Transformer area abnormity monitoring method based on data analysis

Abstract

The invention discloses a transformer area anomaly monitoring method based on data analysis, and the method comprises the steps: taking the power supply amount as the y axis, taking the power sale amount as the x axis, and building a plane coordinate system; in a plane-coordinate system, taking an oblique line with the slope equal to 1 as a boundary line, and taking a region above the left of the oblique line as a high-loss region; the slope range of the high-quality transformer area is determined; setting data points formed by the power supply quantity and the power sale quantity of the plurality of transformer areas in a certain time period into a plane-coordinate system, and calculating data slopes of the transformer areas in two adjacent days; from the high-loss area, screening out a target area with the data slope not in the slope range, and taking the area with the most corresponding data points and the maximum data slope difference in the target area as an abnormal area; and fitting the power loss of the transformer area and the power consumption of the users in the abnormal transformer area, and taking the users of which the fitting degrees are greater than a fitting degree threshold as abnormal users. According to the method, the accuracy and the efficiency of monitoring the abnormity of the transformer area can be remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of electric power monitoring, and in particular to a method for monitoring abnormalities in a substation based on data analysis. Background Art

[0002] At present, the basic management of the power environment in many areas is weak, and the use of a large number of old meters and meter boxes leads to frequent metering failures and errors, and station and line failures also occur from time to time. This not only increases the difficulty of line loss management, but also brings a bad power experience to users. It can be seen from the service work orders that there are many fault reports and complaints, and the line losses in the stations where users belong are affected.

[0003] The current abnormal monitoring of the substation mainly relies on manual screening to monitor the current, voltage, power and line loss of each user and gateway electric energy meter. The monitoring range is large, the monitoring frequency is low, the overall efficiency is low, and there are even monitoring blind spots. There are problems such as low accuracy and efficiency of abnormal monitoring of the substation. Therefore, in order to solve the above problems, a method for abnormal monitoring of the substation based on data analysis is needed, which can significantly improve the accuracy and efficiency of abnormal monitoring of the substation. Summary of the invention

[0004] In view of this, an object of the present invention is to overcome the defects in the prior art and provide a method for monitoring abnormalities in substations based on data analysis, which can significantly improve the accuracy and efficiency of monitoring abnormalities in substations.

[0005] The data analysis-based abnormal monitoring method of the substation area of ​​the present invention comprises:

[0006] With the power supply as the y-axis and the power sales as the x-axis, a plane coordinate system is established;

[0007] In the plane coordinate system, a slash line with a slope equal to 1 is used as the boundary, and the area to the upper left of the slash line is used as the high-loss area; wherein the slash line starts from the origin of the coordinate system;

[0008] Determine the slope range where the high-quality areas are located;

[0009] The data points formed by the power supply and power sales of several substations in a certain period of time are set in a plane coordinate system, and the data slope of the substations between two adjacent days is calculated;

[0010] From the high-loss area, select the target area whose data slope is not within the slope range, and take the area with the most corresponding data points and the largest data slope difference in the target area as the abnormal area;

[0011] For users in abnormal areas, the power loss in the area and the power consumption of the users are fitted, and users whose fitting degree is greater than the fitting degree threshold are regarded as abnormal users.

[0012] Furthermore, the slope range of the high-quality area is determined according to the following method:

[0013] Select a number of high-quality areas;

[0014] Calculate the slopes corresponding to several high-quality areas and count the number of different slopes;

[0015] The slope with the largest number is taken as the central value K0, and the slope range is (K0-r, K0+r).

[0016] Furthermore, the center value K0 is 1, and the slope range is (0.9, 1.1).

[0017] Furthermore, the slope of the target area is greater than 1.2 or less than 0.8.

[0018] Further, the data points corresponding to the area where the abnormality has been found are set in the plane coordinate system, and the data points are subjected to linear regression processing to obtain the linear regression equation corresponding to the area where the abnormality has been found;

[0019] Taking the slope corresponding to the linear regression equation as a reference, the slope of the target area is adjusted to (1.2,∞)U(0.4,0.8).

[0020] The beneficial effects of the present invention are as follows: the present invention discloses a method for monitoring abnormalities in substations based on data analysis, and an analysis method based on big data, which significantly improves the accuracy and efficiency of monitoring abnormalities in substations. Compared with traditional means of monitoring abnormalities in substations, the data analysis method of the present invention can realize automated and intelligent data monitoring and analysis, greatly improving the efficiency of abnormal substation management. In the face of the continuous advancement of future science and technology and the increasing diversification of electricity theft methods, it can flexibly respond to new electricity theft methods and continuously improve the response capabilities of anti-electricity theft work. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0022] Figure 1 It is a schematic diagram of the process of the abnormal monitoring method of the substation area of ​​the present invention;

[0023] Figure 2 It is the slope distribution diagram of the high-quality area of ​​the present invention;

[0024] Figure 3 It is a schematic diagram of the slope characteristic model of the present invention;

[0025] Figure 4 It is a schematic diagram of the characteristic line of the station area in the high loss area of ​​the present invention;

[0026] Figure 5It is a schematic diagram comparing the linear simulation line of the abnormal area data point of the present invention and the set slope line. DETAILED DESCRIPTION

[0027] The present invention is further described below in conjunction with the accompanying drawings, as shown in the drawings:

[0028] This embodiment discloses a method for monitoring abnormality in a substation area based on data analysis, comprising the following steps:

[0029] With the power supply as the y-axis and the power sales as the x-axis, a plane coordinate system is established;

[0030] In the plane coordinate system, a slash line with a slope equal to 1 is used as the boundary, and the area to the upper left of the slash line is used as the high-loss area; wherein the slash line starts from the origin of the coordinate system;

[0031] Determine the slope range where the high-quality areas are located;

[0032] The data points formed by the power supply and power sales of several substations in a certain period of time are set in a plane coordinate system, and the data slope of the substations between two adjacent days is calculated;

[0033] From the high-loss area, select the target area whose data slope is not within the slope range, and take the area with the most corresponding data points and the largest data slope difference in the target area as the abnormal area;

[0034] For users in abnormal areas, the power loss in the area and the power consumption of the users are fitted, and users whose fitting degree is greater than the fitting degree threshold are regarded as abnormal users.

[0035] The present invention uses the ratio between the change in power supply and the change in power sales as the slope in the plane coordinate system, and uses the slope to express the relationship between the three data of power supply, power sales and line loss rate, and uses the slope to objectively reflect the real-time changes and fluctuation trends of the data. The slope itself is the derivative of the plane graph, which reflects the change trend. By observing the change of the slope (or derivative), the overall trend of the change in power sales with the power supply can be understood, and then the abnormal monitoring of the substation area can be realized.

[0036] In this embodiment, the slope range of the high-quality area is determined according to the following method:

[0037] Select several high-quality substations; calculate the slopes corresponding to the several high-quality substations, wherein each high-quality substation may correspond to multiple data points, and multiple slopes may be obtained by calculating the slopes between data points of adjacent dates; the data points are data points formed by the power supply and the power sales, and the slopes may be calculated by applying them to the plane coordinate system constructed by the present invention.

[0038] In the power system, high-quality transformer areas usually refer to transformer power supply ranges or areas that have excellent performance in line loss management, power supply quality, equipment maintenance, etc., and can provide power services to users stably and efficiently. These transformer areas often have lower line loss rates, higher power supply reliability and good user satisfaction.

[0039] According to the above processing, if Figure 2 As shown, the number of different slopes is counted;

[0040] The slope with the largest number is taken as the central value K0, and the slope range is (K0-r, K0+r). Wherein, the central value K0 is 1, and the slope range can be (0.9, 1.1).

[0041] In this embodiment, the oblique line with a slope equal to 1 is used as the boundary to divide into three areas, namely, the area with loss, the area with zero line loss, and the area with negative loss. Figure 3 As shown in the figure, the daily power supply and sales volume of the substation corresponds to a point on the coordinate system. Figure 3 There are 6 points marked in the figure. Without considering the reasons of substation management, the power supply and sales points in the high-loss substation will only fall in the area with slope > 1. Therefore, in the later process of judging the theft power area, the points with line loss of 0 and negative loss area must be excluded first.

[0042] By connecting the points of power supply and sales between two adjacent days, we can obtain 29 slope data for the past 30 days and compare them with the slope range of high-quality substations. The line loss slope of high-quality substations has been obtained to be between 0.9 and 1.1, so the screening range of substations is set to be greater than 1.2 or less than 0.8. That is, the slope of the target substation is greater than 1.2 or less than 0.8. The substation with the highest frequency of data points and the largest difference in data slope is regarded as an abnormal substation. Then, the users in the abnormal substation are fitted with the power loss of the substation (or line loss rate curve) and the power consumption of the users. Users with a fitting degree greater than the fitting degree threshold are regarded as abnormal users. For abnormal users with a higher fitting degree, on-site investigation is carried out. The fitting degree threshold can be set according to the actual working conditions.

[0043] Fitting here refers to comparing and analyzing the user's power consumption data with the power loss data (or line loss rate) of the substation through mathematical methods (such as regression analysis, curve fitting, etc.) to find the correlation between them. If a user's power consumption data and the power loss data (or line loss rate) of the substation show a high correlation or similar trend of change, then it can be considered that the user's power consumption behavior has a greater impact on the substation loss and is marked as an abnormal user.

[0044] In this embodiment, only the slope between two adjacent days is referred to, which has great limitations for the slope feature model. Although the described points are derived from the historical daily data, the essence of the slope feature model has nothing to do with time. The purpose of gathering multiple points is to more accurately describe the characteristics of the radio station stealing area.

[0045] In order to further improve the accuracy of the slope value of the target area, and thus more accurately screen out abnormal areas, many discrete supply and sales points can be gathered together and uniformly drawn as follows: Figure 4 The characteristic lines of the substations shown are used to establish a substation characteristic slope library linked to the stolen substations, and the characteristic range of the stolen substations is obtained, and the abnormal substations are found by linear regression.

[0046] like Figure 5 As shown, the data points corresponding to the substations where abnormalities (such as electricity theft) have been found are set in the plane coordinate system, and linear regression fitting is performed on these data points to obtain the linear regression equation corresponding to the substations where abnormalities have been found, as shown in FIG. Figure 5 The optimized model threshold line shown in the figure; the regression equation is a mathematical expression used in statistics to describe the relationship between one or more independent variables and a dependent variable. In this scenario, the regression equation is used to fit or simulate the changing trend of the slope graph curve. Through regression analysis, an optimal regression equation can be found to minimize the difference between the predicted slope and the actual observed slope.

[0047] The slope corresponding to the linear regression equation is used as a reference. Figure 5 The regression k value line of the middle-stage radio area is 0.76. Since the fitting degree is 0.6261, the slope range should be between 0.4 and 0.8, that is, the slope of the adjusted target area is (1.2,∞)U(0.4,0.8).

[0048] In order to better understand the abnormal monitoring method of the station area of ​​the present invention, the following examples are further described:

[0049] The slope characteristic model method of the present invention was used to screen 1,805 key areas. As shown in Table 1, the results of the investigation and punishment of electricity theft by each pilot unit are statistically analyzed. Among them, Company A generated 34 key areas, verified 1 household for electricity theft, and the economic effect was 478.4 yuan; Company B generated 49 key areas, verified 13 households for electricity theft, and the economic effect was 104,900 yuan; Company C generated 87 key areas, verified 22 households for electricity theft, and the economic effect was 276,000 yuan; Company D generated 29 key areas, verified 2 households for electricity theft, and the economic effect was 36,300 yuan. From the perspective of user type, this method mainly seized customers who directly bypassed the metering device, accounting for 71.42%.

[0050] Table 1

[0051]

[0052] In addition to preventing electricity theft, we established a record of power distribution areas with fluctuations within 30 days through daily cycles, and promptly dealt with the areas that suddenly appeared in the record; as shown in Table 2, from September to December, a high-loss area achieved the established target value of less than 10 units. At the same time, we went the other way and dealt with 2 distribution network fault points and 2 metering errors in advance, recovering more than 12,000 yuan in audit results.

[0053] Table 2

[0054] Management Unit September high loss Loss in September High loss in October Loss in October High loss in November November loss High loss in December Loss in December a place 2 0 7 2 3 0 6 0 b place 0 0 0 1 0 0 1 0 c place 0 0 0 0 0 0 0 0 D 0 0 1 0 1 0 0 0 e-land 1 0 0 0 1 0 0 0 f 0 0 2 0 0 0 2 0 g land 0 0 1 1 0 0 0 0 h 0 0 1 0 0 0 2 0 i place 0 0 2 0 0 0 0 0 j land 1 0 0 0 1 0 1 0

[0055] Table 3 is a table of daily line loss curves of a certain substation. Two days before the substation abnormality, an abnormal value with a slope of 1.74 appeared. It can be seen that the method of the present invention plays a predictive role in monitoring the abnormality of the substation.

[0056] Table 3

[0057]

[0058] Table 4 shows the slope record of a certain substation. This substation began to be monitored on June 23. Before July 5, its slope was out of the recognized range (k∈(1.2,∞)U(0.4,0.8)) for only one day. After the investigation on July 5, the slope value returned to the high-quality substation range and was close to 1.

[0059] Table 4

[0060]

[0061] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

Claims

1. A method for monitoring abnormality in a substation based on data analysis, characterized in that: include: With the power supply as the y-axis and the power sales as the x-axis, a plane coordinate system is established; In the plane coordinate system, a slash line with a slope equal to 1 is used as the boundary, and the area to the upper left of the slash line is used as the high-loss area; wherein the slash line starts from the origin of the coordinate system; Determine the slope range where the high-quality areas are located; The data points formed by the power supply and power sales of several substations in a certain period of time are set in a plane coordinate system, and the data slope of the substations between two adjacent days is calculated; From the high-loss area, select the target area whose data slope is not within the slope range, and take the area with the most corresponding data points and the largest data slope difference in the target area as the abnormal area; For users in abnormal areas, the power loss in the area and the power consumption of the users are fitted, and users whose fitting degree is greater than the fitting degree threshold are regarded as abnormal users.

2. The method for monitoring abnormality in a substation area based on data analysis according to claim 1, characterized in that: Determine the slope range of the high-quality area according to the following method: Select a number of high-quality areas; Calculate the slopes corresponding to several high-quality areas and count the number of different slopes; The slope with the largest number is taken as the central value K0, and the slope range is (K0-r, K0+r).

3. The method for monitoring abnormality in a substation area based on data analysis according to claim 2, characterized in that: The center value K0 is 1, and the slope range is (0.9, 1.1).

4. The method for monitoring abnormality in a substation area based on data analysis according to claim 3 is characterized in that: The slope of the target platform is greater than 1.2 or less than 0.

8.

5. The method for monitoring abnormality in a substation area based on data analysis according to claim 4 is characterized in that: The data points corresponding to the area where the abnormality has been found are set in the plane coordinate system, and the linear regression processing is performed on these data points to obtain the linear regression equation corresponding to the area where the abnormality has been found; Taking the slope corresponding to the linear regression equation as a reference, the slope of the target area is adjusted to (1.2,∞)U(0.4,0.8).

Images