CN-122017720-A - Ammeter misalignment detection method and system based on high-frequency sampling data of electricity acquisition system

CN122017720ACN 122017720 ACN122017720 ACN 122017720ACN-122017720-A

Abstract

The application relates to the technical field of electric variable measurement, and discloses an ammeter misalignment detection method and system based on high-frequency sampling data of an electricity acquisition system, wherein the method comprises the steps of acquiring the high-frequency acquisition data; the method comprises the steps of performing voltage range verification, electric quantity abnormality screening and time continuity verification, calculating to achieve single-phase sub-table phase inversion identification, establishing a linear regression model between a total table and an in-phase sub-table, introducing regularization constraint to solve correction coefficients of the sub-tables, and performing threshold judgment and result output. Compared with the prior art relying on the manual spot check ammeter anomaly identification method, the method has the technical problem that high-precision misalignment detection is difficult to realize under the condition that the sub-meter phase access is ambiguous. According to the application, by introducing a high-frequency data unified time reference construction mechanism and an automatic phase-returning identification mechanism based on voltage correlation, the precise modeling and stable solving of the metering deviation of each sub-meter are realized, and the accuracy of ammeter misalignment detection is improved.

Inventors

LI QUAN
XIA YUBAO
CHEN YOUFENG
LIU GUOPENG
SUN YAN
SHI JIAFENG
SHI MENGYUN
CUI MENGHAN
WANG YIYU
ZHANG XUAN
WANG JING
HE CHENGJIAN
WANG GUOYU

Assignees

南京米特科技股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260410

Claims (10)

1. An ammeter misalignment detection method based on high-frequency sampling data of an electricity acquisition system is characterized by comprising the following steps: Step S10, acquiring high-frequency acquisition data of a total table and each sub table in a target platform area, executing an original detection data set construction task by adopting a time synchronization correction and same window mapping construction mechanism based on the high-frequency acquisition data, and outputting a synchronous original detection data set; Step S20, performing an abnormal data screening task by adopting a multidimensional validity checking mechanism based on the synchronous original detection data set, and outputting an effective detection data set; Step S30, performing a sub-table phase-returning preprocessing task by adopting a voltage correlation phase recognition mechanism based on the effective detection data set, and outputting a phase-separated detection data set; Step S40, performing a correction coefficient solving task by adopting a sliding window sample construction and regularization linear regression solving mechanism based on the split-phase detection data set, and outputting each sub-table correction coefficient set; And S50, calculating the misalignment proportion according to each sub-table correction coefficient set, judging the threshold value, reporting the result, and outputting an ammeter misalignment detection result set.
2. The method for detecting the misalignment of an ammeter based on high-frequency sampling data of an electricity acquisition system as claimed in claim 1, wherein in step S10, high-frequency acquisition data of a total table and each sub table in a target platform area is acquired, an original detection data set construction task is executed by adopting a time synchronization correction and same window mapping construction mechanism based on the high-frequency acquisition data, and a synchronous original detection data set is output, which specifically comprises the following steps: s101, establishing communication connection with a summary list and each sub-list through a preset misalignment detection module via an RS-485 bus, and reading high-frequency acquisition data, wherein the high-frequency acquisition data comprises a list address, a phase type, a multiplying power parameter and sampling time scale information of each ammeter; Step S102, acquiring total surface electric quantity data, total surface voltage data, sub-meter electric quantity data and sub-meter voltage data according to preset high frequency time granularity based on high frequency acquisition data to form an original reading data set; Step 103, checking consistency of each sub-meter reading time in the original reading data set and a total standard time slice, and when time deviation exists, establishing a time deviation correction model based on a calendar extracts from history reading rule, wherein the time deviation correction model corrects the original reading data set into a unified time window by adopting a linear interpolation or time interval electric quantity sharing mode to obtain a synchronous original detection data set.
3. The method for detecting the misalignment of an electric meter based on high-frequency sampling data of an electric power consumption acquisition system as claimed in claim 1, wherein in step S20, an abnormal data screening task is executed by using a multidimensional validity checking mechanism based on a synchronous original detection data set, and the step of outputting the valid detection data set specifically comprises the following steps: Step S201, reading voltage data of the total table and the sub table at each moment in the synchronous original detection data set, performing range constraint verification on the voltage data, and marking the corresponding moment data as invalid data and eliminating the invalid data when any voltage value in the voltage data exceeds a preset normal range; step S202, reading electric quantity data of a total table and a sub table at each moment in a synchronous original detection data set, performing continuity check on the electric quantity data of adjacent freezing time points, and eliminating corresponding window data when electric quantity mutation, deletion, reverse abnormality or time point discontinuity are screened; Step 203, screening out the data with the electric quantity variation amplitude lower than the preset electric quantity variation threshold value in the synchronous original detection data set; and S204, outputting the synchronous original detection data set after abnormal data screening as a valid detection data set.
4. A method for detecting misalignment of an electricity meter based on high-frequency sampling data of an electricity collection system as claimed in claim 3, wherein in step S202, continuity check is performed on the electricity quantity data of adjacent freezing time points, specifically, whether the absolute value of the deviation between the time interval of two adjacent freezing time points and the preset sampling period exceeds the continuity tolerance threshold is taken as a criterion, when: And determining the corresponding section as a discontinuous data section, wherein, Represent the first The time point of the freezing is set to be the same, Represent the first +1 Of the time points of freezing, Representing a preset sampling period of time, Representing a continuity tolerance threshold.
5. The method for detecting misalignment of an electric meter based on high-frequency sampling data of an electric power consumption acquisition system as claimed in claim 1, wherein in step S30, a voltage correlation phase recognition mechanism is adopted to perform a sub-meter phase-returning preprocessing task based on an effective detection data set, and the step of outputting a phase-separated detection data set specifically comprises: step S301 extracting a total three-phase voltage sequence from the active detection dataset Voltage sequence of single-phase sub-meter to be identified Constructing a candidate phase identification data set; Step S302, calculating the voltage sequence And total three-phase voltage sequence Pearson correlation coefficient therebetween Pearson correlation coefficient The method meets the following conditions: ; Wherein, the Representing a voltage sequence corresponding to any candidate phase in the total table, wherein the candidate phase comprises a total three-phase voltage sequence Any one of the following; Representing the number of sample points participating in the correlation calculation; representing the voltage value of the kth voltage sequence representing the single-phase sub-table to be identified at the kth sampling time; Representing the average voltage value of the single-phase sub-table to be identified in a sample interval; Representing average voltage values of the voltage sequences corresponding to the total surface candidate phases in the sample interval; Step S303, selecting a candidate phase with the largest Pearson correlation coefficient as a home phase of a corresponding single-phase sub-table, and only reserving total electric quantity data under the home phase and corresponding sub-table electric quantity data to form a split-phase detection data set.
6. The method for detecting the misalignment of an ammeter based on high-frequency sampling data of an electricity acquisition system as claimed in claim 1, wherein in step S40, a sliding window sample construction and regularization linear regression solution mechanism is adopted to execute a correction coefficient solution task based on a split-phase detection data set, and the step of outputting each sub-table correction coefficient set specifically comprises the following steps: Step S401, respectively establishing a linear regression model between total electric quantity and the same sub-meter electric quantity for each home phase in the split-phase detection data set, setting the total electric quantity as Y, setting the same sub-meter electric quantity vector as X= [ X 1 ,X 2 ,…,X n ], wherein n is the same sub-meter electric quantity vector length, X n is the electric quantity value of the n-th in-phase sub-meter in a corresponding sampling window, and setting a linear relation model: Wherein e represents a comprehensive error term; the correction coefficient corresponding to the nth sub-table is used for representing the proportional relation between the electric quantity of the sub-table and the total electric quantity; constructing M groups of samples by adopting a sliding window strategy based on a linear relation model to form an overdetermined equation set, wherein M is greater than n; Step S402, performing normalization processing on each column of the sample input matrix in the overdetermined equation set, outputting a normalization matrix, and constructing a normal equation with a regular term based on the normalization matrix: Wherein, the method comprises the steps of, Representing a sample input matrix formed by M groups of sliding window samples, wherein each row of the sample input matrix corresponds to an electric quantity observation value of each sub-meter under one sampling window; Representation matrix Is a transposed matrix of (a); Represents regularization parameters, and lambda is more than or equal to 0; representing the identity matrix; representing a normalized coefficient vector obtained based on the normalized matrix solution; Representing a total power observation vector corresponding to the sample input matrix; And S403, solving a normal equation with a regular term by adopting a Gaussian elimination method to obtain correction coefficient sets of each sub-table by normalization.
7. The method for detecting the misalignment of the ammeter based on the high-frequency sampling data of the electricity acquisition system according to claim 1, wherein in the step S50, the misalignment proportion calculation, the threshold judgment and the result reporting are carried out according to each sub-meter correction coefficient set, and the step of outputting the misalignment detection result set of the ammeter specifically comprises the following steps: step S501, obtaining the ith sub-meter correction coefficient from the electric meter misalignment detection result set Correction coefficients according to the sub-table Calculating the misalignment ratio of the corresponding sub-tables , ; When (when) When the corresponding sub-meter measurement is judged to be normal; When (when) When the corresponding sub-table is judged to have slow character moving; When (when) When the corresponding sub-table is judged to have quick character passing; Step S502, the misalignment ratio With a preset misalignment threshold Comparing when meeting When the corresponding sub-meter is marked as a misalignment fault ammeter, a misalignment ammeter list and suggested correction coefficient information are generated; and S503, uploading the misalignment ammeter list, the suggested correction coefficient information and the misalignment proportion to the transportation terminal through a preset remote communication unit, and outputting an ammeter misalignment detection result set.
8. An ammeter misalignment detection system based on high-frequency sampling data of an electricity acquisition system, which is applied to the ammeter misalignment detection method based on high-frequency sampling data of the electricity acquisition system as claimed in any one of claims 1 to 7, and is characterized in that the ammeter misalignment detection system based on high-frequency sampling data of the electricity acquisition system comprises: the data construction module is used for acquiring high-frequency acquisition data of the total table and each sub table in the target platform area, executing an original detection data set construction task by adopting a time synchronization correction and same window mapping construction mechanism based on the high-frequency acquisition data, and outputting a synchronous original detection data set; The validity screening module is used for executing an abnormal data screening task by adopting a multidimensional validity checking mechanism based on the synchronous original detection data set and outputting an effective detection data set; The phase identification module is used for executing a sub-table phase-returning preprocessing task by adopting a voltage correlation phase identification mechanism based on the effective detection data set and outputting a phase-splitting detection data set; The coefficient solving module is used for executing a correction coefficient solving task by adopting a sliding window sample construction and regularization linear regression solving mechanism based on the split-phase detection data set and outputting each sub-table correction coefficient set; and the result judging module is used for carrying out misalignment proportion calculation, threshold judgment and result reporting according to each sub-table correction coefficient set and outputting an ammeter misalignment detection result set.
9. An ammeter misalignment detection device based on high-frequency sampling data of an electricity acquisition system is characterized by comprising a memory, a processor and an ammeter misalignment detection program which is stored on the memory and can run on the processor and is based on the high-frequency sampling data of the electricity acquisition system, wherein the ammeter misalignment detection program based on the high-frequency sampling data of the electricity acquisition system is executed by the processor to realize the ammeter misalignment detection method based on the high-frequency sampling data of the electricity acquisition system according to any one of claims 1 to 7.
10. A computer program product, characterized in that the computer program product comprises an ammeter misalignment detection program based on high-frequency sampling data of an electricity acquisition system, which when executed by a processor implements an ammeter misalignment detection method based on high-frequency sampling data of an electricity acquisition system according to any one of claims 1 to 7.

Description

Ammeter misalignment detection method and system based on high-frequency sampling data of electricity acquisition system Technical Field The invention relates to the technical field of electric variable measurement, in particular to an ammeter misalignment detection method and system based on high-frequency sampling data of an electricity acquisition system. Background Currently, in an electricity consumption information acquisition system, the evaluation of the running state of an ammeter mainly depends on periodic meter reading data or low-frequency statistical data, and manual spot check, experience judgment or a simple total electricity distribution amount comparison method is adopted for identifying the misalignment problem of the ammeter. Such methods have significant drawbacks in practical applications. For example, under the conditions that load fluctuation of a platform area is frequent, electricity utilization behaviors of users have randomness and access phase information of sub-tables is lost or changed, the traditional statistical analysis method based on low-frequency data is difficult to accurately reflect fine granularity characteristics of electric quantity change, and cannot effectively distinguish influence relations between metering errors and load fluctuation, and meanwhile, when strong correlation exists between the sub-tables or problems of noise, loss and time asynchronism exist in the data, the traditional method is difficult to construct a stable and reliable analysis model, so that the fluctuation of misalignment identification results is large, and the misjudgment rate is high. In addition, the prior art generally lacks automatic recognition capability for the sub-meter phase-returning problem, often relies on manual standing accounts or field checking, and once phase wiring adjustment or data identification errors occur in actual operation, analysis basic distortion is easily caused, so that accuracy of ammeter misalignment judgment is further affected. Especially in a high-density area or a distributed load access scene, the prior art cannot fully meet the requirements of carrying out fine, automatic and quantifiable evaluation on the metering deviation of the ammeter. Therefore, there is a need for an ammeter misalignment detection method which can still realize automatic phase returning of the sub-meter, stable solving of metering deviation and quantitative judgment of misalignment proportion under the conditions that the sub-meter phase is not clear, noise interference exists in data and load fluctuation is remarkable, so as to improve the accuracy and stability of ammeter anomaly identification and the feasibility of engineering application. Disclosure of Invention Aiming at the technical defects, the invention aims to provide an ammeter misalignment detection method based on high-frequency sampling data of an electricity acquisition system, and aims to solve the technical problem that in the prior art, the method relies on a manual spot-checking ammeter anomaly identification method, and particularly under the condition of ambiguous sub-meter phase access, high-precision misalignment detection is difficult to realize. In order to solve the technical problems, the invention adopts the following technical scheme that the invention provides an ammeter misalignment detection method based on high-frequency sampling data of an electricity acquisition system. The ammeter misalignment detection method based on the high-frequency sampling data of the electricity acquisition system comprises the following steps: Step S10, acquiring high-frequency acquisition data of a total table and each sub table in a target platform area, executing an original detection data set construction task by adopting a time synchronization correction and same window mapping construction mechanism based on the high-frequency acquisition data, and outputting a synchronous original detection data set; Step S20, performing an abnormal data screening task by adopting a multidimensional validity checking mechanism based on the synchronous original detection data set, and outputting an effective detection data set; Step S30, performing a sub-table phase-returning preprocessing task by adopting a voltage correlation phase recognition mechanism based on the effective detection data set, and outputting a phase-separated detection data set; Step S40, performing a correction coefficient solving task by adopting a sliding window sample construction and regularization linear regression solving mechanism based on the split-phase detection data set, and outputting each sub-table correction coefficient set; And S50, calculating the misalignment proportion according to each sub-table correction coefficient set, judging the threshold value, reporting the result, and outputting an ammeter misalignment detection result set. Preferably, in step S10, high frequency acquisition data of a total table and each sub table in the tar