to the power consumption area is determined. The power control module is used to determine five power control zones based on various power supply anomaly events, corresponding abnormal power supply nodes, and the real-time status of the distribution network. Based on the five power control zones, the virtual operating scenario of the distribution network, and previous power control events, the corresponding power control system is determined, and power control events of the distribution network under different operating conditions are marked. The key power control factor module is used to determine the power control events of the distribution network under different operating conditions based on the identification of the power control system, and to determine multiple key power control factors based on the identification of each power control event. The autonomous control module is used to determine the dynamic optimization projects of the distribution network based on multiple key power control factors and digital twin models of each power consumption area; and to determine the autonomous control system of the distribution network based on the dynamic optimization projects of the distribution network, the peak power consumption events of each power consumption area and the load status of the distribution network.

Images