Low-voltage transformer area hidden danger identification method and system based on degradation risk analysis

Abstract

The invention discloses a low-voltage transformer area hidden danger identification method and system based on degradation risk analysis, and relates to the technical field of data processing, and the method comprises the steps: constructing an initial data set according to measurement data; constructing an abnormal data set according to the instantaneous drop event criterion in combination with the initial data set; according to a joint melting identification algorithm, determining meter front-end abnormity and a corresponding front-end abnormity degree; according to a tail end abnormity identification algorithm, meter tail end abnormity and a corresponding tail end abnormity degree are determined; combining and determining a front-end risk weight and a tail-end risk weight through a linear regression model; determining a front-end hidden danger risk value according to the meter front-end abnormity, the front-end abnormity degree and the front-end risk weight or / and determining a tail-end hidden danger risk value according to the meter tail-end abnormity, the tail-end risk weight and the tail-end abnormity degree; and determining a hidden danger emergency strategy corresponding to the transformer area according to the front-end hidden danger risk value or / and the tail-end hidden danger risk value. According to the scheme, the low-voltage transformer area hidden danger identification accuracy and efficiency are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and specifically, to a method and system for identifying hidden dangers in low-voltage power distribution areas based on deterioration risk analysis. Background Art

[0002] During the operation of the power system, the accurate identification of hidden dangers at the front end and the end of the low-voltage power distribution area meters is a key link to ensure power supply reliability. The existing technologies mainly rely on manual regular inspections or threshold warning mechanisms based on single parameters, and have core defects such as multi-dimensional data fusion faulting and insufficient decoupling of abnormal features. The specific manifestations are as follows: At the data acquisition level, traditional methods only use current and voltage sampling at discrete time points, and fail to construct a dynamic data screening mechanism by combining transient event characteristics, resulting in abnormal data being submerged in steady-state noise; at the hidden danger identification level, existing algorithms mostly adopt a single-dimensional judgment criterion with a fixed threshold, neither establishing a quantitative analysis model for the characteristics of meter joint melting or oxidation, nor having an associated analysis of abnormal fluctuations in the end load, and being unable to distinguish similar voltage dips caused by front-end contact deterioration and end-user electricity consumption; at the risk assessment level, current methods generally adopt a static weight allocation strategy, ignoring the dynamic impact of different abnormal types on the equipment deterioration path, resulting in a significant deviation between the risk level assessment and the actual evolution trend of hidden dangers. The traditional system fails to construct a two-dimensional evaluation system for the risk of front-end joint melting and abnormal end impedance, resulting in the inability to accurately locate the hidden danger source point when formulating maintenance strategies, and often adopting the method of rush repair after the hidden danger evolves into a fault causing user power outage to solve, seriously affecting power supply continuity.

[0003] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present application, and therefore it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0004] The object of the present invention is to solve the problems in the prior art that rely on manual inspections or simple analysis of a single data source, resulting in inaccurate and inefficient identification of hidden dangers in low-voltage power distribution areas, and proposes a method and system for identifying hidden dangers in low-voltage power distribution areas based on deterioration risk analysis; through transient drop event criteria screening, a two-path abnormal identification algorithm (joint melting identification / end abnormal identification) and a dynamic regression model, accurate distinction between front-end contact deterioration and end-load abnormality of the meter, quantitative calibration of the hidden danger risk level, and directional positioning of the hidden danger source point are realized, significantly improving the accuracy and efficiency of hidden danger identification in low-voltage power distribution areas, and providing a reliable data basis for the prediction of hidden dangers in low-voltage power distribution areas and the formulation of response strategies.

[0005] In a first aspect, a technical solution provided in an embodiment of the present invention is a method for identifying hidden dangers in a low-voltage power distribution area based on deterioration risk analysis, including the following steps: Obtain the measurement data of the low-voltage power distribution area according to the sampling period in sequence to construct an initial data set; screen the initial data set according to the momentary drop event criterion in sequence to construct an abnormal data set; Analyze the abnormal data set according to the joint melting recognition algorithm to determine the abnormality at the front end of the meter and its corresponding front-end abnormality degree; when the trigger condition is met, analyze the abnormal data set according to the end-end abnormality recognition algorithm to determine the end-end abnormality of the meter and its corresponding end-end abnormality degree; Determine the front-end risk weight corresponding to the front-end abnormality of the meter and the end-end risk weight corresponding to the end-end abnormality of the meter through a linear regression model combined with the abnormal situation data corresponding to the front-end abnormality degree and the end-end abnormality degree; Determine the front-end hidden danger risk value based on the front-end abnormality of the meter, the front-end abnormality degree, and the front-end risk weight or / and the end-end hidden danger risk value determined based on the end-end abnormality of the meter, the end-end risk weight, and the end-end abnormality degree; Determine the corresponding hidden danger emergency strategy for the power distribution area based on the front-end hidden danger risk value or / and the end-end hidden danger risk value.

[0006] Preferably, the measurement data includes the main meter voltage, user voltage, and user current.

[0007] Preferably, the momentary drop event criterion includes: Criterion A1: Extract data using a sliding window with a width of 2 measurement points, and the voltage drop between the two measurement points within the window is greater than 5V; Criterion A2: The absolute value of the voltage difference between the main meter voltage and the user voltage is greater than 5V; Criterion A3: The current at the voltage low point is greater than 0.1A, and (current at voltage low point - current at voltage high point) / current at voltage high point > 50%, and (current at voltage low point - current at voltage high point) > the average value of the largest 5 currents at the same measurement point within the unit time 10%; Criterion A4: If the three-phase user voltages drop instantaneously at the same time, the momentary drop event is excluded and classified as an end-user; Criterion A5: Obtain the current at the moment when the voltage of the momentary drop event instantaneously drops to the low voltage. If the current disappears within 12 measurement moments at the same measurement point after the momentary drop, or the current at the moment of the start of the momentary drop is less than 20% of the average value of the current at the voltage low point, or the current is less than 0.1A and the voltage has not recovered to the high point voltage, the momentary drop event is excluded; Criterion A6: User voltage difference Current > 1000w, the momentary drop event is excluded; If all the above criteria are met simultaneously, the momentary drop event is established; Among them, the momentary drop event includes the voltage and current data of the two measurements before and after, and the voltage and current data includes the current at the voltage low point, the voltage at the voltage low point, the current at the voltage high point, and the voltage at the voltage high point.

[0008] Preferably, analyzing the abnormal data set according to the joint melting recognition algorithm to determine the abnormality at the front end of the meter and its corresponding front-end abnormality degree, including the following steps: Extract the measurement data in the abnormal data set and calculate to obtain the front-end characteristic data, and compare and analyze the front-end characteristic data through the joint melting recognition algorithm to determine the abnormality at the front end of the meter; determine the front-end abnormality degree according to the abnormal situation corresponding to the abnormality at the front end of the meter; wherein, the front-end characteristic data at least includes the instantaneous voltage drop difference, voltage-current inverse correlation coefficient, current average value, and instantaneous voltage high point average value of each measurement point.

[0009] Preferably, comparing and analyzing the front-end characteristic data through the joint melting recognition algorithm to determine the abnormality at the front end of the meter, including the following steps: When the single instantaneous voltage drop difference Current > 5W, and the voltage-current inverse correlation coefficient in the 7 days before the melting judgment is less than -0.5, and the single current continuously exceeds 80% of the current at the low point of the instantaneous voltage drop, and the voltage recovers to the high point of the voltage before the instantaneous voltage drop, then it is judged that there is a melting abnormality in the abnormality at the front end of the meter; After a single instantaneous voltage drop ends, if there are m (m can be taken as 3) or more sampling moments corresponding to the current greater than the average current of all low points of the instantaneous voltage drop within the unit time at n sampling points (n can be taken as 96) for the user, And at the same time the voltage is greater than the average value of the high points of the instantaneous voltage drop, then it is judged that there is a melting abnormality in the abnormality at the front end of the meter.

[0010] Preferably, determining the front-end abnormality degree according to the abnormal situation corresponding to the abnormality at the front end of the meter, including the following steps: Extract the front-end risk factors corresponding to the abnormal situation, and accumulate according to the weight coefficients corresponding to the data in each dimension of the front-end risk factors to determine the front-end abnormality degree; wherein, the weight coefficients of the data in each dimension of the front-end risk factors are dynamically adjusted according to the actual electricity consumption of the user and the voltage fluctuation situation of the substation area; the data in each dimension of the front-end risk factors includes: instantaneous voltage drop difference, low voltage, theoretical resistance, theoretical resistance change trend, theoretical resistance standard deviation, number of instantaneous voltage drops, synchronous power outage event, synchronous daily line loss sudden increase event, voltage-current inverse correlation coefficient, theoretical power, instantaneous voltage drop event occurrence period, joint melting event; Among them, determining the front-end abnormality degree according to the front-end risk factors includes the following steps: The weight coefficient w1 corresponding to the instantaneous voltage drop difference = the average value of the voltage drop differences of all instantaneous voltage drop events of the current user in 7 days - 5, and the upper limit of the weight coefficient w1 is set to 20. It can be seen that all weights and upper limits can change according to specific events. Among them, when the instantaneous voltage drop difference is the high point voltage of the instantaneous voltage drop event - the low point voltage of the instantaneous voltage drop event; The weight coefficient w2 corresponding to the low voltage = 205v - the average low point voltage of the lowest 5 voltage dip events among the voltage dip events of the current user within 7 days; the upper limit of the set weight coefficient w2 is 20; The weight coefficient w3 corresponding to the theoretical resistance = 10 The average voltage difference of the voltage dip events of the current user within 7 days / the average current difference of the voltage dip events, the upper limit of the set weight coefficient w3 is 40, where the voltage difference is the high point voltage of the voltage dip event - the low point voltage of the voltage dip event, and the current difference is the current at the low point voltage of the voltage dip event - the current at the high point voltage of the voltage dip event; The weight coefficient w4 corresponding to the theoretical resistance change trend = (resistance change rate - 1) 20; where the resistance change rate = the average theoretical resistance of the last 3 voltage dip events of the current user within 7 days / the average theoretical resistance of the first 3 voltage dip events; The weight coefficient corresponding to the theoretical resistance standard deviation , where, is the average resistance, is the maximum resistance, is the minimum resistance, N is the total number of voltage dip events, is the resistance of the nth voltage dip event; The weight coefficient w6 corresponding to the number of voltage dips on the same day = (the number of voltage dip events of the current user on the same day - 1) 10; the upper limit of the set weight coefficient w6 is 20; The weight coefficient w7 corresponding to the number of voltage dips in seven days = (the number of voltage dip events of the current user in seven days - 2) 5; the upper limit of the set weight coefficient w7 is 40; The weight coefficient w8 corresponding to the synchronous power outage event = the number of power outages of the current user within one month when a voltage dip event occurs 10; the upper limit of the set weight coefficient w8 is 40; The weight coefficient corresponding to the voltage - current anti - correlation coefficient ; The weight coefficient w10 corresponding to the theoretical power = (the average theoretical power - 300) / 10; where the weight coefficient w10 is rounded down, and the upper limit of the reduction value of the set weight coefficient w10 is - 20; the theoretical power = the voltage drop △U the current value at the low point voltage; The weight coefficient w11 corresponding to the time period when the voltage dip event occurs = when the proportion of the time stamps of the weekly voltage dip events of the current user between 17:00 - 20:00 is greater than 60%, it is reduced by the proportion - 10 to - 30, where the proportion of 60% is reduced by - 10, the proportion of 80% is reduced by - 20, and the proportion of 100% is reduced by - 30; The weight coefficient w11 corresponding to the connector melting event is 60; the weight coefficient w12 for the sudden increase in the daily line loss of the synchronous daily line: if the total risk value of the user is greater than 100, and there is a sudden increase in the daily line loss within 1 month in the user's substation area where the user is located, on the day when the daily line loss of the synchronous daily line suddenly increases, if the total risk value sum (w1~w11) of the user has an increase of more than 40, w12 = 20. Specifically, the sudden increase in the line loss event includes at least a sudden increase in the line loss exceeding 2%.

[0011] Preferably, when the trigger condition is met, the abnormal data set is analyzed according to the end anomaly recognition algorithm to determine the end anomaly of the meter and its corresponding end anomaly degree; the steps are as follows: Among the users whose front-end hidden danger risk is determined to be greater than the set threshold H, end-user identification is carried out, the measurement data in the abnormal data set is extracted and the end characteristic data is calculated and obtained, and the end anomaly of the meter is determined by comparing and analyzing the end characteristic data through the end anomaly recognition algorithm; the end anomaly degree is determined according to the abnormal situation corresponding to the end anomaly of the meter; among them, the end characteristic data includes at least the constant resistance standard deviation, theoretical voltage drop power, voltage-current anti-correlation coefficient, voltage-drop users, and theoretical resistance of each measurement point.

[0012] Preferably, the end anomaly of the meter is determined by comparing and analyzing the end characteristic data through the end anomaly recognition algorithm; the steps are as follows: When the constant resistance standard deviation of the user < 0.3, it is included in the analysis of the end anomaly of the meter. When the voltage-current anti-correlation coefficient range of the user is -0.7~-1, it is included in the analysis of the end anomaly of the meter. When the single-phase theoretical power range of the user's momentary voltage drop event is greater than 300w, it is included in the analysis of the end anomaly of the meter; among them, when a momentary voltage drop event occurs for a certain user in the same substation area, if the voltage fluctuations of other users are close to that of this user, these users are judged as voltage-drop users together. When more than 2 phases of a three-phase user show front-end anomalies, and |maximum theoretical resistance - minimum theoretical resistance| / maximum theoretical resistance < 0.3 for the phase with the momentary voltage drop, it is included in the analysis of the end anomaly of the meter.

[0013] Preferably, the end anomaly degree is determined according to the abnormal situation corresponding to the end anomaly of the meter, and the steps are as follows: In the list where the front-end anomaly risk value of the meter is greater than the set threshold H, the end risk factors corresponding to the abnormal situation are extracted, and the end anomaly degree is determined by summing the weight coefficients corresponding to the data in each dimension of the end risk factors; among them, the weight coefficients of the data in each dimension of the end risk factors are dynamically adjusted according to the actual electricity consumption of the user and the voltage fluctuation situation of the substation area; the data in each dimension of the end risk factors includes: constant resistance standard deviation, theoretical voltage drop power, voltage-current anti-correlation coefficient, front-end anomaly degree, theoretical resistance, and user voltage-drop event. Among them, determining the terminal abnormal degree according to the terminal risk factors includes the following steps: B1. For the weight coefficient h1 corresponding to the constant standard deviation of resistance, when 0.3 > constant standard deviation of resistance > 0, the weight coefficient h1 is linearly distributed numerically between 40 and 10; B2. For the weight coefficient h2 corresponding to the voltage-current anti-correlation coefficient, when -0.7 > voltage-current anti-correlation coefficient > -1, the weight coefficient h2 is linearly distributed numerically between 10 and 40; B3. For the weight coefficient h3 corresponding to the theoretical voltage drop power, when 300w < theoretical voltage drop power < 600w, the weight coefficient h3 is linearly distributed numerically between 10 and 40, and when the theoretical voltage drop power ≥ 600w, h3 is fixed at 40; B4. For the weight coefficient h4 corresponding to the front-end abnormal degree, in the case of the front-end abnormal situation where more than 2 phases of three-phase users have voltage dips, and the difference between the maximum and minimum values of the corresponding front-end abnormal degree is within the set threshold of 40, and there is a corresponding voltage dip , the weight coefficient h4 is set to 20; B5. For the weight coefficient h5 corresponding to the user co-drop event, when it is judged as multi-user co-drop; the corresponding weight coefficient h5 is set to 20; On the basis that the front-end hidden danger risk value is greater than the set threshold H, based on the analysis dimensions B1 - B5, calculate the terminal hidden danger risk value. When the terminal hidden danger risk value is greater than 40, it is determined that there is a terminal hidden danger.

[0014] Preferably, the training process of the linear regression model is: Obtain the front-end risk factors corresponding to the front-end abnormal degree as the input of the linear regression model, train the linear regression model with the maximum front-end hidden danger risk value as the optimization target, and obtain the front-end risk weights of each front-end risk factor; when the front-end hidden danger risk value is greater than the set threshold H, use the terminal samples and pre-metering fault samples classified by on-site judgment in the front-end risk factor list as the input of the linear regression model, and train the linear regression model with the maximum terminal hidden danger risk value as the optimization target to obtain the terminal risk weights of each terminal risk factor; Among them, the front-end risk factors include multi-dimensional data, specifically including: voltage dip difference, low voltage, theoretical resistance, theoretical resistance change trend, theoretical resistance standard deviation, number of voltage dips, synchronous power outage event, synchronous sudden increase in daily line loss event, voltage-current anti-correlation coefficient, theoretical power, voltage dip event occurrence time period, joint melting event; The terminal risk factors include multi-dimensional data, specifically including: constant standard deviation of resistance, theoretical voltage drop power, voltage-current anti-correlation coefficient, front-end abnormal degree, theoretical resistance, and user co-drop event.

[0015] In a second aspect, a technical solution provided in an embodiment of the present invention is a hidden danger identification system for a low-voltage power distribution area, including: A first construction module: sequentially obtain measurement data of the low-voltage power distribution area according to a sampling period to construct an initial data set; A second construction module: sequentially screen the initial data set according to a momentary drop event criterion to construct an abnormal data set; A first analysis module: determine the abnormal conditions at the front end of the meter and their corresponding front-end abnormal degrees according to analysis dimensions such as a connection melting identification algorithm based on the momentary drop event; A second analysis module: on the basis that the front-end hidden danger risk value is greater than a set threshold H, analyze the abnormal data set according to an end abnormal identification algorithm to determine the abnormal conditions at the end of the meter and their corresponding end abnormal degrees; A calculation module: determine the front-end risk weight corresponding to the abnormal conditions at the front end of the meter and the end risk weight corresponding to the abnormal conditions at the end of the meter through the analytic hierarchy process; A first fusion module: determine the front-end hidden danger risk value according to the abnormal conditions at the front end of the meter, the front-end abnormal degree, and the front-end risk weight; A second fusion module: determine the end hidden danger risk value according to the abnormal conditions at the end of the meter, the end risk weight, and the end abnormal degree; An execution module: determine a hidden danger emergency strategy corresponding to the power distribution area according to the front-end hidden danger risk value or / and the end hidden danger risk value.

[0016] In a third aspect, a technical solution provided in an embodiment of the present invention is an electronic device, including a memory and a processor. A computer program is stored in the memory. When the processor calls the computer program in the memory, the steps of the method for identifying hidden dangers in a low-voltage power distribution area based on deterioration risk analysis are implemented.

[0017] In a fourth aspect, a technical solution provided in an embodiment of the present invention is a storage medium. Computer-executable instructions are stored in the storage medium. When the computer-executable instructions are loaded and executed by a processor, the steps of the method for identifying hidden dangers in a low-voltage power distribution area based on deterioration risk analysis are implemented.

[0018] Advantages of the present invention: Based on the momentary voltage dip event criterion, this application analyzes the obtained substation area data to construct an abnormal data set. By using the joint melting recognition algorithm to quantify the abnormal degree of the front-end contact resistance for the key degradation features (including front-end feature data and end feature data) in the abnormal data set, and combining with the end abnormal recognition algorithm to analyze the load fluctuation characteristics, the front-end / end fault modes are separated. Then, a linear regression model is used to associate the abnormal degrees of the front-end / end with the historical degradation data, dynamically allocate the risk weights of each risk factor, and further calculate the front-end hidden danger risk value and the end hidden danger risk value according to the abnormal degree, abnormal type and the corresponding risk weights, thus constructing a two-dimensional evaluation system to solve the technical problems of insufficient decoupling of abnormal features (i.e., inability to distinguish front-end contact degradation from end overload), static risk weights (i.e., ignoring the dynamic evolution of hidden dangers), and fuzzy hidden danger positioning (i.e., relying on manual inspection) in the hidden danger identification of low-voltage substations, significantly improving the accuracy and efficiency of hidden danger identification, and providing a reliable data basis for the prediction of hidden dangers and the formulation of response strategies in low-voltage substations.

[0019] The above invention content is only an overview of the technical solution of the present invention. In order to be able to more clearly understand the technical means of the present invention, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present invention more obvious and understandable, the following specifically gives the specific implementation manners of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] By reading the detailed description of the non-restrictive embodiments with reference to the following drawings, other features, purposes and advantages of the present invention will become more obvious. The drawings are only for the purpose of showing the preferred embodiments and are not considered as a limitation of the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components.

[0021] Figure 1 It is a flowchart of the method for identifying hidden dangers in low-voltage substations based on degradation risk analysis of the present invention.

[0022] Figure 2 It is a schematic structural diagram of a system for identifying hidden dangers in low-voltage substations of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0023] In order to make the purpose, technical solution and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific implementation manners described herein are only the best embodiments of the present invention, only for explaining the present invention, and do not limit the protection scope of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0024] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts depict the operations (or steps) as sequential processes, many of the operations (or steps) can be implemented in parallel, concurrently, or simultaneously. In addition, the order of the operations can be rearranged. The process can be terminated when its operations are completed, but it can also have additional steps not included in the drawings; the process can correspond to a method, function, procedure, subroutine, subprogram, and so on.

[0025] Embodiment 1: As Figure 1 shown, a hidden danger identification method for low-voltage power distribution areas based on deterioration risk analysis includes the following steps: Obtain the measurement data of the low-voltage power distribution area in sequence according to the sampling period to construct an initial data set; screen the initial data set in sequence according to the momentary drop event criterion to construct an abnormal data set.

[0026] It can be understood that due to the huge amount of data of power users, a data acquisition period can be set. For example, in this embodiment, the sampling period can be set to 15 minutes. For each user, there are 96 measurement time points per day, and the measurement data corresponding to each user end in the low-voltage power distribution area is collected respectively. Among them, the measurement data can be the main meter voltage, user voltage, and user current; the momentary drop event includes the voltage and current data of two consecutive measurements, and the voltage and current data include the current at the voltage low point, the voltage at the voltage low point, the current at the voltage high point, and the voltage at the voltage high point; by using the original measurement data to participate in subsequent data analysis operations, derivative data such as the user voltage difference and current average value involved in the momentary drop event criterion can be obtained; through the momentary drop event criterion, the abnormal events in the low-voltage power distribution area can be preliminarily screened out, so as to focus the abnormal events on the key deterioration data to improve the efficiency and accuracy of the abnormal analysis of the low-voltage power distribution area.

[0027] As an alternative embodiment, the momentary drop event criterion includes: Criterion A1: Extract data using a sliding window with a width of 2 measurement points, and the voltage drop within the window between two measurement points is greater than 5V; Criterion A2: The absolute value of the voltage difference between the main meter voltage and the user voltage is greater than 5V; Criterion A3: The current at the voltage low point is greater than 0.1A, and (current at voltage low point - current at voltage high point) / current at voltage high point > 50% (current at voltage low point - current at voltage high point) > the average value of the largest 5 currents at the same measurement point within the unit time 10%; Criterion A4: If the three-phase user voltages drop simultaneously, the momentary drop event is excluded and classified as an end user; Criterion A5: Obtain the current at the moment when the voltage of the voltage dip event instantaneously drops to a low voltage. If the current disappears within 12 measurement moments at the same measurement point after the voltage dip, or the current at the start moment of the voltage dip is less than 20% of the average current at the voltage low point, or the current is less than 0.1 A and the voltage does not recover to the high point voltage, the voltage dip event is excluded; Criterion A6: User voltage difference Current > 1000 W. Since a single-point fault cannot withstand excessive power, the voltage dip event is excluded; If the above criteria are all met simultaneously, the voltage dip event is established.

[0028] It can be understood that by capturing the voltage mutation between two measurement points within an adjacent 15 minutes through a 2-measurement-point sliding window, and dynamically filtering short-term voltage fluctuations in combination with a 5V threshold, the sensitivity of the captured data for voltage dip events is ensured. Detecting the disappearance or sudden drop of current within 12 measurement points effectively excludes pseudo-voltage dip events caused by instantaneous poor contact or measurement errors, improving the data quality of the abnormal data set; excluding high-energy voltage dip events (such as interference from users at the end of a high-load line) through "user voltage difference × current > 1000 W" can focus on potential hazards of overheating joints or sudden changes in terminal impedance. Through the above criteria, the efficiency of the preliminary all-round investigation of potential hazards is significantly improved.

[0029] According to the joint melting recognition algorithm, analyze the abnormal data set to determine the abnormality at the front end of the meter and its corresponding front-end abnormality degree; when the trigger condition is met, according to the end abnormality recognition algorithm, analyze the abnormal data set to determine the abnormality at the end of the meter and its corresponding end-end abnormality degree.

[0030] As an alternative embodiment, analyzing the abnormal data set according to the joint melting recognition algorithm to determine the abnormality at the front end of the meter and its corresponding front-end abnormality degree includes the following steps: Extract the measurement data in the abnormal data set and calculate to obtain the front-end characteristic data. Through the joint melting recognition algorithm, compare and analyze the front-end characteristic data to determine the abnormality at the front end of the meter; determine the front-end abnormality degree according to the abnormal situation corresponding to the abnormality at the front end of the meter; wherein, the front-end characteristic data at least includes the voltage dip voltage difference at each measurement point, the voltage-current anti-correlation coefficient, the average current, and the average value of the voltage high point of the voltage dip.

[0031] In this embodiment, extracting the measurement data in the abnormal data set and calculating to obtain the front-end characteristic data representing the front-end abnormality can focus on key deteriorating data and thus improve the efficiency of data processing. Then, through the joint melting recognition algorithm, identify whether there is a melting abnormality from multiple dimensions, improving the reliability of the recognition result.

[0032] As an alternative embodiment, comparing and analyzing the front-end characteristic data through the joint melting recognition algorithm to determine the abnormality at the front end of the meter includes the following steps: When the single voltage dip voltage difference The current > 5W, and the voltage-current anti-correlation coefficient in the 7 days before melting judgment is less than -0.5, and the single current duration is greater than the current at the low point of the transient voltage drop 80%, and the voltage recovers to the high voltage point before the transient voltage drop, then it is judged that there is a melting abnormality in the front end of the meter; When a single transient voltage drop ends, within the range of n sampling points (n can be taken as 96), there are m (m can be taken as 3) or more sampling moments corresponding to the current greater than the average current of all transient voltage drop low points within the unit time 120%, and at the same time the voltage is greater than the average value of the transient voltage drop high points, then it is judged that there is a melting abnormality in the front end of the meter.

[0033] In this embodiment, considering that copper and aluminum will reach the metal softening point and melting point under the condition of temperature rise, after reaching, due to the temporary increase in the contact area of the joint (but it will intensify oxidation and even fall off after cooling), combined with the metal melting characteristics, the above-mentioned joint melting recognition scheme is designed, which can comprehensively analyze whether there is a melting situation for the current user, so as to achieve early warning.

[0034] As an alternative embodiment, determining the front-end abnormality degree according to the abnormal situation corresponding to the front end of the meter includes the following steps: Extract the front-end risk factors corresponding to the abnormal situation, and determine the front-end abnormality degree by accumulating according to the weight coefficients corresponding to the data in each dimension of the front-end risk factors; among them, the weight coefficients of the data in each dimension of the front-end risk factors are dynamically adjusted according to the actual user's electricity consumption and the voltage fluctuation situation of the power grid area; among them, the data in each dimension of the front-end risk factors include: transient voltage difference, low voltage, theoretical resistance, theoretical resistance change trend, theoretical resistance standard deviation, number of transient voltage drops, synchronous power outage events, synchronous sudden increase events in daily line losses, voltage-current anti-correlation coefficient, theoretical power, time period when the transient voltage drop event occurs, joint melting event.

[0035] In this embodiment, since the abnormal situations leading to the front-end abnormality are relatively complex, the original data is obtained by classifying the front-end risk factors corresponding to the abnormal situations, and by analyzing each kind of original data separately, and comprehensively judging the analysis results to further determine the front-end abnormality degree.

[0036] Further, determining the front-end abnormality degree according to the front-end risk factors includes the following steps: The weight coefficient w1 corresponding to the transient voltage difference = the average value of the voltage differences of all transient voltage drop events of the current user in 7 days - 5. Since the greater the voltage difference, the worse the user's electricity consumption experience and the higher the maintenance priority, the upper limit of the weight coefficient w1 is set to 20. The upper limit of the weight is set to prevent generalization in extreme cases. Among them, when the transient voltage difference is the high point voltage of the transient voltage drop event - the low point voltage of the transient voltage drop event.

[0037] The weight coefficient w2 corresponding to the low voltage = 205v - the average low point voltage of the lowest 5 voltage dip events among the voltage low points in the current user's voltage dip events within 7 days; since the lower the low point voltage, the worse the user's electricity consumption experience, in order to prevent generalization in extreme cases, the upper limit of the weight coefficient w2 is set to 20.

[0038] The weight coefficient w3 corresponding to the theoretical resistance = 10 The average voltage difference of the current user's voltage dip events within 7 days / the average current difference of the voltage dip events; the upper limit of the weight coefficient w3 is set to 40, where the voltage difference is the high point voltage of the voltage dip event - the low point voltage of the voltage dip event, and the current difference is the current at the low point voltage of the voltage dip event - the current at the high point voltage of the voltage dip event. It can be understood that the copper-aluminum oxidation resistance is between 0.1 ohm and 50 ohms, and the upper limit is set to prevent generalization in extreme cases.

[0039] The weight coefficient w4 corresponding to the theoretical resistance change trend = (resistance change rate - 1) 20; where the resistance change rate = the average theoretical resistance of the last 3 voltage dip events of the current user within 7 days / the average theoretical resistance of the first 3 voltage dip events; it can be understood that as the resistance increases, the power value of the same user's current causing a fault and burnout increases.

[0040] The weight coefficient corresponding to the theoretical resistance standard deviation , where is the average resistance, is the maximum resistance, is the minimum resistance, N is the total number of voltage dip events, is the resistance of the nth voltage dip event; it can be understood that the smaller the theoretical resistance standard deviation, the greater the possibility of a high-resistance contact point, but when it is less than the threshold (0.3), the resistance is surprisingly constant, and it is closer to the impedance caused by fixed line parameters.

[0041] The weight coefficient w6 corresponding to the number of voltage dips on the same day = (the number of voltage dip events of the current user on the same day - 1) 10; the upper limit of the weight coefficient w6 is set to 20; it can be understood that the more voltage dips, the more frequent the voltage drop.

[0042] The weight coefficient w7 corresponding to the number of voltage dips in seven days = (the number of voltage dip events of the current user in seven days - 2) 5; the upper limit of the weight coefficient w7 is set to 40; it can be understood that the more voltage dips, the more frequent the voltage drop.

[0043] The weight coefficient w8 corresponding to the synchronous power outage event = the number of power outages of the current user within one month after the voltage dip event 10; Set the upper limit of the weight coefficient w8 to 40; It can be understood that the more times of synchronous power outages, the higher the possibility of the user having a high-resistance problem.

[0044] The weight coefficient corresponding to the voltage-current inverse correlation coefficient ; It can be understood that the weight coefficient w9 corresponding to the voltage-current inverse correlation coefficient is calculated using the corr correlation coefficient function (for example, it can be based on a python compilation environment), with the input being the time series of voltage and current, optimized through the sequence of current crossing the dead zone, and the output range is between -1 and 1. For the case where the weight coefficient is greater than -0.8 but less than -0.2, a risk value calculation is performed to obtain a dynamic weight coefficient, avoiding the problem of inaccurate hidden danger identification caused by ignoring the dynamic evolution of hidden dangers.

[0045] The weight coefficient w10 corresponding to the theoretical power = (average theoretical power - 300) / 10; where the weight coefficient w10 is rounded downwards, and the upper limit of the reduction value of the weight coefficient w10 is set to -20; Theoretical power = voltage drop △U The current value at the voltage low point; It can be understood that due to the voltage at the head end of the transformer substation area or the three-phase imbalance, the voltage will also decrease at the moment of small current, so it is necessary to reduce the risk of such users to adapt to the specific working conditions.

[0046] The weight coefficient w11 corresponding to the time period when the momentary drop event occurs = the proportion of the time stamp of the current user's weekly momentary drop event between 17:00 - 20:00 is greater than 60% and is reduced by -10 to -30 according to the proportion, where the proportion of 60% is reduced by -10, the proportion of 80% is reduced by -20, and the proportion of 100% is reduced by -30; It can be understood that since the oxidation and heating of a single node cannot support high power, the possibility of hidden dangers at the end increases as the theoretical voltage drop power increases. It can be understood that since this period is the peak period of evening electricity consumption, the normal voltage drop of the line increases due to the simultaneous electricity consumption of the users in the transformer substation area, which may be miscounted into the momentary drop event. A too high proportion means that the user's voltage will only drop during the high-load time of synchronous electricity consumption in the evening. Such problems are caused by synchronous large-load electricity consumption and are not latent defect faults. Risk reduction treatment is carried out. It should be noted that when the computing power is sufficient, the situation of a large number of residential users can be comprehensively considered and included in the calculation.

[0047] The weight coefficient w11 corresponding to the joint melting event = 60; See specifically the situation where the user's melting abnormality is identified from multiple dimensions through the joint melting identification algorithm.

[0048] Synchronous daily line loss sudden increase weight coefficient w12: If the total risk value of a user is greater than 100, and there is a daily line loss sudden increase event within 1 month in the user's area, on the day when the daily line loss suddenly increases synchronously, if the total risk value sum (w1~w11) of the user has an increase of more than 40, w12 = 20. Specifically, the line loss sudden increase event includes at least a line loss sudden increase exceeding 2%.

[0049] As an alternative embodiment, when the trigger condition is met, analyze the abnormal data set according to the end anomaly recognition algorithm to determine the meter end anomaly and its corresponding end anomaly degree; the steps are as follows: Among the users whose front-end hidden danger risk value is determined to be greater than 80, end-user identification is required. Extract the measurement data in the abnormal data set and calculate to obtain the end characteristic data. Compare and analyze the end characteristic data through the end anomaly recognition algorithm to determine the meter end anomaly; determine the end anomaly degree according to the abnormal situation corresponding to the meter end anomaly; among them, the end characteristic data includes at least the constant standard deviation of resistance, theoretical voltage drop power, voltage-current anti-correlation coefficient, voltage-drop users, and theoretical resistance of each measurement point.

[0050] In this embodiment, whether to perform end anomaly situation discrimination is triggered according to the range of the front-end hidden danger risk value. By analyzing the end characteristics of users causing high risks, the computing resources involved in the operation can be reduced. By comprehensively analyzing multi-dimensional end characteristic data, the problem of high omission rate of end hidden dangers caused by single-dimensional judgment can be avoided.

[0051] Preferably, compare and analyze the end characteristic data through the end anomaly recognition algorithm to determine the meter end anomaly; the steps are as follows: When the constant standard deviation of the user's resistance < 0.3, it is included in the analysis of the meter end anomaly; When the range of the voltage-current anti-correlation coefficient of the user is -0.7~-1, it is included in the analysis of the meter end anomaly; it can be understood that compared with latent faults such as same joints and cable bulges, the end impedance increase caused by line parameters shows a stronger voltage-current anti-correlation; When the single-phase theoretical power range of the user's instantaneous voltage drop event is greater than 300w, it is included in the analysis of the meter end anomaly; among them, when the instantaneous voltage drop event of a certain user in the same area occurs, if the voltage fluctuations of other users are close to that user, these users are judged as co-drop users; in on-site case analysis, the theoretical voltage drop power above 300w is generally for end users; When more than 2 phases of a three-phase user have front-end anomalies, and |maximum theoretical resistance - minimum theoretical resistance| / maximum theoretical resistance < 0.3 for the phase with instantaneous voltage drop, it is included in the analysis of the meter end anomaly; it can be understood that compared with latent faults such as same joints and cable bulges, the end impedance increase caused by line parameters shows a stronger voltage-current anti-correlation.

[0052] In this embodiment, when the substation area topology is not perfect (since the power PMS system is deployed in the intranet and new users cannot enter data on the mobile Internet, and after bringing it back by on-site drawing and supplementing it, the supplementing personnel are often different from the on-site construction personnel, resulting in inaccurate topology data), the resistance of problems such as joint oxidation, cable bulging, and clamp will not remain constant due to external wind blowing and shaking, temperature, thermal expansion and contraction, heat softening, melting, etc., but only relatively constant. However, the line parameters at the end of the line and the line impedance caused by the power supply radius are relatively stable within a certain season. By analyzing multi-dimensional characteristic data through the above-mentioned end anomaly recognition algorithm to identify the end high-resistance problem, it is possible to distinguish urgent joint and cable heating faults, and the end high-resistance problem can be planned before the peak electricity consumption period, making on-site identification and defect elimination more planned.

[0053] Preferably, the end anomaly degree is determined according to the anomaly situation corresponding to the end of the meter, including the following steps: From the list where the front-end anomaly risk value of the meter is greater than the set threshold H, extract the end risk-causing factors corresponding to the anomaly situation, and sum according to the weight coefficients corresponding to each dimension data in the end risk-causing factors to determine the end anomaly degree; among them, the weight coefficients of each dimension data in the end risk-causing factors are dynamically adjusted according to the electricity consumption of actual users and the voltage fluctuation situation of the substation area; each dimension data in the end risk-causing factors includes the standard deviation of resistance constancy, theoretical voltage drop power, voltage-current inverse correlation coefficient, front-end anomaly degree, theoretical resistance, and user co-drop event.

[0054] In this embodiment, the end anomaly is characterized and dynamically evaluated by combining the end risk-causing factors. Through the coupled analysis of multi-dimensional risk-causing factors and dynamic weight adaptation, the accurate quantitative grading and evolution trend prediction of end hidden dangers are realized, significantly improving the identification efficiency of end hidden dangers.

[0055] As an alternative embodiment, determining the end anomaly degree according to the end risk-causing factors includes the following steps: B1. For the weight coefficient h1 corresponding to the standard deviation of resistance constancy, when 0.3 > the standard deviation of resistance constancy > 0, the weight coefficient h1 is linearly distributed in the range of 40 to 10.

[0056] B2. For the weight coefficient h2 corresponding to the voltage-current inverse correlation coefficient, when -0.7 > the voltage-current inverse correlation coefficient > -1, the weight coefficient h2 is linearly distributed in the range of 10 to 40.

[0057] B3. For the weight coefficient h3 corresponding to the theoretical voltage drop power, when 300W < theoretical voltage drop power < 600W, the weight coefficient h3 is linearly distributed in the range of 10 - 40. When the theoretical voltage drop power is greater than or equal to 600W, the weight coefficient is fixed at 40.

[0058] B4. For the weight coefficient h4 corresponding to the front-end anomaly degree, in the case of three-phase users with more than 2 phases experiencing momentary voltage drops, and when the difference between the maximum and minimum values of the corresponding front-end anomaly degree is within the set threshold of 40, and corresponding to the momentary voltage drop , the weight coefficient h4 is set to 20.

[0059] B5. For the weight coefficient h5 corresponding to the users' simultaneous voltage drop event, when it is determined to be multi-user simultaneous voltage drop; the corresponding weight coefficient h5 is set to 20.

[0060] It can be understood that on the basis that the front-end hidden danger risk value is greater than the set threshold H, based on the analysis dimensions B1 - B5, the end hidden danger risk value is calculated. When the end hidden danger risk value is greater than the set threshold Z, for example, the threshold = 40, it is determined that there is an end hidden danger. By comprehensively analyzing multi-dimensional data, the reliability of the analysis results can be improved.

[0061] The front-end risk weight corresponding to the front-end anomaly of the meter and the end risk weight corresponding to the end anomaly of the meter are determined through a linear regression model combined with the anomaly situation data corresponding to the front-end anomaly degree and the end anomaly degree.

[0062] Specifically, the training process of the linear regression model is as follows: Obtain the front-end risk factors corresponding to the front-end anomaly degree as the input of the linear regression model, and train the linear regression model with the maximum front-end hidden danger risk value as the optimization target to obtain the front-end risk weight of each front-end risk factor; when the front-end hidden danger risk value is greater than the set threshold H (for example, the threshold H = 80), use the end samples and pre-meter fault samples after on-site judgment and classification in the front-end risk sample list (including all end risk factors of each sample) as the input of the linear regression model, and train the linear regression model with the maximum end hidden danger risk value as the optimization target to obtain the end risk weight of each end risk factor; Among them, the front-end risk factors include multi-dimensional data, specifically including: momentary voltage drop difference, low voltage, theoretical resistance, theoretical resistance change trend, theoretical resistance standard deviation, momentary voltage drop times, synchronous power outage event, synchronous daily line loss sudden increase event, voltage-current anti-correlation coefficient, theoretical power, momentary voltage drop event occurrence period, joint melting event; The end risk factors include multi-dimensional data, specifically including: resistance constant standard deviation, theoretical voltage drop power, voltage-current anti-correlation coefficient, front-end anomaly degree, theoretical resistance, and users' simultaneous voltage drop event.

[0063] In this embodiment, the target of pre-meter fault training is front-end problems (including end problems), that is, the hidden risks at the front end are the greatest; after the front-end risk weights of the front-end risk factors corresponding to the pre-meter faults are trained, according to the on-site investigation of the pre-meter faults, collect the end conditions in the pre-meter fault data set, classify the samples of the pre-meter fault data set into pre-meter faults and end faults, and use the classified samples of the pre-meter fault data set combined with the end risk factors corresponding to the end anomaly degree to train the linear regression model at the end to obtain the end risk weights of the end risk factors. After training, in the list where the pre-meter fault risk value is greater than the set threshold H (for example, the threshold H = 80), use the trained end recognition model to identify the hidden end users among the high-risk users (calculate the risk value by combining the end risk weights with 5 dimensions of end risk factors, and when the risk value is greater than the set threshold, determine the hidden end users among the high-risk users). Through 12 front-end risk factors such as the number of momentary drops and the trend of theoretical resistance change, quantify the dynamic evolution law of hidden dangers such as joint oxidation and loosening, make the calculation of the front-end risk value strongly correlated with the actual deterioration speed, and improve the accuracy of front-end hidden danger identification. When the front-end hidden danger risk > 80, combine 5 end risk factors such as the user's simultaneous drop event and the constant standard deviation of resistance to dynamically correct the end weight allocation and avoid the problem that the front-end high risk masks the end hidden danger.

[0064] Determine the front-end hidden danger risk value according to the front-end anomaly of the meter, the front-end anomaly degree, and the front-end risk weight or / and the end hidden danger risk value determined according to the end anomaly of the meter, the end risk weight, and the end anomaly degree.

[0065] It can be understood that the front-end hidden danger risk value can be determined by weighted summation through the front-end anomaly of the meter, the front-end anomaly degree, and the front-end risk weight; similarly, the end hidden danger risk value determined by weighted summation through the end anomaly of the meter, the end risk weight, and the end anomaly degree.

[0066] Determine the corresponding hidden danger emergency strategy for the low-voltage power distribution area according to the front-end hidden danger risk value or / and the end hidden danger risk value.

[0067] It can be understood that pre-meter faults (front-end hidden dangers) need to be urgently eliminated according to the risk value, and the end with a long power supply radius (end hidden dangers) needs to be eliminated before the arrival of high-load periods such as summer peak load or winter peak load. It should be noted that the elimination means is only related to the type of hidden danger and has nothing to do with the size of the risk value.

[0068] Embodiment 2: Another technical solution provided in the embodiment of the present invention is a low-voltage power distribution area hidden danger identification system, as Figure 2 shown, including: The first construction module 101: sequentially obtain the measurement data of the low-voltage power distribution area according to the sampling period to construct an initial data set; The second construction module 102: Screen the initial data set in sequence according to the momentary drop event criterion to construct an abnormal data set; The first analysis module 103: Analyze dimensions such as the connector melting recognition algorithm according to the momentary drop event to determine the abnormality at the front end of the meter and its corresponding front-end abnormality degree; The second analysis module 104: On the basis that the front-end hidden danger risk value is greater than the set threshold H, analyze the abnormal data set according to the end abnormality recognition algorithm to determine the abnormality at the end of the meter and its corresponding end abnormality degree; The calculation module 105: Determine the front-end risk weight corresponding to the front-end abnormality of the meter and the end risk weight corresponding to the end abnormality of the meter through the analytic hierarchy process; The first fusion module 106: Determine the front-end hidden danger risk value according to the front-end abnormality of the meter, the front-end abnormality degree, and the front-end risk weight; The second fusion module 107: Determine the end hidden danger risk value according to the end abnormality of the meter, the end risk weight, and the end abnormality degree; The execution module 108: Determine the corresponding hidden danger emergency strategy for the power distribution area according to the front-end hidden danger risk value or / and the end hidden danger risk value.

[0069] This embodiment has at least the following substantial technical effects: In this embodiment, the momentary drop event criterion is used to analyze the data of the power distribution area obtained to construct an abnormal data set. For the key degradation characteristics in the abnormal data set (including: front-end characteristic data and end characteristic data), the abnormal degree of the front-end contact resistance is quantified through the connector melting recognition algorithm, and the load fluctuation characteristics are analyzed in combination with the end abnormality recognition algorithm to separate the front-end / end fault modes; and a linear regression model is used to associate the front-end / end abnormality degree with historical degradation data, dynamically allocate the risk weights of each risk factor, and then calculate the front-end hidden danger risk value and the end hidden danger risk value respectively according to the abnormality degree, abnormality type, and corresponding risk weights, constructing a two-dimensional evaluation system to solve the technical problems of insufficient decoupling of abnormal characteristics (that is, unable to distinguish front-end contact degradation from end overload), static risk weights (that is, ignoring the dynamic evolution of hidden dangers), and fuzzy hidden danger positioning (that is, relying on manual investigation) in the hidden danger identification of low-voltage power distribution areas, significantly improving the accuracy and efficiency of hidden danger identification, and providing a reliable data basis for the hidden danger prediction and response strategy formulation of low-voltage power distribution areas.

[0070] Embodiment 3: Thirdly, a technical solution provided in an embodiment of the present invention is an electronic device, including a memory and a processor. A computer program is stored in the memory, and when the processor calls the computer program in the memory, the steps of the method for identifying hidden dangers in low-voltage power distribution areas based on degradation risk analysis are implemented.

[0071] Embodiment 4: Fourthly, a technical solution provided in an embodiment of the present invention is a storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are loaded and executed by a processor, the steps of the method for identifying hidden dangers in a low-voltage power distribution area based on deterioration risk analysis are implemented.

[0072] Through the description of the above embodiments, those skilled in the art can understand that for the convenience and brevity of description, only the division of the above function modules is used as an example. In actual applications, the above functions can be allocated to different function modules according to needs, that is, the internal structure of a specific device is divided into different function modules to complete all or part of the functions described above.

[0073] In the embodiments provided in the present application, it should be understood that the disclosed structure and method can be implemented in other ways. For example, the embodiments of the structure described above are only illustrative. For example, the division of modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another structure, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces, and the indirect coupling or communication connection of structures or units can be in an electrical, mechanical or other forms.

[0074] The units described as separate components may or may not be physically separated. The components displayed as units may be a physical unit or multiple physical units, that is, they may be located in one place, or they may be distributed to multiple different places. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0075] In addition, each functional unit in the embodiments of the present application can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0076] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solution of the embodiments of the present application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions for causing a device (which can be a single-chip microcomputer, a chip, etc.) or a processor to execute all or part of the steps of the methods of the various embodiments of the present application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0077] The above-described specific implementation manners are the preferred implementation manners of the method and system for identifying hidden dangers in low-voltage power distribution areas based on deterioration risk analysis of the present invention, and do not limit the specific scope of the present invention. The scope of the present invention includes but is not limited to this specific implementation manner. All equivalent changes made according to the shape and structure of the present invention are within the protection scope of the present invention.

Claims

1. A method for identifying hidden dangers in low-voltage areas based on degradation risk analysis, characterized by: The steps include: Acquire the measured data of the low-voltage area in sequence according to the sampling period to construct the initial data set; filter the initial data set in sequence according to the instantaneous drop event judgment criteria to construct the abnormal data set; The abnormal data set is analyzed according to the joint melting recognition algorithm to determine the front-end abnormality of the meter and its corresponding front-end abnormality degree; when the trigger condition is met, the abnormal data set is analyzed according to the terminal abnormality recognition algorithm to determine the terminal abnormality of the meter and its corresponding terminal abnormality degree; The front-end risk weight corresponding to the front-end abnormality of the meter and the terminal risk weight corresponding to the terminal abnormality of the meter are determined by combining the abnormal situation data corresponding to the front-end abnormality degree and the terminal abnormality degree through a linear regression model; Determine the front-end hidden danger risk value based on the front-end abnormality of the meter, the degree of the front-end abnormality and the front-end risk weight, or / and determine the terminal hidden danger risk value based on the terminal abnormality of the meter, the terminal risk weight and the degree of the terminal abnormality; Determine the corresponding hidden danger emergency strategy for the substation based on the front-end hidden danger risk value and / or the terminal hidden danger risk value.

2. The method for identifying hidden dangers in low voltage substations based on degradation risk analysis according to claim 1 is characterized in that: The measured data include total meter voltage, user voltage, and user current.

3. The method for identifying hidden dangers in low voltage substations based on degradation risk analysis according to claim 1 or 2, characterized in that: The instantaneous drop event criterion includes: Criterion A1: Use a sliding window with a width of 2 measuring points to extract data, and the voltage drop between the two measuring points in the window is greater than 5V; Criterion A2: The absolute value of the voltage difference between the total meter voltage and the user voltage is greater than 5V; Criteria A3: The current at the low voltage point is greater than 0.1A, and (current at the low voltage point - current at the high voltage point) / current at the high voltage point > 50%, and (current at the low voltage point - current at the high voltage point) > the average value of the maximum 5 currents at the same measurement point in unit time 10%; Criteria A4: If the voltage of three-phase users drops simultaneously, the voltage drop event is excluded and classified as the end user; Criteria A5: Obtain the current when the voltage drops to a low voltage in a transient sag event. If the current disappears within 12 measurement moments at the same measurement point after the transient sag, or the current at the start of the transient sag is less than 20% of the average current at the low voltage point, or the current is less than 0.1A and the voltage has not recovered to the high voltage point, the transient sag event is ruled out; Criterion A6: User voltage difference Current>1000w, instantaneous drop events are excluded; If the above criteria are met at the same time, the instantaneous drop event is established; The instantaneous drop event includes the voltage and current data of the two measurements before and after, and the voltage and current data include the current at the low voltage point, the voltage at the low voltage point, the current at the high voltage point, and the voltage at the high voltage point.

4. The method for identifying hidden dangers in low voltage substations based on degradation risk analysis according to claim 1 is characterized in that: Analyzing the abnormal data set according to the joint melting identification algorithm to determine the front-end abnormality of the meter and its corresponding front-end abnormality degree includes the following steps: Extract the measurement data in the abnormal data set and calculate the front-end feature data. Use the joint melting recognition algorithm to compare and analyze the front-end feature data to determine the front-end abnormality of the meter. The degree of front-end abnormality is determined according to the abnormal situation corresponding to the front-end abnormality of the meter; wherein the front-end characteristic data at least includes the instantaneous voltage drop difference, the voltage-current anti-correlation coefficient, the current average value and the instantaneous voltage drop high point average value of each measuring point.

5. The method for identifying hidden dangers in low voltage substations based on degradation risk analysis according to claim 4 is characterized in that: The front-end feature data is compared and analyzed by the joint melting identification algorithm to determine the front-end abnormality of the meter, including the following steps: When a single instantaneous voltage drop The current is >5W, and the voltage-current anti-correlation coefficient is less than -0.5 7 days before the melting judgment, and the single current is continuously greater than the current at the low point of the instantaneous voltage drop. 80%, and the voltage returns to the high point before the instantaneous drop, it is judged that there is a melting abnormality at the front end of the meter; When a single instantaneous drop ends, the current corresponding to m or more sampling moments within the range of n sampling points is greater than the average current of all instantaneous voltage low points in unit time. 120%, and at the same time the voltage is greater than the average value of the high point of the instantaneous voltage drop, it is judged that there is a melting abnormality at the front end of the meter.

6. The method for identifying hidden dangers in low voltage substations based on degradation risk analysis according to claim 4 is characterized in that: Determining the degree of front-end abnormality according to the abnormal situation corresponding to the front-end abnormality of the meter includes the following steps: The front-end risk factors corresponding to the abnormal situation are extracted, and the degree of front-end abnormality is determined by accumulating the weight coefficients corresponding to the data of each dimension in the front-end risk factors; among which, the weight coefficients of the data of each dimension in the front-end risk factors are dynamically adjusted according to the user's electricity consumption and the voltage fluctuation of the substation; the data of each dimension of the front-end risk factors include: instantaneous sag voltage difference, low voltage, theoretical resistance, theoretical power group change trend, theoretical power group standard deviation, number of instantaneous sags, synchronous power outage events, synchronous daily line loss sudden increase events, voltage and current anti-correlation coefficient, theoretical power, instantaneous sag event occurrence time period, and joint melting events.

7. The method for identifying hidden dangers in low voltage substations based on degradation risk analysis according to claim 1 is characterized in that: When the trigger condition is met, the abnormal data set is analyzed according to the terminal abnormality recognition algorithm to determine the terminal abnormality of the meter and its corresponding terminal abnormality degree; the steps include: End user identification is performed within the user range where the risk of front-end hidden dangers is determined to be greater than the set threshold H, and the measurement data in the abnormal data set is extracted and the terminal feature data is calculated and obtained. The terminal feature data is compared and analyzed by the terminal abnormality identification algorithm to determine the terminal abnormality of the meter; The degree of terminal abnormality is determined according to the abnormal situation corresponding to the abnormality of the meter terminal; wherein the terminal characteristic data at least includes the constant standard deviation of resistance, theoretical voltage drop power, voltage-current anti-correlation coefficient, voltage drop users and theoretical resistance of each measuring point.

8. The method for identifying hidden dangers in low voltage substations based on degradation risk analysis according to claim 7 is characterized in that: The terminal abnormality recognition algorithm is used to compare and analyze the terminal characteristic data to determine the terminal abnormality of the meter; The steps include: When the user resistance constant standard deviation is <0.3, it is included in the meter terminal abnormal analysis; When the user voltage and current anti-correlation coefficient ranges from -0.7 to -1, it is included in the abnormal analysis of the meter terminal; When the single-phase theoretical power range of the user's instantaneous drop event is greater than 300w, it is included in the abnormal analysis of the meter terminal; among them, when a user in the same area has an instantaneous drop event, and the voltage fluctuations of other users are close to that of the user, these users are judged to be the same drop users; When more than two phases of a three-phase user have front-end abnormalities, and the instantaneous drop phase has |maximum theoretical resistance - minimum theoretical resistance| / maximum theoretical resistance < 0.3, it will be included in the meter terminal abnormality analysis.

9. The method for identifying hidden dangers in low voltage substations based on degradation risk analysis according to claim 7 is characterized in that: Determining the degree of terminal abnormality according to the abnormal situation corresponding to the terminal abnormality of the meter includes the following steps: In the list of meters whose front-end abnormal risk values ​​are greater than the set threshold value H, the terminal risk factor corresponding to the abnormal situation is extracted, and the terminal abnormality degree is determined by summing up the weight coefficients corresponding to the various dimensional data in the terminal risk factor; among which, the weight coefficients of the various dimensional data in the terminal risk factor are dynamically adjusted according to the actual power consumption of the user and the voltage fluctuation of the substation; the various dimensional data in the terminal risk factor include: constant standard deviation of resistance, theoretical voltage drop power, voltage-current anti-correlation coefficient, front-end abnormality degree, theoretical resistance and user same-drop events.

10. The method for identifying hidden dangers in low voltage substations based on degradation risk analysis according to claim 1, characterized in that: The training process of the linear regression model is: The front-end risk factor corresponding to the front-end abnormality degree is obtained as the input of the linear regression model, and the linear regression model is trained with the maximum front-end hidden danger risk value as the optimization target to obtain the front-end risk weight of each front-end risk factor; When the front-end hidden danger risk value is greater than the set threshold H, the terminal samples and pre-table fault samples in the front-end risk sample list that have been classified by on-site judgment are used as the input of the linear regression model, and the maximum terminal hidden danger risk value is used as the optimization target to train the linear regression model to obtain the terminal risk weight of each terminal risk factor; Among them, the front-end risk factors include data of multiple dimensions, including: instantaneous voltage drop difference, low voltage, theoretical resistance, theoretical power group change trend, theoretical power group standard deviation, instantaneous drop times, synchronous power outage events, synchronous daily line loss sudden increase events, voltage and current anti-correlation coefficient, theoretical power, instantaneous drop event time period, and joint melting events; The terminal risk factors include data in multiple dimensions, including: constant standard deviation of resistance, theoretical voltage drop power, voltage-current anti-correlation coefficient, front-end abnormality, theoretical resistance, and user same-drop events.

11. A low voltage area hidden danger identification system, applicable to the low voltage area hidden danger identification method based on degradation risk analysis as claimed in any one of claims 1 to 10, characterized in that: include: The first construction module: sequentially obtains the measurement data of the low-voltage area according to the sampling period to construct an initial data set; The second construction module: according to the instantaneous drop event judgment criteria, the initial data set is screened in turn to construct an abnormal data set; The first analysis module: according to the instantaneous drop event, the joint melting identification algorithm analysis and other analysis dimensions are performed to determine the front-end abnormality of the meter and its corresponding front-end abnormality degree; The second analysis module: on the basis that the risk of hidden dangers at the front end is greater than the set threshold value H, the abnormal data set is analyzed according to the terminal abnormality recognition algorithm to determine the terminal abnormality of the meter and its corresponding terminal abnormality degree; Calculation module: Determine the front-end risk weight corresponding to the front-end abnormality of the meter and the terminal risk weight corresponding to the terminal abnormality of the meter through the hierarchical analysis method; The first fusion module: determines the front-end hidden danger risk value according to the front-end abnormality of the meter, the degree of the front-end abnormality and the front-end risk weight; The second fusion module: the terminal hidden danger risk value is determined according to the terminal abnormality of the meter, the terminal risk weight, and the degree of terminal abnormality; Execution module: Determine the hidden danger emergency strategy corresponding to the substation according to the front-end hidden danger risk value and / or the terminal hidden danger risk value.

12. An electronic device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and when the processor calls the computer program in the memory, it implements the steps of the low-voltage substation hidden danger identification method based on degradation risk analysis as described in any one of claims 1 to 10.

13. A storage medium, characterized in that: The storage medium stores computer executable instructions, which, when loaded and executed by a processor, implement the steps of a method for identifying hidden dangers in a low-voltage substation based on degradation risk analysis as described in any one of claims 1 to 10.

Images