CN105203924A - Electricity usage trend abnormity suspicion analysis method and anti-electric-larceny monitoring system - Google Patents

Abstract

The invention discloses a method for analyzing suspicions of abnormal electricity consumption trends and an anti-stealing electricity monitoring system, including electricity consumption information state estimation, analysis of abnormal electricity consumption suspicions in low-voltage public substation areas, filtering and positioning of suspected users, monitoring of key suspects and anti-stealing. Power Stealing Knowledge Base. The state estimation of the electricity consumption information is used to solve the abnormal problems such as abnormality, lack of data of electric energy collection, and asymmetric ledger information. The above-mentioned abnormal power consumption suspicion analysis in the low-voltage public substation area is used for analyzing abnormal power consumption of users in the low-voltage public substation area. The suspect user filtering and locating are used to filter suspected large power users and locate suspected large power users. The monitoring of key suspects is used to track users with abnormal power consumption. The anti-stealing knowledge base is used to accumulate basic knowledge of anti-stealing. The ultimate purpose of the present invention is to automatically locate suspected users and form an investigation list by analyzing abnormal power consumption of users in the low-voltage public transformer station area, reduce the technical analysis intensity of anti-electricity theft staff, and improve analysis efficiency.

Description

A Method for Analyzing Suspicion of Abnormal Power Consumption Trend and Anti-Stealing Electricity Monitoring System

technical field

The invention relates to the technical field of power grid anti-stealing electricity, in particular to a method for analyzing suspicions of abnormal electricity consumption trends and an anti-stealing electricity monitoring system.

Background technique

Driven by economic interests, the situation of power theft in the national power grid is becoming increasingly severe. Among them, the low-voltage station area, especially the low-voltage station area with a complicated power environment, is the hardest hit area for electricity theft. Since the implementation of the SG186 marketing business application, the marketing management system and the electricity consumption information collection system have accumulated rich and large amounts of user information data, providing a data basis for power companies to carry out anti-stealing analysis and monitoring work. However, in the analysis process, each data Information asymmetry exists between systems, data has bad points, shortcomings, etc., which increases the difficulty of analysis, and the overall application of effective mining and analysis of these massive data appears to be relatively lagging behind. At present, the traditional anti-stealing technology analysis method generally performs threshold screening on massive data according to mature empirical rules to narrow the scope of suspect analysis, and then locks up suspect users through curve comparison, or directly conducts spot checks on large target groups, and The audit results depend to a certain extent on the skills and quality of the inspectors, and it is easy to miss the fish that slipped through the net. It is time-consuming and laborious, and the implementation effect is not ideal, forming the reality of "rich data, mature experience, but poor application". With the continuous maturity of big data mining theory, data mining technology has been continuously applied in practice by virtue of the expansion of big data space, which has brought considerable economic benefits and provided beneficial technical application reference for the field of anti-stealing technology. Accordingly, it is necessary to sort out and summarize the anti-stealing process, experience, technical achievements and typical cases, and use the development achievements of information technology to promote the informatization of anti-stealing work and realize the "teaching and learning" of anti-stealing work in marketing. All-round support, training, and research can effectively improve marketing anti-stealing technology analysis and positioning capabilities, and provide technical support and guarantee for better combating electricity theft.

Contents of the invention

For the deficiencies in the above anti-stealing technology analysis, the technical solution adopted in the present invention is:

According to the characteristics of this project, on the basis of studying the data types and characteristics of the power distribution information collection system and marketing management system, this project conducts a follow-up analysis of the abnormal electricity consumption of suspected users, mainly including the following aspects:

■Status estimation of power consumption information

The state estimation of power consumption information is obtained from the marketing SG186 system and the power consumption information collection system, etc.: power data (electricity, voltage, current, etc.), measurement abnormal information (meter abnormal information), (line, station area and user) ledger information Etc. power consumption information, using rules such as threshold elimination, replacement, and filling to estimate power data and ledger data for verification, to reduce the abnormality of power collection and display, lack of data collected at time points, and asymmetry of user ledger information Analytical errors caused by data problems.

■Analysis of abnormal power consumption in low-voltage station areas

After estimating the state of power consumption information, firstly, within the scope of each low-voltage public transformer station, the trend anomaly analysis algorithm is used to calculate the possible power theft corresponding to the abnormal power consumption of each user; secondly, the user’s suspicion of power consumption is calculated: The ratio of the electricity to the user’s electricity consumption is used as the suspicion of electricity theft in the user’s calculation cycle, and the numerical calculation result is used as one of the important basis for the inspection and sorting; finally, according to the suspected electricity and the level of suspicion, combined with the suspected electricity filter rule Identify the suspected large power users, and then delineate the users in the suspected high-level area according to the degree of suspicion, and form an audit list, which can be handed over to the inspectors for on-site evidence collection.

The above-mentioned trend anomaly analysis algorithm is the change trend obtained by comparing the average value of the total power of the same kind of users in the station area with the historical temperature close to the daily power consumption of the same kind of users on the calculation day, and it is used as the power consumption of the same kind of users in the station area for the calculation day Trend level, each user's historical temperature is close to the average level of electricity in the day, the trend electricity value obtained by this trend change, and then subtracting the electricity data of the user's calculation day, and the difference obtained is used as the suspected electricity of the user's calculation day.

■Key suspect user tracking

The monitoring function of key suspects specifically includes the monitoring of users who have been verified for stealing electricity in the past and the monitoring of some users who have not been verified recently but are still in the high-level area of suspicion. The monitoring process can be canceled if the key suspects have no electricity stealing records within a certain monitoring period or the degree of suspicion gradually decreases.

Owing to adopting above-mentioned technical scheme, compared with prior art, the beneficial effect of the present invention is:

1. In the anti-stealing electricity monitoring system, based on the state estimation of user electricity consumption information, it solves the problem of bad data points in different systems and data asymmetry between systems, improves the credibility of the original data, and provides a basis for technical analysis of power consumption suspicions. A reliable data environment;

2. In the anti-stealing power monitoring system, based on the analysis of suspected power and suspicion degree under the trend anomaly analysis algorithm, it overcomes the work intensity of manual data screening, improves work efficiency, and has the advantages of rapid analysis and reliable results. Based on a large number of sample verification, dynamically determine the threshold value of the suspect electric quantity screening and the threshold value of the high suspect area, which can quickly locate the suspect user;

3. In the anti-stealing electricity monitoring system, the technical curve (suspected power consumption, suspicious degree curve) tracking of key users can be effectively tracked and observed for users with abnormal electricity consumption in unconfirmed suspected high-level areas, so as not to miss the fish that slipped through the net. Improve inspection efficiency.

At the same time, the present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

Description of drawings

FIG. 1 is a flow chart of system hardware configuration implementation in an embodiment of the present invention.

Fig. 2 is a flow chart of the implementation of the system software architecture in an embodiment of the present invention.

Fig. 3 is a comparison of power analysis curves of a verified power stealing user according to the present invention.

Detailed ways

Example:

As shown in Figure 1, a method for analyzing abnormal electricity consumption trends and an anti-stealing electricity monitoring system includes four types of working systems: a marketing system server, an electricity consumption collection system server, an anti-electricity theft analysis server, and an anti-electricity theft monitoring workstation. Coordinating work, and jointly completing the anti-stealing suspected analysis and monitoring tasks. The data required by this system comes from the "marketing system server" and "power consumption collection system server", and then through network communication, the data is sent to the "anti-power-stealing analysis server" for centralized storage, analysis, forwarding, and finally sent to "Anti-electricity theft monitoring workstation" displays and alarms. The specific functions of each part are as follows:

"Marketing system server": It is the data storage server of the working system of the power marketing department, which transmits historical data and real-time data to the intermediate server for storage, mainly including the ledger data of lines, substations and substations.

"Power collection system server": it is a data storage server for power users' electricity collection work. It transmits historical data and real-time data to an intermediate server for storage, mainly including meter indications, meter abnormal information data, etc.

"Anti-power stealing analysis server": mainly used to store historical data, real-time data, and the results of abnormal power consumption analysis, and used for historical and real-time data transmitted by servers such as marketing systems and power consumption collection systems after state estimation The results are saved and provided for the "anti-power-stealing monitoring workstation" to obtain data.

"Anti-electricity theft monitoring workstation": It mainly displays the analysis results of suspected users graphically, and is used to track key suspected users at the same time, and can also export the list of audit lists for reference by the audit working group.

As shown in Figure 2, a method for analyzing suspicions of abnormal electricity consumption trends and an anti-stealing electricity monitoring system. In the figure, the software implementation scheme of the system platform is designed, and the development of the entire system platform is realized according to the software implementation process. Ensure that the human-computer interaction interface is convenient and practical, easy to learn and understand.

The software part of this system mainly realizes the accurate analysis of the suspected electric quantity and suspicion degree of the users in the low-voltage public substation area, screens the suspected large users in combination with the suspicion electric quantity threshold, and locates the users in the high-suspect area in combination with the suspicion degree threshold; Tracking and monitoring of key suspected users and confirmed suspected users; rolling accumulation of anti-stealing knowledge base.

As shown in Figure 3, a method for analyzing suspicions of abnormal electricity consumption trends and an anti-stealing electricity monitoring system. The figure shows a comparison of electricity analysis curves of users who have verified electricity theft. Among them, the green line represents its normal trend change, the red line is the actual power consumption change, and the blue line is the electricity stealing suspicion curve. The user is a non-resident user. The suspicion analysis results show that from January 2014 to the end of February 2014, the user was suspected of electricity theft 23 times, the average level of suspected electricity was 84.23 kWh, and the suspected electricity was conservatively estimated at 1937.29 kWh. The suspected date of electricity theft occurred on February 27, 2014. The curve trend shows that the user's electricity consumption level continued to decline during the period from January 18 to February 10, which deviated from the normal trend by 61.1%. According to the audit data, the hidden jumper of the fuse in front of the user's meter steals electricity.

Claims (8)

1. an electricity consumption trend anomaly suspicion analytical approach and supervisory system of opposing electricity-stealing, it is characterized in that, comprise power information state estimation, the multiplexing electric abnormality suspicion analysis of low pressure Gong Biantai district, suspicion user filtering and location, the monitoring of emphasis suspicion family, knowledge base of opposing electricity-stealing four modules; Wherein:

(1) power information state estimation: for solving the data problem such as electric energy acquisition registration exception, time point image data disappearance, user's account information asymmetry;

(2) low pressure Gong Biantai district multiplexing electric abnormality suspicion is analyzed: for analysing the multiplexing electric abnormality mutual affection of user under low pressure Gong Biantai district;

(3) suspicion user filtering and location: for filtering suspicion electricity rich and influential family, and location suspicion rich and influential family;

(4) emphasis suspicion family monitoring: do not check and verify high suspicion position multiplexing electric abnormality user for following the tracks of;

(5) to oppose electricity-stealing knowledge base: to oppose electricity-stealing ABC for accumulation, and record is previously opposed electricity-stealing typical case experience, for study and reference.

2. a kind of electricity consumption trend anomaly suspicion analytical approach according to claim 1 and supervisory system of opposing electricity-stealing, it is characterized in that, described power information state estimation is specifically from acquisitions such as marketing SG186 system, power information acquisition systems: the power informations such as energy data (electricity, voltage, electric current etc.), metering abnormal information (table meter abnormal information), (circuit, platform district and user) account information, rules such as adopting threshold values to reject, substitute, fill up is carried out verification estimate energy data, account data.

3. a kind of electricity consumption trend anomaly suspicion analytical approach according to claim 1 and supervisory system of opposing electricity-stealing, it is characterized in that, described low pressure Gong Biantai district multiplexing electric abnormality suspicion is analyzed, and specifically comprises user's suspicion electricity and the analysis of user power utilization suspicion degree: suspicion electricity computing method adopt trend anomaly analytical algorithm to calculate the possible power-steeling quantity of user corresponding to its multiplexing electric abnormality; The analysis of user power utilization suspicion degree is as the stealing suspicion degree in user's computation period using the ratio of the power consumption of user's suspicion electricity and user.

4. a kind of electricity consumption trend anomaly suspicion analytical approach according to claim 3 and supervisory system of opposing electricity-stealing, it is characterized in that, described trend anomaly analytical algorithm, under Shi Jiangtai district, each fellow users historical temperature is close to the electricity summation average of day, the variation tendency obtained compared with calculating day each fellow users electricity summation, the electricity trend level of day is calculated as fellow users under this district, the trend SOC values that each user's historical temperature obtains by this Long-term change trend close to the electricity average level of day, deduct the electric quantity data that this user calculates day again, the difference obtained calculates the suspicion electricity of day as this user.

5. a kind of electricity consumption trend anomaly suspicion analytical approach according to claim 1 and supervisory system of opposing electricity-stealing, it is characterized in that, described suspicion user filtering and locating module, specifically comprise filtration suspicion electricity rich and influential family and location electricity consumption suspicion rich and influential family: the former filters suspicion electricity rich and influential family according to presetting suspicion electricity threshold values, the latter delimit the high-order district user of suspicion according to suspicion degree size.

6. a kind of electricity consumption trend anomaly suspicion analytical approach according to claim 1 and supervisory system of opposing electricity-stealing, it is characterized in that, described emphasis suspicion family monitoring module, specifically comprise previously stealing check and verify user monitoring and in the recent period part do not check and verify the monitoring of the high-order district user of suspicion, in certain monitoring period, there is not the monitor procedure cancelled that stealing record or suspicion degree reduce gradually in emphasis suspicion family.

7. a kind of electricity consumption trend anomaly suspicion analytical approach according to claim 1 and supervisory system of opposing electricity-stealing, it is characterized in that, described base module of opposing electricity-stealing, specifically comprise stealing rudimentary knowledge technical ability part and stealing case study section, wherein, stealing rudimentary knowledge part comprises conventional (special type) stealing gimmick, sends out the application of stealing technology, and stealing case analysis then contains the overall processes such as the analysis of previously stealing user, location, inspection, evidence obtaining and process, for study and reference.

8. a kind of electricity consumption trend anomaly suspicion analytical approach according to claim 1 and supervisory system of opposing electricity-stealing, it is characterized in that: described a kind of electricity consumption trend anomaly suspicion analytical approach and supervisory system of opposing electricity-stealing use trend anomaly data mining technology, by comparison fellow users trend difference situation, finally calculate user's suspicion electricity and suspicion degree level.A kind of electricity consumption trend anomaly suspicion analytical approach and the low pressure Gong Biantai district multiplexing electric abnormality suspicion analysis of opposing electricity-stealing in supervisory system and suspicion user filtering and location suspicion analytic function instead of manual analysis process, solve the large efficiency of data volume low, the insecure problem of analysis result.