Low-voltage power failure range analysis method and system based on machine learning

Abstract

The invention relates to the technical field of power supply systems, and particularly discloses a low-voltage power failure range analysis method and system based on machine learning, and the method comprises the steps: collecting equipment operation parameters of a low-voltage power supply region, calculating an equipment state abnormality index and a power failure abnormality index, carrying out the dynamic adjustment and power early warning, and generating an adjustment scheme of the low-voltage power supply region, through dynamic adjustment of four monitoring periods, risk hierarchical management and control are realized, power failure abnormal indexes and multi-dimensional features are input into a machine learning model, and through fault propagation path identification and power grid topology mapping, a power failure geographical range is delimited and influence indexes are quantified. A command strategy including fault positioning, material scheduling and path planning is automatically generated based on the emergency level, and the risk prevention and control timeliness and the emergency repair resource scheduling accuracy of the low-voltage distribution network are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power supply systems, and in particular to a method and system for analyzing the range of low-voltage power outages based on machine learning. Background Art

[0002] The power supply reliability of low-voltage distribution networks directly affects the order of social production and life. Existing technologies divide power outage areas through the fixed topology of the distribution automation system, trigger alarms through manual inspections or fault indicators, and initiate scope analysis only after the actual power outage occurs. The existing regulation mechanism is mainly based on electrical quantity monitoring.

[0003] For example, Chinese invention patent publication number CN117578422A discloses a method for analyzing power outage areas in a distribution network. The method involves obtaining distribution network data, which includes static and dynamic data; obtaining a distribution network area with a power outage based on the dynamic data; obtaining a power network topology model based on the static data; and converting the power network topology model into a graph model. The method determines the distribution network area with a power outage and the power network topology model containing accompanying outage equipment based on the distribution network data. The method then converts the power network topology model into a graph model, identifies the graph model using a breadth-first algorithm, and determines the specific power outage area based on the identification results and the distribution network area with a power outage.

[0004] For example, a Chinese invention patent with announcement number CN115347570B discloses a method for analyzing the scope of regional power outages based on main and distribution network coordination. The analysis method includes: obtaining the operating information of the power grid, and performing an N-1 risk scan on the power grid, and assessing the event risk level based on the scan results; judging whether the main equipment is in a risky state based on the event risk level, and if so, conducting a regional power outage scope analysis on the main equipment to obtain the analysis results; the main and distribution network coordination service system receives the analysis results, and calls the model verification function to splice the main network equipment model and the distribution network equipment model to determine the correctness of the analysis results.

[0005] The above technologies have at least the following technical problems: the existing technologies have weak dynamic tracking and prediction capabilities for power outage processes, making it difficult to meet rapid response requirements; the existing technologies overly rely on static data, single-point failure logic, solidified topology models, and lack multi-dimensional correlation analysis. The existing technologies use breadth-first algorithms or model splicing verification, which have high computational complexity and are difficult to meet the real-time analysis needs of large-scale distribution networks. Summary of the Invention

[0006] In response to the deficiencies of the prior art, the present invention provides a low-voltage power outage range analysis method and system based on machine learning, which can effectively solve the problems involved in the above-mentioned background technology.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: The first aspect of the present invention provides a low-voltage power outage range analysis method based on machine learning, including: step one, analyzing the operating status of the low-voltage power supply area through a machine learning model, collecting the power operating parameters of the low-voltage power supply area, and obtaining the low-voltage power outage analysis results of the low-voltage power supply area by analyzing the power operating parameters of the low-voltage power supply area, and generating a power regulation plan for the power supply area; step two, adjusting the power parameters of the power supply area through the power regulation plan of the power supply area, monitoring the adjustment process parameters of the power supply area, and determining whether to issue a power warning to the power supply area; step three, analyzing the power outage range of the low-voltage power supply area based on the low-voltage power outage analysis results of the low-voltage power supply area and the machine learning model, generating a power outage emergency command strategy based on the power outage range of the low-voltage power supply area, and performing command feedback.

[0008] The second aspect of the present invention provides a low-voltage power outage range analysis system based on machine learning, including: an equipment operation status analysis module, which is used to collect the power operation parameters of the low-voltage power supply area through a machine learning model, and obtain the low-voltage power outage analysis results of the low-voltage power supply area by analyzing the power operation parameters of the low-voltage power supply area, and generate a power regulation plan for the power supply area; a power early warning module, which is used to adjust the power parameters of the power supply area through the power regulation plan of the power supply area, monitor the adjustment process parameters of the power supply area, and thus determine whether to issue a power early warning for the power supply area; a low-voltage power outage analysis module, which is used to analyze the power outage range of the low-voltage power supply area based on the low-voltage power outage analysis results of the low-voltage power supply area and the machine learning model, generate a power outage emergency command strategy based on the power outage range of the low-voltage power supply area, and provide command feedback.

[0009] Compared with the prior art, the present invention has at least the following advantages or beneficial effects: (1) The present invention provides a low-voltage power outage range analysis method and system based on machine learning, realizing a full-process closed-loop design of "data collection-analysis-adjustment-warning-range analysis-strategy generation", and realizing dynamic calibration of parameters through four iterative optimization adjustments, significantly improving the adjustment accuracy. The physical state parameters are integrated with the machine learning model, and the power outage characteristic parameters are combined to construct the equipment state abnormality index and the power outage abnormality index through the weighted fusion algorithm, breaking through the limitations of "one-sided data and lack of causality"; according to the power outage range, the number of users affected, the line length and user distribution data are quantified, and differentiated emergency instructions are generated, transforming the low-voltage power outage analysis from "passive result tracing" to "active risk prediction"; according to the adjustment effect of four adjustment cycles, the first to third level power warning is determined, and it is upgraded from "experience decision-making" to "data-driven", significantly improving the power supply reliability and fault response efficiency of the distribution network, and providing a systematic solution for the intelligent operation and maintenance of the low-voltage distribution network.

[0010] (2) The present invention innovatively couples the power regulation scheme with the power early warning mechanism: after optimizing the power parameters according to the power regulation scheme of the power supply area, the regulation process parameters of the power supply area are monitored, the regulation effect is accurately analyzed, and it is determined whether the power early warning is triggered. This breaks through the limitation of traditional power early warning that only relies on thresholds, realizes active regulation of low-voltage power supply areas, and effectively solves the core defects of the existing technology of one-sided causal analysis and delayed regulation.

[0011] (3) The present invention relies on the power outage analysis results and machine learning models to deeply integrate the power grid topology, real-time fault data and historical handling cases to accurately define the scope of power outage impact; on this basis, the system automatically generates an emergency command strategy including fault location, affected user list, optimal path planning and material dispatch suggestions, and feeds back the command to the emergency repair terminal through instructions, providing scientific, accurate and operational strategic recommendations for power outage command and emergency handling. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The present invention is further described with reference to the accompanying drawings. However, the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative effort.

[0013] Figure 1 Schematic diagram of the method steps of the present invention; Figure 2 This is a schematic diagram of system module connections of the present invention; Figure 3 It is a schematic diagram of the low-voltage power outage analysis and adjustment solution flow of the present invention; Figure 4 This is a schematic diagram of the power parameter adjustment and early warning determination process of the present invention; Figure 5 This is a flowchart of the power outage scope analysis and emergency strategy generation process of the present invention. DETAILED DESCRIPTION

[0014] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0015] Reference Figure 1As shown, the first aspect of the present invention provides a low-voltage power outage range analysis method based on machine learning, including: step one, analyzing the operating status of the low-voltage power supply area through a machine learning model, collecting the power operating parameters of the low-voltage power supply area, and obtaining the low-voltage power outage analysis results of the low-voltage power supply area by analyzing the power operating parameters of the low-voltage power supply area, and generating a power regulation plan for the power supply area; step two, adjusting the power parameters of the power supply area through the power regulation plan of the power supply area, monitoring the adjustment process parameters of the power supply area, and determining whether to perform a power warning for the power supply area; step three, analyzing the power outage range of the low-voltage power supply area based on the low-voltage power outage analysis results of the low-voltage power supply area and the machine learning model, generating a power outage emergency command strategy based on the power outage range of the low-voltage power supply area, and performing command feedback.

[0016] Specifically, the power regulation scheme of the power supply area is generated in the following way: the power regulation scheme of the power supply area includes power regulation strategy 1 and power regulation strategy 2; power regulation strategy 1 refers to matching the inverter modulation ratio increase and reactive compensation increase based on the equipment status abnormality index, thereby increasing the inverter modulation ratio and reactive compensation amount of the low-voltage power supply area. Specifically, the current inverter modulation ratio is added to the inverter modulation ratio increase, and the result is the inverter modulation ratio of the low-voltage power supply area after the increase; the current reactive compensation amount is added to the reactive compensation increase, and the result is the completed The reactive compensation amount of the low-voltage power supply area after the increase; the second power regulation strategy refers to matching the increase in regional power supply and the increase in adjustable load shedding ratio in the low-voltage power supply area based on the equipment status abnormality index, thereby increasing the regional power supply and the adjustable load shedding ratio in the low-voltage power supply area; specifically, the current regional power supply is added to the increase in regional power supply, and the result is the regional power supply of the low-voltage power supply area after the increase; the current adjustable load shedding ratio is added to the increase in adjustable load shedding ratio, and the result is the adjustable load shedding ratio of the low-voltage power supply area after the increase.

[0017] The database stores a mapping table of equipment status abnormality index-inverter modulation ratio increase, a mapping table of equipment status abnormality index-reactive compensation increase, a mapping table of equipment status abnormality index-regional power supply increase in low-voltage power supply area, and a mapping table of equipment status abnormality index-adjustable load shedding ratio increase. By directly querying the equipment status abnormality index in the database, the inverter modulation ratio increase corresponding to the inverter modulation ratio, the reactive compensation increase corresponding to the reactive compensation amount, the regional power supply increase in the low-voltage power supply area corresponding to the regional power supply amount of the low-voltage power supply area, and the adjustable load shedding ratio increase corresponding to the adjustable load shedding ratio can be obtained; by increasing and adjusting the inverter modulation ratio, reactive compensation amount, regional power supply in the low-voltage power supply area, and adjustable load shedding ratio, the equipment operation status in the low-voltage power supply area during the monitoring period is improved.

[0018] Furthermore, the low-voltage power outage analysis results of the low-voltage power supply area specifically include: the power outage anomaly index of the low-voltage power supply area during the monitoring period and the equipment status anomaly index of the low-voltage power supply area during the monitoring period. In the low-voltage distribution network, the power outage anomaly index can reflect the regional power outage probability, and the equipment status anomaly index can reflect the health of the equipment. The power operation parameters of the low-voltage power supply area include the vibration spectrum anomaly in the parameters of the equipment status anomaly index of the low-voltage power supply area. The power regulation plan of the power supply area is to match the increase in the power parameters according to the equipment status anomaly index. The power outage anomaly index can match the increase coefficient of the equipment status anomaly index to increase the adjustment, thereby achieving an increase adjustment of the equipment status anomaly index, thereby achieving further increase adjustment of the power parameters. There is a close causal relationship between the two, and together constitute the core logic chain of "equipment status-risk warning-regulation decision"; obtain the vibration spectrum anomaly, partial discharge pulse rising edge slope and line load rate of the equipment belonging to the low-voltage power supply area during the monitoring period.

[0019] The vibration spectrum abnormality quantifies the proportion of abnormal harmonic energy in the mechanical vibration signals of equipment (such as transformers) within the low-voltage power supply area, reflecting the degree of wear, looseness, or defects in the equipment's mechanical components. Specifically, the mechanical vibration signal of the equipment is collected, decomposed into the amplitude spectrum of each frequency component through a fast Fourier transform, and the energy of the abnormal characteristic frequency band is extracted compared with the average energy of the frequency band in a healthy state. The abnormality is calculated using a baseline comparison method and then multiplied by a normalization factor calibrated by technical personnel to obtain the vibration spectrum abnormality. The partial discharge pulse rising edge slope reflects the severity of insulation defects (such as air gap discharge or surface discharge) in equipment (such as cable terminals) within the low-voltage power supply area. The partial discharge pulse signal is collected from the equipment, the pulse rising edge is extracted, and its rising slope is calculated to obtain the partial discharge pulse rising edge slope. The line load rate is the ratio of the actual operating power of the line to the rated power, indicating the degree of overload of the distribution line within the low-voltage power supply area. The three-phase current and voltage of the line are collected in real time, and the three-phase average current and average voltage are calculated. The actual operating power of the line is calculated and the line load rate is calculated by dividing the actual operating power by the rated power.

[0020] Weighted proportional values ​​are introduced from the database to quantify the degree of influence of the proportional relationship between the vibration spectrum abnormality and the defined vibration spectrum abnormality, the proportional relationship between the rising edge slope of the partial discharge pulse and the defined rising edge slope of the partial discharge pulse, and the proportional relationship between the line load rate and the defined line load rate on the equipment status abnormality index of the low-voltage power supply area during the monitoring period. The various influence degrees are summarized to obtain the equipment status abnormality index of the low-voltage power supply area during the monitoring period.

[0021] The device status abnormality index in the low-voltage power supply area during the monitoring period is used to digitally represent the degree of abnormality of the device status in the low-voltage power supply area during the monitoring period. The specific expression is: ; Wherein, EI is the equipment status abnormality index of the low-voltage power supply area during the monitoring period, EA is the vibration spectrum abnormality of the equipment belonging to the low-voltage power supply area during the monitoring period, ES is the rising edge slope of the partial discharge pulse of the equipment belonging to the low-voltage power supply area during the monitoring period, EL is the line load rate of the equipment belonging to the low-voltage power supply area during the monitoring period, E_EA is the defined vibration spectrum abnormality preset in the database, E_ES is the defined partial discharge pulse rising edge slope preset in the database, E_EL is the defined line load rate preset in the database, ea1 is the weighted proportion value of the vibration spectrum abnormality preset in the database, ea2 is the weighted proportion value of the partial discharge pulse rising edge slope preset in the database, and ea3 is the weighted proportion value of the line load rate preset in the database.

[0022] Among them, the definition of vibration spectrum abnormality is the maximum value allowed by the vibration spectrum abnormality; the definition of partial discharge pulse rising edge slope is the maximum value allowed by the partial discharge pulse rising edge slope; the definition of line load rate is the maximum value allowed by the line load rate.

[0023] Higher vibration spectrum anomaly levels and a greater proportion of abnormal energy indicate more severe wear and loosening of the equipment's mechanical components, directly leading to increased vibration and noise, accelerated aging and damage of components like bearings and cores, and a corresponding increase in vibration spectrum anomaly levels, indicating a more severe level of equipment operational anomaly. A greater slope of the PD pulse's rising edge indicates more severe insulation defects (such as air gaps or cracks) and higher discharge energy, leading to decomposition or carbonization of the insulation material, further exacerbating insulation degradation. This significantly increases the vibration spectrum anomaly level. Higher line load rates increase the heat generated by equipment (such as transformers or conductors). This temperature increase accelerates insulation aging (for example, every 10°C increase in transformer oil temperature halves the insulation life). Overload can also increase voltage drop, impacting the equipment's power supply quality. The vibration spectrum anomaly level increases with increasing load rate. The vibration spectrum anomaly level, the PD pulse's rising edge slope, and the line load rate are all positively correlated with the equipment status anomaly index.

[0024] The weighted ratio value of the vibration spectrum abnormality quantifies the degree of influence of the proportional relationship between the vibration spectrum abnormality and the defined vibration spectrum abnormality on the equipment status abnormality index; the weighted ratio value of the partial discharge pulse rising edge slope quantifies the degree of influence of the proportional relationship between the partial discharge pulse rising edge slope and the defined partial discharge pulse rising edge slope on the equipment status abnormality index; the weighted ratio value of the line load rate quantifies the degree of influence of the proportional relationship between the line load rate and the defined line load rate on the equipment status abnormality index; the database stores the mapping relationship between the vibration spectrum abnormality, the partial discharge pulse rising edge slope, and the line load rate and their corresponding weighted ratio values. For example, by inputting the vibration spectrum abnormality, the partial discharge pulse rising edge slope, and the line load rate into the database, the database can retrieve the weighted ratio value of the vibration spectrum abnormality, the weighted ratio value of the partial discharge pulse rising edge slope, and the weighted ratio value of the line load rate. The value range of the weighted ratio value of the vibration spectrum abnormality, the weighted ratio value of the partial discharge pulse rising edge slope, and the weighted ratio value of the line load rate are all between 0 and 1.

[0025] Specifically, the power outage anomaly index of the low-voltage power supply area during the monitoring period, the specific analysis process is: through the machine learning model, the equipment operating status parameters and power outage characteristic parameters of the low-voltage power supply area during the monitoring period are analyzed to obtain the power operation parameters of the low-voltage power supply area. The power operation parameters of the low-voltage power supply area include the vibration spectrum anomaly degree of the low-voltage power supply area during the monitoring period, the grounding impedance change rate of the grounding grid, and the insulation resistance degradation rate.

[0026] The above-mentioned machine learning model is used to analyze the equipment operating status parameters and power outage characteristic parameters in the low-voltage power supply area within the monitoring period. The specific analysis process is as follows: through training, the model finds that there is a strong statistical correlation between specific equipment status parameter combinations and power outage event characteristics, automatically discovers and models these complex nonlinear relationships and interaction effects between features, and uses the time series model to analyze the time series data of equipment status parameters, effectively capturing the dynamic patterns and long-term dependencies of equipment status parameters evolving over time. Through the built-in feature importance of the model, it quantitatively analyzes which equipment operating status parameters have the greatest impact on the final prediction results, as well as the direction of the impact, to derive the power operating parameters of the low-voltage power supply area.

[0027] The grounding impedance change rate of the grounding grid reflects the downward trend in grounding reliability caused by corrosion, loosening of the grounding body or changes in the soil environment. The grounding impedance is regularly measured by ground resistance testers deployed at key nodes of the grounding grid, and the rate of change of the grounding impedance in the time series is calculated to obtain the grounding impedance change rate of the grounding grid. The insulation resistance degradation rate refers to an indicator of how fast the resistance value of the insulating material decreases over time, reflecting the rate of insulation performance degradation. The insulation resistance value of the equipment is regularly measured, and the measured resistance value and recording time are recorded for each time. The insulation resistance is calculated by the voltage applied to the insulation resistance and the measured current. The insulation resistance is subtracted from the insulation resistance value at the previous measurement time point and divided by the time interval between the two measurements to obtain the insulation resistance degradation rate.

[0028] Weighted proportional values ​​are introduced from the database to quantify the degree of influence of the proportional relationship between the vibration spectrum abnormality of the equipment belonging to the low-voltage power supply area during the monitoring period and the defined equipment vibration spectrum abnormality, the proportional relationship between the grounding impedance change rate of the grounding grid and the defined grounding impedance change rate, and the proportional relationship between the insulation resistance degradation rate and the defined insulation resistance degradation rate on the power outage anomaly index of the low-voltage power supply area during the monitoring period. The various influence degrees are summarized to obtain the power outage anomaly index of the low-voltage power supply area during the monitoring period.

[0029] The power outage anomaly index of the low-voltage power supply area during the monitoring period is used to digitally represent the power outage probability of the low-voltage power supply area. The specific expression is: ; Wherein, PI is the power outage anomaly index of the low-voltage power supply area during the monitoring period, EA is the vibration spectrum anomaly of the equipment in the low-voltage power supply area during the monitoring period, PR is the grounding impedance change rate of the grounding grid of the equipment in the low-voltage power supply area during the monitoring period, PF is the insulation resistance degradation rate of the equipment in the low-voltage power supply area during the monitoring period, E_EA is the defined vibration spectrum anomaly preset in the database, P_PR is the defined grounding impedance change rate of the grounding grid preset in the database, P_PF is the defined insulation resistance degradation rate preset in the database, pa1 is the weighted proportion value of the vibration spectrum anomaly preset in the database, pa2 is the weighted proportion value of the grounding impedance change rate of the grounding grid preset in the database, and pa3 is the weighted proportion value of the insulation resistance degradation rate preset in the database.

[0030] Among them, defining the vibration spectrum abnormality is the upper limit value set for the vibration spectrum abnormality; defining the grounding impedance change rate of the grounding grid is the upper limit value set for the grounding impedance change rate of the grounding grid; defining the insulation resistance degradation rate is the upper limit value set for the insulation resistance degradation rate.

[0031] The higher the vibration spectrum abnormality and the greater the proportion of abnormal energy, the more severe the wear and looseness of the equipment's mechanical components, which will directly lead to intensified equipment vibration and noise, accelerated aging and damage of components such as bearings and iron cores. Mechanical failures may directly cause equipment shutdown (such as a transformer winding short circuit due to a loose iron core) or indirectly cause overload (such as an increase in current due to a stuck bearing in a motor), increasing the probability of power outages. The greater the grounding impedance change rate, the more severe the grounding corrosion or loosening, and the lower the reliability of the grounding system. This may cause equipment insulation breakdown and reduce the equipment's insulation tolerance, thereby increasing the insulation resistance degradation rate. The greater the insulation resistance degradation rate, the higher the probability of equipment failure per unit time. The vibration spectrum abnormality, the grounding impedance change rate of the grounding grid, and the insulation resistance degradation rate are all positively correlated with the power outage abnormality index.

[0032] The weighted proportional value of the vibration spectrum abnormality quantifies the degree of influence of the proportional relationship between the vibration spectrum abnormality and the defined vibration spectrum abnormality on the power outage abnormality index; the weighted proportional value of the grounding grid grounding impedance change rate quantifies the degree of influence of the proportional relationship between the grounding grid grounding impedance change rate and the defined grounding grid grounding impedance change rate on the power outage abnormality index; the weighted proportional value of the insulation resistance degradation rate quantifies the degree of influence of the proportional relationship between the insulation resistance degradation rate and the defined insulation resistance degradation rate on the power outage abnormality index; the database stores the mapping relationship between the vibration spectrum abnormality, the grounding grid grounding impedance change rate, and the insulation resistance degradation rate and their corresponding weighted proportional values. For example, by inputting the vibration spectrum abnormality, the grounding grid grounding impedance change rate, and the insulation resistance degradation rate into the database, the database can retrieve the weighted proportional value of the vibration spectrum abnormality, the weighted proportional value of the grounding grid grounding impedance change rate, and the weighted proportional value of the insulation resistance degradation rate. The value range of the weighted proportional value of the vibration spectrum abnormality, the weighted proportional value of the grounding grid grounding impedance change rate, and the weighted proportional value of the insulation resistance degradation rate are all between 0 and 1.

[0033] Specifically, the power parameters of the power supply area are adjusted through the power regulation plan of the power supply area. The specific adjustment process is: compare the power outage anomaly index of the low-voltage power supply area during the monitoring period with the power outage anomaly threshold stored in the database. If the power outage anomaly index of the low-voltage power supply area during the monitoring period is less than the power outage anomaly threshold, the power operation parameters of the low-voltage power supply area are continuously collected; if the power outage anomaly index of the low-voltage power supply area during the monitoring period is greater than or equal to the power outage anomaly threshold, the power parameters of the power supply area are adjusted once through the power regulation strategy 1 in the power regulation plan of the power supply area; the power parameters include inverter modulation ratio, reactive compensation amount, regional power supply amount and adjustable load removal ratio.

[0034] After one adjustment is completed, the adjustment process parameters of the power supply area are monitored. The adjustment process parameters of the power supply area include the power outage abnormality index of the low-voltage power supply area within one adjustment monitoring cycle and the equipment status abnormality index of the low-voltage power supply area within one adjustment monitoring cycle, which represent the power outage abnormality index and the equipment status abnormality index within the monitoring cycle after one adjustment.

[0035] If the power outage anomaly index of the low-voltage power supply area within an adjustment monitoring cycle is still greater than or equal to the power outage anomaly threshold, then based on the equipment status anomaly index of the low-voltage power supply area within an adjustment monitoring cycle and the equipment status anomaly index of the low-voltage power supply area within the monitoring cycle, the reduction rate of the equipment status anomaly index of the low-voltage power supply area within an adjustment monitoring cycle is obtained, and compared with the defined reduction rate preset in the database; the equipment status anomaly index of the low-voltage power supply area within the monitoring cycle refers to the equipment status anomaly index within the monitoring cycle before one adjustment, and the above-mentioned reduction rate of the equipment status anomaly index of the low-voltage power supply area within an adjustment monitoring cycle is obtained by subtracting the difference between the equipment status anomaly index of the low-voltage power supply area within the monitoring cycle and the equipment status anomaly index within an adjustment monitoring cycle, divided by the equipment status anomaly index of the low-voltage power supply area within the monitoring cycle, and the defined reduction rate preset in the database is the minimum value of the reduction rate of the equipment status anomaly index.

[0036] If the reduction rate of the device status abnormality index of the low-voltage power supply area within an adjustment monitoring cycle is greater than or equal to the defined reduction rate, the power parameters of the power supply area are adjusted secondary, and the device status abnormality index of the low-voltage power supply area within the adjustment monitoring cycle is increased based on the power outage abnormality index of the low-voltage power supply area within the adjustment monitoring cycle. The device status abnormality index of the low-voltage power supply area within the adjustment monitoring cycle after the increase is used to update the power regulation strategy one, and the updated power regulation strategy one is marked as the optimized power regulation strategy one. The inverter modulation ratio and the reactive compensation amount in the power parameters of the power supply area are further increased and adjusted by the optimized power regulation strategy one; the inverter modulation ratio of the low-voltage power supply area is further increased by adding the inverter modulation ratio to the increase in the inverter modulation ratio mapped in the database by the increased device status abnormality index within the adjustment monitoring cycle after the increase, to obtain an optimized inverter modulation ratio; the reactive compensation amount of the low-voltage power supply area is further increased by adding the reactive compensation amount to the increase in the reactive compensation amount mapped in the database by the increased device status abnormality index within the adjustment monitoring cycle after the increase, to obtain an optimized reactive compensation amount.

[0037] The above-mentioned increase and adjustment of the equipment status abnormality index of the low-voltage power supply area within one adjustment monitoring cycle based on the power outage abnormality index of the low-voltage power supply area within one adjustment monitoring cycle refers to matching the equipment status abnormality index increase coefficient within one adjustment monitoring cycle by the power outage abnormality index within one adjustment monitoring cycle. The database stores a power outage abnormality index-equipment status abnormality index increase coefficient mapping table. The power outage abnormality index can be directly queried in the database to obtain the equipment status abnormality index increase coefficient corresponding to the power outage abnormality index. The equipment status abnormality index within one adjustment monitoring cycle is multiplied by the equipment status abnormality index increase coefficient within one adjustment monitoring cycle to obtain the increased and adjusted equipment status abnormality index within one adjustment monitoring cycle. The power outage abnormality index after one adjustment is greater than the power outage threshold, indicating that the power outage probability in the low-voltage power supply area after one adjustment does not achieve the expected adjustment effect. During the secondary adjustment, it is necessary to increase the equipment status abnormality index to further increase the inverter modulation ratio and the increase in reactive compensation.

[0038] If the reduction rate of the equipment status abnormality index in the low-voltage power supply area within an adjustment monitoring cycle is less than the defined reduction rate, the power parameters of the power supply area are adjusted for the second time, and the power regulation strategy 2 is updated based on the equipment status abnormality index in the low-voltage power supply area within an adjustment monitoring cycle. The regional power supply amount and the adjustable load shedding ratio of the low-voltage power supply area in the power parameters of the power supply area are adjusted through the power regulation strategy 2, and the power regulation strategy 2 is marked as the initial power regulation strategy 2.

[0039] The two different directions of secondary regulation mentioned above are the regulation strategies of secondary regulation.

[0040] Figure 3 This is a flow chart of the low-voltage power outage analysis and adjustment scheme of the present invention, which specifically involves collecting the power operation parameters of the low-voltage power supply area through a machine learning model, analyzing and generating the equipment status abnormality index and the power outage abnormality index of the low-voltage power supply area within the monitoring period; generating a power adjustment scheme for the dynamic power supply area based on the comparison results of the power outage abnormality index and the power outage abnormality threshold of the low-voltage power supply area within the monitoring period; if the power outage abnormality index is greater than the threshold, performing one adjustment in sequence, and determining whether a secondary adjustment is needed by monitoring the adjustment process parameters of the power supply area.

[0041] After the secondary adjustment is completed, the adjustment process parameters of the power supply area are monitored. The adjustment process parameters of the power supply area include the power outage abnormality index of the low-voltage power supply area during the secondary adjustment monitoring period and the equipment status abnormality index of the low-voltage power supply area during the secondary adjustment monitoring period, which represent the power outage abnormality index and the equipment status abnormality index during the monitoring period after the secondary adjustment.

[0042] If the power outage anomaly index of the low-voltage power supply area during the secondary adjustment monitoring period is less than or equal to the power outage anomaly threshold, the power operation parameters of the low-voltage power supply area will be continuously collected; if the power outage anomaly index of the low-voltage power supply area during the secondary adjustment monitoring period is greater than the power outage anomaly threshold, the power parameters of the power supply area will be adjusted three times through the power adjustment plan of the power supply area.

[0043] Furthermore, the power parameters of the power supply area are adjusted three times through the power regulation plan of the power supply area. The specific adjustment process is: obtaining the adjustment strategy in the secondary adjustment process; if the adjustment strategy of the secondary adjustment is to optimize power regulation strategy one, that is, the secondary adjustment is to further increase the inverter modulation ratio and reactive compensation amount in the power parameters of the power supply area through power regulation strategy one, then the power regulation strategy two is updated based on the equipment status abnormality index of the low-voltage power supply area within the secondary adjustment monitoring period, and the power regulation strategy two is marked as the initial power regulation strategy two, so that the regional power supply amount and the adjustable load shedding ratio of the low-voltage power supply area in the power parameters of the power supply area are adjusted through the initial power regulation strategy two.

[0044] If the adjustment strategy of the secondary adjustment is the initial power adjustment strategy 2, the equipment status abnormality index of the low-voltage power supply area within the secondary adjustment monitoring period is increased and adjusted based on the power outage abnormality index of the low-voltage power supply area within the secondary adjustment monitoring period, and the power adjustment strategy 2 is updated using the increased and adjusted equipment status abnormality index of the low-voltage power supply area within the secondary adjustment monitoring period, so that the regional power supply of the low-voltage power supply area and the adjustable load shedding ratio in the power parameters of the power supply area are further increased and adjusted through the power adjustment strategy 2, and the power adjustment strategy 2 is marked as the optimized power adjustment strategy 2.

[0045] The two different directions of the above three adjustments are the adjustment strategies of the three adjustments.

[0046] The regional power supply in the low-voltage power supply area is further increased and regulated by adding the regional power supply in the low-voltage power supply area plus the increase in the regional power supply in the low-voltage power supply area mapped in the database by the abnormal equipment status index in the secondary adjustment monitoring period after the increase to obtain the optimized regional power supply in the low-voltage power supply area. The adjustable load shedding ratio in the low-voltage power supply area is further increased and regulated by adding the increase in the adjustable load shedding ratio plus the increase in the adjustable load shedding ratio mapped in the database by the abnormal equipment status index in the secondary adjustment monitoring period after the increase to obtain the optimized adjustable load shedding ratio.

[0047] Furthermore, based on the power outage anomaly index of the low-voltage power supply area within the secondary adjustment monitoring period, the equipment status anomaly index of the low-voltage power supply area within the secondary adjustment monitoring period is increased and adjusted, which means that the power outage anomaly index within the secondary adjustment monitoring period is matched with the equipment status anomaly index increase coefficient within the secondary adjustment monitoring period. A power outage anomaly index-equipment status anomaly index increase coefficient mapping table is stored in the database. The power outage anomaly index can be directly queried in the database to obtain the equipment status anomaly index increase coefficient corresponding to the power outage anomaly index. The equipment status anomaly index within the secondary adjustment monitoring period is multiplied by the equipment status anomaly index increase coefficient within the secondary adjustment monitoring period to obtain the increased and adjusted equipment status anomaly index within the secondary adjustment monitoring period. The power outage anomaly index after the secondary adjustment is greater than the power outage threshold, indicating that the power outage probability in the low-voltage power supply area after the secondary adjustment does not achieve the expected adjustment effect. During the third adjustment, it is necessary to increase the equipment status anomaly index to further increase the regional power supply of the low-voltage power supply area and the increase in the adjustable load shedding ratio.

[0048] After the three adjustments are completed, the adjustment process parameters of the power supply area are monitored. The adjustment process parameters of the power supply area include the power outage abnormality index of the low-voltage power supply area within the three adjustment monitoring cycles and the equipment status abnormality index of the low-voltage power supply area within the three adjustment monitoring cycles, which represent the power outage abnormality index and the equipment status abnormality index within the monitoring cycle after the three adjustments; based on the power outage abnormality index of the low-voltage power supply area within the three adjustment monitoring cycles and the equipment status abnormality index of the low-voltage power supply area within the three adjustment monitoring cycles, it is determined whether to issue a power warning to the power supply area.

[0049] Specifically, it is determined whether to issue a power warning to the power supply area. The specific analysis process is: if the power outage anomaly index of the low-voltage power supply area within the three adjustment monitoring cycles is less than or equal to the power outage anomaly threshold, it means that the power outage probability of the low-voltage power supply area has reached the expected adjustment effect after three adjustments, and then it is determined that no power warning will be issued to the power supply area; if the power outage anomaly index of the low-voltage power supply area within the three adjustment monitoring cycles is greater than the power outage anomaly threshold, the adjustment strategy during the three adjustment processes is obtained; if the adjustment strategy of the three adjustments is the optimized power adjustment strategy 2, it means that the power outage probability of the low-voltage power supply area has not reached the expected adjustment effect after three adjustments, and then it is determined that a first-level power warning will be issued to the power supply area.

[0050] If the adjustment strategy of the three adjustments is the initial power adjustment strategy two, then the specific four adjustment processes of the power parameters of the power supply area based on the power adjustment scheme of the power supply area are: based on the power outage abnormality index of the low-voltage power supply area within the three adjustment monitoring cycles, the equipment status abnormality index of the low-voltage power supply area within the three adjustment monitoring cycles is increased and adjusted, and the equipment status abnormality index of the low-voltage power supply area within the three adjustment monitoring cycles after the increase and adjustment is used to update the power adjustment strategy two, so that the regional power supply of the low-voltage power supply area and the adjustable load shedding ratio in the power parameters of the power supply area are further increased and adjusted through the power adjustment strategy two, and the power adjustment strategy two is marked as the optimized power adjustment strategy two.

[0051] The regional power supply in the low-voltage power supply area is further increased and adjusted by adding the regional power supply in the low-voltage power supply area to the increase in the regional power supply in the low-voltage power supply area mapped in the database by the abnormal equipment status index within the three adjustment monitoring cycles after the increase, to obtain the optimized regional power supply in the low-voltage power supply area. The adjustable load shedding ratio in the low-voltage power supply area is further increased and adjusted by adding the increased adjustable load shedding ratio to the increase in the adjustable load shedding ratio mapped in the database by the abnormal equipment status index within the three adjustment monitoring cycles after the increase, to obtain the optimized adjustable load shedding ratio.

[0052] Furthermore, based on the power outage anomaly index of the low-voltage power supply area within the three adjustment monitoring cycles, the equipment status anomaly index of the low-voltage power supply area within the three adjustment monitoring cycles is increased and adjusted, which means that the power outage anomaly index within the three adjustment monitoring cycles is matched with the equipment status anomaly index increase coefficient within the three adjustment monitoring cycles. A power outage anomaly index-equipment status anomaly index increase coefficient mapping table is stored in the database. The power outage anomaly index can be directly queried in the database to obtain the equipment status anomaly index increase coefficient corresponding to the power outage anomaly index. The equipment status anomaly index within the three adjustment monitoring cycles is multiplied by the equipment status anomaly index increase coefficient within the three adjustment monitoring cycles to obtain the increased and adjusted equipment status anomaly index within the three adjustment monitoring cycles. The power outage anomaly index after three adjustments is greater than the power outage threshold, indicating that the power outage probability in the low-voltage power supply area after three adjustments has not achieved the expected adjustment effect. During the fourth adjustment, it is necessary to increase the equipment status anomaly index to further increase the regional power supply of the low-voltage power supply area and the increase in the adjustable load shedding ratio.

[0053] After the four adjustments are completed, the adjustment process parameters of the power supply area are monitored. The adjustment process parameters of the power supply area include the power outage abnormality index of the low-voltage power supply area within the four adjustment monitoring cycles and the equipment status abnormality index of the low-voltage power supply area within the four adjustment monitoring cycles, which represent the power outage abnormality index and the equipment status abnormality index within the monitoring cycle after four adjustments.

[0054] If the power outage anomaly index of the low-voltage power supply area within the four adjustment monitoring cycles is less than or equal to the power outage anomaly threshold, it means that the power outage probability of the low-voltage power supply area has achieved the expected adjustment effect after four adjustments, and the power operation parameters of the low-voltage power supply area are continuously collected; if the power outage anomaly index of the low-voltage power supply area within the four adjustment monitoring cycles is greater than the power outage anomaly threshold, it means that the power outage probability of the low-voltage power supply area has not achieved the expected adjustment effect after four adjustments, and the fluctuation duration of the equipment status anomaly index of the low-voltage power supply area within the four adjustment monitoring cycles is obtained. Specifically, the equipment status anomaly index of the low-voltage power supply area within the four adjustment monitoring cycles is subtracted from the equipment anomaly threshold preset in the database, and the difference is divided by the reduction rate of the equipment status anomaly index of the low-voltage power supply area within the four adjustment monitoring cycles, so as to obtain the fluctuation duration of the equipment status anomaly index of the low-voltage power supply area within the four adjustment monitoring cycles. The reduction rate of the equipment status anomaly index of the low-voltage power supply area within the four adjustment monitoring cycles is obtained by the low-voltage power supply area. The equipment status abnormality index of the domain within three adjustment monitoring cycles is subtracted from the equipment status abnormality index of the low-voltage power supply area within four adjustment monitoring cycles, and the difference is divided by the equipment status abnormality index of the low-voltage power supply area within three adjustment monitoring cycles; if the fluctuation duration of the equipment status abnormality index of the low-voltage power supply area within four adjustment monitoring cycles is less than or equal to the fluctuation duration threshold preset in the database, it means that the power outage probability of the low-voltage power supply area has not achieved the expected adjustment effect after four adjustments, but the equipment operation status has improved, and it is determined that a second-level power warning is issued to the power supply area; if the fluctuation duration of the equipment status abnormality index of the low-voltage power supply area within four adjustment monitoring cycles is greater than the fluctuation duration threshold, it means that the power outage probability of the low-voltage power supply area has not achieved the expected adjustment effect after four adjustments, but the equipment operation status has not improved, the fluctuation duration of the equipment status abnormality index of the low-voltage power supply area within the monitoring cycle is too long, and the equipment operation status has not shown a significant improvement trend, and it is determined that a third-level power warning is issued to the power supply area.

[0055] Figure 4 This is a schematic diagram of the power parameter adjustment and early warning determination process of the present invention. On the basis of the first adjustment, by monitoring the adjustment process parameters of the power supply area, according to the comparison result of the reduction rate of the equipment status abnormality index and the limit value, the optimization strategy one or the initial strategy two is selected for secondary adjustment; the adjustment process parameters after the secondary adjustment are monitored and it is analyzed whether a third adjustment is needed. If the secondary adjustment does not achieve the expected adjustment effect, a third adjustment is performed and the adjustment process parameters after the third adjustment are monitored; if the third adjustment does not achieve the expected adjustment effect, the power early warning level is determined.

[0056] Conduct power warning for the power supply area. The specific warning process is as follows: Conduct a first-level warning for the power supply area, which means initiating a high-level emergency response and notification, pushing a first-level warning information to the system, and pushing an "emergency power outage warning + temporary power supply plan" to users via SMS / APP; Conduct a second-level warning for the power supply area, which means predicting potential risk areas, pushing a second-level warning information to the system, and not taking forced intervention for the time being; Conduct a third-level warning for the power supply area, which means continuously monitoring the equipment status and expanding the data collection scope, pushing the third-level warning information to the system, pushing the equipment operation synchronization information to the work order system, and notifying the operation and maintenance personnel to verify on-site and feedback the results.

[0057] The present invention innovatively deeply couples the power regulation scheme with the power early warning mechanism: after optimizing the power parameters according to the power regulation scheme of the power supply area, it monitors the regulation process parameters of the power supply area, accurately analyzes the regulation effect, and determines whether to trigger the power early warning. This breaks through the limitation of traditional power early warning that only relies on thresholds, realizes active regulation of low-voltage power supply areas, and effectively solves the core defects of existing technologies such as one-sided causal analysis and delayed regulation.

[0058] Specifically, the power outage scope of the low-voltage power supply area is analyzed. The specific analysis process is: obtain multi-dimensional data of the low-voltage power supply area, pre-process the data, covering historical power outage event records, real-time monitoring parameters, equipment status data and external environment data, and denoise the data to eliminate dimensional differences and noise interference; extract target features from the pre-processed data, perform correlation analysis on the data in the database, screen data with features that are strongly correlated with power outage events, and form a feature vector containing multi-dimensional risk indicators. The correlation analysis needs to combine mutual information and random forests to analyze linear relationships, nonlinear relationships, and spatiotemporal associations to screen data; input the power outage anomaly index and feature vector of the low-voltage power supply area within four adjustment monitoring cycles into the machine learning model The model uses graph neural networks or spatiotemporal convolutional networks to analyze the power outage characteristics. Graph neural networks are used to capture the spatial correlation of power grid topology (such as fault propagation between lines), and spatiotemporal convolutional networks are used to model the dynamic trend of time series. The model identifies the mapping relationship between high-risk combinations such as "high vibration abnormality + high load rate" and the power outage propagation path, and outputs the possible diffusion direction and impact range of the fault. Combined with the physical topology of the power grid or distribution automation data, the predicted propagation path is mapped to a specific geographical area, the power outage range is delineated, and the key indicators of the number of affected users and line length are quantified. The number of affected users is obtained based on the analysis of quantitative topological data and user distribution data (such as regional population density or electricity load), thereby analyzing the power outage range in the low-voltage power supply area.

[0059] Furthermore, a power outage emergency command strategy is generated. The specific analysis process is as follows: integrating multi-dimensional data of the low-voltage power supply area and constructing a virtual mirror of the distribution network. Through data fusion and simulation technology, the abstract power outage risk is converted into an operational and verifiable digital scenario, thereby constructing a virtual mirror of the distribution network, simulating the fault propagation path and impact range, and combining historical power outage faults with the power outage anomaly index of the low-voltage power supply area during the monitoring period to predict the emergency level of the low-voltage power supply area; based on the emergency level of the low-voltage power supply area, a power outage emergency command strategy corresponding to the emergency level is generated, and command feedback is provided to the power outage emergency command strategy.

[0060] For extremely high-risk areas, "power outage warning + temporary power supply plan" will be pushed via SMS / APP; for medium-risk areas, a "peak-shifting electricity consumption recommendation" reminder will be sent; for low-risk areas, only "pay attention to subsequent notifications" will be pushed; at the same time, "demand response rewards" will be introduced to encourage users to actively participate in load regulation and improve user cooperation.

[0061] Different implementation instructions are issued to low-voltage power supply areas based on the different causes of power outages. If a region is predicted to experience power outages due to overload, priority is given to discharging the regional energy storage system to supplement power supply, while coordinating with distributed photovoltaic / wind power to reduce output to avoid reverse power supply exacerbating the fault. If it is a transient fault, energy storage is used to quickly fill the gap and shorten the outage duration. Based on real-time load distribution, the interconnection switch transfer capacity is dynamically adjusted to prioritize power supply to high-priority users. Non-critical loads are automatically cut off, and "peak power consumption" instructions are sent to adjustable loads through smart meters to reduce peak loads and alleviate line pressure.

[0062] Figure 5 This is a schematic diagram of the power outage scope analysis and emergency strategy generation process of the present invention. Specifically, it combines the power outage anomaly index and multi-dimensional feature data of four adjustment cycles, analyzes the fault propagation path through a graph neural network or a spatiotemporal convolutional network, maps it to a geographical area to delineate the power outage scope, and quantifies indicators such as the number of affected users and line length; generates a graded emergency strategy based on the warning level: for extremely high / high levels, mobile power generation vehicles are activated and non-critical loads are cut off; for medium levels, peak power consumption is dispatched; and for low levels, continuous monitoring is performed.

[0063] Relying on power outage analysis results and machine learning models, the present invention deeply integrates power grid topology, real-time fault data and historical handling cases to accurately define the scope of power outage impact; on this basis, the system automatically generates an emergency command strategy including fault location, affected user list, optimal path planning and material dispatch suggestions, and feeds back to the emergency repair terminal through instructions, providing scientific, accurate and operational strategic recommendations for power outage command and emergency handling.

[0064] The present invention realizes the closed-loop design of the whole process of "data collection-analysis-adjustment-early warning-range analysis-strategy generation" by providing a low-voltage power outage range analysis method and system based on machine learning. It combines the deep integration and dynamic modeling of multi-source data with the machine learning model, breaking through the limitations of "one-sided data and missing causality". The core advantage lies in transforming low-voltage power outage analysis from "passive result tracing" to "active risk prediction", and upgrading from "experience-based decision-making" to "data-driven", which significantly improves the power supply reliability and fault response efficiency of the distribution network, and provides a systematic solution for the intelligent operation and maintenance of the low-voltage distribution network.

[0065] Reference Figure 2 As shown, the second aspect of the present invention provides a low-voltage power outage range analysis system based on machine learning, including: an equipment operation status analysis module, a power early warning module, a low-voltage power outage analysis module and a database.

[0066] The equipment operation status analysis module is connected to the power early warning module, the power early warning module is connected to the low-voltage power outage analysis module, and the equipment operation status analysis module is connected to the low-voltage power outage analysis module.

[0067] A database is used to store parameters designed in a low-voltage power outage range analysis system based on machine learning.

[0068] The above content is merely an example and explanation of the structure of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the present invention, they should all fall within the scope of protection of the present invention.

Claims

1. A low-voltage power outage range analysis method based on machine learning is characterized by: include: Step 1: Analyze the operating status of the low-voltage power supply area through a machine learning model, collect the power operating parameters of the low-voltage power supply area, analyze the power operating parameters of the low-voltage power supply area, obtain the low-voltage power outage analysis results of the low-voltage power supply area, and generate a power regulation plan for the power supply area; Step 2: Adjust the power parameters of the power supply area through the power regulation plan of the power supply area, monitor the adjustment process parameters of the power supply area, and determine whether to issue a power warning for the power supply area; Step 3: Based on the low-voltage power outage analysis results and machine learning models in the low-voltage power supply area, analyze the power outage scope of the low-voltage power supply area, generate a power outage emergency command strategy based on the power outage scope of the low-voltage power supply area, and provide command feedback.

2. The low-voltage power outage range analysis method based on machine learning according to claim 1 is characterized in that: The low-voltage power outage analysis results of the low-voltage power supply area specifically include: The power outage anomaly index of the low-voltage power supply area during the monitoring period and the equipment status anomaly index of the low-voltage power supply area during the monitoring period; Obtain the vibration spectrum anomaly, partial discharge pulse rising edge slope, and line load rate of equipment in the low-voltage power supply area during the monitoring period; Weighted proportional values ​​are introduced from the database to quantify the degree of influence of the proportional relationship between the vibration spectrum abnormality and the defined vibration spectrum abnormality, the proportional relationship between the partial discharge pulse rising edge slope and the defined partial discharge pulse rising edge slope, and the proportional relationship between the line load rate and the defined line load rate on the equipment status abnormality index in the low-voltage power supply area during the monitoring period. The various influence degrees are summarized to obtain the equipment status abnormality index in the low-voltage power supply area during the monitoring period. The device status anomaly index of the low-voltage power supply area during the monitoring period is used to digitally represent the degree of abnormality of the device status in the low-voltage power supply area during the monitoring period.

3. The low-voltage power outage range analysis method based on machine learning according to claim 2 is characterized in that: The specific analysis process of the power outage abnormality index of the low-voltage power supply area during the monitoring period is as follows: The power operation parameters of the low-voltage power supply area include the vibration spectrum anomaly degree, grounding impedance change rate of the grounding grid, and insulation resistance degradation rate within the monitoring period. Weighted proportional values ​​are introduced from the database to quantify the degree of influence of the proportional relationship between the vibration spectrum abnormality of the equipment in the low-voltage power supply area during the monitoring period and the defined equipment vibration spectrum abnormality, the proportional relationship between the grounding impedance change rate of the grounding grid and the defined grounding impedance change rate, and the proportional relationship between the insulation resistance degradation rate and the defined insulation resistance degradation rate on the power outage abnormality index of the low-voltage power supply area during the monitoring period. The various influence degrees are summarized to obtain the power outage abnormality index of the low-voltage power supply area during the monitoring period. The power outage anomaly index of the low-voltage power supply area during the monitoring period is used to digitally represent the power outage probability of the low-voltage power supply area.

4. The method for analyzing low-voltage power outage range based on machine learning according to claim 1, characterized in that: The specific generation process of the power regulation scheme for the power supply area is as follows: The power regulation scheme of the power supply area includes power regulation strategy 1 and power regulation strategy 2; The power regulation strategy 1 refers to matching the increase in inverter modulation ratio and reactive power compensation based on the device status abnormality index, thereby increasing the inverter modulation ratio and reactive power compensation in the low-voltage power supply area; The second power regulation strategy refers to matching the increase in regional power supply and the increase in adjustable load shedding ratio in the low-voltage power supply area based on the equipment status abnormality index, thereby increasing the regional power supply and the adjustable load shedding ratio in the low-voltage power supply area.

5. The low-voltage power outage range analysis method based on machine learning according to claim 1 is characterized in that: The power parameters of the power supply area are adjusted by the power adjustment scheme of the power supply area. The specific adjustment process is as follows: Power parameters include inverter modulation ratio, reactive power compensation, regional power supply, and adjustable load shedding ratio; The power outage anomaly index of the low-voltage power supply area during the monitoring period is compared with the power outage anomaly threshold stored in the database. If the power outage anomaly index of the low-voltage power supply area during the monitoring period is less than the power outage anomaly threshold, the power operation parameters of the low-voltage power supply area are continuously collected; If the power outage anomaly index of the low-voltage power supply area during the monitoring period is greater than or equal to the power outage anomaly threshold, the power parameters of the power supply area are adjusted once using the power regulation strategy 1 in the power regulation scheme of the power supply area; After one adjustment is completed, the adjustment process parameters of the power supply area are monitored, wherein the adjustment process parameters of the power supply area include a power outage anomaly index of the low-voltage power supply area within one adjustment monitoring cycle and an equipment status anomaly index of the low-voltage power supply area within one adjustment monitoring cycle; If the power outage anomaly index of the low-voltage power supply area during one adjustment monitoring cycle is still greater than or equal to the power outage anomaly threshold, then based on the device status anomaly index of the low-voltage power supply area during one adjustment monitoring cycle and the device status anomaly index of the low-voltage power supply area during the monitoring cycle, obtain the reduction rate of the device status anomaly index of the low-voltage power supply area during one adjustment monitoring cycle, and compare it with the defined reduction rate preset in the database; If the reduction rate of the device status abnormality index of the low-voltage power supply area within one adjustment monitoring cycle is greater than or equal to the defined reduction rate, the power parameters of the power supply area are adjusted for the second time, and the device status abnormality index of the low-voltage power supply area within one adjustment monitoring cycle is increased based on the power outage abnormality index of the low-voltage power supply area within one adjustment monitoring cycle. The device status abnormality index of the low-voltage power supply area within one adjustment monitoring cycle after the increase is used to update the power regulation strategy one, and the updated power regulation strategy one is marked as the optimized power regulation strategy one. The inverter modulation ratio and the reactive compensation amount in the power parameters of the power supply area are further increased and adjusted by the optimized power regulation strategy one; If the reduction rate of the equipment status abnormality index of the low-voltage power supply area within one adjustment monitoring cycle is less than the defined reduction rate, the power parameters of the power supply area are adjusted for the second time, and the power regulation strategy 2 is updated based on the equipment status abnormality index of the low-voltage power supply area within one adjustment monitoring cycle. The regional power supply amount and the adjustable load shedding ratio of the low-voltage power supply area in the power parameters of the power supply area are adjusted by the power regulation strategy 2, and the power regulation strategy 2 is marked as the initial power regulation strategy 2; After the secondary adjustment is completed, the adjustment process parameters of the power supply area are monitored, wherein the adjustment process parameters of the power supply area include a power outage anomaly index of the low-voltage power supply area during the secondary adjustment monitoring period and an equipment status anomaly index of the low-voltage power supply area during the secondary adjustment monitoring period; If the power outage anomaly index of the low-voltage power supply area during the secondary adjustment monitoring period is less than or equal to the power outage anomaly threshold, the power operation parameters of the low-voltage power supply area are continuously collected; If the power outage anomaly index of the low-voltage power supply area within the secondary adjustment monitoring period is greater than the power outage anomaly threshold, the power parameters of the power supply area are adjusted three times through the power regulation plan of the power supply area.

6. The low-voltage power outage range analysis method based on machine learning according to claim 5 is characterized in that: The power parameters of the power supply area are adjusted three times using the power regulation scheme of the power supply area. The specific adjustment process is as follows: Obtaining the adjustment strategy during the secondary adjustment process; If the regulation strategy is to optimize the power regulation strategy 1, the power regulation strategy 2 is updated based on the abnormal equipment status index of the low-voltage power supply area within the secondary adjustment monitoring period, and the power regulation strategy 2 is marked as the initial power regulation strategy 2, so as to adjust the regional power supply amount and the adjustable load shedding ratio of the low-voltage power supply area in the power parameters of the power supply area through the initial power regulation strategy 2; If the regulation strategy is the initial power regulation strategy 2, then the equipment status abnormality index of the low-voltage power supply area within the secondary adjustment monitoring period is increased and adjusted based on the power outage abnormality index of the low-voltage power supply area within the secondary adjustment monitoring period, and the equipment status abnormality index of the low-voltage power supply area within the secondary adjustment monitoring period after the increase and adjustment is used to update the power regulation strategy 2, so that the regional power supply of the low-voltage power supply area and the adjustable load shedding ratio in the power parameters of the power supply area are further increased and adjusted through the power regulation strategy 2, and the power regulation strategy 2 is marked as the optimized power regulation strategy 2; After the three adjustments are completed, the adjustment process parameters of the power supply area are monitored, and the adjustment process parameters of the power supply area include the power outage anomaly index of the low-voltage power supply area within the three adjustment monitoring cycles and the equipment status anomaly index of the low-voltage power supply area within the three adjustment monitoring cycles; Based on the power outage abnormality index of the low-voltage power supply area within three adjustment monitoring cycles and the equipment status abnormality index of the low-voltage power supply area within three adjustment monitoring cycles, it is determined whether to issue a power warning to the power supply area.

7. The method for analyzing low-voltage power outage range based on machine learning according to claim 6, characterized in that: The specific analysis process for determining whether to issue a power warning for the power supply area is as follows: If the power outage anomaly index of the low-voltage power supply area within three adjustment monitoring cycles is less than or equal to the power outage anomaly threshold, it is determined that no power warning will be issued for the power supply area; If the power outage anomaly index of the low-voltage power supply area during the three adjustment monitoring cycles is greater than the power outage anomaly threshold, the adjustment strategy during the three adjustment processes is obtained; If the regulation strategy is the optimized power regulation strategy 2, a first-level power warning is issued to the power supply area; If the regulation strategy is the initial power regulation strategy 2, the specific regulation process of the power parameters of the power supply area based on the power regulation scheme of the power supply area is: based on the power outage anomaly index of the low-voltage power supply area within the three adjustment monitoring cycles, the equipment status anomaly index of the low-voltage power supply area within the three adjustment monitoring cycles is increased and adjusted, and the equipment status anomaly index of the low-voltage power supply area within the three adjustment monitoring cycles after the increase and adjustment is used to update the power regulation strategy 2, so as to further increase the regional power supply amount and the adjustable load shedding ratio of the low-voltage power supply area in the power parameters of the power supply area through the power regulation strategy 2, and mark the power regulation strategy 2 as the optimized power regulation strategy 2; After the four adjustments are completed, the adjustment process parameters of the power supply area are monitored, and the adjustment process parameters of the power supply area include the power outage anomaly index of the low-voltage power supply area within the four adjustment monitoring cycles and the equipment status anomaly index of the low-voltage power supply area within the four adjustment monitoring cycles; If the power outage anomaly index of the low-voltage power supply area within four adjustment monitoring cycles is less than or equal to the power outage anomaly threshold, the power operation parameters of the low-voltage power supply area will be continuously collected; If the power outage anomaly index of the low-voltage power supply area within the four adjustment monitoring cycles is greater than the power outage anomaly threshold, the fluctuation duration of the device status anomaly index of the low-voltage power supply area within the four adjustment monitoring cycles is obtained; if the fluctuation duration of the device status anomaly index of the low-voltage power supply area within the four adjustment monitoring cycles is less than or equal to the fluctuation duration threshold preset in the database, it is determined that a second-level power warning is issued for the power supply area; If the fluctuation duration of the equipment status abnormality index in the low-voltage power supply area within four adjustment monitoring cycles is greater than the fluctuation duration threshold, a third-level power warning will be issued to the power supply area.

8. The method for analyzing low-voltage power outage range based on machine learning according to claim 1, characterized in that: The power outage range of the low-voltage power supply area is analyzed, and the specific analysis process is as follows: Acquire multi-dimensional data of the low-voltage power supply area and pre-process the data to eliminate dimensional differences and noise interference; Extract target features from preprocessed data to form a feature vector containing multi-dimensional risk indicators; The power outage anomaly index and feature vector of the low-voltage power supply area within four adjustment monitoring cycles are input into the machine learning model. The model outputs the fault diffusion direction and impact range through feature correlation analysis. Combined with the physical topology of the power grid or distribution automation data, the predicted propagation path is mapped to a specific geographical area, the power outage scope is delineated, and the number of affected users and line lengths are quantified, thereby analyzing the power outage scope in the low-voltage power supply area.

9. The method for analyzing low-voltage power outage range based on machine learning according to claim 1, characterized in that: The specific analysis process of generating the power outage emergency command strategy is as follows: Integrate multi-dimensional data of low-voltage power supply areas and build a virtual image of the distribution network to simulate the fault propagation path and impact range, and predict the emergency level of low-voltage power supply areas; Based on the emergency level of the low-voltage power supply area, a power outage emergency command strategy corresponding to the emergency level is generated, and command feedback is provided for the power outage emergency command strategy.

10. A system using the low-voltage power outage range analysis method based on machine learning as described in any one of claims 1 to 9, characterized in that: include: The equipment operation status analysis module is used to collect power operation parameters of the low-voltage power supply area through machine learning models, analyze the power operation parameters of the low-voltage power supply area, obtain low-voltage power outage analysis results in the low-voltage power supply area, and generate a power regulation plan for the power supply area; The power early warning module is used to adjust the power parameters of the power supply area through the power regulation plan of the power supply area, monitor the adjustment process parameters of the power supply area, and thus determine whether to issue a power early warning for the power supply area; The low-voltage power outage analysis module is used to analyze the power outage scope of the low-voltage power supply area based on the low-voltage power outage analysis results and machine learning models in the low-voltage power supply area, generate a power outage emergency command strategy based on the power outage scope of the low-voltage power supply area, and provide command feedback.

Images