An operation data acquisition terminal based on power grid special transformer

Abstract

The present invention discloses an operation data acquisition terminal based on a special transformer of a power grid, which relates to the field of data processing technology, and includes an acquisition module for analyzing and detecting the specific noise conditions in each type of data, and determining the main noise and noise source therein. A denoising module is used to set a denoising strategy according to the main noise conditions and noise sources of each type of data, thereby improving the adaptability of noise removal and ensuring the reliability of subsequent data analysis. A construction module is used to analyze the electricity consumption history of special transformer users to accurately formulate the user's electricity consumption pattern. An analysis module is used to set the boundaries of each time period according to the user's electricity consumption pattern, thereby determining the abnormal deviation part, constructing an abnormal chain, and comparing the abnormal chain with the data curve to determine the abnormality. A calculation module is used to integrate each type of data, and finally obtain a power grid operation index, thereby improving the accuracy and adaptability of the description of the power grid operation status and ensuring the safe and stable operation of the power grid.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an operation data acquisition terminal based on a power grid special transformer. Background Art

[0002] The operation data collection of power grid dedicated transformers is a method used to collect power information from specific industrial users or commercial users. These users usually require a large amount of power supply and have higher requirements for power quality.

[0003] In the prior art, the grid operation status is often described only based on the power load conditions, without considering other power parameters and the power consumption habits of dedicated transformer users, resulting in poor accuracy and adaptability of the grid operation status description.

[0004] Therefore, how to improve the accuracy and adaptability of the description of power grid operation status is a technical problem that needs to be solved. Summary of the invention

[0005] The purpose of the present invention is to solve the problem of poor accuracy and adaptability of the description of the power grid operation status in the prior art, and propose an operation data acquisition terminal based on a power grid special transformer, wherein the terminal comprises:

[0006] The acquisition module is used to collect the power-related data of the dedicated transformer users over a period of time and detect the noise in the power-related data;

[0007] A construction module is used to construct the power consumption pattern of the dedicated transformer user through the historical power related data of the dedicated transformer user;

[0008] A denoising module is used to determine a denoising strategy according to the noise situation in the power-related data, and remove the noise in the power-related data through the denoising strategy;

[0009] The analysis module is used to convert the power-related data within a period of time into a power-related data curve that changes with time, and to set the boundaries of the power-related data in different time periods according to the power consumption mode of the dedicated transformer user, to time-align the boundaries of the power-related data in different time periods with the power-related data curve, to find the abnormal parts on the power-related data curve, to connect all the abnormal parts on the power-related data curve in chronological order, to construct an abnormal chain, and to evaluate the abnormality of each abnormal chain by comparing the abnormal chain with the corresponding power-related data curve;

[0010] The calculation module is used to determine the power grid operation state according to the abnormality degree of the abnormal chain of each type of data in the power related data and the reflection weight of each type of data in the power related data on the power grid state.

[0011] In some embodiments of the present application, the acquisition module is specifically used to:

[0012] The power-related data include basic power data, power quality data, equipment operation data and load data. The noise type, noise intensity and noise frequency of each type of data in the basic power data, power quality data, equipment operation data and load data are determined through a preset noise identification model. The noise intensity of all noise types under the same category of data is compared to determine the relative noise intensity of each noise type, and the noise types are divided into main noise types according to the relative noise intensity. The noise source of each type of data is determined based on the characteristics of the power-related data and the noise frequency of the main noise type.

[0013] In some embodiments of the present application, the building blocks are specifically used for:

[0014] Preprocess the historical power-related data of the dedicated transformer users, and convert the historical power-related data of the dedicated transformer users into a time series of power-related data, identify the fixed change features in the time series of power-related data, and remove the components of the fixed change features in the time series of power-related data based on the fixed change features to obtain the remaining time series of power-related data;

[0015] Split the remaining power related data time series into multiple sub-time series, calculate the coefficient of variation of each sub-time series, and integrate the coefficient of variation of each sub-time series to obtain the coefficient of variation of the remaining power related data time series;

[0016] The limit value is set according to the length of the time series of the residual power related data, the time series of the residual power related data is calculated by the autocorrelation function to obtain the series value, the latter part of the series value is intercepted by the limit value to obtain the latter series value, the random fluctuation correction value is obtained by the size and change of the latter series value, the random fluctuation correction value is adjusted according to the coefficient of variation of the time series of the residual power related data, and the adjusted random fluctuation correction value is used to determine whether there is random fluctuation in the time series of the residual power related data;

[0017] If there is random fluctuation, the random fluctuation characteristics of the remaining power related data time series are identified, and the power consumption pattern of the dedicated transformer users is constructed by combining the fixed change characteristics;

[0018] Otherwise, the electricity consumption pattern of dedicated transformer users is constructed based on the fixed change characteristics.

[0019] In some embodiments of the present application, the denoising module is specifically used to:

[0020] A denoising algorithm is selected according to the main noise type and noise source of each type of data in basic power data, power quality data, equipment operation data and load data, and the denoising parameters in the denoising algorithm are set according to the noise intensity, so as to remove the noise in the power related data.

[0021] In some embodiments of the present application, the analysis module includes a first unit, a second unit and a third unit, and the first unit is used to:

[0022] According to the historical power-related data of the dedicated transformer users, the frequent points of the power-related data in each period are screened, and the frequent point set of the power-related data in each period is constructed;

[0023] The power consumption patterns of the dedicated transformer users are used to predict the power related prediction data in each period, and the adjustment coefficient in the period is determined according to the overlapping and non-overlapping parts of the frequent point sets of the energy related data in the same period on the power related prediction data;

[0024] ;

[0025] in, is the adjustment factor, is the conversion factor, is the amount of data that does not overlap between the frequent point set and the power-related prediction data, For the The size of non-overlapping data, is the amount of overlap between the frequent point set and the power related prediction data, is a preset constant;

[0026] The value range of the electric energy related prediction data in the time period is adjusted according to the adjustment coefficient, and the minimum value and the maximum value of the value range of the adjusted energy related prediction data are used as the boundaries in the time period.

[0027] In some embodiments of the present application, the second unit is used for:

[0028] Determine the part of the electric energy-related data curve that exceeds the boundary of each time period, and regard this part as the abnormal part. Evenly select several marking points on the abnormal part, and calculate the distances from several markings to the boundary respectively. Integrate the distances from all marking points to the boundary to determine a deviation, and mark the deviation on the abnormal part. Connect all abnormal parts in time or time sequence to construct an abnormal chain.

[0029] In some embodiments of the present application, the third unit is used for:

[0030] The abnormality of the abnormal chain is evaluated according to the deviation of each abnormal part in the abnormal chain, the length of the abnormal chain, and the length of the electric energy related data curve;

[0031] ;

[0032] in, is the abnormality degree of the abnormal chain, , are the weight of the abnormal chain itself, the relative weight of the abnormal chain and the power-related data curve, is the number of abnormal parts in the abnormal chain, The first The deviation of the abnormal part, is the average rate of change of the deviation of the abnormal part in the abnormal chain, is the preset constant, is the length of the exception chain, is the length of the electric energy related data curve.

[0033] In some embodiments of the present application, the computing module is specifically used to:

[0034] According to the abnormality degree of the abnormal chain of each type of data and the weight of each type of data reflecting the power grid status, all categories of data in the power-related data are weighted and summed to obtain a power grid operation index, which is used to describe the power grid operation status.

[0035] By applying the above technical solutions, the acquisition module is used to analyze and detect the specific noise conditions in each type of data, and determine the main noise and noise sources. The denoising module is used to set the denoising strategy according to the main noise conditions and noise sources of each type of data, so as to improve the adaptability of noise removal and ensure the reliability of subsequent data analysis. The construction module is used to analyze the electricity consumption history of special transformer users to accurately formulate the user's electricity consumption pattern. The analysis module is used to set the boundaries of each time period according to the user's electricity consumption pattern, so as to determine the abnormal deviation part, build an abnormal chain, and compare the abnormal chain with the data curve to determine the degree of abnormality. The calculation module is used to integrate each type of data and finally obtain a power grid operation indicator, thereby improving the accuracy and adaptability of the description of the power grid operation status and ensuring the safe and stable operation of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is a structural schematic diagram of an operation data acquisition terminal based on a power grid special transformer proposed by the present invention;

[0037] Figure 2 This is a structural schematic diagram of the analysis module in the operation data acquisition terminal based on the power grid special transformer proposed by the present invention. DETAILED DESCRIPTION

[0038] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0039] Reference Figure 1 , an operation data acquisition terminal based on power grid special transformer, including the following modules:

[0040] The acquisition module is used to collect the power-related data of the dedicated transformer users over a period of time and detect the noise in the power-related data.

[0041] In this embodiment, different data types in the power-related data represent different situations, and it is necessary to detect the specific noise situation of each type of data. The noise situation includes the noise type, intensity, frequency, etc.

[0042] In some embodiments of the present application, the acquisition module is specifically used to:

[0043] The power-related data include basic power data, power quality data, equipment operation data and load data. The noise type, noise intensity and noise frequency of each type of data in the basic power data, power quality data, equipment operation data and load data are determined through a preset noise identification model. The noise intensity of all noise types under the same category of data is compared to determine the relative noise intensity of each noise type, and the noise types are divided into main noise types according to the relative noise intensity. The noise source of each type of data is determined based on the characteristics of the power-related data and the noise frequency of the main noise type.

[0044] In this embodiment, the basic power data includes real-time values ​​and trends such as voltage, current, and frequency. The power quality data includes harmonic content, voltage flicker, three-phase imbalance, etc. The equipment operation data includes temperature, sound, vibration, etc. The preset noise recognition model here can be a fast Fourier transform or a machine learning model. There may be multiple types of noise in a type of data, and the main noise needs to be identified. The relative intensity of noise is the relative intensity obtained after normalizing the noise intensity of different noise types. The noise type with higher relative noise intensity is regarded as the main noise, and the source of noise is determined in combination with the characteristics of the data. For example, the system noise in the equipment operation data is strong, which is often caused by equipment calibration errors or sensor offsets (noise source).

[0045] A building module is used to build the power consumption pattern of dedicated transformer users through their historical power-related data.

[0046] In this embodiment, because different dedicated transformer users have different electricity usage habits and electricity demands, the electricity usage differences between different users are relatively large. Therefore, the electricity usage pattern of the user is constructed based on the previous electricity usage data of the dedicated transformer user. The electricity usage pattern is a prediction model that can predict electricity-related data.

[0047] In some embodiments of the present application, the building blocks are specifically used for:

[0048] Preprocess the historical power-related data of the dedicated transformer users, and convert the historical power-related data of the dedicated transformer users into a time series of power-related data, identify the fixed change features in the time series of power-related data, and remove the components of the fixed change features in the time series of power-related data based on the fixed change features to obtain the remaining time series of power-related data;

[0049] Split the remaining power related data time series into multiple sub-time series, calculate the coefficient of variation of each sub-time series, and integrate the coefficient of variation of each sub-time series to obtain the coefficient of variation of the remaining power related data time series;

[0050] The limit value is set according to the length of the time series of the residual power related data, the time series of the residual power related data is calculated by the autocorrelation function to obtain the series value, the latter part of the series value is intercepted by the limit value to obtain the latter series value, the random fluctuation correction value is obtained by the size and change of the latter series value, the random fluctuation correction value is adjusted according to the coefficient of variation of the time series of the residual power related data, and the adjusted random fluctuation correction value is used to determine whether there is random fluctuation in the time series of the residual power related data;

[0051] If there is random fluctuation, the random fluctuation characteristics of the remaining power related data time series are identified, and the power consumption pattern of the dedicated transformer users is constructed by combining the fixed change characteristics;

[0052] Otherwise, the electricity consumption pattern of dedicated transformer users is constructed based on the fixed change characteristics.

[0053] In this embodiment, the characteristics of time series data are extracted, such as fixed change characteristics such as trend, seasonality, and periodicity. The trend characteristics reflect the long-term change trend of the data over time, the seasonal characteristics reflect the law of data changing with seasons, and the periodic characteristics reflect the law of data changing periodically over time. Often, the power consumption situation will be affected by these fixed change characteristics, so in general, the impact of these fixed changes will be considered to construct the power consumption model. However, in the prior art, the impact of random fluctuations on the power consumption model is not considered. Here, the fixed change characteristics are first identified, and the fixed change characteristic components are removed from the time series of power-related data. After removing these components, random fluctuations can be easily detected.

[0054] It should be noted that the methods for processing trend components include differences (such as first-order differences, second-order differences, etc.) and trend fitting (such as linear trends, exponential trends, etc.). The methods for processing seasonal components include seasonal adjustments (such as seasonal differences, seasonal smoothing, etc.) and seasonal models (such as seasonal ARIMA models, seasonal neural networks, etc.). The methods for processing periodic components include periodic decomposition (such as Fourier analysis) and periodic models (such as periodic ARIMA models, periodic neural networks, etc.).

[0055] In this embodiment, because the stability of the data will affect the judgment of random fluctuations, the stability of the time series of the remaining power related data is first determined, and the coefficient of variation is used here to describe it. Time series analysis techniques, such as autocorrelation function (ACF) or partial autocorrelation function (PACF), are applied to detect random fluctuations. If the ACF still maintains a high value when the lag number is large, there may be random fluctuations. The autocorrelation function describes the correlation between a time series and itself at different time points. The calculation method of ACF is to calculate the covariance of the time series and its own delayed k time points, and then divide it by the variance of the time series. The value of ACF is usually 1 when k=0, and gradually decreases as k increases, and eventually tends to 0. If the ACF still maintains a high value when the lag number is large, this may indicate that there are random fluctuations in the data, that is, the data is not stationary, but is affected by random noise. The calculation result of the autocorrelation function (ACF) is a series of values ​​that reflect the correlation between the time series and itself at different lag points. Specifically, the calculation result of the ACF function is a sequence that changes with the k value, where k represents the number of lag points in the time series data.

[0056] In this embodiment, a limit value is set according to the length of the time series of the remaining power-related data, and the latter part of the series value of the autocorrelation function calculation result is intercepted by the limit value. The fluctuation correction value (indicating whether there is random fluctuation in the sequence) is determined by the size and change of the latter part of the series value, and the fluctuation correction value is adjusted according to the coefficient of variation. Different coefficients of variation correspond to different adjustment coefficients, and the fluctuation correction value is adjusted by multiplying the adjustment coefficient and the fluctuation correction value. Identify the random fluctuation characteristics, and combine the fixed change characteristics to construct a prediction model, which can predict the change of power-related data.

[0057] The denoising module is used to determine a denoising strategy according to the noise situation in the power-related data, and remove the noise in the power-related data through the denoising strategy.

[0058] In some embodiments of the present application, the denoising module is specifically used to:

[0059] A denoising algorithm is selected according to the main noise type and noise source of each type of data in basic power data, power quality data, equipment operation data and load data, and the denoising parameters in the denoising algorithm are set according to the noise intensity, so as to remove the noise in the power related data.

[0060] In this embodiment, the source of noise can be determined by the noise type and noise frequency distribution, etc. The denoising parameters in the denoising algorithm are set according to the noise intensity. For example, for median filtering, a suitable filter window size can be selected according to the noise intensity; for wavelet denoising, a suitable threshold can be selected according to the noise intensity, etc.

[0061] The analysis module is used to convert the power-related data within a period of time into a power-related data curve that changes with time, and set the boundaries of the power-related data in different time periods according to the power consumption pattern of the special transformer users, align the boundaries of the power-related data in different time periods with the power-related data curve, find the abnormal parts on the power-related data curve, connect all the abnormal parts on the power-related data curve in chronological order, construct an abnormal chain, and evaluate the abnormality of each abnormal chain by comparing the abnormal chain with the corresponding power-related data curve.

[0062] In this embodiment, the power consumption patterns of the dedicated transformer users are used to predict the power-related data in different time periods, thereby setting a boundary, and the portion of the power-related data curve that exceeds the boundary is integrated to determine the degree of abnormality.

[0063] In some embodiments of the present application, Figure 2 As shown, the analysis module includes a first unit, a second unit and a third unit, and the first unit is used for:

[0064] According to the historical power-related data of the dedicated transformer users, the frequent points of the power-related data in each period are screened, and the frequent point set of the power-related data in each period is constructed;

[0065] The power consumption patterns of the dedicated transformer users are used to predict the power related prediction data in each period, and the adjustment coefficient in the period is determined according to the overlapping and non-overlapping parts of the frequent point sets of the energy related data in the same period on the power related prediction data;

[0066] ;

[0067] in, is the adjustment factor, is the conversion factor, is the amount of data that does not overlap between the frequent point set and the power-related prediction data, For the The size of non-overlapping data, is the amount of overlap between the frequent point set and the power related prediction data, is a preset constant;

[0068] The value range of the electric energy related prediction data in the time period is adjusted according to the adjustment coefficient, and the minimum value and the maximum value of the value range of the adjusted energy related prediction data are used as the boundaries in the time period.

[0069] In this embodiment, the frequent points (value size, frequent occurrence) of the power-related data in each time period are screened according to the historical power-related data of the dedicated transformer users. Indicates that the impact of non-overlapping data is corrected by the ratio of the non-overlapping part to the total number, and then the class average is taken to represent the adjustment coefficient. The minimum and maximum values ​​are adjusted by multiplying the adjustment coefficients.

[0070] In some embodiments of the present application, the second unit is used for:

[0071] Determine the part of the electric energy-related data curve that exceeds the boundary of each time period, and regard this part as the abnormal part. Evenly select several marking points on the abnormal part, and calculate the distances from several markings to the boundary respectively. Integrate the distances from all marking points to the boundary to determine a deviation, and mark the deviation on the abnormal part. Connect all abnormal parts in time or time sequence to construct an abnormal chain.

[0072] In this embodiment, the distances from all marking points to the boundary are integrated to determine a deviation, where the distance refers to the distance to a boundary to which the marking point is closer.

[0073] In some embodiments of the present application, the third unit is used for:

[0074] The abnormality of the abnormal chain is evaluated according to the deviation of each abnormal part in the abnormal chain, the length of the abnormal chain, and the length of the electric energy related data curve;

[0075] ;

[0076] in, is the abnormality degree of the abnormal chain, , are the weight of the abnormal chain itself, the relative weight of the abnormal chain and the power-related data curve, is the number of abnormal parts in the abnormal chain, The first The deviation of the abnormal part, is the average rate of change of the deviation of the abnormal part in the abnormal chain, is the preset constant, is the length of the exception chain, is the length of the electric energy related data curve.

[0077] In this embodiment, the abnormal amount of the abnormal chain itself is determined based on the deviation of the abnormal part of the abnormal chain and the change in the deviation, the relative abnormal amount is determined based on the ratio of the length of the abnormal chain and the electric energy related data curve, and the abnormal degree of the abnormal chain is determined by combining the abnormal amount of the abnormal chain itself and the relative abnormal amount.

[0078] The calculation module is used to determine the power grid operation state according to the abnormality degree of the abnormal chain of each type of data in the power related data and the reflection weight of each type of data in the power related data on the power grid state.

[0079] In some embodiments of the present application, the computing module is specifically used to:

[0080] According to the abnormality degree of the abnormal chain of each type of data and the weight of each type of data reflecting the power grid status, all categories of data in the power-related data are weighted and summed to obtain a power grid operation index, which is used to describe the power grid operation status.

[0081] In this embodiment, the load data has a relatively greater weight in reflecting the power grid status than other types of data.

[0082] By applying the above technical solutions, the acquisition module is used to analyze and detect the specific noise conditions in each type of data, and determine the main noise and noise sources. The denoising module is used to set the denoising strategy according to the main noise conditions and noise sources of each type of data, so as to improve the adaptability of noise removal and ensure the reliability of subsequent data analysis. The construction module is used to analyze the electricity consumption history of special transformer users to accurately formulate the user's electricity consumption pattern. The analysis module is used to set the boundaries of each time period according to the user's electricity consumption pattern, so as to determine the abnormal deviation part, build an abnormal chain, and compare the abnormal chain with the data curve to determine the degree of abnormality. The calculation module is used to integrate each type of data and finally obtain a power grid operation indicator, thereby improving the accuracy and adaptability of the description of the power grid operation status and ensuring the safe and stable operation of the power grid.

[0083] This solution also includes the following functions or modules: Data storage: The processed data is stored in the internal memory of the terminal for subsequent use or transmission. The stored data usually includes raw data and processed data. Data upload: The terminal uploads the data to the master station management system through the uplink communication interface according to the preset schedule or event trigger. The uploaded data can be used for remote meter reading, power management, fault diagnosis, etc. Communication connection: The terminal establishes a communication connection with the connected power meter through the downlink communication interface (such as RS-485, PLC, etc.). At the same time, a communication connection is established with the master station management system through the uplink communication interface (such as GPRS / CDMA, Ethernet, optical fiber, etc.).

[0084] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present invention can be implemented by hardware, or by software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present invention.

[0085] Those skilled in the art will appreciate that the accompanying drawings are merely schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily required for implementing the present invention.

[0086] Those skilled in the art will appreciate that the modules in the system in the implementation scenario can be distributed in the system of the implementation scenario according to the implementation scenario description, or can be changed accordingly and located in one or more systems different from the implementation scenario. The modules in the above implementation scenario can be combined into one module, or can be further split into multiple submodules.

[0087] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An operation data acquisition terminal based on a power grid special transformer, characterized in that: include: The acquisition module is used to collect the power-related data of the dedicated transformer users over a period of time and detect the noise in the power-related data; A construction module is used to construct the power consumption pattern of the dedicated transformer user through the historical power related data of the dedicated transformer user; A denoising module is used to determine a denoising strategy according to the noise situation in the power-related data, and remove the noise in the power-related data through the denoising strategy; The analysis module is used to convert the power-related data within a period of time into a power-related data curve that changes with time, and to set the boundaries of the power-related data in different time periods according to the power consumption mode of the dedicated transformer user, to time-align the boundaries of the power-related data in different time periods with the power-related data curve, to find the abnormal parts on the power-related data curve, to connect all the abnormal parts on the power-related data curve in chronological order, to construct an abnormal chain, and to evaluate the abnormality of each abnormal chain by comparing the abnormal chain with the corresponding power-related data curve; A calculation module, used to determine the power grid operation state according to the abnormality degree of the abnormal chain of each type of data in the power-related data and the reflection weight of each type of data in the power-related data on the power grid state; The acquisition module is specifically used for: The power-related data include basic power data, power quality data, equipment operation data and load data. The noise type, noise intensity and noise frequency of each type of data in the basic power data, power quality data, equipment operation data and load data are determined by a preset noise recognition model. The noise intensity of all noise types under the same category of data is compared to determine the relative noise intensity of each noise type. The noise types are divided into main noise types according to the relative noise intensity, and the noise source of each type of data is determined based on the characteristics of the power-related data and the noise frequency of the main noise type. The analysis module includes a first unit, a second unit and a third unit, wherein the first unit is used for: According to the historical power-related data of the dedicated transformer users, the frequent points of the power-related data in each period are screened, and the frequent point set of the power-related data in each period is constructed; The power consumption patterns of the dedicated transformer users are used to predict the power related prediction data in each time period, and the adjustment coefficient in the time period is determined according to the overlapping and non-overlapping parts of the frequent point sets of the power related data in the same time period on the power related prediction data; ； in, is the adjustment factor, is the conversion factor, is the amount of data that does not overlap between the frequent point set and the power-related prediction data, For the The size of non-overlapping data, is the amount of overlap between the frequent point set and the power related prediction data, is a preset constant; The value range of the electric energy related prediction data in the time period is adjusted according to the adjustment coefficient, and the minimum value and the maximum value of the value range of the adjusted electric energy related prediction data are used as the boundaries in the time period.

2. The operation data acquisition terminal based on power grid special transformer according to claim 1 is characterized in that: The building blocks are specifically used for: Preprocess the historical power-related data of the dedicated transformer users, and convert the historical power-related data of the dedicated transformer users into a time series of power-related data, identify the fixed change features in the time series of power-related data, and remove the components of the fixed change features in the time series of power-related data based on the fixed change features to obtain the remaining time series of power-related data; Split the remaining power related data time series into multiple sub-time series, calculate the coefficient of variation of each sub-time series, and integrate the coefficient of variation of each sub-time series to obtain the coefficient of variation of the remaining power related data time series; The limit value is set according to the length of the time series of the residual power related data, the time series of the residual power related data is calculated by the autocorrelation function to obtain the series value, the latter part of the series value is intercepted by the limit value to obtain the latter series value, the random fluctuation correction value is obtained by the size and change of the latter series value, the random fluctuation correction value is adjusted according to the coefficient of variation of the time series of the residual power related data, and the adjusted random fluctuation correction value is used to determine whether there is random fluctuation in the time series of the residual power related data; If there is random fluctuation, the random fluctuation characteristics of the remaining power related data time series are identified, and the power consumption pattern of the dedicated transformer users is constructed by combining the fixed change characteristics; Otherwise, the electricity consumption pattern of dedicated transformer users is constructed based on the fixed change characteristics.

3. The operation data acquisition terminal based on power grid special transformer according to claim 1 is characterized in that: The denoising module is specifically used for: A denoising algorithm is selected according to the main noise type and noise source of each type of data in basic power data, power quality data, equipment operation data and load data, and the denoising parameters in the denoising algorithm are set according to the noise intensity, so as to remove the noise in the power related data.

4. The operation data acquisition terminal based on power grid special transformer according to claim 1 is characterized in that: The second unit is used for: Determine the part of the electric energy-related data curve that exceeds the boundary of each time period, and regard this part as the abnormal part. Evenly select several marking points on the abnormal part, and calculate the distances from several markings to the boundary respectively. Integrate the distances from all marking points to the boundary to determine a deviation, and mark the deviation on the abnormal part. Connect all abnormal parts in time or time sequence to construct an abnormal chain.

5. The operation data acquisition terminal based on power grid special transformer according to claim 4 is characterized in that: The third unit is used for: The abnormality of the abnormal chain is evaluated according to the deviation of each abnormal part in the abnormal chain, the length of the abnormal chain, and the length of the electric energy related data curve; ； in, is the abnormality degree of the abnormal chain, , are the weight of the abnormal chain itself, the relative weight of the abnormal chain and the power-related data curve, is the number of abnormal parts in the abnormal chain, The first The deviation of the abnormal part, is the average rate of change of the deviation of the abnormal part in the abnormal chain, is the preset constant, is the length of the exception chain, is the length of the electric energy related data curve.

6. The operation data acquisition terminal based on power grid special transformer according to claim 1 is characterized in that: The computing module is specifically used for: According to the abnormality degree of the abnormal chain of each type of data and the weight of each type of data reflecting the power grid status, all categories of data in the power-related data are weighted and summed to obtain a power grid operation index, which is used to describe the power grid operation status.

Images