Electric energy meter wrong wiring analysis method based on electricity utilization information acquisition system

Abstract

The invention discloses an electric energy meter wrong wiring analysis method based on an electricity utilization information acquisition system, belongs to the technical field of electric energy meters, and aims to solve the problem of carrying out electric energy meter wrong wiring judgment and early warning according to the type of the electric energy meter under a complex power grid condition. The method comprises the following steps: acquiring key electric energy parameters corresponding to an on-site electric energy meter through an electricity utilization information acquisition system, recording timestamps, and constructing a basic data set; performing data preprocessing on the key electric energy parameters of the basic data set, establishing a white list for the special electricity users, and based on the key electric energy parameters after data preprocessing, rejecting the white list users to construct a discrimination data set; classifying and dividing the discrimination data set according to the type of the electric energy meter, performing wrong wiring diagnosis by adopting multiple diagnosis rules, and identifying the electric energy meter with wrong wiring abnormity; and when the electric energy meter is diagnosed as a wrong wiring abnormity, an early warning mechanism is triggered, field defect elimination information is generated, the early warning information is pushed in real time to generate a defect elimination work order, and operation and maintenance personnel are notified.

Description

Technical Field

[0001] The present invention relates to the technical field of electric energy meters, and more particularly to an electric energy meter miswiring analysis method based on an electricity consumption information acquisition system. Background Art

[0002] With the continuous development of smart grid and power informatization, power consumption information collection systems play an increasingly important role in the daily operations of power companies. These systems can not only monitor users' power consumption in real time, but also discover potential problems through data analysis to improve power supply quality and service level. In actual operation, due to various reasons, such as equipment installation errors, equipment aging failures, improper system maintenance or human operation errors, on-site electric energy meters may be miswired. Miswiring refers to errors that occur during the installation of electric energy metering equipment, resulting in inaccurate electric energy metering. Specifically, the electric energy meter is not correctly connected to the circuit according to the standard specifications, and there may be problems such as omissions, miscounts or excessive metering errors. This error will not only directly affect the accuracy of electric energy metering, but may also result in unfair electricity bill calculations for users, resulting in excessive or low electricity bill expenditures for users. At the same time, miswiring problems may also damage the profits and image of power companies. Miswiring is often not easy to detect because it usually does not cause significant failures or power outages, and the misconnected electric energy meter may show normal operating conditions in the initial stage. Therefore, how to effectively detect the miswiring problem of electricity meters has become a technical challenge that power companies are facing and need to solve urgently.

[0003] The Chinese patent application with the announcement number CN113740795B discloses a method for judging the misconnection of a three-phase four-wire electric energy meter, including: S100, selecting an abnormal window where the current abnormality occurs; S200, using the cosine similarity algorithm to respectively obtain the phase sequence current correlation coefficient in the abnormal window, the power factor corresponding to the abnormal phase sequence, and the cosine similarity between the electric energy representation value and the horizontal line; S300, comparing the calculated result with the corresponding threshold value to determine whether there is a misconnection abnormality. Make full use of big data analysis methods, use the 95th percentile, select the abnormal output window through the high-load working current, use the cosine similarity algorithm to analyze the abnormal window, obtain the current similarity between the phase sequences, the similarity of the electric energy representation value horizontal line, etc., so as to make a judgment on the misconnection abnormality.

[0004] Although the existing technology has greatly reduced the manual workload and eliminated the influence of human factors, thus realizing intelligent and automatic rapid miswiring abnormality judgment, greatly reducing the difficulty of on-site verification, it still fails to effectively cope with the complex and changeable power grid environment and the diverse needs of different types of electric energy meters, especially in dealing with large-scale deployment and high real-time requirements. There are still problems such as slow response speed and insufficient accuracy. Therefore, in order to overcome these limitations, the present invention proposes a method for analyzing electric energy meter miswiring based on an electricity consumption information collection system. Summary of the invention

[0005] In view of the shortcomings of the prior art, the purpose of the present invention is to provide an electric energy meter miswiring analysis method based on an electricity consumption information collection system. Based on the massive data of the electricity consumption information collection system, the method solves the problem of electric energy meter miswiring identification and early warning under complex power grid conditions according to electric energy meter files, collected data, operating parameters, etc.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for analyzing wrong wiring of electric energy meters based on an electric energy information collection system, comprising:

[0008] The electricity consumption information collection system automatically collects the key electricity parameters corresponding to each electricity meter, records the timestamp, and builds a basic data set;

[0009] Perform data preprocessing on the key power parameters of the basic data set, including identification of missing points, identification of duplicate records, data standardization, timestamp alignment and resampling, and establish a whitelist. Based on the key power parameters after data preprocessing, remove the whitelist users to construct a discriminant data set;

[0010] Classifying and dividing the discriminant data set according to the type of electric energy meter to obtain a first discriminant data set, a second discriminant data set and a third discriminant data set;

[0011] Using a plurality of diagnostic rules to perform miswiring diagnosis on the first discrimination data set, the second discrimination data set and the third discrimination data set, and identifying the electric energy meter with miswiring abnormality;

[0012] When the electric energy meter is diagnosed with a miswiring abnormality, the early warning mechanism is triggered, on-site fault elimination information is generated, the early warning information is pushed in real time to generate a fault elimination work order and notify the operation and maintenance personnel.

[0013] Specifically, the specific steps of data preprocessing of key electric energy parameters include:

[0014] Identify the earliest timestamp and the latest timestamp of the key power parameters in all basic data sets, define the time period between the earliest timestamp and the latest timestamp as a common time axis, and convert the timestamp of the key power parameter data into an offset relative to the common time axis;

[0015] Set the target sampling frequency f t , get the sampling timestamp according to the public time axis. If the start time of the public time axis is T start , then the sampling timestamp is: T * n =T start +n×f t , where n is a positive integer;

[0016] The original sampling frequency of each electric energy meter in the basic data set is obtained, and the resampling factor is calculated by the ratio of the target sampling frequency to the original sampling frequency, that is:

[0017]

[0018] Where RF is the resampling factor, f t is the target sampling frequency, f o is the original sampling frequency;

[0019] Configure the resampling threshold T RF , if the resampling factor is greater than the resampling threshold, the data is downsampled, otherwise the data is upsampled.

[0020] Specifically, the specific steps of data preprocessing of key electric energy parameters also include:

[0021] When the data is upsampled, the sampling timestamp is T * n When configuring the interpolation time interval [T * n -τ,T * n +τ], where τ is the interpolation time interval, which is used to measure the time range that affects the current sampling timestamp. For any key power parameter x, the interpolation value of its sampling timestamp is calculated as follows:

[0022]

[0023] Among them, x(T * n ) is the key power parameter x at the sampling time stamp T * n , m is the number of original sampling points in the interpolation time interval, t k is the original sampling timestamp in the interpolation time interval, x(t k) The key power parameter x at the original sampling time stamp t k The value of ω(t k ) is the non-negative weight corresponding to the original sampling timestamp in the interpolation time interval.

[0024] Specifically, the specific steps of data preprocessing of key electric energy parameters also include:

[0025] When downsampling data, configure the sampling window The sampling time stamp is T * n The sampling window in which the For any key electric energy parameter y, the calculation method of the sampling timestamp value is:

[0026]

[0027] Among them, y(T * n ) is the key power parameter y at the sampling time stamp T * n The value at time t i is the original sampling timestamp within the sampling window, y(t i ) The key power parameter y at the original sampling time stamp t i The value of ω ′ (t i ) is the non-negative weight corresponding to the original sampling timestamp in the interpolation time interval.

[0028] Specifically, the specific steps of miswiring diagnosis include:

[0029] According to the first diagnostic rule, a preliminary abnormality detection is performed on the key electric energy parameters of each electric energy meter in the discrimination data set, including reverse power consumption diagnosis and current and voltage abnormality diagnosis, to identify electric energy meters with miswiring abnormalities, and remove the electric energy meters marked as miswiring abnormalities in the second and third discrimination data sets and add them to the whitelist;

[0030] According to the second diagnostic rule, voltage imbalance diagnosis and current imbalance diagnosis are performed on the second discrimination data set and the third discrimination data set, and miswiring diagnosis is performed by constructing the second diagnostic rule score to identify the electric energy meters with miswiring abnormalities, and the electric energy meters marked as miswiring abnormalities in the third discrimination data set are removed and added to the whitelist;

[0031] According to the third diagnostic rule, the third discrimination data set is analyzed for abnormality of the neutral current to identify the electric energy meter with the abnormality of the wrong wiring.

[0032] Specifically, the first diagnostic rule includes:

[0033] The active power data of each electric energy meter in the discriminant data set is traversed, and when the active power is detected to be negative, it is marked as an active power abnormal point;

[0034] Extract the timestamps of the abnormal active power points and perform cluster analysis on the timestamps;

[0035] The length of the time period in each cluster category, the number of active power anomaly points contained, and the distance between cluster categories are calculated to construct the reverse power consumption diagnosis score. The calculation formula for the reverse power consumption diagnosis score of each electric energy meter is as follows:

[0036]

[0037] Among them, S f is the reverse power consumption diagnosis score, Q is the number of timestamp clustering categories, is the number of abnormal active power points in the qth category, is the duration of the timestamp of the qth category, is the distance measure between the qth category and the pth category, η N , η T and η D They are all non-negative weights, corresponding to the non-negative weight coefficients of the number of active power anomaly points, the length of the timestamp duration period, and the distance between cluster categories.

[0038] Specifically, the first diagnostic rule also includes:

[0039] Calculate the current-voltage ratio corresponding to each sampling timestamp of each electric energy meter in the discriminant data set, perform a current-voltage ratio anomaly test, and obtain a current-voltage anomaly diagnosis score;

[0040] For each electric energy meter, a first diagnostic rule score is constructed based on the reverse power consumption diagnostic score and the current and voltage abnormality diagnostic score. The calculation method of the first diagnostic rule score is as follows:

[0041] S1=γ f S f +γ d S d

[0042] Among them, S1 is the first diagnostic rule score, S f and S d are the reverse power consumption diagnosis score and the current and voltage abnormality diagnosis score, γ f and γ d are the non-negative weight coefficients of the reverse power consumption diagnosis score and the current and voltage anomaly diagnosis score respectively;

[0043] A first diagnosis threshold is configured, and when the first diagnosis rule score of the electric energy meter is greater than the first diagnosis threshold, it is marked as a miswiring abnormality, and the discrimination data set is traversed to filter the electric energy meters marked as miswiring abnormalities.

[0044] Specifically, the second diagnostic rule includes:

[0045] Obtain the three-phase voltage sequence corresponding to the sampling timestamp of each electric energy meter in the discriminant data set, and for each sampling timestamp, calculate the average value of the three-phase voltage;

[0046] Calculate the deviation of each phase voltage from the average value of the three-phase voltage, and calculate the instantaneous voltage unbalance based on the three-phase voltage deviation, that is:

[0047]

[0048] Among them, U uf is the instantaneous voltage unbalance, ΔU a is the voltage deviation of phase A, ΔU b is the voltage deviation of phase B, ΔU c is the phase C voltage deviation, is the average value of three-phase voltage;

[0049] Configure the measurement threshold T c , used to measure the smoothing time interval, for each sampling timestamp T * n , calculate the voltage smoothing imbalance within the measurement threshold, that is:

[0050]

[0051] in, is the sampling timestamp T * n Voltage smoothing unbalance, U uf (t) is the instantaneous voltage unbalance at the sampling time stamp t;

[0052] Configure the voltage threshold. For each sampling timestamp, when the voltage smoothing imbalance is greater than the voltage threshold and is the instantaneous voltage imbalance, mark it as voltage imbalance. Count the number of sampling timestamps marked as voltage imbalance in each electric energy meter.

[0053] Specifically, the second diagnostic rule also includes:

[0054] Obtain the three-phase current sequence corresponding to the sampling timestamp of each electric energy meter in the discriminant data set, calculate the average value of the three-phase current for each sampling timestamp, and calculate the deviation of each phase current from the average value of the three-phase current;

[0055] Calculate the instantaneous voltage imbalance and current smoothing imbalance based on the three-phase current deviation, make a current imbalance judgment for each sampling timestamp, obtain the sampling timestamps with current imbalance, and count the number of sampling timestamps with current imbalance in each electric energy meter;

[0056] According to the number of voltage unbalance and current unbalance sampling timestamps marked by each electric energy meter, a second diagnostic rule score is constructed. The calculation method of the second diagnostic rule score is as follows:

[0057]

[0058] Among them, S1 is the second diagnostic rule score, N u is the number of times the voltage is unbalanced, N is the number of times the current is unbalanced, N T is the total number of sampling timestamps of the electric energy meter; μ u and μ I are the non-negative weight coefficients of the marking times of voltage unbalance and the marking times of current unbalance respectively;

[0059] A second diagnostic threshold is configured. When the second diagnostic rule score of the electric energy meter is greater than the second diagnostic threshold, it is marked as a miswiring abnormality. The second discrimination data set and the third discrimination data set are traversed to filter the electric energy meters marked as miswiring abnormalities.

[0060] Specifically, the third diagnostic rule includes:

[0061] Configure the zero line mean threshold and the zero line standard deviation threshold, obtain the zero line current sequence corresponding to the sampling timestamp of each electric energy meter in the third discrimination data set, calculate its mean and standard deviation, and mark it as a potential abnormality when the mean and standard deviation of the electric energy meter are both greater than the zero line mean threshold and the zero line standard deviation threshold;

[0062] The electric energy meters marked as potential abnormalities are screened out, and the zero-line current series is smoothed by sliding average. The smoothed zero-line current series is subjected to autocorrelation analysis, and the autocorrelation coefficient of the zero-line current series is calculated, namely:

[0063]

[0064] Among them, R ξ is the autocorrelation coefficient at lag ξ, is the zero-line current value at the ιth moment, μ0 is the mean value of the zero-line current, and N0 is the total number of data points in the zero-line current sequence;

[0065] Set the delay time T ξ For each electric energy meter marked as potentially abnormal, calculate the zero line current sequence lag time T ξ The autocorrelation coefficient, configure the correlation threshold, when the lag time Tξ When the autocorrelation coefficient is greater than the correlation threshold, it is marked as a miswiring anomaly.

[0066] Beneficial effects of the present invention:

[0067] 1. The power consumption information collection system collects key power parameters of the power meter in real time, and applies multiple diagnostic rules to detect miswiring, so as to timely discover potential miswiring anomalies. When miswiring anomalies are detected, the early warning mechanism is automatically triggered, and early warning information is generated and pushed to the operation and maintenance personnel in real time. Quick response and measures are taken to repair the problem, reducing the system downtime and fault duration, thereby improving the reliability of the power system.

[0068] 2. After detecting a miswiring anomaly, an early warning message is automatically generated and pushed to the operation and maintenance personnel in the form of a work order. The early warning message contains detailed information such as the device number, error type, and occurrence time, helping the operation and maintenance personnel to quickly locate the problem. This helps to shorten troubleshooting time and improve maintenance efficiency.

[0069] 3. Accurately identify the energy meters with miswiring anomalies by classifying and dividing the discriminant data set and applying multiple diagnostic rules. Regularly generated miswiring anomaly statistical reports can help managers gain an in-depth understanding of the occurrence of miswiring anomalies in energy meters of different regions or types, thereby formulating more effective preventive measures and management strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 A schematic diagram of a method for analyzing wrong wiring of an electric energy meter based on an electric energy meter information collection system;

[0071] Figure 2 A flowchart of the specific steps of data preprocessing;

[0072] Figure 3 A flowchart of the specific steps for miswiring diagnosis;

[0073] Figure 4 is a schematic diagram of the first diagnostic rule;

[0074] Figure 5 is a schematic diagram of the second diagnostic rule. DETAILED DESCRIPTION

[0075] See also Figure 1 This embodiment introduces a method for analyzing wrong wiring of electric energy meters based on an electric energy information collection system, including:

[0076] Step S1: The key power parameters corresponding to each power meter are collected through the power consumption information collection system, and the timestamp is recorded to construct a basic data set for subsequent real-time monitoring and analysis;

[0077] Among them, the types of energy meters include: single-phase energy meter, three-phase three-wire energy meter and three-phase four-wire energy meter;

[0078] Key power parameters include: power, voltage, current, power consumption, neutral current, power factor;

[0079] The communication protocols of the electric energy meter include DL / T 645 protocol and DLT-698.45 protocol to ensure stable data transmission.

[0080] Step S2: Data preprocessing is performed on the key power parameters of the basic data set, including identification of missing points, identification of duplicate records, data standardization, timestamp alignment and resampling, and a whitelist is established to exclude users who do not perform abnormal analysis. Based on the key power parameters after data preprocessing, whitelist users are excluded to construct a discriminant data set;

[0081] Step S3: classifying and dividing the discrimination data set according to the type of electric energy meter to obtain a first discrimination data set, a second discrimination data set and a third discrimination data set;

[0082] Step S4: using a plurality of diagnostic rules to perform miswiring diagnosis on the first discrimination data set, the second discrimination data set and the third discrimination data set to identify the electric energy meter with miswiring abnormality;

[0083] Step S5: When the electric energy meter is diagnosed as having a wrong wiring abnormality, the early warning mechanism is triggered to generate early warning information, including the device number, error type, and occurrence time, and the early warning information is pushed to the operation and maintenance personnel's work platform in real time.

[0084] In this embodiment, the power consumption information collection system uses advanced sensing, communication, automatic control and other technologies to realize data collection, data management, power quality data statistics, line loss statistical analysis, timely collect and grasp user power consumption information, discover abnormal power consumption, monitor and control the power load of power users, and provide technical support for the implementation of marketing business strategies such as tiered electricity prices and intelligent fee control. The power consumption information collection system collects key power parameters of single-phase power meters, three-phase three-wire power meters and three-phase four-wire power meters on site in real time every day. The key power parameters include voltage, current, power, power factor, power consumption and neutral current, and records the timestamps of key power parameters. Automatically identify and remove missing points and duplicate records in the collected key power parameters. Create a white list containing specific users, whose power meters do not participate in abnormal analysis. Based on the pre-processed key power parameters, the data of whitelist users is removed to construct a discriminant data set for discriminant analysis. The discriminant data set is divided into three subsets according to the type of power meter: the first discriminant data set, the second discriminant data set and the third discriminant data set. Through multiple diagnostic rules, including reverse power consumption diagnosis, phase current reversal, current imbalance limit, reverse active power, etc., the three discriminant data sets are analyzed to identify the power meters with miswiring anomalies and record relevant information for further analysis. Once the diagnostic results show that the power meter has a miswiring problem, the early warning mechanism is triggered immediately. Detailed early warning information including device number, error type and occurrence time is generated. The early warning information is pushed out in real time in the form of work orders to ensure that relevant personnel can respond and handle the problem in a timely manner.

[0085] See also Figure 2 Preferably, the specific steps of data preprocessing of the basic data set include:

[0086] Step 2.1: Identify the missing data points of the basic data set, obtain the missing timestamps where the missing data points are located, and delete the key power parameter data where the missing timestamps are located;

[0087] Step 2.2: Identify duplicate records of the base data set and remove duplicate entries, retaining one of each set of duplicate records;

[0088] Step 2.3: Perform data standardization on the basic data set, convert the key power parameter data of different dimensions to the same scale, and ensure that the key power parameter data use consistent units;

[0089] Step 2.4: Identify the earliest timestamp and the latest timestamp of the key power parameters in all basic data sets, define the time period between the earliest timestamp and the latest timestamp as the common time axis, and convert the timestamp of the key power parameter data into an offset relative to the common time axis to align the basic data sets in time;

[0090] Step 2.5: Set the target sampling frequency f t , get the sampling timestamp according to the public time axis. If the start time of the public time axis is T start , then the sampling timestamp is: T * n =T start +n×f t , where n is a positive integer;

[0091] Step 2.6: Get the original sampling frequency of each electric energy meter in the basic data set, and calculate the resampling factor by the ratio of the target sampling frequency to the original sampling frequency, that is:

[0092]

[0093] Where RF is the resampling factor, f t is the target sampling frequency, f o is the original sampling frequency;

[0094] Step 2.7: Configure the resampling threshold T RF , if the resampling factor is greater than the resampling threshold, the data is downsampled, otherwise the data is upsampled;

[0095] Step 2.8: When upsampling the data, at the sampling timestamp T * n When configuring the interpolation time interval [T * n -τ,T * n +τ], where τ is the interpolation time interval, which is used to measure the time range that affects the current sampling timestamp. For any key power parameter x, the interpolation value of its sampling timestamp is calculated as follows:

[0096]

[0097] Among them, x(T * n ) is the key power parameter x at the sampling time stamp T * n , m is the number of original sampling points in the interpolation time interval, t k is the original sampling timestamp in the interpolation time interval, x(t k ) The key power parameter x at the original sampling time stamp t k The value of ω(t k ) is the non-negative weight corresponding to the original sampling timestamp in the interpolation time interval, exemplarily, Among them, σ is the standard deviation, which determines the smoothness of the weight; weighted interpolation significantly reduces the interference of outliers in the data on the interpolation results, especially when there is noise or mutation in the signal, while maintaining the smoothness and accuracy of the interpolation.

[0098] Step 2.9: Configure the sampling window when downsampling the data The sampling time stamp is T * n The sampling window in which the For any key electric energy parameter y, the calculation method of the sampling timestamp value is:

[0099]

[0100] Among them, y(T * n ) is the key power parameter y at the sampling time stamp T * n The value at time t i is the original sampling timestamp within the sampling window, y(t i ) The key power parameter y at the original sampling time stamp t i The value of ω ′ (t i ) is the non-negative weight corresponding to the original sampling timestamp in the interpolation time interval, exemplarily, Among them, σ is the standard deviation, which determines the degree of smoothing of the weight. The weight is assigned to each data point in the sampling window, and then the weighted average is calculated to generate a new downsampled data point. The data point with a larger weight, that is, the data point closer to the target time point, has a greater impact on the result. It can reduce the impact of short-term noise and retain the long-term trend of the data.

[0101] Step 2.10: Create a whitelist, remove whitelist users from the basic data set, resample the basic data set after removal according to the target sampling frequency, and construct a discriminant data set;

[0102] In this embodiment, by analyzing each key electric energy parameter in the basic data set, it is identified whether there are missing values ​​in the data. Common missing value markers, such as NaN, null values, etc., can be used to detect missing data. When a missing value is identified, the timestamp where the missing value is located is recorded. For electric energy data, the timestamp is usually a time identifier in hours, minutes or seconds, so the timestamp of the missing data needs to be recorded for subsequent processing. Delete the electric energy parameter data corresponding to the timestamp containing the missing data. You can choose to delete the missing data of the entire row of data or certain key fields. In order to ensure the integrity of the data, delete the entire row of data. In all records of the basic data set, there may be duplicate rows or duplicate timestamps. It is necessary to compare the data of each row to determine whether there are duplicate records. Duplicate judgment can be performed by a combination of fields such as timestamps and key electric energy parameters. For the identified duplicate records, delete the redundant records, keep only one, keep the latest record, or keep the first and most representative record according to actual needs. The basic data set is subjected to data standardization, and the key power parameter data of different dimensions are converted to the same scale to ensure that the key power parameter data use consistent units. Since different parameters in the power data have different dimensions (units), the data is standardized. Exemplarily, the unit of power is uniformly converted to kilowatt (kW), and the unit of voltage is uniformly converted to volt (V) to ensure that all power parameters use the same unit. By analyzing all power parameter data sets, the earliest timestamp and the latest timestamp in each data set are found. The earliest timestamp refers to the first valid record time in the data set, and the latest timestamp is the last valid record time in the data set. Based on the range of the earliest and latest timestamps, a common time axis is defined. The common time axis will serve as the time base for all data sets to ensure that the data of different power meters can be aligned on the same time scale. The timestamp in each record is converted to an offset or index relative to the common time axis. Exemplarily, assuming that the earliest timestamp is T0 and the timestamp of a certain record is T1, then the offset of T1 relative to the common time axis is T1-T0. Unifying the data into a standard time frame helps with subsequent data alignment and analysis. Generate a timestamp sequence starting from the earliest timestamp and with an interval of the target sampling frequency based on the target sampling frequency. Obtain the original sampling frequency of each electric energy meter and calculate it by analyzing the timestamp interval. Exemplarily, calculate the time difference between two consecutive timestamps to obtain the original sampling frequency. Calculate the resampling factor by the ratio of the target sampling frequency to the original sampling frequency to accurately calculate the resampling requirements, which helps to avoid oversampling or undersampling of data in subsequent processing; set the resampling threshold to ensure that the sampling frequency of the data meets the target requirements while avoiding information loss. Improve data quality through reasonable upsampling and downsampling operations to provide more valuable data for subsequent analysis and modeling.A whitelist is established to indicate that the electric energy meters of whitelist users do not participate in abnormal analysis, and the basic data set after elimination is resampled according to the target sampling frequency, and a discriminant data set is constructed; by eliminating whitelist users who do not need to be paid attention to, more resources and attention can be focused on users who may require additional monitoring, thereby improving operational efficiency and reducing costs.

[0103] See also Figure 3 Preferably, the specific steps of miswiring diagnosis include:

[0104] Step S4.1: According to the first diagnostic rule, a preliminary abnormality detection is performed on the key electric energy parameters of each electric energy meter in the discrimination data set, including reverse power consumption diagnosis and current and voltage abnormality diagnosis, to identify electric energy meters with miswiring abnormalities, and remove the electric energy meters marked as miswiring abnormalities in the second discrimination data set and the third discrimination data set, and add them to the whitelist;

[0105] Step S4.2: According to the second diagnostic rule, voltage imbalance diagnosis and current imbalance diagnosis are performed on the second discrimination data set and the third discrimination data set, and miswiring diagnosis is performed by constructing the second diagnostic rule score to identify the electric energy meters with miswiring abnormalities, and the electric energy meters marked as miswiring abnormalities in the third discrimination data set are removed and added to the whitelist;

[0106] Step S4.3: According to the third diagnostic rule, the third discrimination data set is subjected to a neutral current anomaly analysis to identify the electric energy meter with the wrong wiring anomaly.

[0107] See also Figure 4 , preferably, the first diagnostic rule includes:

[0108] The active power data of each electric energy meter in the discriminant data set is traversed. When the active power is detected to be negative, it is marked as an active power abnormal point. The timestamp of the active power abnormal point is extracted, and the timestamp is clustered. The length of the time period of the timestamp in each cluster category, the number of active power abnormal points included, and the distance between cluster categories are calculated to construct a reverse power consumption diagnosis score. The calculation formula for the reverse power consumption diagnosis score of each electric energy meter is as follows:

[0109]

[0110] Among them, S f is the reverse power consumption diagnosis score, Q is the number of timestamp clustering categories, is the number of abnormal active power points in the qth category, is the duration of the timestamp of the qth category, is the distance measure between the qth category and the pth category, η N , η T and ηD All are non-negative weights, corresponding to the non-negative weight coefficients of the number of active power anomaly points, the duration of the timestamp, and the distance between cluster categories; further diagnosis is based on the distribution of active power anomaly points, the continuity of timestamps, and other factors. If the active power anomaly points occur in a continuous period of time, it may be an equipment failure or a continuous measurement error, rather than an occasional anomaly; if the active power anomaly points appear alone, it may be an error in data collection, or an instantaneous fluctuation; if a certain electric energy meter is abnormal at multiple times or multiple dates, it may be that the electric energy meter has a systemic problem and needs further investigation. The construction of the reverse power consumption diagnostic score is based on the number of active power anomaly points, the duration of the timestamp, and the distance between cluster categories. When active power anomaly points appear continuously, the reverse power consumption diagnostic score increases, and the length of the time period of the anomaly points is calculated as part of the score. If a large number of active power anomaly points appear on the same electric energy meter within a period of time, the reverse power consumption diagnostic score also increases; the closer the time periods of active power anomaly points are, the greater the reverse power consumption diagnostic score;

[0111] Configure the current-voltage ratio threshold range, calculate the current-voltage ratio corresponding to each sampling timestamp of each electric energy meter in the discriminant data set, and when the detected current-voltage ratio exceeds the current-voltage ratio threshold range, mark it as a current-voltage ratio abnormal point, extract the timestamp of the current-voltage ratio abnormal point, and perform cluster analysis on the timestamps. Calculate the time period length of the timestamp in each cluster category, the number of current-voltage ratio abnormal points contained, and the distance between cluster categories to construct a current-voltage abnormality diagnosis score;

[0112] For each electric energy meter, a first diagnostic rule score is constructed based on the reverse power consumption diagnostic score and the current and voltage abnormality diagnostic score. The calculation method of the first diagnostic rule score is as follows:

[0113] S1=γ f S f +γ d S d

[0114] Among them, S1 is the first diagnostic rule score, S f and S d They are reverse power consumption diagnosis score, and current and voltage abnormality diagnosis score, γ f and γ d are the non-negative weight coefficients of the reverse power consumption diagnosis score and the current and voltage anomaly diagnosis score respectively;

[0115] A first diagnosis threshold is configured, and when the first diagnosis rule score of the electric energy meter is greater than the first diagnosis threshold, it is marked as a miswiring abnormality, and the discrimination data set is traversed to filter the electric energy meters marked as miswiring abnormalities.

[0116] See also Figure 5 , preferably, the second diagnostic rule includes:

[0117] The three-phase voltage sequence corresponding to the sampling timestamp of each electric energy meter of the second discriminant data set and the third discriminant data set is obtained, and for each sampling timestamp, the three-phase voltage average value is calculated, that is:

[0118]

[0119] in, is the average value of three-phase voltage, U a , U b and U c Represents the three phase voltages of the three-phase voltage, which are the voltages of phase A, phase B and phase C;

[0120] Calculate the deviation of each phase voltage from the average value of the three-phase voltage at that moment, that is:

[0121]

[0122] Among them, ΔU a is the voltage deviation of phase A, ΔU b is the voltage deviation of phase B, ΔU c is the voltage deviation of phase C;

[0123] The instantaneous voltage unbalance is calculated based on the three-phase voltage deviation, that is:

[0124]

[0125] Among them, U uf is the instantaneous voltage unbalance, configure the measurement threshold T c , used to measure the smoothing time interval, for each sampling timestamp T * n , calculate the voltage smoothing imbalance within the measurement threshold, that is:

[0126]

[0127] in, is the sampling timestamp T * n Voltage smoothing unbalance, U uf (t) is the instantaneous voltage unbalance at the sampling time stamp t;

[0128] Configure the voltage threshold. For each sampling timestamp, when the voltage smoothing imbalance is greater than the voltage threshold and the instantaneous voltage imbalance, mark it as voltage imbalance; count the number of sampling timestamps marked as voltage imbalance in each electric energy meter;

[0129] Obtain the three-phase current sequence corresponding to the sampling timestamp of each electric energy meter in the discriminant data set, calculate the average value of the three-phase current for each sampling timestamp, and calculate the deviation of each phase current from the average value of the three-phase current;

[0130] The instantaneous voltage unbalance is calculated based on the three-phase current deviation, that is:

[0131]

[0132] Among them, I uf is the instantaneous voltage unbalance, ΔI a is the voltage deviation of phase A, ΔI b is the voltage deviation of phase B, ΔI c is the phase C voltage deviation, is the average value of three-phase voltage;

[0133] For each sampling time stamp T * n , calculate the current smoothing imbalance within the measurement threshold, that is:

[0134]

[0135] in, is the sampling timestamp T * n Voltage smoothing imbalance, I uf (t) is the instantaneous voltage unbalance at the sampling time stamp t;

[0136] Configure the current threshold. For each sampling timestamp, when the current smoothing imbalance is greater than the voltage threshold and is the instantaneous current imbalance, mark it as current imbalance; count the number of sampling timestamps marked as current imbalance in each electric energy meter;

[0137] According to the number of voltage unbalance and current unbalance sampling timestamps marked by each electric energy meter, a second diagnostic rule score is constructed. The calculation method of the second diagnostic rule score is as follows:

[0138]

[0139] Among them, S1 is the second diagnostic rule score, N u is the number of times the voltage is unbalanced, N is the number of times the current is unbalanced, N T is the total number of sampling timestamps of the electric energy meter; μ u and μ I are the non-negative weight coefficients of the marking times of voltage unbalance and the marking times of current unbalance respectively;

[0140] A second diagnostic threshold is configured. When the second diagnostic rule score of the electric energy meter is greater than the second diagnostic threshold, it is marked as a miswiring abnormality. The second discrimination data set and the third discrimination data set are traversed to filter the electric energy meters marked as miswiring abnormalities.

[0141] Preferably, the third diagnostic rule includes:

[0142] Configure the zero-line mean threshold and the zero-line standard deviation threshold, obtain the zero-line current sequence corresponding to the sampling timestamp of each electric energy meter in the third discrimination data set, calculate its mean and standard deviation, and when the mean and standard deviation of the electric energy meter are both greater than the zero-line mean threshold and the zero-line standard deviation threshold, mark it as a potential anomaly; perform a preliminary screening of the zero-line current of each electric energy meter, if the mean and standard deviation of the zero-line current of the electric energy meter are both greater than the configured threshold, it is considered that the zero-line current of the electric energy meter may have abnormal fluctuations or interference, and is marked as a potential anomaly, and automatically screen out those electric energy meters that may have faults or abnormal behaviors as objects for further analysis.

[0143] The electric energy meters marked as potential abnormalities are screened out, and the zero-line current series is smoothed by sliding average. The smoothed zero-line current series is subjected to autocorrelation analysis, and the autocorrelation coefficient of the zero-line current series is calculated, namely:

[0144]

[0145] Among them, R ξ is the autocorrelation coefficient at lag ξ, is the zero-line current value at the first moment, μ0 is the mean value of the zero-line current, and N0 is the total number of data points in the zero-line current sequence; when the autocorrelation coefficient remains at a high value for a long time, it means that the zero-line current has repeated fluctuations or is affected by incorrect wiring;

[0146] Set the delay time T ξ For each electric energy meter marked as potentially abnormal, calculate the zero line current sequence lag time T ξ The autocorrelation coefficient, configure the correlation threshold, when the lag time T ξ When the autocorrelation coefficient of the zero-line current sequence is greater than the correlation threshold, it is marked as a miswiring anomaly. The autocorrelation coefficient of the zero-line current sequence at the lag moment is calculated to identify the correlation on a longer time scale in the current sequence and capture possible long-term abnormal patterns. When the calculated autocorrelation coefficient at the lag moment is greater than the preset threshold, it means that there is obvious repetitive fluctuation in the zero-line current, which may be caused by wrong wiring. If the autocorrelation coefficient exceeds the threshold, the electric energy meter will be marked as a miswiring anomaly, thus providing clues for troubleshooting and repairing the electric energy meter.

[0147] Preferably, the early warning mechanism includes:

[0148] Analyze the key power parameters of the power meter through a series of diagnostic rules to determine whether there is any miswiring anomaly;

[0149] When the diagnosis result shows that the electric energy meter has a wrong wiring abnormality, an early warning message is automatically generated, and the early warning message includes:

[0150] Equipment number: unique identifier of the energy meter, used for tracking and management;

[0151] Error type: including reverse current, abnormal power factor, voltage imbalance or abnormal neutral current, etc.

[0152] Occurrence time: record the specific timestamp when the miswiring anomaly is detected;

[0153] The constructed warning information will be sent to the operation and maintenance personnel in the form of a work order to ensure that the operation and maintenance personnel can receive the warning information in a timely manner;

[0154] After receiving the warning information, the operation and maintenance personnel will conduct on-site inspection to confirm whether there is indeed an abnormal miswiring. Once confirmed, the operation and maintenance personnel will carry out necessary fault elimination work to ensure that the wiring of the energy meter complies with relevant standards and specifications;

[0155] After the repair is completed, the operation and maintenance personnel will feedback the repair situation to the system and record the results of the abnormal handling. In addition, statistical reports on miswiring anomalies are generated regularly to analyze the frequency and causes of miswiring anomalies in different regions or different types of electric energy meters, providing data support for subsequent optimization and maintenance.

[0156] Working principle and its effect:

[0157] The method for analyzing the abnormality of wrong wiring of an electric energy meter of the present invention is based on an electric energy meter information collection system and is implemented by the following steps:

[0158] The key power parameters corresponding to each power meter are collected through the power consumption information collection system, and the timestamp is recorded to build a basic data set. The basic data set is preprocessed, including identifying missing points, duplicate records, data standardization, timestamp alignment, and resampling. A whitelist is established to exclude users who do not perform abnormal analysis.

[0159] The discriminant data set is classified and divided according to the type of electric energy meter to form the first discriminant data set, the second discriminant data set and the third discriminant data set. A variety of diagnostic rules are used to perform wrong wiring diagnosis on the three discriminant data sets. The specific steps are as follows:

[0160] The first diagnostic rule: includes reverse power consumption diagnosis and current and voltage anomaly diagnosis, calculates the reverse power consumption diagnosis score and the current and voltage anomaly diagnosis score, and finally constructs the first diagnostic rule score. Configure the first diagnostic threshold, and when the score is greater than the threshold, it is marked as a wrong wiring anomaly.

[0161] Second diagnostic rule: Perform voltage unbalance diagnosis and current unbalance diagnosis, calculate instantaneous voltage unbalance and voltage smoothing unbalance, and current unbalance judgment, and finally construct the second diagnostic rule score. Configure the second diagnostic threshold. When the score is greater than the threshold, it is marked as a wrong wiring abnormality.

[0162] The third diagnostic rule: perform neutral current anomaly analysis, calculate the mean and standard deviation of the neutral current sequence, use sliding average to smooth the data, perform autocorrelation analysis, and calculate the autocorrelation coefficient. Set the lag time, and when the autocorrelation coefficient at the lag time is greater than the correlation threshold, mark it as a miswiring anomaly.

[0163] When the energy meter is diagnosed with a miswiring anomaly, the early warning mechanism is triggered, and an early warning message containing the device number, error type, and occurrence time is generated, and pushed to the operation and maintenance personnel in real time through SMS or application notifications. After receiving the early warning message, the operation and maintenance personnel conduct on-site inspections and troubleshooting, and enter the on-site troubleshooting situation into the system. The system regularly generates statistical reports on miswiring anomalies, analyzes the frequency and causes of miswiring anomalies in different regions or different types of energy meters, and provides data support for subsequent optimization.

[0164] In summary, the electric energy meter miswiring anomaly analysis method of the present invention effectively monitors the wiring status of the electric energy meter through a series of steps, promptly discovers and handles miswiring problems, thereby significantly improving the safety, reliability and operation and maintenance efficiency of the power system, while promoting data-driven decision-making and management.

[0165] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A method for analyzing wrong wiring of electric energy meters based on an electric energy information collection system, characterized in that: include: The key power parameters corresponding to each power meter are collected through the power consumption information collection system, and the timestamp is recorded to build a basic data set; Data preprocessing is performed on the key power parameters of the basic data set, including identification of missing points, identification of duplicate records, data standardization, timestamp alignment and resampling, and a whitelist is established for special electricity users. Based on the key power parameters after data preprocessing, whitelist users are removed to construct a discriminant data set; Classifying and dividing the discriminant data set according to the type of electric energy meter to obtain a first discriminant data set, a second discriminant data set and a third discriminant data set; Using a plurality of diagnostic rules to perform miswiring diagnosis on the first discrimination data set, the second discrimination data set and the third discrimination data set, and identifying the electric energy meter with miswiring abnormality; When the electric energy meter is diagnosed with a miswiring abnormality, the early warning mechanism is triggered, and on-site fault elimination information is generated. The early warning information is pushed in real time to generate a fault elimination work order and notify the operation and maintenance personnel.

2. The method for analyzing wrong wiring of electric energy meters based on the electric energy information collection system according to claim 1, characterized in that: The specific steps of data preprocessing of the key electric energy parameters include: Identify the earliest timestamp and the latest timestamp of the key power parameters in all basic data sets, define the time period between the earliest timestamp and the latest timestamp as a common time axis, and convert the timestamp of the key power parameter data into an offset relative to the common time axis; Set the target sampling frequency f t , get the sampling timestamp according to the public time axis. If the start time of the public time axis is T start , then the sampling timestamp is: T * n =T start +n×f t , where n is a positive integer; The original sampling frequency of each electric energy meter in the basic data set is obtained, and the resampling factor is calculated by the ratio of the target sampling frequency to the original sampling frequency, that is: Where RF is the resampling factor, f t is the target sampling frequency, f o is the original sampling frequency; Configure the resampling threshold T RF , if the resampling factor is greater than the resampling threshold, the data is downsampled, otherwise the data is upsampled.

3. A method for analyzing wrong wiring of electric energy meters based on an electric energy information collection system according to claim 2, characterized in that: The specific steps of data preprocessing of the key electric energy parameters also include: When the data is upsampled, the sampling timestamp is T * n When configuring the interpolation time interval [T * n -τ,T * n +τ], where τ is the interpolation time interval, which is used to measure the time range that affects the current sampling timestamp. For any key power parameter x, the interpolation value of its sampling timestamp is calculated as follows: Among them, x(T * n ) is the key power parameter x at the sampling time stamp T * n , m is the number of original sampling points in the interpolation time interval, t k is the original sampling timestamp in the interpolation time interval, x(t k ) The key power parameter x at the original sampling time stamp t k The value of ω(t k ) is the non-negative weight corresponding to the original sampling timestamp in the interpolation time interval.

4. The method for analyzing wrong wiring of electric energy meters based on the electric energy information collection system according to claim 3, characterized in that: The specific steps of data preprocessing of the key electric energy parameters also include: When downsampling data, configure the sampling window The sampling time stamp is T * n The sampling window in which the For any key electric energy parameter y, the calculation method of the sampling timestamp value is: Among them, y(T * n ) is the key power parameter y at the sampling time stamp T * n The value at time t i is the original sampling timestamp within the sampling window, y(t i ) The key power parameter y at the original sampling time stamp t i The value of ω ′ (t i ) is the non-negative weight corresponding to the original sampling timestamp in the interpolation time interval.

5. The method for analyzing wrong wiring of electric energy meters based on the electric energy information collection system according to claim 1, characterized in that: The specific steps of the wrong wiring diagnosis include: According to the first diagnostic rule, a preliminary abnormality detection is performed on the key electric energy parameters of each electric energy meter in the discrimination data set, including reverse power consumption diagnosis and current and voltage abnormality diagnosis, to identify electric energy meters with miswiring abnormalities, and remove the electric energy meters marked as miswiring abnormalities in the second and third discrimination data sets and add them to the whitelist; According to the second diagnostic rule, the second discrimination data set and the third discrimination data set are subjected to secondary diagnosis, including voltage imbalance diagnosis and current imbalance diagnosis. The miswiring diagnosis is performed by constructing the second diagnostic rule score to identify the electric energy meters with miswiring abnormalities, and the electric energy meters marked as miswiring abnormalities in the third discrimination data set are removed and added to the whitelist. According to the third diagnostic rule, the third discrimination data set is analyzed for abnormality of the neutral current to identify the electric energy meter with the abnormality of the wrong wiring.

6. A method for analyzing wrong wiring of electric energy meters based on an electric energy information collection system according to claim 5, characterized in that: The first diagnostic rule includes: The active power data of each electric energy meter in the discriminant data set is traversed, and when the active power is detected to be negative, it is marked as an active power abnormal point; Extract the timestamps of the abnormal active power points and perform cluster analysis on the timestamps; The length of the time period in each cluster category, the number of active power anomaly points contained, and the distance between cluster categories are calculated to construct the reverse power consumption diagnosis score. The calculation formula for the reverse power consumption diagnosis score of each electric energy meter is as follows: Among them, S f is the reverse power consumption diagnosis score, Q is the number of timestamp clustering categories, is the number of abnormal active power points in the qth category, is the duration of the timestamp of the qth category, is the distance measure between the qth category and the pth category, η N , η T and η D They are all non-negative weights, corresponding to the non-negative weight coefficients of the number of active power anomaly points, the length of the timestamp duration period, and the distance between cluster categories.

7. A method for analyzing wrong wiring of electric energy meters based on an electric energy information collection system according to claim 6, characterized in that: The first diagnostic rule also includes: Calculate the current-voltage ratio corresponding to each sampling timestamp of each electric energy meter in the discriminant data set, perform a current-voltage ratio anomaly test, and obtain a current-voltage anomaly diagnosis score; For each electric energy meter, a first diagnostic rule score is constructed based on the reverse power consumption diagnostic score and the current and voltage abnormality diagnostic score. The calculation method of the first diagnostic rule score is as follows: S1=γ f S f +g d S d Among them, S1 is the first diagnostic rule score, S f and S d are the reverse power consumption diagnosis score and the current and voltage abnormality diagnosis score, γ f and γ d are the non-negative weight coefficients of the reverse power consumption diagnosis score and the current and voltage anomaly diagnosis score respectively; A first diagnosis threshold is configured, and when the first diagnosis rule score of the electric energy meter is greater than the first diagnosis threshold, it is marked as a miswiring abnormality, and the discrimination data set is traversed to filter the electric energy meters marked as miswiring abnormalities.

8. The method for analyzing wrong wiring of electric energy meters based on the electric energy information collection system according to claim 5, characterized in that: The second diagnostic rule includes: Obtain the three-phase voltage sequence corresponding to the sampling timestamp of each electric energy meter in the discriminant data set, and for each sampling timestamp, calculate the average value of the three-phase voltage; Calculate the deviation of each phase voltage from the average value of the three-phase voltage, and calculate the instantaneous voltage unbalance based on the three-phase voltage deviation, that is: Among them, U uf is the instantaneous voltage unbalance, ΔU a is the voltage deviation of phase A, ΔU b is the voltage deviation of phase B, ΔU c is the phase C voltage deviation, is the average value of three-phase voltage; Configure the measurement threshold T c , used to measure the smoothing time interval, for each sampling timestamp T * n , calculate the voltage smoothing imbalance within the measurement threshold, that is: in, is the sampling timestamp T * n Voltage smoothing unbalance, U uf (t) is the instantaneous voltage unbalance at the sampling time stamp t; Configure the voltage threshold. For each sampling timestamp, when the voltage smoothing imbalance is greater than the voltage threshold and is the instantaneous voltage imbalance, mark it as voltage imbalance. Count the number of sampling timestamps marked as voltage imbalance in each electric energy meter.

9. The method for analyzing wrong wiring of electric energy meters based on the electric energy information collection system according to claim 8, characterized in that: The second diagnostic rule also includes: Obtain the three-phase current sequence corresponding to the sampling timestamp of each electric energy meter in the discriminant data set, calculate the average value of the three-phase current for each sampling timestamp, and calculate the deviation of each phase current from the average value of the three-phase current; Calculate the instantaneous voltage imbalance and current smoothing imbalance based on the three-phase current deviation, make a current imbalance judgment for each sampling timestamp, obtain the sampling timestamps with current imbalance, and count the number of sampling timestamps with current imbalance in each electric energy meter; According to the number of voltage unbalance and current unbalance sampling timestamps marked by each electric energy meter, a second diagnostic rule score is constructed. The calculation method of the second diagnostic rule score is as follows: Among them, S1 is the second diagnostic rule score, N u is the number of times the voltage is unbalanced, N is the number of times the current is unbalanced, N T is the total number of sampling timestamps of the electric energy meter; μ u and μ I are the non-negative weight coefficients of the marking times of voltage unbalance and the marking times of current unbalance respectively; A second diagnostic threshold is configured. When the second diagnostic rule score of the electric energy meter is greater than the second diagnostic threshold, it is marked as a miswiring abnormality. The second discrimination data set and the third discrimination data set are traversed to filter the electric energy meters marked as miswiring abnormalities.

10. The method for analyzing wrong wiring of electric energy meters based on the electric energy information collection system according to claim 5, characterized in that: The third diagnostic rule includes: Configure the zero line mean threshold and the zero line standard deviation threshold, obtain the zero line current sequence corresponding to the sampling timestamp of each electric energy meter in the third discrimination data set, calculate its mean and standard deviation, and mark it as a potential abnormality when the mean and standard deviation of the electric energy meter are both greater than the zero line mean threshold and the zero line standard deviation threshold; The electric energy meters marked as potential abnormalities are screened out, and the zero-line current series is smoothed by sliding average. The smoothed zero-line current series is subjected to autocorrelation analysis, and the autocorrelation coefficient of the zero-line current series is calculated, namely: Among them, R ξ is the autocorrelation coefficient at lag ξ, I l 0 is the zero-line current value at the lth moment, μ0 is the mean value of the zero-line current, and N0 is the total number of data points in the zero-line current sequence; Set the delay time T ξ For each electric energy meter marked as potentially abnormal, calculate the zero line current sequence lag time T ξ The autocorrelation coefficient, configure the correlation threshold, when the lag time T ξ When the autocorrelation coefficient is greater than the correlation threshold, it is marked as a miswiring anomaly.

Images