Electricity utilization information acquisition system and method for identifying faults of electric energy meter

Abstract

The invention relates to the technical field of electric energy meter fault identification, and discloses an electricity utilization information acquisition system and method for identifying electric energy meter faults, and the system comprises a time-frequency fingerprint extraction module which extracts time-frequency fingerprints through three-layer discrete wavelet transform; the periodic feature construction module is used for constructing a periodic feature vector; the periodic feature analysis module obtains whether the electric energy meter has a fault or not through a first fault detection model; the instantaneous feature construction module is used for constructing an instantaneous feature vector; the electrical diagram construction module is used for constructing a sparse electrical diagram; and the electrical diagram analysis module obtains the fault type of the electric energy meter through a second fault detection model. According to the method, the electric energy meters with the suspected faults are screened out according to the periodic characteristics, then the sparse electrical diagrams are constructed according to the physical topology connection between the electric energy meters and the instantaneous characteristics, and correlation analysis is performed on the sparse electrical diagrams, so that the fault types of the electric energy meters are identified, and the overall detection efficiency and precision are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electric energy meter fault identification, and more particularly to an electricity consumption information collection system and method for identifying electric energy meter faults. Background Art

[0002] Electricity meters are instruments used to measure and record electrical energy usage. They are widely used for electricity metering in residential, commercial, and industrial settings. In power systems, they are not only an important data source for recording user electricity usage but also crucial for bill settlement and fault identification. Traditionally, fault identification of electricity meters relies primarily on regular manual inspections and proactive user reporting. This approach is not only inefficient but also suffers from issues such as delayed identification, incomplete coverage, and human error, making it difficult to meet the high real-time and accuracy requirements of modern power systems.

[0003] With the development of big data and artificial intelligence technologies, existing methods collect electricity meter usage data to form time series and use time series analysis models (such as LSTM and ARIMA) to model the meter's operating status and detect anomalies. While this approach improves the automation of fault identification, it generally ignores the potential electrical connections between meters, resulting in suboptimal fault identification accuracy. For example, if multiple households' electricity meters are connected to the same line, and an abnormal power consumption event at one household (such as a sudden power outage of a high-power device or voltage fluctuation) affects the power parameters of other adjacent meters, it can easily misidentify a normal meter as abnormal if time series analysis is performed solely on a single meter. Summary of the Invention

[0004] The present invention provides an electricity consumption information collection system and method for identifying electric energy meter faults, which solve the technical problems in the above-mentioned background technology.

[0005] The present invention provides an electricity consumption information collection system for identifying electric energy meter faults, comprising:

[0006] The time-frequency fingerprint extraction module is used to collect the active power of each electric energy meter to construct an active power sequence, and extract the time-frequency fingerprint through a three-layer discrete wavelet transform;

[0007] The time-frequency fingerprint consists of low-frequency energy values, mid-frequency energy values, and high-frequency energy values;

[0008] A periodic feature construction module is used to calculate the day-night power ratio of each electric energy meter and form a periodic feature vector with the time-frequency fingerprint;

[0009] a periodic feature analysis module, configured to analyze the periodic feature vector of each electric energy meter using the first fault detection model, output a value indicating whether the electric energy meter has a fault, and mark the electric energy meter with a fault;

[0010] The instantaneous feature construction module is used to calculate the power factor, three-phase voltage imbalance and three-phase current mean of each electric energy meter at the current moment, and form an instantaneous feature vector with the current power consumption and active power;

[0011] An electrical diagram construction module, which is used to construct a sparse electrical diagram based on the instantaneous feature vector of each electric energy meter and the physical topological connections between the electric energy meters;

[0012] The electrical diagram analysis module is used to analyze the sparse electrical diagram through the second fault detection model, and the output value represents the fault type of the electric energy meter with the fault mark.

[0013] Furthermore, the length of the active power sequence is equal to the acquisition time period divided by the acquisition time interval, where the acquisition time period and the acquisition time interval are both custom parameters; the active power sequence is subjected to wavelet transform through three-layer discrete wavelet transform to obtain three-layer detail coefficients, and the sum of the squares of the first-layer detail coefficients, the second-layer detail coefficients and the third-layer detail coefficients are calculated respectively as the low-frequency energy value, the medium-frequency energy value and the high-frequency energy value.

[0014] Furthermore, the day-night power ratio is equal to the ratio of the average power of the daytime period to the average power of the nighttime period, wherein the daytime period and the nighttime period are both custom parameters.

[0015] Furthermore, the first fault detection model is composed of a nonlinear transformation layer and a first classifier;

[0016] The nonlinear transformation layer is used to transform the periodic feature vector into a first update vector, wherein the number of dimensions of the first update vector is a custom parameter;

[0017] The calculation formula of the nonlinear transformation layer is as follows:

[0018] Where 1≤i≤L, Represents the first update vector output by the nonlinear transformation layer, L represents the maximum number of layers of the nonlinear transformation layer, which is a custom parameter. and Represent the first intermediate vectors of the output of the i-1th layer and the i-th layer respectively, represents the periodic eigenvector of the input nonlinear transformation layer, and They represent the first weight parameter and the first bias parameter of the i-th layer of the nonlinear transformation layer, GELU represents the GELU activation function, and LaryerNorm represents the inter-layer normalization function;

[0019] The first classifier inputs the first update vector, and the output value indicates whether the electric energy meter has a fault;

[0020] The first classifier is built based on a multi-layer perceptron, and the corresponding activation function is the Sigmoid activation function.

[0021] Furthermore, during the training process of the first fault detection model, a reconstruction mapping layer is added after the nonlinear transformation layer;

[0022] The reconstruction mapping layer is used to convert the first update vector into a second update vector, wherein the number of dimensions of the second update vector is the same as the number of dimensions of the periodic feature vector;

[0023] The calculation formula for reconstructing the mapping layer is as follows:

[0024] Where 1≤j≤J, represents the second update vector of the reconstruction mapping layer output, J represents the maximum number of layers of the reconstruction mapping layer, J = L, and Represent the second intermediate vectors output by the j-1th layer and the jth layer respectively, represents the first update vector of the input reconstruction mapping layer, and denote the second weight parameter and the second bias parameter of the jth layer of the reconstruction mapping layer, respectively. and They represent the third weight parameter and the third bias parameter of the jth layer of the reconstruction mapping layer, σ represents the Sigmoid activation function, and ⊙ represents the Hadamard product;

[0025] The calculation formula of the loss function Loss of the first fault detection model is as follows:

[0026] Loss = λ rec ×Loss rec +λ clf ×Loss clf , where Loss rec Represents the reconstruction loss, Loss clf represents the classification loss, λ rec represents the reconstruction loss coefficient, λ clf Represents the classification loss coefficient, the sum of the reconstruction loss coefficient and the classification loss coefficient is 1;

[0027] Where 1≤k≤K, K represents the number of training samples, Represents the first update vector output by the nonlinear transformation layer after inputting the sample data of the kth training sample, represents the second update vector of the reconstructed mapping layer output after inputting the sample data of the kth training sample, and ||·||2 represents the L2 norm;

[0028] Where yk represents the sample label of the kth training sample, Represents the value of the output of the first fault detection model after inputting the sample data of the kth training sample.

[0029] Furthermore, the power factor is equal to the phase difference between the line voltage and the line current;

[0030] Three-phase voltage unbalance U unb The calculation formula is as follows:

[0031] Among them U a 、U b and U c Represents phase A voltage, phase B voltage and phase C voltage respectively, U avg Indicates the three-phase average voltage, max means taking the maximum value;

[0032] The calculation formula for the three-phase current mean is as follows: Among them I a , I b and I c Represent the A phase current, B phase current and C phase current respectively.

[0033] Furthermore, the sparse electrical graph consists of nodes and edges between nodes, where nodes correspond one-to-one to electricity meters. If there is a physical topological connection between the electricity meters, edges are constructed between the corresponding nodes. If the cosine similarity of the instantaneous feature vectors between the electricity meters is greater than or equal to a preset similarity threshold, edges are constructed between the corresponding nodes, where the preset similarity threshold is a custom parameter.

[0034] Furthermore, the second fault detection model consists of a graph data analysis layer and a second classifier;

[0035] The graph data analysis layer is used to update the instantaneous feature vectors of all nodes in the sparse electrical graph and output a third update vector, wherein the number of dimensions of the third update vector is a custom parameter;

[0036] The second classifier inputs the third update vector of the node corresponding to the electric energy meter with the fault mark, and outputs a value indicating the fault type of the electric energy meter with the fault mark;

[0037] The second classifier is built based on a multi-layer perceptron, and the corresponding activation function is the Softmax activation function.

[0038] Furthermore, the calculation formula of the graph data analysis layer includes:

[0039] Where 1≤u≤U, U represents the number of nodes in the sparse electrical graph, represents the third update vector of the u-th node, hu and h v Represent the instantaneous feature vectors of the u-th node and the v-th node respectively, N u represents the set of nodes that have edges with the u-th node, W1 and W2 represent the first weight parameter and the second weight parameter respectively, b1 and b2 represent the first bias parameter and the second bias parameter respectively;

[0040] represents the correlation coefficient between the u-th node and the v-th node, W3 represents the third weight parameter, || represents the concatenation operation, T represents the transposition operation, and σ represents the Sigmoid activation function.

[0041] The present invention provides a method for collecting electricity consumption information for identifying an electric energy meter fault, comprising the following steps:

[0042] Step S201: collect the active power of each electric energy meter to construct an active power sequence, and extract the time-frequency fingerprint through a three-layer discrete wavelet transform;

[0043] Step S202, calculating the daytime and nighttime power ratio of each electric energy meter and forming a periodic feature vector with the time-frequency fingerprint;

[0044] Step S203: Analyze the periodic characteristic vector of each electric energy meter using the first fault detection model, output a value indicating whether the electric energy meter has a fault, and mark the electric energy meter with a fault;

[0045] Step S204: Calculate the power factor, three-phase voltage imbalance, and three-phase current mean of each electric energy meter at the current moment, and form an instantaneous feature vector with the current power consumption and active power;

[0046] Step S205 , constructing a sparse electrical graph based on the instantaneous feature vector of each electric energy meter and the physical topological connections between the electric energy meters;

[0047] Step S206 : Analyze the sparse electrical diagram using the second fault detection model, and output a value indicating the fault type of the electric energy meter with the fault mark.

[0048] The beneficial effects of the present invention are as follows: the present invention first screens out suspected faulty electricity meters based on periodic characteristics through a two-stage analysis, then constructs a sparse electrical diagram based on the physical topological connections and transient characteristics between the electricity meters, and performs correlation analysis on the sparse electrical diagram to identify the fault type of the electricity meter, avoiding complex fault type analysis for all electricity meters. While reducing computing resource consumption, it can also reduce the false alarm rate, thereby improving the efficiency and accuracy of the overall detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1It is a schematic diagram of an electricity consumption information collection system for identifying electric energy meter faults according to the present invention;

[0050] Figure 2 The present invention is a flow chart of a method for collecting electricity consumption information for identifying electric energy meter faults.

[0051] In the figure: time-frequency fingerprint extraction module 101, periodic feature construction module 102, periodic feature analysis module 103, instantaneous feature construction module 104, electrical diagram construction module 105, electrical diagram analysis module 106. DETAILED DESCRIPTION

[0052] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed solely to enable those skilled in the art to better understand and implement the subject matter described herein, and that the functions and arrangements of the elements discussed may be varied without departing from the scope of this specification. Various examples may omit, substitute, or add various processes or components as needed. In addition, features described with respect to some examples may also be combined in other examples.

[0053] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in one or more embodiments of the present invention should have the usual meanings understood by people with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in one or more embodiments of the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprising" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0054] like Figures 1 and 2 As shown, a power consumption information collection system for identifying power meter faults includes:

[0055] The time-frequency fingerprint extraction module 101 is used to collect the active power of each electric energy meter to construct an active power sequence, and extract the time-frequency fingerprint through a three-layer discrete wavelet transform;

[0056] The time-frequency fingerprint consists of low-frequency energy values, mid-frequency energy values, and high-frequency energy values;

[0057] A periodic feature construction module 102 is used to calculate the day-night power ratio of each electric energy meter and form a periodic feature vector with the time-frequency fingerprint;

[0058] The periodic feature analysis module 103 is used to analyze the periodic feature vector of each electric energy meter using the first fault detection model, output a value indicating whether the electric energy meter has a fault, and mark the electric energy meter with a fault;

[0059] The instantaneous feature construction module 104 is used to calculate the power factor, three-phase voltage imbalance and three-phase current mean of each electric energy meter at the current moment, and form an instantaneous feature vector with the current power consumption and active power;

[0060] An electrical diagram construction module 105 is configured to construct a sparse electrical diagram based on the instantaneous feature vector of each electric energy meter and the physical topological connections between the electric energy meters;

[0061] The electrical diagram analysis module 106 is configured to analyze the sparse electrical diagram using a second fault detection model, and output a value indicating a fault type of the electric energy meter with a fault mark.

[0062] It should be noted that the two-stage analysis of the present invention first screens out electricity meters with suspected faults, avoiding complex fault type analysis of all electricity meters. While reducing computing resource consumption, it can also reduce the false alarm rate, thereby improving the overall detection efficiency and accuracy.

[0063] In one embodiment of the present invention, the length of the active power sequence is equal to the collection time period divided by the collection time interval, where the collection time period and the collection time interval are both custom parameters. Preferably, the collection time period is set to 7 days and the collection time interval is set to 15 minutes, then the length of the active power sequence is equal to 672; the active power sequence is subjected to wavelet transform through three-layer discrete wavelet transform to obtain three-layer detail coefficients, and the sum of the squares of the first-layer detail coefficients, the second-layer detail coefficients and the third-layer detail coefficients are calculated respectively as the low-frequency energy value, the medium-frequency energy value and the high-frequency energy value.

[0064] In one embodiment of the present invention, the day-night power ratio is equal to the ratio of the average power of the daytime period to the average power of the nighttime period, wherein the daytime period and the nighttime period are both custom parameters. Preferably, the daytime period is set to 7 a.m. to 10 p.m., and the nighttime period is set to 10 p.m. to 7 a.m.

[0065] In one embodiment of the present invention, the first fault detection model is composed of a nonlinear transformation layer and a first classifier;

[0066] The nonlinear transformation layer is used to transform the periodic feature vector into a first update vector, wherein the number of dimensions of the first update vector is a custom parameter. Preferably, the number of dimensions of the first update vector is set to 16;

[0067] The calculation formula of the nonlinear transformation layer is as follows:

[0068] Where 1≤i≤L, represents the first update vector output by the nonlinear transformation layer, L represents the maximum number of layers of the nonlinear transformation layer, which is a custom parameter. Preferably, L is set to 3. and Represent the first intermediate vectors of the output of the i-1th layer and the i-th layer respectively, represents the periodic eigenvector of the input nonlinear transformation layer, and They represent the first weight parameter and the first bias parameter of the i-th layer of the nonlinear transformation layer, GELU represents the GELU activation function, and LaryerNorm represents the inter-layer normalization function;

[0069] The first classifier inputs the first update vector, and the output value indicates whether the electric energy meter has a fault;

[0070] The first classifier is built based on a multi-layer perceptron, and the corresponding activation function is the Sigmoid activation function.

[0071] It should be noted that, based on the above content, the first weight parameters of the three layers in the nonlinear transformation layer can be designed as matrices of 4×8, 8×12, and 12×16 sizes, respectively. The dimensions of the first intermediate vectors of the three layers are 8, 12, and 16, respectively. The first bias parameters of the three layers can be designed as vectors of 1×8, 1×12, and 1×16 sizes, respectively. In addition, residual connections can be introduced to prevent gradient disappearance, thereby improving the stability of subsequent model training.

[0072] In one embodiment of the present invention, during the training process of the first fault detection model, a reconstruction mapping layer is added after the nonlinear transformation layer;

[0073] The reconstruction mapping layer is used to convert the first update vector into a second update vector, wherein the number of dimensions of the second update vector is the same as the number of dimensions of the periodic feature vector;

[0074] The calculation formula for reconstructing the mapping layer is as follows:

[0075] Where 1≤j≤J, represents the second update vector of the reconstruction mapping layer output, J represents the maximum number of layers of the reconstruction mapping layer, J = L, and Represent the second intermediate vectors output by the j-1th layer and the jth layer respectively, represents the first update vector of the input reconstruction mapping layer, and denote the second weight parameter and the second bias parameter of the jth layer of the reconstruction mapping layer, respectively. and They represent the third weight parameter and the third bias parameter of the jth layer of the reconstruction mapping layer, σ represents the Sigmoid activation function, and ⊙ represents the Hadamard product;

[0076] The calculation formula of the loss function Loss of the first fault detection model is as follows:

[0077] Loss = λ rec ×Loss rec +λ clf ×Loss clf , where Loss rec Represents the reconstruction loss, Loss clf represents the classification loss, λ rec represents the reconstruction loss coefficient, λ clf Represents the classification loss coefficient. The sum of the reconstruction loss coefficient and the classification loss coefficient is 1. Preferably, the reconstruction loss coefficient is set to 0.3 and the classification loss coefficient is set to 0.7.

[0078] Where 1≤k≤K, K represents the number of training samples, Represents the first update vector output by the nonlinear transformation layer after inputting the sample data of the kth training sample, represents the second update vector of the reconstructed mapping layer output after inputting the sample data of the kth training sample, and ||·||2 represents the L2 norm;

[0079] where y k represents the sample label of the kth training sample, Represents the value of the output of the first fault detection model after inputting the sample data of the kth training sample.

[0080] It should be noted that, according to the above content, the second weight parameters and the third weight parameters of the three layers in the reconstruction mapping layer can be designed as matrices of 16×12, 12×8 and 8×4 sizes, respectively. The dimensions of the second intermediate vectors of the three layers are 12, 8 and 4, respectively. The second bias parameters and the third bias parameters of the three layers can be designed as vectors of 1×12, 1×8 and 1×4 sizes, respectively. In addition, the Sigmoid activation function of the first classifier is used for binary classification, while the Sigmoid activation function of the reconstruction mapping layer is used for information screening to prevent noise propagation, enhance the nonlinear expression ability of the model, and thus improve the robustness of the model.

[0081] It should be noted that a training sample consists of sample data and sample labels. The sample labels of the training samples used to train the first fault detection model are obtained through manual labeling, that is, the sample data is a periodic feature vector, and the sample label is whether the electricity meter has a fault. In addition, the periodic feature vector needs to be normalized before being input into the first fault detection model to eliminate dimensional differences.

[0082] In one embodiment of the present invention, the power factor is equal to the phase difference between the line voltage and the line current;

[0083] Three-phase voltage unbalance U unb The calculation formula is as follows:

[0084] Among them U a 、U b and U c Represents phase A voltage, phase B voltage and phase C voltage respectively, U avg Indicates the three-phase average voltage, max means taking the maximum value;

[0085] The calculation formula for the three-phase current mean is as follows: Among them I a , I b and I c Represent the A phase current, B phase current and C phase current respectively.

[0086] It should be noted that line voltage refers to the voltage between two live wires, and line current refers to the current flowing through the load. The above power factor calculation refers to its application in a symmetrical three-phase system. In an asymmetrical three-phase system, the power factor is equal to the ratio of the sum of the three-phase active power to the sum of the three-phase apparent power. The power factor can be directly obtained through a smart energy meter, that is, by directly calculating the phase difference output power factor through high-frequency sampling of voltage and current waveforms. This will not be elaborated here.

[0087] In one embodiment of the present invention, a sparse electrical graph is composed of nodes and edges between nodes, wherein the nodes correspond one-to-one to the electricity meters, and edges are constructed between the corresponding nodes if there is a physical topological connection between the electricity meters. If the cosine similarity of the instantaneous feature vectors between the electricity meters is greater than or equal to a preset similarity threshold, edges are constructed between the corresponding nodes, wherein the preset similarity threshold is a custom parameter. Preferably, the preset similarity threshold is set to 0.6.

[0088] It should be noted that according to the above operation, there may be multiple edges between nodes. For the convenience of calculation, a "pruning" operation is directly performed here, that is, only one edge is retained between the nodes. In addition, the number of edges between nodes can also be used as a weight parameter in the calculation, which will not be elaborated here.

[0089] In one embodiment of the present invention, the second fault detection model consists of a graph data analysis layer and a second classifier;

[0090] The graph data analysis layer is used to update the instantaneous feature vectors of all nodes in the sparse electrical graph and output a third update vector, wherein the number of dimensions of the third update vector is a custom parameter. Preferably, the number of dimensions of the third update vector is set to 32;

[0091] The second classifier inputs the third update vector of the node corresponding to the electric energy meter with the fault mark, and outputs a value indicating the fault type of the electric energy meter with the fault mark;

[0092] The second classifier is built based on a multi-layer perceptron, and the corresponding activation function is the Softmax activation function.

[0093] In one embodiment of the present invention, the calculation formula of the graph data analysis layer includes:

[0094] Where 1≤u≤U, U represents the number of nodes in the sparse electrical graph, represents the third update vector of the u-th node, h u and h v Represent the instantaneous feature vectors of the u-th node and the v-th node respectively, N u represents the set of nodes that have edges with the u-th node, W1 and W2 represent the first weight parameter and the second weight parameter respectively, b1 and b2 represent the first bias parameter and the second bias parameter respectively;

[0095] represents the correlation coefficient between the u-th node and the v-th node, W3 represents the third weight parameter, || represents the concatenation operation, T represents the transposition operation, and σ represents the Sigmoid activation function.

[0096] It should be noted that according to the calculation formula of the graph data analysis layer, the first weight parameter and the second weight parameter both need to be designed as matrices of size 5×32, the first bias parameter and the second bias parameter both need to be designed as vectors of size 1×32, and the third weight parameter needs to be designed as a vector of size 1×10.

[0097] It should be noted that the sample labels of the training samples used to train the second fault detection model can also be obtained through manual labeling, that is, the fault types of the electricity meter may include: abnormal power factor, three-phase voltage imbalance, uneven current, power mutation, electricity theft, etc.; similarly, the instantaneous feature vectors of all nodes in the sparse electrical diagram need to be normalized before being input into the second fault detection model to eliminate dimensional differences.

[0098] In one embodiment of the present invention, Figure 2 As shown, a method for collecting electricity consumption information to identify electric energy meter failures includes the following steps:

[0099] Step S201: collect the active power of each electric energy meter to construct an active power sequence, and extract the time-frequency fingerprint through a three-layer discrete wavelet transform;

[0100] Step S202, calculating the daytime and nighttime power ratio of each electric energy meter and forming a periodic feature vector with the time-frequency fingerprint;

[0101] Step S203: Analyze the periodic characteristic vector of each electric energy meter using the first fault detection model, output a value indicating whether the electric energy meter has a fault, and mark the electric energy meter with a fault;

[0102] Step S204: Calculate the power factor, three-phase voltage imbalance, and three-phase current mean of each electric energy meter at the current moment, and form an instantaneous feature vector with the current power consumption and active power;

[0103] Step S205 , constructing a sparse electrical graph based on the instantaneous feature vector of each electric energy meter and the physical topological connections between the electric energy meters;

[0104] Step S206 : Analyze the sparse electrical diagram using the second fault detection model, and output a value indicating the fault type of the electric energy meter with the fault mark.

[0105] It should be noted that the intervals and thresholds are set for ease of comparison. The threshold size depends on the amount of sample data and the cardinality set by those skilled in the art for each set of sample data, as long as it does not affect the proportional relationship between the parameter and the quantized value. Furthermore, the above formulas are all dimensionless numerical calculations. These formulas are derived from software simulations of the most recent real-world conditions using large amounts of data. The preset parameters in these formulas are set by those skilled in the art based on actual conditions.

[0106] The above describes the embodiments of this embodiment, but this embodiment is not limited to the above specific implementation methods. The above specific implementation methods are merely illustrative and not restrictive. Ordinary technicians in this field can also make many forms based on the inspiration of this embodiment, all of which are protected by this embodiment.

Claims

1. An electricity consumption information collection system for identifying electric energy meter faults, characterized in that: include: The time-frequency fingerprint extraction module is used to collect the active power of each electric energy meter to construct an active power sequence, and extract the time-frequency fingerprint through a three-layer discrete wavelet transform; The time-frequency fingerprint consists of low-frequency energy values, mid-frequency energy values, and high-frequency energy values; A periodic feature construction module is used to calculate the day-night power ratio of each electric energy meter and form a periodic feature vector with the time-frequency fingerprint; a periodic feature analysis module, configured to analyze the periodic feature vector of each electric energy meter using the first fault detection model, output a value indicating whether the electric energy meter has a fault, and mark the electric energy meter with a fault; The instantaneous feature construction module is used to calculate the power factor, three-phase voltage imbalance and three-phase current mean of each electric energy meter at the current moment, and form an instantaneous feature vector with the current power consumption and active power; An electrical diagram construction module, which is used to construct a sparse electrical diagram based on the instantaneous feature vector of each electric energy meter and the physical topological connections between the electric energy meters; The electrical diagram analysis module is used to analyze the sparse electrical diagram through the second fault detection model, and the output value represents the fault type of the electric energy meter with the fault mark.

2. The power consumption information collection system for identifying power meter faults according to claim 1, characterized in that: The length of the active power sequence is equal to the acquisition time period divided by the acquisition time interval, where the acquisition time period and the acquisition time interval are both custom parameters; the active power sequence is subjected to wavelet transform through three-layer discrete wavelet transform to obtain three-layer detail coefficients, and the sum of the squares of the first-layer detail coefficient, the second-layer detail coefficient, and the third-layer detail coefficient are calculated respectively as the low-frequency energy value, the medium-frequency energy value, and the high-frequency energy value.

3. The power consumption information collection system for identifying power meter faults according to claim 1, characterized in that: The day-night power ratio is the ratio of the average power during the daytime period to the average power during the nighttime period. Both the daytime period and the nighttime period are user-defined parameters.

4. The power consumption information collection system for identifying power meter faults according to claim 1, characterized in that: The first fault detection model consists of a nonlinear transformation layer and a first classifier; The nonlinear transformation layer is used to transform the periodic feature vector into a first update vector, wherein the number of dimensions of the first update vector is a custom parameter; The calculation formula of the nonlinear transformation layer is as follows: Where 1≤i≤L, Represents the first update vector output by the nonlinear transformation layer, L represents the maximum number of layers of the nonlinear transformation layer, which is a custom parameter. and Represent the first intermediate vectors of the output of the i-1th layer and the i-th layer respectively, represents the periodic eigenvector of the input nonlinear transformation layer, and They represent the first weight parameter and the first bias parameter of the i-th layer of the nonlinear transformation layer, GELU represents the GELU activation function, and LaryerNorm represents the inter-layer normalization function; The first classifier inputs the first update vector, and the output value indicates whether the electric energy meter has a fault; The first classifier is built based on a multi-layer perceptron, and the corresponding activation function is the Sigmoid activation function.

5. The power consumption information collection system for identifying power meter faults according to claim 4, characterized in that: During the training of the first fault detection model, a reconstruction mapping layer is added after the nonlinear transformation layer; The reconstruction mapping layer is used to convert the first update vector into a second update vector, wherein the number of dimensions of the second update vector is the same as the number of dimensions of the periodic feature vector; The calculation formula for reconstructing the mapping layer is as follows: Where 1≤j≤J, represents the second update vector of the reconstruction mapping layer output, J represents the maximum number of layers of the reconstruction mapping layer, J = L, and Represent the second intermediate vectors output by the j-1th layer and the jth layer respectively, represents the first update vector of the input reconstruction mapping layer, and denote the second weight parameter and the second bias parameter of the jth layer of the reconstruction mapping layer, respectively. and They represent the third weight parameter and the third bias parameter of the jth layer of the reconstruction mapping layer, σ represents the Sigmoid activation function, and ⊙ represents the Hadamard product; The calculation formula of the loss function Loss of the first fault detection model is as follows: Loss = λ rec ×Loss rec +λ clf ×Loss clf , where Loss rec Represents the reconstruction loss, Loss clf represents the classification loss, λ rec represents the reconstruction loss coefficient, λ clf Represents the classification loss coefficient, the sum of the reconstruction loss coefficient and the classification loss coefficient is 1; Where 1≤k≤K, K represents the number of training samples, Represents the first update vector output by the nonlinear transformation layer after inputting the sample data of the kth training sample, represents the second update vector of the reconstructed mapping layer output after inputting the sample data of the kth training sample, and ||·||2 represents the L2 norm; where y k represents the sample label of the kth training sample, Represents the value of the output of the first fault detection model after inputting the sample data of the kth training sample.

6. The power consumption information collection system for identifying power meter faults according to claim 1, characterized in that: The power factor is equal to the phase difference between the line voltage and the line current; Three-phase voltage unbalance U unb The calculation formula is as follows: Among them U a 、U b and U c Represents phase A voltage, phase B voltage and phase C voltage respectively, U avg Indicates the three-phase average voltage, max means taking the maximum value; The calculation formula for the three-phase current mean is as follows: Among them I a , I b and I c Represent the A phase current, B phase current and C phase current respectively.

7. The power consumption information collection system for identifying power meter faults according to claim 1, characterized in that: A sparse electrical graph consists of nodes and edges between nodes, where nodes correspond one-to-one to electricity meters. If there is a physical topological connection between electricity meters, edges are constructed between the corresponding nodes. If the cosine similarity of the instantaneous feature vectors between the electricity meters is greater than or equal to the preset similarity threshold, edges are constructed between the corresponding nodes, where the preset similarity threshold is a custom parameter.

8. The power consumption information collection system for identifying power meter faults according to claim 7, characterized in that: The second fault detection model consists of a graph data analysis layer and a second classifier; The graph data analysis layer is used to update the instantaneous feature vectors of all nodes in the sparse electrical graph and output a third update vector, wherein the number of dimensions of the third update vector is a custom parameter; The second classifier inputs the third update vector of the node corresponding to the electric energy meter with the fault mark, and outputs a value indicating the fault type of the electric energy meter with the fault mark; The second classifier is built based on a multi-layer perceptron, and the corresponding activation function is the Softmax activation function.

9. The power consumption information collection system for identifying power meter faults according to claim 8, characterized in that: Calculation formula for graph data analysis layer include: Where 1≤u≤U, U represents the number of nodes in the sparse electrical graph, represents the third update vector of the u-th node, h u and h v Represent the instantaneous feature vectors of the u-th node and the v-th node respectively, N u represents the set of nodes that have edges with the u-th node, W1 and W2 represent the first weight parameter and the second weight parameter respectively, b1 and b2 represent the first bias parameter and the second bias parameter respectively; represents the correlation coefficient between the u-th node and the v-th node, W3 represents the third weight parameter, || represents the concatenation operation, T represents the transposition operation, and σ represents the Sigmoid activation function.

10. A method for collecting electricity usage information to identify electric energy meter failures, characterized in that: Executing the power consumption information collection system for identifying power meter faults according to any one of claims 1 to 9 comprises the following steps: Step S201: collect the active power of each electric energy meter to construct an active power sequence, and extract the time-frequency fingerprint through a three-layer discrete wavelet transform; Step S202, calculating the daytime and nighttime power ratio of each electric energy meter and forming a periodic feature vector with the time-frequency fingerprint; Step S203: Analyze the periodic characteristic vector of each electric energy meter using the first fault detection model, output a value indicating whether the electric energy meter has a fault, and mark the electric energy meter with a fault; Step S204: Calculate the power factor, three-phase voltage imbalance, and three-phase current mean of each electric energy meter at the current moment, and form an instantaneous feature vector with the current power consumption and active power; Step S205 , constructing a sparse electrical graph based on the instantaneous feature vector of each electric energy meter and the physical topological connections between the electric energy meters; Step S206 : Analyze the sparse electrical diagram using the second fault detection model, and output a value indicating the fault type of the electric energy meter with the fault mark.

Images