Electricity stealing detection method and system based on correlation between voltage difference and power loss

Abstract

This invention proposes a method and system for detecting electricity theft based on the correlation between voltage difference and power loss. The method includes the following steps: S1, acquiring power consumption monitoring data of a distribution substation through smart meters, including three-phase voltage data collected by the main meter of the substation at a preset acquisition frequency, three-phase voltage data synchronously collected by user meters within the substation at the preset acquisition frequency, and real-time three-phase power loss data corresponding to each acquisition time; S2, calculating the voltage difference between the single-phase voltage of each user meter within the substation and the corresponding phase voltage of the main meter of the substation at each acquisition time; S3, calculating the correlation between the single-phase voltage difference of each user and the corresponding phase power loss; S4, judging the probability of electricity theft based on the correlation calculation results of each user, with users having a higher probability of electricity theft due to lower correlation. This invention provides technical support for power companies' anti-electricity theft work.

Description

Technical Field

[0001] This invention belongs to the field of power system and its automation technology, specifically relating to a method and system for detecting electricity theft based on the correlation between voltage difference and power loss. Background Technology

[0002] Traditional electricity theft detection relies primarily on analyzing abnormal electricity consumption. However, with increasingly sophisticated methods of electricity theft, such as directly breaking into branch lines and stealing electricity without meter readings, traditional methods are finding it increasingly difficult to solve the current anti-theft challenges. With the deepening of smart grid construction and the widespread adoption of smart meters, data is collected every 15 minutes, including user meter readings, voltage, and electricity consumption information, providing a new perspective for combating electricity theft.

[0003] Based on voltage data from 96 points in a transformer substation per day (collected every 15 minutes), this invention proposes a method and system for detecting electricity theft by analyzing the correlation between the voltage difference between the main meter and user meters and power loss. Summary of the Invention

[0004] The purpose of this invention is to propose a method and system for detecting electricity theft based on the correlation between voltage difference and power loss, providing technical support for power companies' anti-electricity theft efforts.

[0005] To achieve the above objectives, the technical solution of the present invention is as follows:

[0006] A method for detecting electricity theft based on the correlation between voltage difference and power loss, specifically including the following steps:

[0007] S1. Obtain power consumption monitoring data of the distribution area through smart meters, including three-phase voltage data collected by the main meter of the distribution area at a preset collection frequency, three-phase voltage data collected synchronously by user meters in the distribution area at the preset collection frequency, and real-time three-phase power loss data corresponding to each collection time.

[0008] S2. Calculate the voltage difference between the single-phase voltage of each user's meter in the distribution area and the corresponding phase voltage of the distribution area's main meter at each acquisition time according to phase.

[0009] S3. Calculate the correlation between the single-phase voltage difference of each user and the corresponding phase power loss according to phase;

[0010] S4. Based on the correlation calculation results of each user, the probability of electricity theft is judged. The greater the correlation, the greater the probability of electricity theft.

[0011] Preferably, the preset sampling frequency is once every 15 minutes, with 96 sampling points throughout the day.

[0012] Preferably, the Pearson correlation coefficient between the single-phase voltage difference of each user and the corresponding phase power loss is calculated for each phase to characterize the correlation between the two:

[0013]

[0014] in: The Pearson correlation coefficient represents the single-phase voltage difference of a user and the corresponding phase power loss. The value range of is [-1, 1]. The closer the absolute value is to 1, the stronger the correlation. Indicates a time index. This indicates the number of all data collection points in a day. Indicates the first Single-phase voltage difference at each time point Indicates the whole day The average value of the single-phase voltage difference at each time point Indicates the first The corresponding phase power loss at each time point Indicates the whole day The average value of the corresponding phase power loss at each time point.

[0015] Preferably, the method further includes, after obtaining the power consumption monitoring data of the distribution area through the smart meter, performing time-sharing and phase-by-phase line loss calculation, identifying the phases suspected of electricity theft by comparing the three-phase line losses, and calculating the correlation between the single-phase voltage difference of each user and the corresponding phase power loss for the phases suspected of electricity theft, so as to assess the possibility of electricity theft.

[0016] A power theft detection system based on the correlation between voltage difference and power loss includes a processor, a memory, and a computer program stored in the memory. When the processor executes the computer program, it specifically performs any of the steps in the above-described power theft detection method.

[0017] Compared with the prior art, the present invention has the following beneficial effects:

[0018] This invention uses the correlation between voltage difference and power loss as the core criterion for judging electricity theft, rather than a single voltage or power threshold. This effectively avoids interference from normal factors such as grid fluctuations and load changes on the identification results. Furthermore, it eliminates the need for additional monitoring equipment, directly reusing the existing 96-point voltage and power data collection from smart meters, allowing for rapid deployment to existing distribution area monitoring systems. The entire identification process is automated, replacing the traditional manual on-site investigation mode, significantly improving the efficiency of electricity theft investigation and reducing maintenance manpower. Distribution areas that normally require 3-7 days of investigation can be accurately located and investigated in just one day using this method, greatly reducing on-site investigation time. Attached Figure Description

[0019] Figure 1 This is a flowchart of the data processing of the present invention;

[0020] Figure 2 This is an image showing the heating of the field connector in one embodiment of the present invention. Detailed Implementation

[0021] The following is in conjunction with the appendix Figure 1-2 The technical solution of the present invention will be described in detail below.

[0022] Within the same transformer substation, the impedance around each meter is the same. If electricity theft occurs, the voltage difference recorded around that meter will be positively correlated with the theft. Therefore, the calculated impedance will show a significant increase compared to historical impedance or neighboring lines. This invention uses this as a basis for judging abnormal electricity consumption by users.

[0023] According to Ohm's law and Kirchhoff's voltage law:

[0024] ΔU = I × R_loss

[0025] P_loss = I² × R_loss

[0026] ⇒ P_loss = (ΔU)² / R_loss

[0027] This indicates a squared relationship between voltage difference (ΔU) and power loss (P_loss). Normal user circuits have stable impedance, while users stealing electricity experience impedance abrupt changes due to illegal wiring, resulting in this correlation.

[0028] Given that voltage difference and power loss exhibit a stable positive correlation between electricity theft users, and that the correlation coefficient can suddenly drop due to impedance anomalies caused by illegal wiring, this invention proposes an electricity theft detection method based on the correlation between voltage difference and power loss, specifically including the following steps:

[0029] S1. Obtain power consumption monitoring data of the distribution area through smart meters, including three-phase voltage data collected by the main meter of the distribution area at a preset collection frequency, three-phase voltage data collected synchronously by user meters in the distribution area at the preset collection frequency, and real-time three-phase power loss data corresponding to each collection time.

[0030] S2. Calculate the voltage difference between the single-phase voltage of each user's meter in the distribution area and the corresponding phase voltage of the distribution area's main meter at each acquisition time according to phase.

[0031] S3. Calculate the correlation between the single-phase voltage difference of each user and the corresponding phase power loss according to phase;

[0032] S4. Based on the correlation calculation results of each user, the probability of electricity theft is judged. The greater the correlation, the greater the probability of electricity theft.

[0033] In this embodiment, the Pearson correlation coefficient between the single-phase voltage difference of each user and the corresponding phase power loss is calculated for each phase to characterize the correlation between the two:

[0034]

[0035] in: The Pearson correlation coefficient represents the single-phase voltage difference of a user and the corresponding phase power loss. The value range of is [-1, 1]. The closer the absolute value is to 1, the stronger the correlation. Indicates a time index. This indicates the number of all data collection points in a day. Indicates the first The single-phase voltage difference at each time point (total meter voltage - user voltage). Indicates the whole day The average value of the single-phase voltage difference at each time point Indicates the first The corresponding phase power loss at each time point Indicates the whole day The average value of the corresponding phase power loss at each time point.

[0036] In this embodiment, the preset sampling frequency is once every 15 minutes (00:00-23:45), with 96 sampling points throughout the day;

[0037] Data dimensions:

[0038] Total 96 points of voltage {V total_1 V total_2 , ...,V total_t , ...,V total_96}

[0039] User meter readings: 96 points voltage {V user_1 Vᵤ ser_2 , ..., V user_t , ..., Vᵤ ser_96}

[0040] Real-time power loss {P1, P2, ..., P} 96}

[0041] Differential pressure calculation:

[0042] ΔV t = V total_t- V user_t (t=1,2,...,96)

[0043] Phase sequence matching:

[0044] Based on the data acquired by the HPLC module, a phase topology map of the low-voltage distribution area is generated. The distribution of the three-phase meters in the distribution area is classified and compared. The three-phase users need to calculate the correlation with the voltage of phases A, B, and C of the main meter separately, and perform the above processing in the partition.

[0045] Correlation calculation:

[0046] The Pearson correlation coefficient is a commonly used statistic used to measure the strength of the linear correlation between two variables. Pearson correlation coefficient:

[0047]

[0048] in, The standard deviation of the voltage difference. This represents the standard deviation of power loss.

[0049] Computational implementation (Python example):

[0050] import numpy as np

[0051] def pearson_correlation(delta_V, power_loss):

[0052] """

[0053] Calculate the Pearson correlation coefficient between pressure drop and power loss.

[0054] :param delta_V: Voltage difference array (length 96)

[0055] :param power_loss: Power loss array (length 96)

[0056] :return: Correlation coefficient r

[0057] """

[0058] # Calculate the mean

[0059] mean_dV = np.mean(delta_V)

[0060] mean_P = np.mean(power_loss)

[0061] # Calculate the numerator: Sum of covariances

[0062] numerator = np.sum((delta_V - mean_dV) * (power_loss - mean_P))

[0063] # Calculate the denominator: product of standard deviations

[0064] std_dV = np.sqrt(np.sum((delta_V - mean_dV)**2))

[0065] std_P = np.sqrt(np.sum((power_loss - mean_P)**2))

[0066] denominator = std_dV * std_P

[0067] # Calculate the correlation coefficient

[0068] r = numerator / denominator if denominator != 0 else 0

[0069] return r

[0070] The criterion for determining electricity theft is: when the correlation r is much greater than the correlation r′ of the next user, the above users are suspected of electricity theft.

[0071] Preferably, the method further includes, after obtaining the power consumption monitoring data of the distribution area through the smart meter, performing time-sharing and phase-by-phase line loss calculation, identifying the phases suspected of electricity theft by comparing the three-phase line losses, and calculating the correlation between the single-phase voltage difference of each user and the corresponding phase power loss for the phases suspected of electricity theft, so as to assess the possibility of electricity theft.

[0072] The present invention also proposes an electricity theft detection system based on the correlation between voltage difference and power loss, including a processor, a memory, and a computer program stored in the memory. When the processor executes the computer program, it specifically performs any of the steps in the above-mentioned electricity theft detection method.

[0073] Through analysis of measured data from the XX City power distribution area, the electricity theft detection method of this invention can narrow down the scope of suspected electricity theft users, providing verification support for power companies' anti-electricity theft work.

[0074] Experiment 1:

[0075] The line loss of phase B in this transformer area was calculated by time-sharing and phase-by-phase analysis. The line loss rate of phase B was much higher than that of the other two phases, suggesting that phase B may be involved in electricity theft.

[0076] A Python program was used to analyze the data for this distribution area. The results showed that the correlation coefficients of the first two users, "Song***** Livestock Breeding Professional Cooperative," dropped sharply after that of other users, significantly higher than other users. On-site investigation revealed that the user had purchased a meter and connected it to phase B of the Song***** Vegetable and Fruit Professional Cooperative's power line, engaging in electricity theft.

[0077] Experiment 2:

[0078]

[0079] Calculations showed that the phase A line loss rate was much higher than that of the other two phases in this transformer area. According to the correlation results, the first two households were suspected of stealing electricity.

[0080]

[0081] Upon on-site inspection, it was found that the incoming line connector of the first meter, "Zhuo*Rui," was overheated and red-hot (reference). Figure 2 The resistance value was much higher than at other locations. After treatment, the line loss rate was reduced to within acceptable limits.

[0082] In summary, this invention proposes an intelligent identification method for electricity theft users based on 96 voltage data points. By collecting voltage data from the main transformer station meter and user meters in real time (every 15 minutes / point), a dynamic analysis model of voltage difference (ΔV = Vtotal - Vuser) and power loss is constructed. The core of this invention lies in the discovery of a significant positive correlation between voltage difference and power loss for electricity theft users, which is disrupted by the impedance anomalies caused by illegal wiring by electricity theft users.

[0083] This invention provides core algorithm support for the anti-electricity theft system of the Internet of Things for power distribution. In the future, it can be combined with deep reinforcement learning to build a one-unit-one-area model active analysis system.

[0084] The above are preferred embodiments of the present invention. Any changes made to the technical solution of the present invention that do not exceed the scope of the technical solution of the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for detecting electricity theft based on the correlation between voltage difference and power loss, characterized in that, Specifically, the following steps are included: S1. Obtain power consumption monitoring data of the distribution area through smart meters, including three-phase voltage data collected by the main meter of the distribution area at a preset collection frequency, three-phase voltage data collected synchronously by user meters in the distribution area at the preset collection frequency, and real-time three-phase power loss data corresponding to each collection time. S2. Calculate the voltage difference between the single-phase voltage of each user's meter in the distribution area and the corresponding phase voltage of the distribution area's main meter at each acquisition time according to phase. S3. Calculate the correlation between the single-phase voltage difference of each user and the corresponding phase power loss according to phase; S4. Based on the correlation calculation results of each user, the probability of electricity theft is judged. The greater the correlation, the greater the probability of electricity theft.

2. The electricity theft detection method based on the correlation between voltage difference and power loss according to claim 1, characterized in that, The preset sampling frequency is once every 15 minutes, with 96 sampling points throughout the day.

3. The electricity theft detection method based on the correlation between voltage difference and power loss according to claim 1, characterized in that, The Pearson correlation coefficient between the single-phase voltage difference of each user and the corresponding phase power loss is calculated for each phase to characterize the correlation between the two: in: The Pearson correlation coefficient represents the single-phase voltage difference of a user and the corresponding phase power loss. The value range of is [-1, 1]. The closer the absolute value is to 1, the stronger the correlation. Indicates a time index. This indicates the number of all data collection points in a day. Indicates the first Single-phase voltage difference at each time point Indicates the whole day The average value of the single-phase voltage difference at each time point Indicates the first The corresponding phase power loss at each time point Indicates the whole day The average value of the corresponding phase power loss at each time point.

4. The electricity theft detection method based on the correlation between voltage difference and power loss according to claim 1, characterized in that, The method further includes, after obtaining power consumption monitoring data of the distribution area through smart meters, performing time-sharing and phase-sharing line loss calculation, identifying the phases suspected of electricity theft by comparing the three-phase line losses, and calculating the correlation between the single-phase voltage difference of each user and the corresponding phase power loss for the phases suspected of electricity theft, in order to assess the possibility of electricity theft.

5. A power theft detection system based on the correlation between voltage difference and power loss, characterized in that, It includes a processor, a memory, and a computer program stored in the memory. When the processor executes the computer program, it specifically performs the steps in the electricity theft detection method as described in any one of claims 1-4.

Images