Area metering device misalignment identification method and system based on EM algorithm

Abstract

The invention relates to the technical field of electric energy metering of a low-voltage power distribution system, and provides an EM algorithm-based district metering device misalignment identification method and system. The method comprises the following steps: establishing a theoretical misalignment value model of each outgoing line in different metering misalignment states, and determining a mapping relation between the different misalignment states and actual electric quantity data; the method comprises the following steps: constructing theoretical deviation values of each outgoing line metering device in different misalignment states on the basis of a transformer area incoming and outgoing line electric energy data and energy conservation relationship, estimating an initial attribution probability of a line state in combination with an index normalization mapping method, and obtaining a posterior attribution probability of each outgoing line line through Gaussian mixture distribution modeling under dynamic prior guidance; and based on the posterior affiliation probability of each outgoing line, constructing an objective function of joint observation value fitting quality, time continuity and misalignment amplitude balance, solving the objective function, and quantifying the electric quantity misalignment proportion of each outgoing line at each moment so as to judge classification and identification of misalignment behaviors.

Description

Technical Field

[0001] This invention relates to the field of power metering technology for low-voltage power distribution systems, and in particular to a method and system for identifying inaccurate metering devices in transformer substations based on the EM algorithm. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] With the continuous improvement of the intelligence level of the power distribution network, low-voltage distribution substations have widely deployed intelligent incoming line switches and power metering devices for each outgoing line, building a visualized and controllable power distribution management system based on collected data. Detailed analysis of 15-minute-level power data for each circuit in the substation has become a core foundation for various scenarios such as intelligent operation and maintenance, energy efficiency assessment, and risk warning.

[0004] However, due to factors such as equipment aging, installation deviations, electromagnetic interference, and human factors, metering devices inevitably experience inaccuracies during long-term operation, leading to abnormal line losses in the distribution area, accumulated data deviations, and distorted operation and maintenance decisions.

[0005] Therefore, existing technologies have at least the following technical problems: existing technologies mostly use line loss anomalies as a single judgment criterion, which cannot identify multiple inaccuracy states or perform fine classification, and lack full utilization of time series data, making it difficult to capture the dynamic characteristics of inaccuracy behavior; in addition, existing clustering or modeling methods are mostly based on static priors or uniform assumptions, which are prone to getting trapped in local optima, resulting in poor recognition accuracy and stability, making it difficult to meet the actual needs of high-precision, interpretable and deployable inaccuracy recognition in smart distribution networks. Summary of the Invention

[0006] To address the technical problems mentioned above, this invention provides a method and system for identifying metering device malfunctions in distribution networks based on the EM algorithm. In the absence of complete ledger information or in the presence of complex wiring environments, this method automatically identifies the malfunctioning state of metering devices using existing power data, thereby improving the perception accuracy and lean management level of the distribution network.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: The first aspect of the present invention provides a method for identifying inaccuracies in metering devices in a transformer substation based on the EM algorithm.

[0008] A method for identifying misalignment of metering devices in a transformer substation based on the EM algorithm, comprising: The actual power data of each outgoing line metering device in the low-voltage distribution substation is obtained. The theoretical error model of each outgoing line under different metering error states is established through the EM algorithm to clarify the mapping relationship between different error states and actual power data. Based on the power data of the substation's incoming and outgoing lines and the energy conservation relationship, the theoretical deviation of each outgoing line metering device under different error states is constructed. The initial assignment probability of the line status is estimated by combining the exponential normalization mapping method. Under the guidance of dynamic prior, the posterior assignment probability of each outgoing line is obtained by modeling through Gaussian mixture distribution. Based on the posterior attribution probability of each outgoing line, an objective function is constructed to assess the fitting quality, temporal continuity, and balance of inaccuracy magnitude of the joint observations. The objective function is solved to quantify the proportion of power inaccuracy of each outgoing line at each time point, in order to classify and identify inaccuracy behaviors.

[0009] Furthermore, the theoretical inaccuracy model is expressed by the following formula:

[0010] in, It is the theoretical misalignment charge value of the i-th outgoing line at sampling time t, assuming it is in the j-th misalignment state, and is used to estimate the fitting effect of the line on the bus loss under different misalignment states; It is the error amplitude coefficient corresponding to the j-th error state, that is, the proportion of the power measurement deviation of the outgoing line under the assumed j-th error state; It is the original metering power data of the i-th outgoing line at sampling time t.

[0011] Furthermore, the method for constructing the theoretical deviation of each outgoing line metering device under different misalignment states based on the power data of the incoming and outgoing lines of the transformer area and the energy conservation relationship includes: considering the actual observed line loss between the incoming and outgoing lines of the transformer area, weighting it according to the power ratio of each outgoing line to construct the theoretical line loss contribution value that each outgoing line should bear at the current moment, and calculating the difference with the theoretical misalignment power. The resulting deviation is the matching error of the outgoing line under different misalignment states at the current moment relative to the actual line loss conservation constraint of the transformer area.

[0012] Furthermore, under the guidance of dynamic priors, the posterior assignment probability of each outgoing line is obtained by modeling with a Gaussian mixture distribution. The method includes: based on the initial assignment probability of the line misalignment state, before entering the expectation step, normalizing and dynamically updating the overall proportion of each misalignment state at each time step; when calculating the prior probability of the misalignment state used in the expectation step, applying time smoothing processing to the average result of the initial assignment probability, transforming it into a dynamic prior parameter reflecting the proportion of each misalignment state at the current time driven by the observation data; and obtaining the posterior assignment probability of each outgoing line by modeling with a Gaussian mixture distribution based on the dynamic prior parameter.

[0013] Furthermore, the objective function is expressed by the following formula:

[0014] in, It is the optimization objective function, which represents the log-likelihood function with regularization terms; Represents the set of all parameters to be optimized, including the mean of all misaligned states. ,variance Prior probability of inaccurate state and the inaccuracy amplitude coefficient ; It is the probability that the i-th outgoing line belongs to the j-th inaccurate state at time t; It is a Gaussian density function; This is the total number of sampling times; and It is the weighted average of the electrical quantities at all sampling points of the i-th outgoing line in the j-th misalignment state; It is the time smoothing regularization coefficient; It is the amplitude regularization coefficient.

[0015] Furthermore, the method for quantifying the power inaccuracy ratio of each outgoing line at each time step includes: after the EM algorithm converges, weighted fusion of the inaccuracy behavior of each line at each time step based on the responsibility probability distribution and inaccuracy amplitude coefficient obtained in the final iteration, to obtain the power inaccuracy ratio of each outgoing line at each time step.

[0016] A second aspect of the present invention provides a misalignment identification system for metering devices in a transformer substation based on the EM algorithm.

[0017] A system for identifying misalignment of metering devices in a transformer substation based on the EM algorithm, comprising: The clustering attribution module is configured to: acquire the actual power data of each outgoing line metering device in the low-voltage distribution substation; establish a theoretical misalignment model for each outgoing line under different metering misalignment states using the EM algorithm; clarify the mapping relationship between different misalignment states and actual power data; construct the theoretical deviation of each outgoing line metering device under different misalignment states based on the power data of the substation's incoming and outgoing lines and the energy conservation relationship; estimate the initial assignment probability of the line status using the exponential normalization mapping method; and obtain the posterior assignment probability of each outgoing line using Gaussian mixture distribution modeling under dynamic prior guidance. The classification and identification module is configured to: construct an objective function for the fitting quality, temporal continuity and error magnitude balance of joint observations based on the posterior probability of each outgoing line; solve the objective function; quantify the power error ratio of each outgoing line at each time moment; and classify and identify the error behavior.

[0018] A third aspect of the present invention provides a computer device comprising: A processor, adapted to execute computer programs; A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in the method for identifying inaccurate metering devices in a transformer substation based on the EM algorithm as described in the first aspect above.

[0019] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program adapted to be loaded by a processor and to execute the steps of the method for identifying misalignment of metering devices based on the EM algorithm as described in the first aspect above.

[0020] The fifth aspect of the present invention provides a computer program product or computer program.

[0021] This invention provides a computer program product or computer program comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps in the EM algorithm-based method for identifying inaccurate metering devices in the first aspect described above.

[0022] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention sets multiple typical levels of inaccuracy and introduces them as latent state variables into the model, which can effectively identify mild, moderate and severe inaccuracy behaviors, has high state discrimination ability, and can classify and respond to different types of measurement deviations and anomalies. By constructing an initial state attribution probability driven by energy conservation deviation, the iterative convergence direction of the responsibility distribution is guided, which effectively avoids the model from getting trapped in local optima or oscillating convergence, and improves the stability and accuracy of the algorithm in large sample complex data scenarios.

[0023] 2. This invention applies a time-smoothing constraint to the mean parameter of the misaligned state to avoid drastic changes in the state parameter over time; at the same time, it applies a distribution balance control to the misalignment amplitude coefficient of each state to prevent the clustering results from deviating from the actual physical meaning, effectively improving the controllability and engineering usability of the model; the final output of the model is the accurate misalignment ratio and misalignment amount of each line at each moment, which has the characteristics of continuity, high precision and strong interpretability, and is suitable for various business scenarios such as long-term monitoring, anomaly detection and decision support.

[0024] 3. This invention analyzes the time series fluctuation characteristics of the inaccurate power of each outgoing line, and combines the inaccuracy rate level and change trend to determine the equipment accuracy deviation and non-technical loss, thereby realizing the automatic classification and labeling of equipment abnormalities and electricity theft, and improving operation and maintenance efficiency. Attached Figure Description

[0025] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0026] Figure 1 This is a flowchart illustrating the method for identifying misalignment of metering devices in a transformer substation based on the EM algorithm, as shown in an embodiment of the present invention. Figure 2 This is a structural diagram of a misalignment identification system for transformer substation metering devices based on the EM algorithm, as shown in an embodiment of the present invention. Figure 3 This is a structural diagram of a computer device shown in an embodiment of the present invention. Detailed Implementation

[0027] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0028] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0029] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0030] Figure 1 This is a flowchart illustrating the method for identifying misalignment of metering devices in a transformer substation based on the EM algorithm, as shown in an embodiment of the present invention; see reference. Figure 1 The method includes: S1. By establishing a multi-misalignment state power deviation model, based on the power data of incoming and outgoing lines of the transformer area and the energy conservation relationship, the theoretical deviation of each outgoing line metering device under different misalignment states is constructed. The initial assignment probability of the line state is estimated by combining the exponential normalization mapping method. Under the guidance of dynamic prior, the posterior assignment probability of each outgoing line is obtained by modeling with Gaussian mixture distribution, so as to achieve the preliminary clustering and attribution of the misalignment characteristics of the metering equipment.

[0031] This embodiment constructs a Gaussian mixture distribution-based model for identifying inaccurate states of metering devices in distribution areas. Using the electrical data from each outgoing metering device as the observed input variable, several typical inaccurate states are constructed as latent latent category variables. The model uses the state mean, state variance, state prior probability, and inaccuracy amplitude factor as parameter sets, and models and optimizes them using an expectation-maximization (EM) framework. The inaccurate state identification model uses liability probability to cluster and interpret the observed data, and suppresses time discontinuities and state skewness through a regularized objective function, ultimately achieving the inversion calculation of inaccurate electrical quantities for each line at each sampling time.

[0032] First, 15-minute granular electricity data is collected from the incoming smart switches deployed in the low-voltage distribution substation and from the outgoing metering devices (such as smart meters) connected to each user side or load unit. The continuous collection period is no less than 30 days to ensure the sample covers various operating conditions. The data at each moment includes the total electricity consumption of one incoming line. and the power consumption of N outgoing lines This forms a complete data matrix. The collected electricity consumption data is the timed upload result from the automatic meter reading system of the distribution area after edge processing. It has the characteristics of unified timestamp, consistent units, and standardized structure, and can be directly used for modeling and analysis. To ensure data validity, only time slices with complete records of both incoming and outgoing lines are selected for subsequent processing.

[0033] To simulate and quantify the theoretically expected energy deviation of each outgoing line under different misalignment states, and thus clarify and compare the fitting ability of each misalignment state to actual observation data, a theoretical misalignment model for each outgoing line under different metering misalignment states is established to clarify the mapping relationship between different misalignment states and actual energy data. Therefore, K metering device misalignment state levels are preset, with each misalignment state corresponding to a fixed misalignment amplitude coefficient, denoted as . , , representing the proportional deviation between the actual recorded electricity and the true electricity in the j-th misalignment state. At sampling time t, the theoretical misalignment electricity corresponding to the j-th misalignment state of the i-th outgoing line can be expressed as:

[0034] in, It is the theoretical misalignment charge value of the i-th outgoing line at sampling time t, assuming it is in the j-th misalignment state, and is used to estimate the fitting effect of the line on the bus loss under different misalignment states; It is the error amplitude coefficient corresponding to the j-th error state, that is, the proportion of the power measurement deviation of the outgoing line under the assumed j-th error state; This represents the original metered electrical data of the i-th outgoing line at sampling time t. The theoretical miscalculation formula establishes a functional mapping relationship between the possible miscalculation of each outgoing line under each assumed miscalculation state and the original observation data.

[0035] Considering the actual observed line losses between the incoming and outgoing lines of the transformer area, they should be weighted according to the power consumption ratio of each outgoing line to construct the theoretical line loss contribution value that each outgoing line should bear at the current moment. This value should then be compared with the theoretical misalignment power consumption to calculate the deviation. The resulting deviation represents the matching error of the outgoing line under different misalignment states at the current moment relative to the actual line loss conservation constraint of the transformer area. A deviation metric is constructed as follows:

[0036]

[0037] in, It is the line loss deviation metric of the i-th outgoing line at sampling time t, assuming it is in the j-th inaccuracy state. It represents the absolute difference between the fitted inaccuracy and the actual line loss allocation value. The smaller the deviation, the more the current inaccuracy state matches the actual line loss contribution. This is the theoretical misalignment quantity; The actual observed line loss represents the difference between the total incoming line power and the outgoing line power.

[0038] To normalize the assignment probability among multiple misalignment states, an exponential mapping based on the deviation value is adopted. The initial assignment probability value of each outgoing line under each misalignment state is generated through Softmax, and the specific definition is as follows:

[0039] in, It represents the initial assignment probability of the i-th outgoing line belonging to the j-th inaccurate state at sampling time t, and its value ranges between (0,1). The assignment sensitivity coefficient is used to adjust the response strength of the state assignment probability to error deviation. This enables the deduction of the initial inaccurate state probability of each line by utilizing the power distribution and conservation relationship, without prior model parameters, thus providing dynamic guidance for the state prior of the subsequent EM algorithm.

[0040] Based on the initial assignment probability of line misalignment states, before entering the expectation step, it is necessary to perform normalization correction and dynamic update on the overall proportion of each misalignment state at each time step. Therefore, when calculating the prior probability of misalignment states used in the expectation step, a static uniform distribution or a fixed prior based solely on historical experience is not directly adopted. Instead, a time smoothing process is applied to the average result of the initial assignment probability, transforming it into a dynamic prior parameter reflecting the proportion of each misalignment state at the current time step driven by the observation data. Its mathematical expression is as follows:

[0041] in, This represents the prior probability value of the j-th misalignment state at time t, which is used in the EM algorithm to estimate the assignment weight of the line in the j-th misalignment state. These are the prior update coefficients, used to balance the fusion of historical inference and current observations. Then, the posterior probability of each outgoing line belonging to each inaccurate state is calculated using a standard Gaussian mixture model:

[0042] in, It is the responsibility probability (posterior attribution probability) of the i-th outgoing line belonging to the j-th inaccurate state at time t. The Gaussian density function is:

[0043] in, and These are the weighted average and weighted variance of the electricity consumption at all sampling points of the i-th outgoing line in the j-th inaccurate state, respectively. With the support of dynamic prior, the mean and variance of the current inaccurate state data are used to accurately update the attribution probability, providing an expectation estimate for the maximization step.

[0044] S2. Based on the known posterior attribution probability, a multi-objective optimization function with regularization term is constructed to jointly model the fitting quality of observations, temporal continuity, and the balance of inaccuracy magnitude. The model parameters are estimated by maximizing the weighted log-likelihood, and the inaccuracy magnitude and state probability are integrated to quantify the power inaccuracy ratio and total deviation value of each outgoing line at each time. Finally, the classification, identification, and source analysis of inaccuracy behavior are realized.

[0045] To further improve the adaptability and stability of the substation misalignment identification model, an interpretable multi-objective optimization function is constructed by maximizing the overall weighted log-likelihood function and jointly introducing multiple regularization constraints. This function includes a log-likelihood term for the observed data, a continuity constraint term for the mean of misaligned states in the time series, and a control term for the balance of the misalignment amplitude distribution for each state. The first term is a standard Gaussian weighted log-likelihood function, using the posterior responsibility probability of each outgoing line belonging to each misalignment state at each time step as the weight, to model the degree of matching between the observed values ​​and the distribution fitted by the current model. The second term constrains the sum of squares of the changes in the mean parameter of each misalignment state between two consecutive time steps, maintaining temporal consistency and trend stability, and avoiding drastic parameter jumps caused by model overfitting and local noise. The third regularization term targets the global distribution of the misalignment amplitude coefficients, constraining the deviation between its mean and individual values, suppressing excessive deviation of cluster centers from the overall distribution, preventing abnormal states from absorbing a large number of non-abnormal points, and ensuring the balance of the structure of each cluster. The optimization objective function is expressed as follows:

[0046] in, It is the optimization objective function, which represents the log-likelihood function with regularization terms; Represents the set of all parameters to be optimized, including the mean of all misaligned states. ,variance Prior probability of inaccurate state and the inaccuracy amplitude coefficient ; This is the total number of sampling times; It is the time smoothing regularization coefficient, which controls the smoothness of the mean of the out-of-range state on the time axis and prevents the model from overfitting to noise; It is the amplitude regularization coefficient, which suppresses excessive divergence between state inaccuracy amplitude coefficients and improves the stability of the cluster structure.

[0047] Optimize objective function The goal of solving this problem is to determine the current probability of liability. Given the given information, optimize the set of model parameters. This allows the misalignment identification model to fit the actual observation data to the maximum extent, while controlling the smoothness of parameters and the rationality of the misalignment amplitude distribution.

[0048] A block-based optimization strategy is adopted, that is, in the M-step of the EM algorithm framework, updates are performed separately according to parameter type, specifically as follows: The closed-form solution (standard EM update formula) can be directly obtained by weighted averaging of the log-likelihood terms, and updated once in each iteration; the inaccuracy magnitude coefficient By extracting the line loss reconstruction error term from the objective function, the following least squares problem is constructed:

[0049] The above problem is a convex optimization problem. After taking the partial derivatives, construct the normal equation (in matrix form), and the closed-form solution can be obtained directly.

[0050] To further achieve a precise quantitative representation of the inaccuracy status of each outgoing metering device in the transformer substation at any given time, after the EM algorithm converges, the inaccuracy behavior of each line at each time point needs to be weighted and fused based on the responsibility probability distribution and inaccuracy amplitude coefficient obtained from the final iteration. The calculation formula is as follows:

[0051]

[0052] in, This represents the percentage of inaccuracy in the predicted unit power volume for the i-th outgoing line at time t, and its value ranges from all... The decision reflects the uncertainty information of the fusion of multiple inaccurate states; It is the estimated inaccurate power. Calculate the average misalignment rate for the time series of misaligned electrical quantities for each line. And variance, set fixed or dynamic thresholds (e.g.) (Or abnormal fluctuation) as identification criteria; lines that meet the threshold conditions are marked as metering device inaccuracy; if the inaccuracy rate is large and the fluctuation is small, it is judged as equipment deviation inaccuracy; if the inaccuracy rate fluctuates drastically, it is judged as suspected electricity theft.

[0053] In summary, a method for identifying inaccuracies in metering devices in transformer substations based on the EM algorithm has been developed.

[0054] The above combination Figure 1 The method for identifying misalignment of metering devices in substations based on the EM algorithm provided in this embodiment of the invention has been described in detail. Next, the system for identifying misalignment of metering devices in substations based on the EM algorithm provided in this embodiment of the invention will be described in conjunction with the accompanying drawings.

[0055] Figure 2 This is a schematic diagram of the structure of a misalignment identification system for transformer substation metering devices based on the EM algorithm, as shown in an embodiment of the present invention. Figure 2 The system described in this invention includes: The clustering attribution module is configured to: establish a multi-misalignment state power deviation model, construct the theoretical deviation of each outgoing line metering device under different misalignment states based on the power data of the incoming and outgoing lines of the transformer area and the energy conservation relationship, estimate the initial assignment probability of the line state by combining the exponential normalization mapping method, and obtain the posterior assignment probability of each outgoing line by modeling through Gaussian mixture distribution under the guidance of dynamic prior, so as to achieve the preliminary clustering attribution of the misalignment characteristics of the metering equipment; The classification and identification module is configured to: construct a multi-objective optimization function with regularization terms based on the known posterior attribution probability; jointly model the fitting quality of observations, temporal continuity, and the balance of inaccuracy magnitude; complete the estimation of model parameters by maximizing the weighted log-likelihood; and integrate the inaccuracy magnitude and state probability to quantify the power inaccuracy ratio and total deviation value of each outgoing line at each time point, ultimately realizing the classification, identification, and source analysis of inaccurate behavior.

[0056] In some embodiments, in order to simulate and quantify the theoretically expected power deviation of each outgoing line under different misalignment states, thereby clarifying and comparing the fitting ability of each misalignment state to the actual observed data, a theoretical misalignment model of each outgoing line under different metering misalignment states is established to clarify the mapping relationship between different misalignment states and actual power data.

[0057] In some embodiments, considering the actual observed line loss between the incoming and outgoing lines of the transformer area, it should be weighted according to the power ratio of each outgoing line to construct the theoretical line loss contribution value that each outgoing line should bear at the current moment, and the difference between it and the theoretical misalignment power should be calculated. The resulting deviation is the matching error of the outgoing line under different misalignment states at the current moment relative to the actual line loss conservation constraint of the transformer area.

[0058] In some embodiments, based on the initial assignment probability of the line misalignment state, before entering the expectation step, it is necessary to perform normalization correction and dynamic update on the overall proportion of each misalignment state at each time. To this end, when calculating the prior probability of the misalignment state used in the expectation step, a static uniform distribution or a fixed prior based solely on historical experience is not directly adopted. Instead, the average result of the initial assignment probability is subjected to time smoothing processing, which transforms it into a dynamic prior parameter that reflects the proportion of each misalignment state at the current time under the driving force of observation data.

[0059] In some embodiments, to further improve the adaptability and stability of the substation misalignment state identification model, an interpretable multi-objective optimization function is constructed by maximizing the overall weighted log-likelihood function and jointly introducing multiple regularization constraints. This function includes the log-likelihood term of the observed data, the continuity constraint term of the misalignment state mean in the time series, and the control term for the balance of the misalignment amplitude distribution of each state.

[0060] In some embodiments, in order to further realize the accurate quantitative expression of the inaccuracy status of each outgoing metering device in the transformer area at any time, after the EM algorithm converges, the inaccuracy behavior of each line at each time needs to be weighted and fused based on the responsibility probability distribution and inaccuracy amplitude coefficient obtained by the final iteration.

[0061] According to embodiments of the present invention, the substation metering device misalignment identification system based on the EM algorithm can correspond to the execution of the method described in the embodiments of the present invention, and the above and other operations and / or functions of each module of the substation metering device misalignment identification system based on the EM algorithm are respectively for implementing Figure 1 For the sake of brevity, the corresponding processes of each method in the code will not be elaborated here.

[0062] See Figure 3 The diagram shows the structure of a computer device, which includes a processor, a communication interface, and a computer-readable storage medium. The processor, communication interface, and computer-readable storage medium are connected via a bus or other means. The communication interface is used to receive and send data. The computer-readable storage medium can be stored in the computer device's memory. The computer-readable storage medium stores computer programs, including program instructions, and the processor executes the program instructions stored in the computer-readable storage medium. The processor (or CPU, Central Processing Unit) is the computing and control core of the computer device, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to implement the corresponding steps in the embodiment of the method for identifying inaccurate metering devices in transformer substations based on the EM algorithm.

[0063] This embodiment provides a computer-readable storage medium (Memory), which is a memory device in a computer device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and extended storage media supported by the computer device. The computer-readable storage medium provides storage space that stores the processing system of the computer device. Furthermore, this storage space also contains one or more instructions suitable for loading and execution by the processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be high-speed RAM memory or non-volatile memory, such as at least one disk storage device; optionally, it can also be at least one computer-readable storage medium located remotely from the aforementioned processor.

[0064] In one embodiment, the computer-readable storage medium stores one or more instructions; the processor loads and executes one or more instructions stored in the computer-readable storage medium to implement the corresponding steps in the above embodiment of the method for identifying inaccurate metering devices in the substation area based on the EM algorithm.

[0065] This embodiment provides a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. The processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the corresponding steps in the above embodiment of the method for identifying inaccurate metering devices based on the EM algorithm.

[0066] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0067] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0068] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0069] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0070] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0071] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for identifying misalignment of a power distribution area metering device based on an EM algorithm, characterized in that, The method comprises the following steps: The actual power data of each outgoing line metering device in a low-voltage distribution area is acquired, and a theoretical error quantity model of each outgoing line in different metering error states is established through an EM algorithm, so as to determine the mapping relationship between different error states and the actual power data; based on the energy conservation relationship between the incoming and outgoing line energy data of the distribution area, a theoretical deviation quantity of each outgoing line metering device in different error states is constructed, the initial attribution probability of the line state is estimated by combining an exponential normalization mapping method, and under the guidance of dynamic priori, the posteriori attribution probability of each outgoing line is acquired through Gaussian mixture distribution modeling; Based on the posteriori attribution probability of each outgoing line, a target function is constructed by combining the fitting quality of the joint observation value, the time continuity and the error amplitude balance, the target function is solved, and the power error proportion of each outgoing line at each time is quantified to judge the classification identification of the error behavior.

2. The EM algorithm-based misalignment identification method for transformer area metering device according to claim 1, characterized in that, The theoretical error quantity model is expressed by the following formula: wherein, is the theoretical misalignment electric quantity value of the ith outgoing line at the sampling time t assuming the jth misalignment state, used to estimate the fitting effect of the line on the bus loss under different misalignment states; is the misalignment amplitude coefficient corresponding to the jth misalignment state, that is, assuming that the electric quantity measurement deviation proportion of the outgoing line in the jth misalignment state; is the original metering electric quantity data of the ith outgoing line at the sampling time t.

3. The EM algorithm-based misalignment identification method for transformer area metering device according to claim 1, characterized in that, The theoretical deviation quantity of each outgoing line metering device in different error states is constructed based on the energy conservation relationship between the incoming and outgoing line energy data of the distribution area; the method comprises the following steps: considering the actual observed line loss quantity between the incoming and outgoing lines of the distribution area, the line loss quantity is weighted and distributed according to the power proportion of each outgoing line, so as to construct the theoretical line loss contribution value of each outgoing line at the current time, and the difference between the theoretical error quantity is calculated, and the deviation is the matching error of each outgoing line in different error states at the current time relative to the real line loss conservation constraint of the distribution area.

4. The EM algorithm-based misalignment identification method for transformer area metering device according to claim 1, characterized in that, Under the guidance of dynamic priori, the posteriori attribution probability of each outgoing line is acquired through Gaussian mixture distribution modeling; the method comprises the following steps: based on the initial attribution probability of the line error state, the overall proportion of each error state at each time is normalized and dynamically updated before entering the expectation step; when the error state priori probability used in the calculation expectation step is formally used, the time smoothing processing is applied to the average result of the initial attribution probability, so as to convert it into a dynamic priori parameter reflecting the proportion of each error state at the current time under the observation data driving; according to the dynamic priori parameter, the posteriori attribution probability of each outgoing line is acquired through Gaussian mixture distribution modeling.

5. The EM algorithm-based misalignment identification method for transformer area metering device according to claim 1, characterized in that, The target function is expressed by the following formula: wherein, is the optimization objective function, representing the log-likelihood function with regularization terms; represents the set of all parameters to be optimized, including all miss state mean , variance , miss state prior probability and miss amplitude coefficient ; is the responsibility probability of the i-th outgoing line at time t belonging to the j-th miss state; is the Gaussian density function; is the total number of sampling times; and is the weighted average of all sampling point electric quantities of the i-th outgoing line under the j-th miss state; is the time smoothing regularization coefficient; is the amplitude regularization coefficient.

6. The EM algorithm-based misalignment identification method for transformer area metering device according to claim 1, characterized in that, The power error proportion of each outgoing line at each time is quantified; the method comprises the following steps: after the EM algorithm converges, the error behavior of each line at each time is weighted and fused according to the responsibility probability distribution and the error amplitude coefficient obtained by the final iteration, so as to obtain the power error proportion of each outgoing line at each time.

7. A system for identifying misalignment of a power distribution area metering device based on an EM algorithm, characterized by, The method comprises the following steps: The clustering attribution module is configured to: acquire the actual power data of each outgoing line metering device in a low-voltage distribution area, establish a theoretical error quantity model of each outgoing line in different metering error states through an EM algorithm, and determine the mapping relationship between different error states and the actual power data; based on the energy conservation relationship between the incoming and outgoing line energy data of the distribution area, a theoretical deviation quantity of each outgoing line metering device in different error states is constructed, the initial attribution probability of the line state is estimated by combining an exponential normalization mapping method, and under the guidance of dynamic priori, the posteriori attribution probability of each outgoing line is acquired through Gaussian mixture distribution modeling; The classification identification module is configured to: based on the posteriori attribution probability of each outgoing line, construct a target function of joint observation value fitting quality, time continuity and misalignment amplitude balance, solve the target function, quantify the power misalignment proportion of each outgoing line at each time, and judge the classification identification of the misalignment behavior. 8.A computer device, comprising: a processor adapted to execute a computer program; a computer readable storage medium having stored therein a computer program, the computer program, when executed by the processor, implements the steps of the EM algorithm based misalignment identification method of the transformer area metering device according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, the computer program is suitable for being loaded and executed by the processor to implement the steps of the EM algorithm based misalignment identification method of the transformer area metering device according to any one of claims 1-6.

10. A computer program product, characterised in that, The computer program product comprises a computer program, the computer program, when executed by the processor, implements the steps of the EM algorithm based misalignment identification method of the transformer area metering device according to any one of claims 1-6.

Images