An analysis method and system for fault probability of an electric energy meter

Abstract

The application provides an analysis method and system for fault probability of an electric energy meter, comprising: obtaining technical parameters of an in-service electric energy meter, and setting multiple in-service durations for the in-service electric energy meter; inputting the technical parameters of the in-service electric energy meter and the multiple in-service durations into a pre-constructed fault analysis model to obtain fault classification probabilities of the in-service electric energy meter under the multiple set in-service durations respectively; taking the in-service duration corresponding to the fault classification probability greater than a fault threshold and the largest number of electric energy meters as a fault replacement inflection point of the in-service electric energy meter, and calculating a fault probability of the in-service electric energy meter at the fault replacement inflection point; wherein the fault analysis model is constructed based on a decision tree algorithm by using current technical parameters of the in-service electric energy meter and technical parameters of historical fault electric energy meters. The application can accurately predict the fault replacement inflection point and the corresponding fault probability of the electric energy meter based on the decision tree algorithm, and effectively reduce the disturbance to the power grid and the abnormal electric quantity caused by the fault electric energy meter not being replaced in time.

Description

Technical Field

[0001] This invention relates to the field of electricity meter fault analysis, and specifically to a method and system for analyzing the probability of electricity meter faults. Background Technology

[0002] Currently, the failure probability analysis of electricity meters mainly relies on calculating the rated failure rate of the meters based on their installation and failure status to determine the failure condition. This method analyzes and explains the historical operation of the meters, but its accuracy is low when used for fault identification of currently operating meters, failing to meet the business needs of electricity meter failure probability analysis. Furthermore, over time, the failure probability of a batch of electricity meters installed at fixed historical times continuously increases. Existing electricity meter failure probability analysis methods struggle to accurately predict the inflection point of large-scale failure and reinstallation of electricity meters, as well as the failure probability of individual meters at that inflection point. This can easily lead to grid fluctuations and abnormal power output caused by untimely reinstallation of faulty meters. Summary of the Invention

[0003] To address the problem of low accuracy in predicting the failure probability of electricity meters in existing technologies, this invention provides a method for analyzing the failure probability of electricity meters, including:

[0004] Obtain the technical parameters of the operating electricity meters and set multiple operating durations for the operating electricity meters;

[0005] The technical parameters of the operating energy meter and multiple operating durations are input into a pre-built fault analysis model to obtain the fault classification probability of the operating energy meter under multiple set operating durations.

[0006] The in-service duration corresponding to the fault classification probability when the number of electricity meters is greater than the fault threshold is taken as the fault reinstallation inflection point of the in-service electricity meters, and the fault probability corresponding to the in-service electricity meters at the fault reinstallation inflection point is calculated.

[0007] The fault analysis model is based on the decision tree algorithm and is constructed using the current technical parameters of the operating electricity meter and the technical parameters of historically faulty electricity meters.

[0008] Preferably, the technical parameters of the operating energy meter and the historically faulty energy meter include: region, supplier, operating time calculated using the freeze date of the operating energy meter or the fault date of the historically faulty energy meter and the corresponding installation date, inventory time calculated using the warehousing date of the operating energy meter or the historically faulty energy meter and the corresponding installation date, and fault and normal labels.

[0009] Preferably, the construction of the fault condition analysis model includes:

[0010] Based on the decision tree algorithm, the information gain rate of each attribute is calculated by using the parameters of each electricity meter under the attributes of region, supplier, operating time, and inventory time respectively.

[0011] The attribute corresponding to the maximum information gain ratio is taken as the root node, and after determining the root node, the information gain ratio corresponding to each attribute is repeatedly calculated to generate internal nodes according to the maximum information gain ratio obtained in each round, until the node whose information gain ratio remains unchanged is taken as the leaf node.

[0012] The fault analysis model is constructed based on the root node, internal nodes, and leaf nodes.

[0013] Preferably, the step of setting the operating time corresponding to the fault classification probability when the number of electricity meters is at its maximum and the fault threshold is greater than the fault threshold as the fault reinstallation inflection point for the electricity meters in operation, and calculating the fault probability corresponding to the electricity meters in operation at the fault reinstallation inflection point, includes:

[0014] All energy meters with a set running time corresponding to the fault threshold are regarded as fault energy meters under the fault reinstallation inflection point.

[0015] The ratio of the number of faulty energy meters under a set operating time to the total number of all operating energy meters is taken as the fault probability of the operating energy meters reaching the fault reinstallation inflection point.

[0016] Preferably, the plurality of set operating durations are determined by adding the current operating duration of the operating energy meter to the duration of the plurality of set changes.

[0017] Preferably, it further includes: the technical parameters of each operating energy meter and the technical parameters of the historically faulty energy meters are obtained using... The result is obtained after outlier removal according to the principle.

[0018] Based on the same inventive concept, the present invention also provides a system for analyzing the probability of electricity meter failures, comprising:

[0019] The acquisition module is used to acquire the technical parameters of the operating electricity meter and set multiple operating durations for the operating electricity meter;

[0020] The fault analysis module is used to input the technical parameters of the operating energy meter and multiple operating durations into a pre-built fault analysis model to obtain the fault classification probability of the operating energy meter under multiple set operating durations.

[0021] The module for determining the reinstallation inflection point and fault probability is used to take the operating time corresponding to the fault classification probability when the fault is greater than the fault threshold and the number of electricity meters is the largest as the fault reinstallation inflection point of the electricity meters in operation, and to calculate the fault probability of the electricity meters in operation at the fault reinstallation inflection point.

[0022] The fault analysis model is based on the decision tree algorithm and is constructed using the current technical parameters of the operating electricity meter and the technical parameters of historically faulty electricity meters.

[0023] Preferably, the technical parameters of the operating energy meters and the technical parameters of the historically faulty energy meters obtained by the acquisition module include: region, supplier, operating time calculated using the freeze date of the operating energy meters or the fault date of the historically faulty energy meters and the corresponding installation date, inventory time calculated using the warehousing date of the operating energy meters or the historically faulty energy meters and the corresponding installation date, and fault and normal labels.

[0024] Preferably, the construction of the fault condition analysis model includes:

[0025] Based on the decision tree algorithm, the information gain rate of each attribute is calculated by using the parameters of each electricity meter under the attributes of region, supplier, operating time, and inventory time respectively.

[0026] The attribute corresponding to the maximum information gain ratio is taken as the root node, and after determining the root node, the information gain ratio corresponding to each attribute is repeatedly calculated to generate internal nodes according to the maximum information gain ratio obtained in each round, until the node whose information gain ratio remains unchanged is taken as the leaf node.

[0027] The fault analysis model is constructed based on the root node, internal nodes, and leaf nodes.

[0028] Preferably, the reinstallation inflection point and failure probability determination module includes:

[0029] The reinstallation inflection point determination submodule is used to identify all energy meters with a set running time corresponding to the fault threshold as fault energy meters under the fault reinstallation inflection point.

[0030] The fault probability determination submodule is used to take the ratio of the number of faulty energy meters under a set operating time to the total number of all operating energy meters as the fault probability of the operating energy meters reaching the fault reinstallation inflection point.

[0031] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0032] This invention provides a method and system for analyzing the probability of electricity meter failures, comprising: acquiring the technical parameters of operating electricity meters and setting multiple operating durations for the operating electricity meters; inputting the technical parameters of the operating electricity meters and the multiple operating durations into a pre-constructed fault analysis model to obtain the fault classification probabilities of the operating electricity meters under the multiple set operating durations; taking the operating duration corresponding to the fault classification probability that is greater than the fault threshold and has the largest number of electricity meters as the fault reinstallation inflection point of the operating electricity meters, and calculating the fault probability corresponding to the operating electricity meters at the fault reinstallation inflection point; wherein, the fault analysis model is based on a decision tree algorithm and is constructed using the current technical parameters of the operating electricity meters and the technical parameters of historically failed electricity meters. This invention, based on the decision tree algorithm, can accurately predict the fault reinstallation inflection point of electricity meters and the corresponding fault probability, effectively reducing the disturbance to the power grid and abnormal power output caused by the untimely replacement of failed electricity meters. Attached Figure Description

[0033] Figure 1 This is a flowchart of a method for analyzing the failure probability of an electricity meter according to the present invention;

[0034] Figure 2 This is a schematic diagram of an analysis system for the failure probability of an electricity meter according to the present invention. Detailed Implementation

[0035] Example 1

[0036] To address the problem of low accuracy in predicting the failure probability of electricity meters in existing technologies, this invention provides a method for analyzing the failure probability of electricity meters, such as... Figure 1 As shown, it includes:

[0037] Step 1: Obtain the technical parameters of the operating energy meters and set multiple operating durations for the operating energy meters;

[0038] Step 2: Input the technical parameters of the operating energy meter and multiple operating durations into the pre-built fault analysis model to obtain the fault classification probability of the operating energy meter under multiple set operating durations.

[0039] Step 3: Take the operating time corresponding to the fault classification probability when the number of electricity meters is the largest and the fault threshold is greater than the fault threshold as the fault reinstallation inflection point of the electricity meters in operation, and calculate the fault probability corresponding to the electricity meters in operation at the fault reinstallation inflection point.

[0040] The fault analysis model is based on the decision tree algorithm and is constructed using the current technical parameters of the operating electricity meter and the technical parameters of historically faulty electricity meters.

[0041] In step 1, the technical parameters of the currently operating electricity meter are obtained, and multiple operating durations are set for the operating electricity meter;

[0042] The technical parameters of the operating energy meters include: region, supplier, operating time calculated using the freeze date and corresponding installation date of the operating energy meters, inventory time calculated using the warehousing date and corresponding installation date of the operating energy meters, and normal label.

[0043] In this embodiment, from the perspective of asset lifecycle, the longer an asset operates and the closer it is to retirement, the higher its failure probability. Therefore, the operating time attribute of an electricity meter is of great significance for predicting the failure probability of the electricity meter. The formula for calculating the operating time of an operating electricity meter is as follows:

[0044] L i =R-AR i

[0045] In the formula, L i AR represents the operating duration of the i-th operating energy meter, R represents the freeze date of the operating data, and AR represents the operating duration of the i-th operating energy meter. i This indicates the installation date of the i-th electricity meter.

[0046] The inventory duration of the electricity meters in operation is determined by the following formula:

[0047] K n =AR n -DR n

[0048] In the formula, K n AR represents the inventory duration of the nth electricity meter in operation. n Indicates the installation date of the nth electricity meter, DR n This indicates the date the nth energy meter was put into storage.

[0049] Since the data for both the "operational duration" and "inventory duration" attributes should be greater than or equal to 0, the calculated operational duration and inventory duration of the operating energy meters are used... Outlier identification is performed in principle. Data less than 0 is identified as outliers and removed.

[0050] The multiple set operating durations for the operating energy meters are determined by adding the operating duration of the operating energy meters to the time length of multiple set changes. They are subsequently used to predict the inflection point and failure probability of a large number of failures and reinstallations of operating energy meters in the next few years.

[0051] The operating durations of the various operating energy meters are determined by the following formula:

[0052] ZY′ i,j =zy i +j (j=1,2,3…n)

[0053] In the formula, ZY′ i,j This indicates that the operating time of the i-th operating energy meter has changed to the operating time up to the j-th time period, zy i The current operating time of the i-th electricity meter is represented by , and the change time is represented by j.

[0054] Before step 2, a fault analysis model is constructed using the decision tree algorithm. The specific steps are as follows:

[0055] Obtain the technical parameters of historically faulty energy meters, including: region, supplier, in-service duration calculated using the fault date and corresponding installation date of the historically faulty energy meters, inventory duration calculated using the warehousing date and corresponding installation date of the historically faulty energy meters, and fault tag;

[0056] The operating time of historically faulty energy meters is calculated using the following formula:

[0057] L′ i =R i -AR′ i

[0058] In the formula, L′ i R represents the operating time of the i-th energy meter in the historical fault energy meters. i AR′ represents the failure date of the i-th energy meter in the historical fault energy meters. i This indicates the installation date of the i-th energy meter in the historical fault energy meters.

[0059] The inventory duration of historical faulty energy meters is determined by the following formula:

[0060] K h =AR h -DR h

[0061] In the formula, K h AR represents the inventory duration of the h-th historical faulty energy meter. h Indicates the installation date of the h-th historical faulty energy meter, DR h This indicates the date the h-th historical faulty energy meter was entered into the inventory.

[0062] The operating time and inventory time of historically faulty energy meters were also utilized. The principle is to identify outliers and then remove them.

[0063] Based on the decision tree algorithm, a wide table for decision tree analysis is formed by using the technical parameters of each in-operation energy meter and historically faulty energy meter according to the parameters of each energy meter under the attributes of region, supplier, in-operation duration, and inventory duration, and the information gain rate of each attribute is calculated accordingly.

[0064] The attribute corresponding to the maximum information gain ratio is taken as the root node, and after determining the root node, the information gain ratio corresponding to each attribute is repeatedly calculated to generate internal nodes according to the maximum information gain ratio obtained in each round, until the node whose information gain ratio remains unchanged is taken as the leaf node.

[0065] The fault analysis model is constructed based on the root node, internal nodes, and leaf nodes.

[0066] In this embodiment, the fault analysis model is constructed based on the C4.5 decision tree algorithm and can be applied to predict the fault probability of all operating energy meters. According to the fault labels contained in the technical parameters of operating energy meters and the technical parameters of historically faulty energy meters, energy meters labeled as normal are classified as label 0, and energy meters labeled as faulty are classified as label 1. The decision tree algorithm classification probability is output, and the data of operating energy meters and historically faulty energy meters with a probability greater than 0.9 in each node are selected from the classification probabilities.

[0067] For the analysis dimension i, it is determined by the following formula:

[0068]

[0069] In the formula, S i Z represents the number of operating and historically faulty energy meters corresponding to the probability that the i-th supplier (or region or installation year) is classified into label 1 is greater than 0.9. d represents the fault classification probability. i This represents the total number of electricity meters in operation from the i-th supplier (or region or year of installation);

[0070] The failure probability of the energy meter under analysis dimension i is calculated using the following formula:

[0071]

[0072] In the formula, P i This represents the probability of meter failure for the i-th supplier (or region or year of installation).

[0073] Based on the calculated energy meter failure probability P of the i-th dimension i This allows us to obtain the current failure probability of electricity meters in various regions and from various suppliers, understand the overall health status of electricity meters in operation, and provide decision support for electricity meter inventory management.

[0074] In step 2, the technical parameters of the operating energy meter obtained in step 1 and the corresponding multiple set operating durations are input into the constructed fault analysis model, and the fault classification probability of the operating energy meter under multiple set operating durations is output.

[0075] In step 3, based on the fault classification probability obtained in step 2, the ratio of the number of operating energy meters with a fault threshold of 0.9 or higher in the label 1 classification probability to the total number of operating energy meters is calculated as the fault probability of the energy meter. The operating time corresponding to the maximum fault probability is set as the fault reinstallation inflection point for a large number of energy meters to fail, and the fault probability of the operating energy meters corresponding to the fault reinstallation inflection point is calculated.

[0076] The failure probability of the in-operation energy meter corresponding to the failure reinstallation inflection point is calculated using the following formula:

[0077]

[0078] In the formula, P j P represents the probability of an in-operation energy meter failing when the operating time is j. j Z represents the number of operating energy meters whose classification probability is greater than the fault threshold when the operating time is j, and Z represents the total number of operating energy meters.

[0079] This invention provides a method for analyzing the probability of electricity meter failures. By predicting the probability of electricity meter failures in future time periods, the method obtains the proportion of electricity meter failures and can accurately predict the inflection point of large-scale replacement of electricity meters and the number of meters to be replaced. This solves the problem that existing technologies, when conducting electricity meter failure analysis based on electricity meter data, can only calculate the total failure rate of electricity meters based on data of electricity meters that have already failed. It is difficult to accurately predict the situation where currently operating electricity meters may have failed, the inflection point of large-scale replacement of electricity meters, and the corresponding failure probability. This reduces the probability of power grid disturbances and abnormal power output caused by the untimely replacement of faulty electricity meters, effectively reducing the disturbances and abnormal power output caused by the untimely replacement of faulty electricity meters, and is helpful for electricity meter inventory management and operation management.

[0080] Example 2

[0081] Based on the same inventive concept, this invention also provides a system for analyzing the probability of electricity meter failures, such as... Figure 2 As shown, it includes:

[0082] The acquisition module is used to acquire the technical parameters of the operating electricity meter and set multiple operating durations for the operating electricity meter;

[0083] The fault analysis module is used to input the technical parameters of the operating energy meter and multiple operating durations into a pre-built fault analysis model to obtain the fault classification probability of the operating energy meter under multiple set operating durations.

[0084] The module for determining the reinstallation inflection point and fault probability is used to take the operating time corresponding to the fault classification probability when the fault is greater than the fault threshold and the number of electricity meters is the largest as the fault reinstallation inflection point of the electricity meters in operation, and to calculate the fault probability of the electricity meters in operation at the fault reinstallation inflection point.

[0085] The fault analysis model is based on the decision tree algorithm and is constructed using the current technical parameters of the operating electricity meter and the technical parameters of historically faulty electricity meters.

[0086] The acquisition module includes a parameter acquisition submodule and a parameter setting submodule;

[0087] The parameter acquisition submodule is used to acquire the technical parameters of the currently operating electricity meters;

[0088] The technical parameters of the operating energy meters obtained by the parameter acquisition submodule include: region, supplier, operating time calculated using the freeze date and corresponding installation date of the operating energy meter, inventory time calculated using the warehousing date and corresponding installation date of the operating energy meter, and normal label.

[0089] In this embodiment, from the perspective of asset lifecycle, the longer an asset operates and the closer it is to retirement, the higher its failure probability. Therefore, the operating time attribute of an electricity meter is of great significance for predicting the failure probability of the electricity meter. The formula for calculating the operating time of an operating electricity meter is as follows:

[0090] L i =R-AR i

[0091] In the formula, L i AR represents the operating duration of the i-th operating energy meter, R represents the freeze date of the operating data, and AR represents the operating duration of the i-th operating energy meter. i This indicates the installation date of the i-th electricity meter.

[0092] The inventory duration of the electricity meters in operation is determined by the following formula:

[0093] K n =AR n -DR n

[0094] In the formula, K n AR represents the inventory duration of the nth electricity meter in operation. n Indicates the installation date of the nth electricity meter, DR nThis indicates the date the nth energy meter was put into storage.

[0095] Since the data for both the "operational duration" and "inventory duration" attributes should be greater than or equal to 0, the calculated operational duration and inventory duration of the operating energy meters are used... Outlier identification is performed in principle. Data less than 0 is identified as outliers and removed.

[0096] The parameter setting submodule is used to set multiple operating durations for the operating energy meter;

[0097] The parameter setting submodule is determined by adding the operating time of the operating energy meter to the time length of multiple setting changes;

[0098] The operating durations of the various operating energy meters are determined by the following formula:

[0099] ZY′ i,j =zy i +j (j=1,2,3…n)

[0100] In the formula, ZY′ i,j This indicates that the operating time of the i-th operating energy meter has changed to the operating time up to the j-th time period, zy i The current operating time of the i-th electricity meter is represented by , and the change time is represented by j.

[0101] Before using the fault analysis module to analyze the operating energy meters, a fault analysis model is constructed using the decision tree algorithm. The specific steps are as follows:

[0102] Obtain the technical parameters of historically faulty energy meters, including: region, supplier, in-service duration calculated using the fault date and corresponding installation date of the historically faulty energy meters, inventory duration calculated using the warehousing date and corresponding installation date of the historically faulty energy meters, and fault tag;

[0103] The operating time of historically faulty energy meters is calculated using the following formula:

[0104] L′ i =R i -AR′ i

[0105] In the formula, L′ i R represents the operating time of the i-th energy meter in the historical fault energy meters. i AR′ represents the failure date of the i-th energy meter in the historical fault energy meters. i This indicates the installation date of the i-th energy meter in the historical fault energy meters.

[0106] The inventory duration of historical faulty energy meters is determined by the following formula:

[0107] K h =AR h -DR h

[0108] In the formula, K h AR represents the inventory duration of the h-th historical faulty energy meter. h Indicates the installation date of the h-th historical faulty energy meter, DR h This indicates the date the h-th historical faulty energy meter was entered into the inventory.

[0109] The operating time and inventory time of historically faulty energy meters were also utilized. The principle is to identify outliers and then remove them.

[0110] Based on the decision tree algorithm, a wide table for decision tree analysis is formed by using the technical parameters of each in-operation energy meter and historically faulty energy meter according to the parameters of each energy meter under the attributes of region, supplier, in-operation duration, and inventory duration, and the information gain rate of each attribute is calculated accordingly.

[0111] The attribute corresponding to the maximum information gain ratio is taken as the root node, and after determining the root node, the information gain ratio corresponding to each attribute is repeatedly calculated to generate internal nodes according to the maximum information gain ratio obtained in each round, until the node whose information gain ratio remains unchanged is taken as the leaf node.

[0112] The fault analysis model is constructed based on the root node, internal nodes, and leaf nodes.

[0113] In this embodiment, the fault analysis model is constructed based on the C4.5 decision tree algorithm and can be applied to predict the fault probability of all operating energy meters. According to the fault labels contained in the technical parameters of operating energy meters and the technical parameters of historically faulty energy meters, energy meters labeled as normal are classified as label 0, and energy meters labeled as faulty are classified as label 1. The decision tree algorithm classification probability is output, and the data of operating energy meters and historically faulty energy meters with a probability greater than 0.9 in each node are selected from the classification probabilities.

[0114] For the analysis dimension i, it is determined by the following formula:

[0115]

[0116] In the formula, S i Z represents the number of operating and historically faulty energy meters corresponding to the probability that the i-th supplier (or region or installation year) is classified into label 1 is greater than 0.9. d represents the fault classification probability.i This represents the total number of electricity meters in operation from the i-th supplier (or region or year of installation);

[0117] The failure probability of the energy meter under analysis dimension i is calculated using the following formula:

[0118]

[0119] In the formula, P i This represents the probability of meter failure for the i-th supplier (or region or year of installation).

[0120] Based on the calculated energy meter failure probability P of the i-th dimension i This allows us to obtain the current failure probability of electricity meters in various regions and from various suppliers, understand the overall health status of electricity meters in operation, and provide decision support for electricity meter inventory management.

[0121] The fault analysis module is used to input the technical parameters of the operating energy meter obtained in step 1 and the corresponding multiple set operating durations into the constructed fault analysis model, and output the fault classification probability of the operating energy meter under multiple set operating durations.

[0122] The reinstallation inflection point and failure probability determination module includes: a failure inflection point determination submodule and a failure probability determination submodule;

[0123] The fault inflection point determination submodule is used to calculate the ratio of the number of operating energy meters that are classified into the label 1 category with a fault threshold of 0.9 or higher to the total number of operating energy meters based on the classification probability. The operating time corresponding to the maximum fault probability is set as the fault reinstallation inflection point when a large number of energy meters fail.

[0124] The fault probability determination submodule is used to calculate the fault probability of the in-operation energy meter corresponding to the fault reinstallation inflection point.

[0125] The failure probability of the in-operation energy meter corresponding to the failure reinstallation inflection point is calculated using the following formula:

[0126]

[0127] In the formula, P j P represents the probability of an in-operation energy meter failing when the operating time is j. j Z represents the number of operating energy meters whose classification probability is greater than the fault threshold when the operating time is j, and Z represents the total number of operating energy meters.

[0128] Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0129] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application 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, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0130] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. 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... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0131] 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.

[0132] 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.

[0133] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of the claims of the present invention pending approval.

Claims

1. A method for analyzing the probability of failure of an electricity meter, characterized in that, include: Obtain the technical parameters of the operating electricity meters and set multiple operating durations for the operating electricity meters; The technical parameters of the operating energy meter and multiple operating durations are input into a pre-built fault analysis model to obtain the fault classification probability of the operating energy meter under multiple set operating durations. The in-service duration corresponding to the fault classification probability when the number of electricity meters is greater than the fault threshold is taken as the fault reinstallation inflection point of the in-service electricity meters, and the fault probability corresponding to the in-service electricity meters at the fault reinstallation inflection point is calculated. The fault analysis model is based on the decision tree algorithm and is constructed using the current technical parameters of the operating energy meters and the technical parameters of historically faulty energy meters. The multiple set operating durations for the operating energy meters are determined by adding the operating duration of the operating energy meters to the time length of multiple set changes. They are subsequently used to predict the inflection point and failure probability of a large number of failures and reinstallations of operating energy meters in the next few years. The operating durations of the various operating energy meters are determined by the following formula: ZY′ i,j =zy i +j (j=1,2,3…n) In the formula, ZY′ i,j This indicates that the operating time of the i-th operating energy meter has changed to the operating time up to the j-th time period, zy i This represents the current operating duration of the i-th energy meter, and j represents the change duration. The failure probability of the in-operation energy meter corresponding to the failure reinstallation inflection point is calculated using the following formula: In the formula, P j P represents the probability of an in-operation energy meter failing when the operating time is j. j Z represents the number of operating energy meters whose classification probability is greater than the fault threshold when the operating time is j, and Z represents the total number of operating energy meters.

2. The method according to claim 1, characterized in that, The technical parameters of the operating energy meters and the historically faulty energy meters include: region, supplier, operating time calculated using the freeze date of the operating energy meters or the fault date of the historically faulty energy meters and the corresponding installation date, inventory time calculated using the warehousing date of the operating energy meters or the historically faulty energy meters and the corresponding installation date, and fault and normal labels.

3. The method according to claim 1, characterized in that, The construction of the fault analysis model includes: Based on the decision tree algorithm, the information gain rate of each attribute is calculated by using the parameters of each electricity meter under the attributes of region, supplier, in-service duration, and inventory duration respectively. The attribute corresponding to the maximum information gain ratio is taken as the root node, and after determining the root node, the information gain ratio corresponding to each attribute is repeatedly calculated to generate internal nodes according to the maximum information gain ratio obtained in each round, until the node whose information gain ratio remains unchanged is taken as the leaf node. The fault analysis model is constructed based on the root node, internal nodes, and leaf nodes.

4. The method according to claim 1, characterized in that, The step of setting the operating time corresponding to the fault classification probability when the number of electricity meters exceeds the fault threshold and is the largest as the fault reinstallation inflection point for the electricity meters in operation, and calculating the fault probability corresponding to the electricity meters in operation at the fault reinstallation inflection point, includes: All energy meters with a set running time corresponding to the fault threshold are regarded as fault energy meters under the fault reinstallation inflection point. The ratio of the number of faulty energy meters under a set operating time to the total number of all operating energy meters is taken as the fault probability of the operating energy meters reaching the fault reinstallation inflection point.

5. The method according to claim 1, characterized in that, The multiple set operating durations are determined by adding the current operating duration of the operating energy meter to the time length of multiple setting changes.

6. The method according to claim 2, characterized in that, Also includes: The technical parameters of each operating electricity meter and the technical parameters of the historically faulty electricity meters are obtained by using... The result is obtained after outlier removal according to the principle.

7. A system for analyzing the probability of electricity meter failure, characterized in that, include: The acquisition module is used to acquire the technical parameters of the operating electricity meter and set multiple operating durations for the operating electricity meter; The fault analysis module is used to input the technical parameters of the operating energy meter and multiple operating durations into a pre-built fault analysis model to obtain the fault classification probability of the operating energy meter under multiple set operating durations. The module for determining the reinstallation inflection point and fault probability is used to take the operating time corresponding to the fault classification probability when the fault is greater than the fault threshold and the number of electricity meters is the largest as the fault reinstallation inflection point of the electricity meters in operation, and to calculate the fault probability of the electricity meters in operation at the fault reinstallation inflection point. The fault analysis model is based on the decision tree algorithm and is constructed using the current technical parameters of the operating energy meters and the technical parameters of historically faulty energy meters. The multiple set operating durations for the operating energy meters are determined by adding the operating duration of the operating energy meters to the time length of multiple set changes. They are subsequently used to predict the inflection point and failure probability of a large number of failures and reinstallations of operating energy meters in the next few years. The operating durations of the various operating energy meters are determined by the following formula: ZY′ i,j =zy i +j (j=1,2,3…n) In the formula, ZY′ i,j This indicates that the operating time of the i-th operating energy meter has changed to the operating time up to the j-th time period, zy i This represents the current operating duration of the i-th energy meter, and j represents the change duration. The failure probability of the in-operation energy meter corresponding to the failure reinstallation inflection point is calculated using the following formula: In the formula, P j P represents the probability of an in-operation energy meter failing when the operating time is j. j Z represents the number of operating energy meters whose classification probability is greater than the fault threshold when the operating time is j, and Z represents the total number of operating energy meters.

8. The system according to claim 7, characterized in that, The technical parameters of the operating energy meters and the technical parameters of the historically faulty energy meters obtained by the acquisition module include: region, supplier, operating time calculated using the freeze date of the operating energy meters or the fault date of the historically faulty energy meters and the corresponding installation date, inventory time calculated using the warehousing date of the operating energy meters or the historically faulty energy meters and the corresponding installation date, and fault and normal labels.

9. The system according to claim 7, characterized in that, The construction of the fault analysis model includes: Based on the decision tree algorithm, the information gain rate of each attribute is calculated by using the parameters of each electricity meter under the attributes of region, supplier, in-service duration, and inventory duration respectively. The attribute corresponding to the maximum information gain ratio is taken as the root node, and after determining the root node, the information gain ratio corresponding to each attribute is repeatedly calculated to generate internal nodes according to the maximum information gain ratio obtained in each round, until the node whose information gain ratio remains unchanged is taken as the leaf node. The fault analysis model is constructed based on the root node, internal nodes, and leaf nodes.

10. The system according to claim 7, characterized in that, The reinstallation inflection point and failure probability determination module includes: The reinstallation inflection point determination submodule is used to identify all energy meters with a set running time corresponding to the fault threshold as fault energy meters under the fault reinstallation inflection point. The fault probability determination submodule is used to determine the fault probability of an operating energy meter reaching the fault reinstallation inflection point by using the ratio of the number of faulty energy meters to the total number of all operating energy meters under a set operating time.

Images