CN-115951109-B - Power theft judgment method and system based on secondary circuit of specially-changed user current

CN115951109BCN 115951109 BCN115951109 BCN 115951109BCN-115951109-B

Abstract

The invention discloses a power stealing judging method based on a private transformer user current secondary circuit, which comprises the steps of obtaining data information of a historical normal user and a power stealing user from the private transformer user, calculating current values, current ratios, current differences and current ratio differences of the power stealing user and the normal user, analyzing step difference data to obtain a ratio dividing threshold value and a current dividing threshold value, obtaining actual user power consumption data, judging an effective section of the actual user, judging an unbalanced section, and finishing power stealing behavior judgment of the actual user according to the unbalanced section. The invention also discloses a system for realizing the electricity larceny judging method based on the secondary circuit of the private transformer user current. The invention calculates the judging threshold value through short historical data and judges the electricity larceny according to the obtained judging threshold value, so that the invention does not need a large amount of historical data, and has high accuracy, good reliability, simplicity and convenience.

Inventors

ZHANG RENDONG
Zhou Shanyan
She Liwei
Zeng Youfeng
Tang Eting
ZENG MINRUI
Zeng Binxiu
LI XIANWEI
ZOU ZHIHUA
WU QUNFEI
ZHOU SHIJUN
LIU LINPING
HUANG WEIYUE

Assignees

国网湖南省电力有限公司
国网湖南省电力有限公司衡阳供电分公司
国家电网有限公司

Dates

Publication Date: 20260505
Application Date: 20221208

Claims (10)

1. A power theft judging method based on a secondary circuit of a private transformer user current comprises the following steps: s1, acquiring data information of a history normal user and data information of a power stealing user from a special transformer user; S2, calculating the ratio of the maximum value to the minimum value of the A/C two-phase current of the electricity stealing user and the normal user at each moment and the maximum current value corresponding to the electricity stealing user and the normal user at each moment according to the data information obtained in the step S1; s3, calculating the difference between the maximum value and the minimum value of the current ratio data and the maximum current value data of the two types of users according to the data obtained by calculation in the step S2; s4, equally dividing the difference data obtained by the calculation in the step S3, and carrying out joint distribution analysis on the electricity stealing users and the normal users according to the division result to obtain a ratio division threshold value and a current division threshold value; s5, acquiring electricity utilization data of an actual user; s6, judging an effective interval of an actual user according to the electricity consumption data obtained in the step S5 and the ratio segmentation threshold value and the current segmentation threshold value obtained in the step S4; S7, judging an unbalanced section according to the maximum value of the current in the effective section obtained in the step S6; s8, according to the unbalanced interval obtained in the step S7, the electricity stealing behavior judgment of the actual user is completed.
2. The electricity larceny judging method based on the private transformer user current secondary circuit according to claim 1, wherein the historical normal user data information and the electricity larceny user data information in step S1 are specifically day current 96 point time sequence data information of the previous month when the electricity larceny user is confirmed to be electricity larceny on site, and day current 96 point time sequence data information of the normal user operating normally for one month.
3. The method for determining the power theft based on the secondary circuit of the private transformer user current according to claim 2, wherein the step S2 is characterized in that the ratio of the maximum value to the minimum value of the a/C two-phase currents of the power theft user and the normal user at each moment and the maximum current value corresponding to the power theft user and the normal user at each moment are calculated, and specifically comprises the following steps: For each electricity stealing user, calculating the ratio of the maximum value to the minimum value of the A/C two-phase current at each moment And maximum current value corresponding to each moment Wherein t represents the time of 96-point time sequence data of daily current, n represents the number of electricity stealing users, and i represents the number of days; for each normal user, calculating the ratio of the maximum value to the minimum value of the A/C two-phase current at each moment And maximum current value corresponding to each moment Wherein t represents the time of 96-point time series data of daily current, m represents the number of normal users, and i represents the number of days.
4. The method for determining fraudulent use of electricity based on a secondary circuit of a private transformer user current according to claim 3, wherein the step S3 of calculating the difference between the maximum value and the minimum value of the current ratio data for two types of users and the difference between the maximum value and the minimum value of the maximum current value data for two types of users specifically comprises the following steps: the difference between the maximum value and the minimum value of the current ratio data for two types of users is calculated by adopting the following formula: Wherein d r is the difference between the maximum value and the minimum value of the current ratio data for two types of users; the difference between maximum current data and minimum values of two types of users is calculated by adopting the following formula: Where d i is the difference between maximum and minimum values of the maximum current value data for the two classes of users.
5. The method for determining the fraudulent use of electricity based on the secondary circuit of the private transformer user current according to claim 4, wherein in the step S4, the difference data calculated in the step S3 is equally divided, and the fraudulent use of electricity and the normal user are subjected to joint distribution analysis according to the division result to obtain a ratio division threshold and a current division threshold, and the method specifically comprises the following steps: Equidistant dividing is carried out on the difference d r between the maximum value and the minimum value of the two types of users on the current ratio data, and the fine granularity is divided into 1 bit after decimal point; equidistant dividing is carried out on the maximum current value data and the difference d i between the maximum value and the minimum value of two types of users, and the fine granularity is divided into 1 bit after decimal point; and then, analyzing the electricity stealing users and the normal users according to the dividing result by adopting a two-dimensional histogram to obtain a final ratio dividing threshold value and a current dividing threshold value.
6. The method for determining fraudulent use of electricity based on a secondary circuit of a private transformer user current according to claim 5, wherein the ratio dividing threshold is 1.3, and the current dividing threshold is 0.5.
7. The method for determining the fraudulent use of electricity based on the secondary circuit of the private transformer user current according to claim 6, wherein the step S6 is characterized in that the effective interval of the actual user is determined according to the electricity consumption data obtained in the step S5 and the ratio dividing threshold and the current dividing threshold obtained in the step S4, and specifically comprises the following steps: Dividing the acquired daily current 96-point time sequence data of the actual user into 12 interval sections, wherein each interval section comprises 8-point time sequence data; For each interval, calculating the maximum current value of each point in the interval and the ratio of the maximum value to the minimum value of the A/C two-phase current; Comparing the calculated maximum current value with a current division threshold value, and simultaneously comparing the ratio of the maximum value to the minimum value of the A/C two-phase current with a ratio division threshold value: If the maximum current value is larger than the current division threshold value and the ratio of the maximum value to the minimum value of the A/C two-phase current is larger than the ratio division threshold value, judging that the current section is an effective section; otherwise, the current interval is judged to be an invalid interval.
8. The method for determining fraudulent use of electricity based on a secondary circuit of a private transformer user current according to claim 7, wherein in the effective interval section obtained in step S6 in step S7, an unbalanced interval section is determined according to a maximum value of the current, specifically comprising the following steps: for a three-phase three-wire meter: acquiring an A-phase current A 1 and a C-phase current C 1 at the moment corresponding to the maximum value point of the current in the effective section, and an A-phase current A 2 and a C-phase current C 2 at the moment corresponding to the minimum value point of the current in the effective section; Calculating the unbalance rate Ib as And judging: if the unbalance rate Ib is larger than the set threshold value, marking the effective section as an unbalance section; otherwise, marking the effective section as a balance section; For a three-phase four-wire meter: Obtaining a current maximum value I ma and a current minimum value I mi in three-phase current at a moment corresponding to a current maximum point in an effective section; Calculating and judging a current value I p to be I p ＝(I ma -I mi ) X, wherein X is the multiplying power of the current transformer, and judging: if the current value I p is larger than the set threshold value, marking the effective section as an unbalanced section; otherwise, the effective section is marked as a balance section.
9. The electricity larceny judging method based on the secondary circuit of the private transformer user current according to claim 8, wherein the unbalanced section obtained according to step S7 in step S8 is used for finishing the electricity larceny behavior judgment of the actual user, and specifically comprises the following steps: the determination is made for each day of data of the actual user: if the number of unbalanced sections is larger than the set number in 12 sections on the day, it is determined that the actual user has electricity larceny on the day.
10. A system for realizing the electricity larceny judging method based on the private transformer user current secondary circuit according to one of claims 1-9, which is characterized by comprising a historical data acquisition module, a current calculation module, a difference calculation module, a threshold calculation module, an actual data acquisition module, an effective section judging module, an unbalanced section judging module and an electricity larceny judging module; the system comprises a history data acquisition module, a current calculation module, a difference calculation module, a threshold calculation module, an effective section judgment module, an unbalanced section judgment module and a power stealing judgment module which are sequentially connected in series, wherein the output end of the actual data acquisition module is connected with the effective section judgment module, the history data acquisition module is used for acquiring data information of a history normal user and data information of a power stealing user in a special transformer user and uploading the data to the current calculation module, the current calculation module is used for calculating the ratio of the maximum value to the minimum value of an A/C two-phase current of the power stealing user and the normal user at each moment and the maximum current value corresponding to the power stealing user and the normal user at each moment according to the acquired data, and uploading the data to the difference calculation module, the difference calculation module is used for calculating the difference of the maximum value and the minimum value of the two types of the user according to the acquired data, and uploading the maximum current value data to the threshold calculation module, the threshold calculation module is used for equally dividing the difference data and carrying out joint distribution on the power stealing user and the user according to the obtained data, obtaining the ratio and the actual threshold analysis module is used for obtaining the actual data of the power stealing user and the threshold analysis module, the system comprises an effective section judging module, an unbalanced section judging module and a power stealing judging module, wherein the effective section judging module is used for judging an effective section of an actual user according to acquired data and according to power consumption data, a ratio dividing threshold value and a current dividing threshold value, and uploading the data to the unbalanced section judging module, the unbalanced section judging module is used for judging an unbalanced section according to the acquired data in the effective section and according to the maximum value of current, and uploading the data to the power stealing judging module, and the power stealing judging module is used for completing the judgment of the power stealing behavior of the actual user according to the acquired data and according to the unbalanced section.

Description

Power theft judgment method and system based on secondary circuit of specially-changed user current Technical Field The invention belongs to the field of electric automation, and particularly relates to a method and a system for judging electricity theft based on a secondary circuit of a private transformer user current. Background Along with the development of economic technology and the improvement of living standard of people, electric energy becomes an indispensable secondary energy source in the production and living of people, and brings endless convenience to the production and living of people. Therefore, ensuring stable and reliable operation of electric energy becomes one of the most important tasks of the electric power system. At present, the electric power system still has the phenomenon of electricity larceny, but the electricity larceny is light, and then the power supply order and the metering data of the electric power system are disturbed, and economic and energy losses are caused for the electric power system, and the topology structure and the circuit connection mode of the electric power system are artificially changed, so that the electric power system is greatly at risk of operation. Therefore, the determination of the electricity larceny behavior of the power system is particularly important. At present, with the rapid development of machine learning algorithms, more methods for analyzing electricity larceny of special transformer users are based on daily electricity consumption data of the special transformer users for nearly 1 year, fusion features are constructed through feature engineering, and then classification learning is performed by using the machine learning algorithms such as decision trees, random forests, XGBoost models and the like, or automatic clustering analysis is performed by using the unsupervised learning algorithms such as density clustering and mean clustering. However, both the supervised learning method based on classification and the unsupervised learning method based on clustering have the following problems that firstly, the method needs a large amount of basic data to construct electricity utilization characteristics, generally more than one year of basic data, the data granularity is larger, and the electricity stealing behavior of electricity stealing users is purposeful subjective behavior which generally exists in a fixed period of a fixed day, the data granularity is smaller, if the fine granularity characteristics of a moment hierarchy are difficult to extract by using coarse-granularity annual electric quantity data, and secondly, compared with common special-variant users, when the machine learning algorithm modeling is adopted for classification recognition, the special-variant electricity stealing users cannot ensure the balance of positive and negative samples, so that the accuracy and the reliability of a model are directly influenced. Disclosure of Invention The invention aims to provide a power theft judging method based on a secondary circuit of a private transformer user current, which has high accuracy, good reliability, simplicity and convenience. The second object of the invention is to provide a system for realizing the electricity larceny judging method based on the secondary loop of the private transformer user current. The invention provides a power theft judging method based on a secondary circuit of a specially-changed user current, which comprises the following steps: s1, acquiring data information of a history normal user and data information of a power stealing user from a special transformer user; S2, calculating the ratio of the maximum value to the minimum value of the A/C two-phase current of the electricity stealing user and the normal user at each moment and the maximum current value corresponding to the electricity stealing user and the normal user at each moment according to the data information obtained in the step S1; s3, calculating the difference between the maximum value and the minimum value of the current ratio data and the maximum current value data of the two types of users according to the data obtained by calculation in the step S2; s4, equally dividing the difference data obtained by the calculation in the step S3, and carrying out joint distribution analysis on the electricity stealing users and the normal users according to the division result to obtain a ratio division threshold value and a current division threshold value; s5, acquiring electricity utilization data of an actual user; s6, judging an effective interval of an actual user according to the electricity consumption data obtained in the step S5 and the ratio segmentation threshold value and the current segmentation threshold value obtained in the step S4; S7, judging an unbalanced section according to the maximum value of the current in the effective section obtained in the step S6; s8, according to the unbalanced interval obtained in the step