Abnormal power consumption analysis method based on integrated power consumption and line loss system big data

Abstract

This method for analyzing abnormal power usage based on big data from an integrated power and line loss system is based on an integrated power and line loss management system. It performs targeted processing and analysis on data collected from the system, combined with real-time data from the D5000 and marketing procurement systems. Starting with theoretical line loss and contemporaneous line loss, it analyzes abnormal power usage that leads to line loss. By analyzing the architecture of the abnormal power usage analysis model, it proposes a method for analyzing and confirming the causes of abnormal power usage through changes in line loss, and verifies this through real-world case studies.

Description

Technical Field

[0001] The present invention relates to an abnormal electricity consumption analysis method, and in particular to an abnormal electricity consumption analysis method based on big data of an integrated electricity quantity and line loss system, belonging to the technical field of smart power networks. Background Art

[0002] my country currently has a large population and a massive electricity consumption base. Consequently, power line losses due to abnormal power consumption are becoming increasingly serious. Power line losses occur at every stage of transmission, transformation, distribution, and sales. As a key comprehensive economic indicator for power companies, the line loss rate reflects not only the economic efficiency of power companies but also their profitability. Subsequently, the rapid development of information technologies such as cloud computing, the Internet of Things, mobile internet, and social networks has led to unprecedented development and increasing maturity of big data technologies, exemplified by distributed storage, distributed computing, and massive data mining.

[0003] Nowadays, the analysis of abnormal power consumption in the distribution network needs to rely on the analysis and comparison of data collected by automation equipment and management systems. The massive data collected has the typical "4V" big data characteristics, namely Volume, Velocity, Variety, and Value, that is, huge data volume, fast generation speed, multiple data types, and low value density.

[0004] Faced with the ever-increasing amount of data, the traditional storage and processing technologies used are still mostly in the process of collection, induction, and statistics, which makes it impossible to effectively utilize the data and to dig out the deep value of the data. For example, the classification and analysis of abnormal electricity data still mostly relies on manual verification, which is inefficient and has a low accuracy rate. There is an urgent need for a new management model to effectively process this type of data. At the same time, there are integrated electricity and line loss management systems that focus on the calculation and statistical display of electricity and line loss data, but the in-depth mining of abnormal data and the analysis and identification of abnormal electricity consumption are still in the initial stage. Therefore, there is an urgent need for an integrated electricity and line loss management system and big data technology to build an effective and direct analysis model for the abnormal data in the massive electricity and line loss data stored in the system, so as to achieve the correct evaluation and classification of abnormal electricity consumption. Summary of the Invention

[0005] The purpose of this invention is to overcome the above-mentioned shortcomings by providing an abnormal power consumption analysis method based on big data from an integrated power and line loss system. This method combines the data collected in the system with real-time data from the D5000 and the marketing procurement system to perform targeted processing and analysis. Starting with theoretical line loss and contemporaneous line loss, this method analyzes abnormal power consumption that leads to line loss. By analyzing the architecture of the abnormal power consumption analysis model, this method proposes a method for analyzing and confirming the causes of abnormal power consumption through changes in line loss, and verifies this method through actual case studies.

[0006] The object of the present invention is achieved like this:

[0007] An abnormal power consumption analysis method based on big data of an integrated power consumption and line loss system is provided, wherein:

[0008] First, build a large database and model for abnormal power consumption analysis:

[0009] Abnormal power consumption analysis database: The server extracts line ledger information, including historical data structures such as user installed capacity, load, power consumption, power consumption information, working conditions, and abnormal event records; at the same time, the server collects the following data: line gateway power consumption, public-private transformer daily power consumption, public-private transformer monthly power consumption, line loss statistical curve, line power consumption details, and metering device data in the line loss statistics data of the same period; line loss calculation value, public-private transformer copper loss and iron loss data, and line topology diagram in the line loss calculation data of the line loss system; active and reactive power, active and reactive power, real-time power factor, terminal calendar clock, and terminal parameters in the power consumption information data collection system; real-time line voltage, line current, and load curve data in the D5000 system;

[0010] The server builds an abnormal power consumption analysis model: The abnormal power consumption analysis model includes two parts: power consumption and line loss data access and abnormal power consumption analysis, such as Figure 2 As shown in the figure, the integrated power and line loss system's big data includes line loss data, user load data, user power consumption information data, and power consumption at line gateways. Big data-based analysis and identification of abnormal power consumption involves comparing data from the same source and type, as well as data from the same source and different types. This comparison includes user load comparison, daily power consumption comparison, current comparison, and active power comparison; while data from the same source and different types includes voltage and current comparison, and current and active power comparison. The various comparison techniques are shown in the figure. The primary comparison target is the characteristic values ​​of abnormal power consumption data in the abnormal power consumption database.

[0011] Then, the abnormal electricity consumption characteristic values ​​are selected: typical abnormal electricity consumption data are selected, including technical failures such as disconnection, power supply failure, line failure, meter error, data transmission error, and electricity theft, and big data comparison is performed with normal electricity consumption data and abnormal electricity consumption. Then, data conversion is performed and the abnormal electricity consumption characteristic values ​​are marked.

[0012] Next, we analyze the abnormal power usage model: Based on the abnormal power usage comparison algorithm, we conduct an analysis and assessment of abnormal power usage, distinguishing between abnormal and normal power usage. Abnormal data from the system is extracted and compared with the abnormal power usage feature values. The resulting comparison data is then imported into the abnormal power usage analysis model, ultimately enabling the diagnosis and classification of abnormal power usage.

[0013] Finally, abnormal power consumption diagnosis is derived.

[0014] Compared with the prior art, the present invention has the following beneficial effects:

[0015] This paper, based on an integrated power consumption and line loss management system, conducts targeted processing and analysis of data collected within the system, combined with real-time data from the D5000 and marketing procurement systems. Starting with theoretical line loss and contemporaneous line loss, it analyzes abnormal power usage that leads to line loss. By analyzing the architecture of the abnormal power usage analysis model, it proposes a method for analyzing and confirming the causes of abnormal power usage through changes in line loss, and verifies this through real-world case studies. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a flow chart of abnormal power consumption analysis of the present invention.

[0017] Figure 2 This is the architecture diagram of the abnormal power consumption analysis model.

[0018] Figure 3 This is a single contact line model diagram of the present invention.

[0019] Figure 4 This is a diagram of the line loss calculation model for a single line of the present invention.

[0020] Figure 5 This is the fault line loss model diagram of the 10kV Jiang B line in the present invention.

[0021] Figure 6 This is a diagram of the electricity theft analysis model of the present invention.

[0022] Figure 7 This is a topological diagram of a 10kV line of the present invention.

[0023] Figure 8 This is a detailed diagram of the patent variation of the present invention.

[0024] Figure 9 This is the current diagram of the three-phase unbalanced current of the user of the present invention.

[0025] Figure 10 This is the current diagram of the measuring point of the electric meter of the present invention. DETAILED DESCRIPTION

[0026] See also Figures 1 to 10 , the present invention relates to an abnormal power consumption analysis method based on big data of an integrated power consumption and line loss system;

[0027] Definition of noun:

[0028] Line loss: Line loss can be divided into statistical line loss, theoretical line loss and management line loss according to its characteristics.

[0029] Statistical line loss: It is the actual line loss of the power grid. Its value is the difference between the power supply and the power sales measured by the electricity meter, reflecting the actual loss of the power grid.

[0030] Theoretical line loss, also known as technical line loss, is calculated theoretically based on the parameters of power supply equipment and real-time grid load data. It reflects the theoretical amount of power loss that should be expected in a given grid structure and operating mode.

[0031] Managed line loss: This refers to actual grid losses minus theoretical line loss. It's calculated as the difference between statistical and theoretical line losses. Theoretical line loss research primarily utilizes measured data such as grid topology and power flow, based on power system knowledge, using theoretical calculation methods. Alternatively, machine learning algorithms can be used to predict line loss using historical data. Managed line loss includes electricity theft, meter errors, and leakage.

[0032] The Line Loss System is a key component of the integrated power and line loss management platform. It integrates data from operational lines, distribution transformers, pole-mounted transformers, interconnecting switches, and station buildings, and displays detailed information on power supply and sales. By integrating this integrated data and analyzing it using big data technology with data from systems like the PowerCai and D5000, it accurately identifies abnormal power usage, significantly improving analysis efficiency and providing strong support for the safe operation of the distribution network.

[0033] Theoretical support:

[0034] A: Abnormal power consumption analysis:

[0035] According to the electricity calculation method, the power P measurement formula in the electric energy meter is:

[0036] P = UIcosα……(1-1);

[0037] Where U is the voltage value of the measuring element, I is the current value of the measuring element, and α is the power factor angle.

[0038] Formula (1-1) shows that the power measured by an energy meter is related to three variables: voltage, current, and the phase relationship between voltage and current. Abnormal power usage analysis monitors and analyzes the normality of electrical data such as voltage, current, power factor (phase angle), line loss, and power consumption, as well as the changing trends of these parameters. This allows analysis of abnormal power usage to determine whether there is power theft or metering failure.

[0039] Abnormal electricity usage is caused by technical failures such as disconnection, power supply failure, line failure, meter errors, data transmission errors, as well as electricity theft and external interference.

[0040] B: Abnormal power consumption analysis and identification process under data background:

[0041] In the context of power big data, a large amount of electricity usage data is collected and integrated within the integrated power consumption and line loss management system. This data is purposefully selected based on actual needs to form a database of abnormal electricity usage. Secondly, the abnormal electricity usage is characterized by comparing the abnormal data with characteristic values ​​in the system to determine the type of abnormal electricity usage.

[0042] The present invention provides an abnormal power consumption analysis method based on big data of an integrated power consumption and line loss system, the method comprising:

[0043] First, build a large database and model for abnormal power consumption analysis:

[0044] Abnormal power consumption analysis database: The server extracts line ledger information, including historical data structures such as user installed capacity, load, power consumption, power consumption information, working conditions, and abnormal event records; at the same time, the server collects the following data: line gateway power consumption, public-private transformer daily power consumption, public-private transformer monthly power consumption, line loss statistical curve, line power consumption details, and metering device data in the line loss statistics data of the same period; line loss calculation value, public-private transformer copper loss and iron loss data, and line topology diagram in the line loss calculation data of the line loss system; active and reactive power, active and reactive power, real-time power factor, terminal calendar clock, and terminal parameters in the power consumption information data collection system; real-time line voltage, line current, and load curve data in the D5000 system;

[0045] The server builds an abnormal power consumption analysis model: The abnormal power consumption analysis model includes two parts: power consumption and line loss data access and abnormal power consumption analysis, such as Figure 2 As shown in the figure, the big data of the integrated power consumption and line loss system includes line loss data, user load data, user power consumption information data, line gateway power consumption, etc. (these existing data can be obtained from the server and internal systems such as the D5000 system). The analysis and identification of abnormal power consumption based on big data includes: comparison of data of the same source and type, and comparison of data of the same source and type. The comparison of data of the same source and type includes: user load comparison, daily power consumption comparison, current comparison, and active power comparison; the comparison of data of the same source and type includes voltage and current comparison, and current and active power comparison. Various comparison technologies are shown in the figure. Its main comparison target is the characteristic value of abnormal power consumption data in the abnormal power consumption database.

[0046] Then, the abnormal electricity consumption characteristic values ​​are selected: typical abnormal electricity consumption data are selected, including technical failures such as disconnection, power supply failure, line failure, meter error, data transmission error, and electricity theft, and big data comparison is performed with normal electricity consumption data and abnormal electricity consumption. Then, data conversion is performed and the abnormal electricity consumption characteristic values ​​are marked.

[0047] Next, we analyze the abnormal power usage model: Based on the abnormal power usage comparison algorithm, we conduct an analysis and assessment of abnormal power usage, distinguishing between abnormal and normal power usage. Abnormal data from the system is extracted and compared with the abnormal power usage feature values. The resulting comparison data is then imported into the abnormal power usage analysis model, ultimately enabling the diagnosis and classification of abnormal power usage.

[0048] Finally, abnormal power consumption diagnosis is derived.

[0049] The following is described in detail with a specific embodiment:

[0050] Analysis and establishment of distribution network line loss model:

[0051] The topology of the distribution network is complex, the distribution network equipment covers a wide range, and the operation mode changes frequently. These situations have a great impact on the calculation results of the distribution network branch line loss. Taking a single interconnection line as an example, the distribution network branch line diagram is as follows Figure 3 Based on this current situation, in order to reduce the impact of planned power outages, temporary power supply, and operating mode changes on line loss calculations, this line loss model analysis and calculation establishes a real-time update model for grid topology changes and operating mode changes. This incorporates changes in line transformer relationships, change time, and power consumption caused by topology changes and operating mode changes into the line loss calculation, achieving accurate calculation of line losses in the model.

[0052] For the convenience of calculation and expression, it is assumed that the average daily load of each part of the 10kV Jiang A line and the 10kV Jiang B line is exactly the same and the load is evenly distributed within 24 hours.

[0053] 1. Single line line loss model

[0054] like Figure 3 As shown in the figure, when the 10kV Jiang A line and the 10kV Jiang B line work independently, the line loss W1 and line loss rate η1 of the 10kV Jiang A line are:

[0055] W1=W A -W 1a -W 2a (2-1)

[0056] η1=W1÷W A (2-2)

[0057] Among them, WA is the power input from the 10kV Jiang A line gateway, W1a is the power collected by the public transformer A metering device, and W2a is the power collected by the private transformer A metering device.

[0058] The line loss W2 and line loss rate η2 of 10kV Jiang A line are:

[0059] W2=W B -W 1b -W2b (2-3)

[0060] η2=W2÷W B (2-4)

[0061] Among them, WB is the power input from the 10kV Jiang B line gateway, W1b is the power collected by the public transformer A metering device, and W2b is the power collected by the dedicated transformer A metering device. Figure 4 shown.

[0062] 2. 10kV Jiang B line fault line loss model

[0063] If the outgoing cable of the 10kV Jiang B substation fails, the repair time is T. That is, within T time, when the interconnector K2001 is closed, the 10kV Jiang B line circuit breaker opens. Because the line and the user metering devices in the line correspond to each other, the portion of electricity supplied by the 10kV Jiang A line to the 10kV Jiang B line cannot be counted by the Jiang A line gateway metering device. This portion of electricity will be classified as line loss. When the 10kV Jiang A line supplies the 10kV Jiang B line, the line gateway input electricity WA-B of the 10kV Jiang A line is expressed as:

[0064] W A-B =W A +W B ×T / 24 (2-5)

[0065] At this time, the line loss rate ηA-B is expressed as:

[0066] η2=(W A-B -W 1a -W 2a )÷W A (2-6)

[0067] It can be seen from equations (2-2) and (2-6) that line A is in a high-loss state at this time.

[0068] The input power WB-A of the 10kV Jiang B line at this time is

[0069] W B-A =W B ×T / 24 (2-7)

[0070] The line loss power W2 and line loss rate η2 are expressed as:

[0071] W2=W B-A -W 1b -W 2b (2-8)

[0072] η B-A =W1÷W A (2-9)

[0073] It can be seen from equations (2-7), (2-8) and (2-9) that line A is in a negative loss state at this time.

[0074] Taking 10kV Jiang A line as reference object, the calculation model is as follows Figure 5 shown.

[0075] 3. Abnormal power loss model for a certain user

[0076] Assume that when a fault occurs on the 10kV Jiang B line, the transformer A is isolated by the user switch B. The fault time is T1. At this time, the tie switch K2001 is disconnected. The line loss WB-b and line loss rate ηB-b of the 10kV Jiang B line are:

[0077] W B-b =W B -W 1a -W 2a ×T1 / 24 (2-10)

[0078] η1=W B-b ÷W B (2-11)

[0079] Assume that at this time, the protection on the outgoing line side of the substation is activated due to a fault in the public transformer user, that is, the 10kV Jiang B circuit breaker is disconnected, the 10kV Jiang B line section switch K2101 is disconnected, and the interconnecting switch K2001 is closed. Within the time T1, the 10kV Jiang A line supplies the rear section of the 10kV Jiang B line.

[0080] At this time, the daily line loss WA-Bb and line loss rate ηA-Bb of the 10kV Jiang A line are expressed as:

[0081] W A-Bb =W1+W Ba ×T1 / 24 (2-12)

[0082] η A-Bb =W A-Bb ÷W A (2-13)

[0083] Among them, WBa is the daily power consumption of the 10kV Jiang B line Zhonggong Substation A.

[0084] The daily line loss WA-Bb of the 10kV Jiang A line is expressed as:

[0085] W B-Bb =W2+W Ba ×T1 / 24 (2-12)

[0086] From formulas (2-3), (2-4) and (2-12), the line loss rate ηB-Bb

[0087] η A-Bb =WB-Bb ÷W B (2-13)

[0088] Abnormal power consumption analysis:

[0089] First, we use the line loss characteristic value comparison to analyze and evaluate abnormal power consumption, distinguish abnormal power consumption from normal power consumption, and then classify the line loss into two categories: communication line loss and line theft loss.

[0090] 1. Communication line loss

[0091] Communication line losses in distribution networks are generally caused by communication limitations and failures in metering devices, which can result in data synchronization failures for a period of time. Once the communication failure is repaired, the metering device resumes data transmission, uploads the communication data, and repairs the data, thereby correcting the line loss anomaly caused by such communication failures.

[0092] Another major cause is permanent failure of the metering device, such as damage from lightning strikes, vandalism, or power theft. This type of failure results in data loss, resulting in abnormal power usage on the line. Operations and maintenance personnel are required to analyze the cause of this anomaly and perform targeted repairs to restore normal metering and communication. The metering device uploads data every 15 minutes on average, and the communication system updates every 30 minutes. If data synchronization fails, the data from the moment before the failure and the data from the same moment the previous day are weighted and averaged to form the metering data at the time of the communication failure. Based on this data, the line loss value is calculated and compared with the actual line loss value to identify any matches with the abnormal power usage characteristics and determine the cause of the abnormal power usage.

[0093] 2. Loss due to wire theft

[0094] By analyzing the line loss and theoretical line loss data for the same period, we can identify users or substations with abnormal electricity consumption characteristics. That is, according to the line loss system data for the same period, the daily electricity consumption of a certain user or substation is very different from the historical data, the line loss rate has increased significantly, and the electricity consumption trend is very different from that of similar users. At the same time, the deviation value from the theoretical line loss exceeds the set threshold. For such electricity users, they are classified as key targets of electricity theft. We use big data technology to build a separate data model to summarize their electricity consumption patterns and conduct targeted analysis. Specific electricity theft analysis models are as follows: Figure 6 shown.

[0095] After identifying users with abnormal line losses, we analyze the causes of these abnormal power losses for each of their public and private transformers. Public transformers can be categorized as residential and non-residential users. Private transformer users can be categorized as those experiencing three-phase imbalance, metering anomalies, or load fluctuations. This allows us to further identify the abnormal power usage type and narrow down the scope.

[0096] Currently, electricity theft often occurs by altering the meter structure and wiring, or adjusting current and voltage, to under- or omit electricity usage. Abnormal user-side electricity metering can lead to high line loss. By comparing line loss and energy data with abnormal electricity usage signatures, users and their types of abnormal usage can be quickly identified.

[0097] The following is an analysis of a practical application case:

[0098] On July 2, 2018, according to the line loss system monitoring for the same period, the line loss of a 10kV line in June 2018 was 8.69%, exceeding the set high loss threshold of 6% for monthly line loss. The line loss on July 2 was 10.05%. By collecting relevant data and inputting it into the model for comparative analysis and evaluation, it was initially determined that a user was stealing electricity, and the characteristic value of the theft was three-phase voltage imbalance. The line topology diagram derived by the system is as follows Figure 7 The details of the public-private transformers on this line are as follows. Figure 8 shown.

[0099] The integrated power and line loss system exports data such as Figure 8 As shown, daily line losses on this line increased from 1.2% to 5.4% starting on December 5, 2016, and have since stabilized at around 6%. The average daily line loss increased from 1,282 kWh to 7,050 kWh, a year-on-year increase of 5,768 kWh. However, due to the high-loss thresholds set at 10% for daily line losses and 6% for monthly line losses, these thresholds were not triggered during 2016 and 2017, and the electricity theft user was not detected. In July 2018, with the onset of summer and increased electricity load, the user's electricity theft increased, leading to abnormally high line losses. Field verification confirmed the accuracy of the model's output.

[0100] verify:

[0101] Through the inspection of the marketing procurement system, it was found that the metered current of this household has been unbalanced in three phases since December 5, 2016. Figure 9 As shown. At around 13:00 on the afternoon of July 18, 2018, the Marketing Department and the Electric Power Law Enforcement Office, a group of more than ten people, went to Jiangyin Tianlian Textile Co., Ltd. for an on-site inspection. The on-site inspection found that the household used a current short-circuit wire to short-circuit the metering transformer, resulting in inaccurate metering, which violated Article 101 of the "Power Supply Business Rules": Deliberately making the power metering device of the power supply company inaccurate or ineffective is an act of electricity theft. Figure 10 Shown is the current curve at the on-site meter measurement point.

[0102] In addition: It should be noted that the above specific implementation is only an optimization scheme of this patent. Any changes or improvements made by technicians in this field based on the above concept are within the scope of protection of this patent.

Claims

1. An abnormal power consumption analysis method based on big data of an integrated power consumption and line loss system, characterized by: The method is: Step 1: Build a large database and model for abnormal power consumption analysis: Abnormal power consumption analysis database: extract line ledger information from the server; The server builds an abnormal power consumption analysis model: The abnormal power consumption analysis model includes: power consumption and line loss data access, and an abnormal power consumption analysis module; Step 2: The comparison method of importing abnormal power consumption analysis module is: comparison of same-source and same-type data, and comparison of same-source and different-type data; Step 3: Select abnormal power consumption feature values: Select typical abnormal power consumption data from the historical database, including data on technical failures such as disconnections, power failures, line failures, meter errors, data transmission errors, and electricity theft. Compare the data with normal power consumption data and abnormal power consumption data, then perform data conversion and mark the abnormal power consumption feature values. Step 4: Analyze the abnormal power consumption analysis model constructed in step 1: Based on the abnormal power consumption comparison algorithm, the power consumption and line loss data are connected and evaluated according to the abnormal power consumption analysis module to distinguish abnormal power consumption from normal power consumption; At the same time, the abnormal data in the abnormal power consumption analysis database is extracted, compared with the abnormal power consumption characteristic value, and the comparison data is extracted. The abnormal power consumption is diagnosed and classified according to the generated comparison results; Step 5: Export abnormal power consumption diagnosis; In step 1, the server collects and saves the following data: line gateway electricity, public-private transformer daily electricity, public-private transformer monthly electricity, line loss statistical curve, line electricity details, and metering device data in the line loss statistical data of the same period; Line loss calculation data of the line loss system includes line loss calculation values, copper loss and iron loss data of public and private transformers, and line topology diagrams; Active and reactive power, active and reactive power, real-time power factor, terminal calendar clock, and terminal parameters in the power consumption information data acquisition system; real-time line voltage, line current, and load curve data in the D5000 system; In step 2: the comparison of the same source and type of data includes: user load comparison, daily power consumption comparison, current comparison, and active power comparison; Homogeneous data includes voltage and current comparison, current and active power comparison.

Images