A method for fault assessment of electric energy metering box

Abstract

The present invention relates to the field of fault assessment technology, and specifically discloses a method for fault assessment of an electric energy meter box. The method selects multiple key assessment indicators from the basic information, operation information, and fault repair information of the electric energy meter box to construct a key component selection model, thereby determining the key components that affect the product quality, operation safety, and usage function of the electric energy meter box; constructs a fault matching neural network model, and performs risk assessment on the key components of the electric energy meter box based on the correlation between faults and influencing factors to obtain the fault parameters of the electric energy meter box; combines the correlations between influencing factors, between faults, and between faults and influencing factors to perform multi-stress cross-correlation analysis on the electric energy meter box, thereby completing the fault assessment of the electric energy meter box. The method solves the problem that the existing technology increases the complexity of subsequent calculations, the accuracy and intelligence of fault assessment need to be improved, and the safety and stability of power supply cannot be ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of fault assessment, and in particular to a method for assessing faults of an electric energy metering box. Background Art

[0002] The energy meter box is the face of State Grid Corporation of China and the first line of defense for metering equipment and key components, including smart energy meters, data collection terminals, and circuit breakers. Currently, energy meter boxes come in a wide variety of sizes and specifications, with varying degrees of workmanship. Some equipment experiences damage, corrosion, and aging within less than two years of operation. This makes it difficult to provide a sound operating environment for the metering equipment and is unable to meet the demands of the modern era for intelligent collection and management of electricity usage information, resulting in adverse social impacts and additional economic losses. Conducting energy meter box fault assessments is an effective measure to ensure that key elements of the production process are effectively controlled, thereby ensuring the safety and stability of the power supply.

[0003] At present, there are three main types of faults in the electricity meter boxes used on-site in China: first, the box body is rusted and the coating is damaged, which affects the aesthetics of the electricity meter box and damages the corporate image; second, electrical components such as wire terminals and protective devices are burned, posing serious safety hazards and easily causing accidents such as electric shock and fire; third, the box door and door lock are damaged, resulting in the electricity meter box being unable to perform normal protection and unable to meet the needs of anti-electricity theft management.

[0004] Chinese patent publication number CN114879124A discloses a fault analysis system for an electric energy metering device, which includes: a fault judgment module, a data acquisition module, and a data processing module; the fault judgment module is used to receive a fault analysis request, identify the corresponding fault type to be analyzed, generate a data acquisition instruction, and send the data acquisition instruction to the data acquisition module; the data acquisition module is used to obtain the actual operating data of the smart electric energy meter, and receive the data acquisition instruction, determine the target operating data from the actual operating data, and package the target operating data with the data acquisition instruction and send it to the data processing module; the data processing module is used to receive the target operating data and the data acquisition instruction, perform data processing on the target operating data based on the data acquisition instruction, obtain a data processing result, and send the data processing result to the fault judgment module; the fault judgment module is also used to receive the data processing result and determine whether the electric energy metering device has a fault to be analyzed.

[0005] However, the above technology has at least the following problems: the existing technology does not evaluate the key components of the electricity meter box in advance, which increases the complexity of subsequent calculations, does not fully consider the cross-relationship of multiple stresses, and cannot fully analyze the fault conditions of the electricity meter box. The accuracy and intelligence of fault assessment need to be improved, and the safety and stability of power supply cannot be ensured. Summary of the Invention

[0006] The present invention provides an electric energy meter box fault assessment method, which is an effective measure to ensure that key factors of the electric energy meter box are effectively controlled during the production process, ensures that the electric energy meter box meets the expected use and requirements, and further ensures the safety and stability of power supply. It solves the problems that the prior art fails to assess the key components of the electric energy meter box in advance, increases the complexity of subsequent calculations, does not fully consider the cross-relationship of multiple stresses, cannot fully analyze the fault conditions of the electric energy meter box, the accuracy and intelligence of the fault assessment need to be improved, and cannot ensure the safety and stability of power supply.

[0007] The present invention specifically includes the following technical solutions:

[0008] A method for evaluating a fault of an electric energy meter box comprises the following steps:

[0009] Step S1. Select multiple key evaluation indicators from the basic information, operation information and troubleshooting information of the energy meter box to build a key component selection model to determine the key components that affect the energy meter box product quality, operation safety and usage function;

[0010] Step S2. Construct a fault matching neural network model, perform risk assessment on key components of the energy meter box based on the correlation between the fault and the influencing factors, and obtain the fault parameters of the energy meter box;

[0011] Step S3: Based on the correlations between influencing factors, between faults, and between faults and influencing factors, a multi-stress cross-correlation analysis is performed on the electric energy metering box to complete the fault assessment of the electric energy metering box.

[0012] Furthermore, the step S1 specifically includes:

[0013] A key component selection model was constructed by selecting multiple key evaluation indicators from the basic information, operational information, and troubleshooting information of the energy meter box to identify key components that affect the product quality, operational safety, and functional use of the energy meter box. The evaluation indicators were primarily established based on five dimensions: safety, electricity usage, metering, operation and maintenance, and anti-theft measures. Key components are those whose operation, connection, data, control, alarm, or failure conditions may have a direct impact on the energy meter box. Non-key components are those whose operation, connection, data, control, alarm, or failure conditions may have an indirect impact or no impact on the energy meter box.

[0014] Furthermore, the step S1 specifically includes:

[0015] The influence entropy of components on the quality parameters of the electric energy meter box in various dimensions after different conditions occur is calculated, and an influence threshold is established. The influence entropy of each component on the quality parameters of the electric energy meter box is ranked, and the top M components that meet the threshold are selected as the key components that affect the product quality, operation safety, and usage function of the electric energy meter box.

[0016] Furthermore, the step S2 specifically includes:

[0017] Obtain the failure modes and causes of failure in the historical operation fault data of the electric energy meter box, and obtain the influencing factors of the current electric energy meter box within a preset time period before the fault occurs; build a fault matching neural network model, select a fixed number of electric energy meter boxes within a preset time period. The specific values ​​of the influencing factors are used as sample data and input into the fault matching neural network. After intelligent training and learning of the neural network, the fault parameters of each electric energy meter box are output.

[0018] Furthermore, the step S2 specifically includes:

[0019] The fault matching neural network model includes input layer, mapping layer, state layer, stability layer and output layer.

[0020] Furthermore, the step S3 specifically includes:

[0021] Study the correlation between influencing factors, between faults, and between faults and influencing factors, and conduct multi-stress cross-correlation analysis.

[0022] Furthermore, the step S2 specifically includes:

[0023] According to the similarity of the influencing factors of each electric energy meter box to a certain cluster in the sample data, which is different from that of other electric energy meter boxes, a fuzzy membership degree is assigned to the influencing factors of each electric energy meter box, so that the electric energy meter boxes with higher similarity of influencing factors are divided into one category.

[0024] Furthermore, the step S2 specifically includes:

[0025] The fuzzy membership threshold is selected to select the electricity metering boxes with similarity lower than the threshold, and further analysis is performed between the influencing factors and the faults; the fault matching neural network model is updated according to the fuzzy factor and fuzzy clustering algorithm, and the update rules are established.

[0026] The present invention has at least the following technical effects or advantages:

[0027] 1. Comprehensively consider the importance of each component of the energy meter box from the five dimensions of safety, electricity consumption, metering, operation and maintenance, and anti-theft electricity, so as to accurately evaluate the key components of the energy meter box and determine the key components that affect the product quality, operational safety, and usage functions of the energy meter box, thereby determining the key areas for the next step of risk assessment;

[0028] 2. Based on the characteristics of the energy meter box and actual operating experience, the fault-related feature quantities are extracted and a fault matching neural network model is constructed. Based on the correlation between the fault and the influencing factors, the risk assessment of the key components of the energy meter box is carried out to obtain the fault parameters of the energy meter box. The effective analysis and judgment of the fault condition of the energy meter box can help staff quickly and comprehensively analyze the line problem, reducing analysis time and improving work efficiency.

[0029] 3. Combined with the correlation between influencing factors, between faults, and between faults and influencing factors, a multi-stress cross-correlation analysis is conducted on the electric energy metering box. By integrating various monitoring data, the fuzzy c-means clustering method is used to classify the electric energy metering boxes. The fault matching neural network model is optimized based on fuzzy membership, and potential faults in the design, production and use processes are evaluated to improve the accuracy and intelligence of fault assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of a method for fault assessment of an electric energy meter box according to the present invention;

[0031] Figure 2 This is a structural block diagram of the fault matching neural network model described in the present invention. DETAILED DESCRIPTION

[0032] The embodiment of the present application provides a method for fault assessment of an electric energy meter box, thereby solving the problem that the prior art fails to assess the key components of the electric energy meter box in advance, increases the complexity of subsequent calculations, fails to fully consider the cross-relationships of multiple stresses, and cannot comprehensively analyze the fault conditions of the electric energy meter box. The accuracy and intelligence of the fault assessment need to be improved, and the safety and stability of the power supply cannot be ensured.

[0033] The technical solution in the embodiments of the present application is to solve the above problems, and the overall idea is as follows:

[0034] The importance of each component of the energy meter box was comprehensively considered from five dimensions: safety, electricity usage, metering, operation and maintenance, and anti-theft electricity. This allowed for an accurate assessment of the key components of the energy meter box, identifying the key components that affect the product quality, operational safety, and functional use of the energy meter box, and determining the key areas for the next risk assessment. Based on the characteristics of the energy meter box and actual operating experience, fault-related feature quantities were extracted, and a fault matching neural network model was constructed. Risk assessment of the key components of the energy meter box was conducted based on the correlation between faults and influencing factors, and the fault parameters of the energy meter box were obtained. Effective analysis and judgment of the fault condition of the energy meter box can help staff quickly and comprehensively analyze line problems, reducing analysis time and improving work efficiency. Combining the correlations between influencing factors, between faults, and between faults and influencing factors, a multi-stress cross-correlation analysis was conducted on the energy meter box. By integrating various monitoring data, a fuzzy c-means clustering method was used to classify the energy meter box. The fault matching neural network model was optimized based on fuzzy membership, assessing potential faults during design, production, and use, and improving the accuracy and intelligence of fault assessment.

[0035] In order to better understand the above technical solution, the above technical solution will be described in detail below with reference to the accompanying drawings and specific implementation methods.

[0036] Refer to the attached Figure 1 The electric energy meter box fault assessment method of the present invention comprises the following steps:

[0037] S1. Select multiple key evaluation indicators from the basic information, operation information and fault repair information of the electricity meter box to build a key component selection model to determine the key components that affect the product quality, operation safety and usage functions of the electricity meter box.

[0038] Each component of an energy meter box is relatively independent, has a clear function, and has varying degrees of impact on its quality parameters. Therefore, in order to conduct a targeted risk assessment of an energy meter box, it is necessary to conduct a component criticality assessment.

[0039] A key component selection model is constructed by selecting multiple key evaluation indicators from the basic information, operational information, and troubleshooting information of the energy meter box to identify key components that affect the product quality, operational safety, and functional use of the energy meter box. The evaluation indicators are mainly established based on five dimensions: safety, electricity usage, metering, operation and maintenance, and anti-theft. Key components refer to components whose operation, connection, data, control, alarm, or failure conditions may have a direct impact on the energy meter box; non-key components refer to components whose operation, connection, data, control, alarm, or failure conditions may have an indirect impact or no impact on the energy meter box.

[0040] As an embodiment of the present invention, the influence entropy of each component of the electric energy meter box on the quality parameters of the electric energy meter box in five dimensions, namely, safety, electricity consumption, metering, operation and maintenance, and prevention of electricity theft, is obtained after different conditions occur. The influence entropy includes the influence of each component on the quality parameters of the electric energy meter box under normal operation, connection, data transmission, control feedback, alarm and failure conditions; further, the influence direction includes personal and equipment safety, normal electricity consumption, correct metering, operation and maintenance, and the risk of electricity theft.

[0041] Specifically, the calculation formula for the entropy of the influence of the current component on the quality parameters of the electric energy meter box in each dimension after different conditions occur is determined respectively under the condition that all components except the current component are operating normally:

[0042]

[0043]

[0044] Among them, S i represents the influence entropy of the i-th component on the quality parameters of the electric energy meter box, that is, the confusion caused by the i-th component on the quality parameters of the electric energy meter box in five dimensions after different conditions occur, ω j represents the weight of the j-th dimension, It represents the initial impact entropy of the quality parameters of the electric energy meter box in the jth dimension when the i-th component has no problem. It represents the entropy of the impact of the i-th component on the quality parameters of the electric energy meter box in the j-th dimension after different conditions occur; represents the impact factor of the i-th component on the quality parameters of the electric energy meter box in the j-th dimension. The impact factor is used to characterize the dynamic deviation caused by the current component on the quality parameters of the electric energy meter box in the j-th dimension. i It represents the user's evaluation value of the impact of different conditions on the use of the electric energy meter box after the i-th component occurs. It represents the threshold of the evaluation value. The higher the evaluation value, the greater the impact on user usage after different conditions occur. N represents the number of all electricity meter boxes participating in the impact entropy test.

[0045] An impact threshold is established, the impact entropy of each component on the quality parameters of the electricity meter box is ranked, and the top M components that meet the threshold are selected as the key components that affect the product quality, operation safety, and usage function of the electricity meter box.

[0046] In a specific embodiment, the key components obtained through the key component selection model include seven components: a box body, a box door, a door lock, an inlet and outlet switch, a terminal block, a wire, and a connector.

[0047] S2. Construct a fault matching neural network model, conduct risk assessment on key components of the electric energy meter box based on the correlation between faults and influencing factors, and obtain the fault parameters of the electric energy meter box.

[0048] Conduct risk assessment on key components of the electricity meter box, study the failure modes of key components of the electricity meter box, that is, failure manifestations, evaluate the risk level of each failure mode, and classify the failures.

[0049] Based on the characteristics of the energy meter box and actual operating experience, factors that affect energy meter box failure can be divided into four categories: environmental, mechanical, electrical, and other factors. Environmental factors include salt spray (corrosive), temperature, humidity, dust, rain, and solar radiation; mechanical factors include impact and vibration; electrical factors include lightning strikes, overcurrent, or short circuits; and other factors include power theft and property management issues.

[0050] In one specific embodiment, various influencing factors are quantified to facilitate research and analysis. A five-level quantification is used for four factors: salt spray, temperature, humidity, and dust. A two-level quantification is used for eight factors: rain intrusion, sunlight radiation, impact, vibration, lightning strike, power overcurrent or short circuit, high incidence of power theft, and property management quality.

[0051] Based on the actual operation data of the electricity meter box, a cross-analysis of faults and their influencing factors was carried out, and the correlation between influencing factors, between faults, and between faults and influencing factors was studied. A multi-stress cross-correlation analysis model including environmental, mechanical, and electrical stress was formed, and the correlation coefficient was obtained.

[0052] Furthermore, to conduct multi-stress cross-correlation analysis, it is first necessary to match the faults with the influencing factors. Based on the historical operation failure experience of the electricity meter box, the corresponding relationship between the matching faults and the influencing factors is analyzed.

[0053] As a specific embodiment, the failure modes and causes of failures in the historical operational fault data of the energy meter box are obtained, and the influencing factors of the current energy meter box within a preset time period before the failure occurs are obtained. A fault matching neural network model is constructed, and the specific values ​​of the influencing factors of a fixed number of energy meter boxes within the preset time period are selected as sample data. This data is input into the fault matching neural network. After intelligent training and learning of the neural network, the fault parameters of each energy meter box are output, which include the failure mode, failure cause, and fault level.

[0054] The fault matching neural network model includes an input layer, a mapping layer, a state layer, a stabilization layer and an output layer.

[0055] The input X of the fault matching neural network is set as:

[0056]

[0057] Among them, 4 represents the number of categories of fault factors, {t1,…,t n} represents the preset time period, and n represents the maximum number of time points divided into the time period. The input layer has a total of 4 neurons. Each category of fault factors is input into each neuron. The input layer sends the input data to the mapping layer.

[0058] The mapping layer maps the input data to the computational space through standardization to facilitate subsequent calculations. The mapping layer sends the standardized data to the state layer in a fully connected form. The state layer realizes the memory of the information state in the neuron. There are t n neurons, and the state vector function of each neuron is:

[0059] q tn T=q tn T-1+qT-1

[0060] qT=f(ω1×q tn T+ω2×(δ×(T-1)))

[0061] Among them, q tn T represents the tth n The state vector function of each neuron is represented by T, qT-1, the hidden state of the neuron at the previous moment, f, the activation function, ω1 and ω2, the weight of the state vector function and the weight of the hidden state, respectively, and δ, the probability of influence. The state layer sends the state vector function of each neuron to the stabilization layer.

[0062] The stabilization layer stabilizes the state according to the temporal changes of the influencing factors in the state layer, forms linear correlations between different neurons, and obtains the fault function in the stable state. The specific calculation is:

[0063]

[0064] Where R represents the hazard function, represents the linear correlation function, θ represents the convergence factor, and γ represents the stability factor. The stability layer sends the fault function to the output layer.

[0065] The output layer outputs the fault parameters, and the calculation of the output layer is:

[0066] y=g(ω3×R)

[0067] Where y is the fault parameter, g represents the output function, and ω3 represents the connection weight of the output layer.

[0068] The fault parameters of the electric energy meter box, namely the failure mode, failure cause and fault level, are obtained according to the output of the fault matching neural network model.

[0069] S3. Combined with the correlations between influencing factors, between faults, and between faults and influencing factors, a multi-stress cross-correlation analysis is performed on the electric energy metering box to complete the fault assessment of the electric energy metering box.

[0070] In order to improve the accuracy of fault assessment of electric energy metering boxes, the correlations among influencing factors, among faults, and between faults and influencing factors were studied, and multi-stress cross-correlation analysis was performed.

[0071] The influencing factors are clustered and divided. The clustering method adopts the existing fuzzy c-means clustering. According to the similarity of the influencing factors of each electric energy meter box to a certain cluster and other electric energy meter boxes in the sample data, a fuzzy membership degree is assigned to the influencing factors of each electric energy meter box, so that the electric energy meter boxes with high similarity in influencing factors are divided into one category. The specific implementation method is as follows:

[0072] The set of factors influencing the establishment of an energy meter box is expressed as x = {x1, x2, ..., x m}, m represents the number of electric energy meter boxes, and the cluster set is represented as v = {v1, v2, ..., v c}, c represents the total number of cluster categories; the fuzzy membership degree of the kth electric energy meter box to the lth cluster is expressed as u kl :

[0073]

[0074] Among them, x k represents the influencing factors of the kth electric energy meter box, k∈[1,m], v l represents the lth cluster, v l' represents the l'th cluster, l,l'∈[1,c], l≠l', ε represents the fuzzy factor.

[0075] Select the fuzzy membership threshold to select the energy meter boxes with similarity lower than the threshold, and conduct further analysis between influencing factors and faults. Update the fault matching neural network model based on the fuzzy factor and fuzzy clustering algorithm, and establish the update rules:

[0076]

[0077]

[0078] in, and They represent the weight of the updated state vector function and the weight of the implicit state, respectively. This enables multi-stress cross-correlation analysis, obtaining more accurate fault parameters, namely failure mode, failure cause, and fault level, and completing the fault assessment of the energy meter box.

[0079] In summary, the electric energy meter box fault assessment method of the present invention is completed.

[0080] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts 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, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0081] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0082] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0083] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A method for evaluating faults in an electric energy meter box, characterized in that: The following steps are involved: Step S1. Select multiple key evaluation indicators from the basic information, operation information and troubleshooting information of the energy meter box to build a key component selection model to determine the key components that affect the energy meter box product quality, operation safety and usage function; Step S2. Construct a fault matching neural network model, perform risk assessment on key components of the energy meter box based on the correlation between the fault and the influencing factors, and obtain the fault parameters of the energy meter box; Step S3: Based on the correlations between influencing factors, between faults, and between faults and influencing factors, a multi-stress cross-correlation analysis is performed on the electric energy metering box to complete the fault assessment of the electric energy metering box.

2. The method for evaluating a fault of an electric energy meter box according to claim 1, wherein: The step S1 specifically includes: A number of key evaluation indicators are selected from the basic information, operation information and fault repair information of the electricity meter box to construct a key component selection model to determine the key components that affect the product quality, operation safety and usage functions of the electricity meter box; the evaluation indicators are mainly established based on the five dimensions of safety, electricity consumption, metering, operation and maintenance, and anti-theft of electricity; key components refer to components whose operation, connection, data, control, alarm, or failure conditions may have a direct impact on the electricity meter box; non-key components refer to components whose operation, connection, data, control, alarm, or failure conditions may have an indirect impact or no impact on the electricity meter box.

3. The method for evaluating a fault of an electric energy meter box according to claim 2, wherein: The step S1 specifically includes: The influence entropy of components on the quality parameters of the electric energy meter box in various dimensions after different conditions occur is calculated, and an influence threshold is established. The influence entropy of each component on the quality parameters of the electric energy meter box is ranked, and the top M components that meet the threshold are selected as the key components that affect the product quality, operation safety, and usage function of the electric energy meter box.

4. The method for evaluating a fault of an electric energy meter box according to claim 1, wherein: The step S2 specifically includes: Obtain the failure modes and causes of failure in the historical operation fault data of the electric energy meter box, and obtain the influencing factors of the current electric energy meter box within a preset time period before the fault occurs; build a fault matching neural network model, select a fixed number of electric energy meter boxes within a preset time period. The specific values ​​of the influencing factors are used as sample data and input into the fault matching neural network. After intelligent training and learning of the neural network, the fault parameters of each electric energy meter box are output.

5. The method for evaluating faults of an electric energy meter box according to claim 2, wherein: The step S2 specifically includes: The fault matching neural network model includes input layer, mapping layer, state layer, stability layer and output layer.

6. The method for evaluating faults of an electric energy meter box according to claim 1, wherein: The step S3 specifically includes: Study the correlation between influencing factors, between faults, and between faults and influencing factors, and conduct multi-stress cross-correlation analysis.

7. The method for evaluating faults of an electric energy meter box according to claim 6, wherein: The step S3 specifically includes: According to the similarity of the influencing factors of each electric energy meter box to a certain cluster in the sample data, which is different from that of other electric energy meter boxes, a fuzzy membership degree is assigned to the influencing factors of each electric energy meter box, so that the electric energy meter boxes with higher similarity of influencing factors are divided into one category.

8. The method for evaluating faults of an electric energy meter box according to claim 7, wherein: The step S3 specifically includes: The fuzzy membership threshold is selected to select the electricity metering boxes with similarity lower than the threshold, and further analysis is performed between the influencing factors and the faults; the fault matching neural network model is updated according to the fuzzy factor and fuzzy clustering algorithm, and the update rules are established.

Images