Gateway electric energy meter health evaluation and intelligent operation and maintenance-based method and device, electronic equipment and storage medium

Abstract

Performing abnormal value and noise data removal on the operation data of the gateway electric energy meter, and then preprocessing the denoised data to generate a corresponding standard transformation matrix; calculating correlation factors among the data of the gateway electric energy meters in the standard transformation matrix, and constructing a correlation factor matrix; performing characteristic decomposition on the correlation factor matrix through a characteristic equation, obtaining characteristics of correlation factors and corresponding characteristic values, calculating contribution degrees of the correlation factors in the correlation factor matrix according to the characteristic values, and determining the fault characteristic set through the contribution degrees; and carrying out relation matching on the fault feature set, calculating the probability of the fault feature set corresponding to different types of fault modes after the relation matching, and judging a final fault type through a probability density function.

Description

Technical Field

[0001] The present invention relates to the technical field of fault diagnosis in electric energy metering devices, and in particular to a method, device, electronic equipment, and storage medium based on health evaluation and intelligent operation and maintenance of gateway electric energy meters. Background Art

[0002] At present, the operation and maintenance of gateway electric energy meters is mainly carried out by professionals carrying calibration instruments and equipment to conduct regular on-site inspections. However, practice has shown that with this operation and maintenance management model, gateway electric energy meters that are operating normally are often "checked without any problems", resulting in a waste of manpower and material resources for electric energy metering management; at the same time, although some gateway electric energy meters have not yet reached the calibration time, due to the lack of accurate and effective supervision methods, the hidden faults that have occurred during their operation have not been discovered in time, and they have not been calibrated and maintained in time, resulting in electric energy metering errors and even the loss of electric energy metering results. Summary of the invention

[0003] In order to improve the above problems, the present invention proposes a method based on health evaluation and intelligent operation and maintenance of gateway electric energy meters, which includes the following steps:

[0004] Step 1: using data acquisition equipment, regularly collect historical energy metering data of multiple gateway energy meters in the past period of time, exemplary historical energy metering data includes on-site voltage, current, on-site temperature and humidity data, and historical energy metering data information such as the number of active energy pulses, current active power, reactive power, power factor, etc. measured by the gateway energy meters under inspection;

[0005] Step 2: Remove outliers and noise data from the collected historical data, then perform preprocessing, and establish a verification and evaluation model for gateway electric energy meters through multi-source data feature fusion and probability judgment;

[0006] In the process of establishing the verification and evaluation model of the gateway electric energy meter, the voltage, current, on-site temperature and humidity data in the multi-source historical data, as well as the number of active electric energy pulses, current active power, reactive power and power factor measured by the gateway electric energy meter under inspection are preprocessed to generate the corresponding standard feature matrix;

[0007] In the preprocessing process, the historical data collected by each gateway electric energy meter is linearly transformed to obtain the corresponding initial sample matrix X mn , that is, the initial sample matrix X is obtained by the following formula mn ;

[0008] X mn =(x ij ) mn

[0009] Among them, xij Represents the initial sample matrix X mn The historical data of each gateway electric energy meter in, 1≤i≤m, 1≤j≤n, m, n represent the initial sample matrix X mn Dimensions;

[0010] The historical data x of each gateway electric energy meter ij Perform standard transformation to obtain the historical data of each gateway electric energy meter after standard transformation And according to Create a standard transformation matrix Among them, x is calculated by the following formula ij Perform standard transformation to obtain

[0011]

[0012] Among them, s j Denotes the initial sample matrix X mn Medium ij The standard value of the jth column of Represents the initial sample matrix X mn Medium ij The average value of the jth column of ;

[0013] By the following formula Create a standard transformation matrix Right now

[0014]

[0015] Among them, 1≤i≤m, 1≤j≤n, m and n represent the standard transformation matrix Dimensions;

[0016] The standard transformation matrix is ​​calculated by the following formula The historical data of each gateway electric energy meter after the standard conversion The correlation factor c ij , establish the correlation factor matrix C mn ;Right now

[0017]

[0018] in, Represents the standard transformation matrix Any two data in Represents the standard transformation matrix The average value of the i-th row, Represents the standard transformation matrix The mean of the jth column;

[0019] The correlation factor matrix C is calculated by the characteristic equation mn Perform eigendecomposition and obtain c ij The characteristic λ ij and the corresponding eigenvalue p ij , calculate the correlation factor matrix C according to the eigenvalue p mn Medium ij Contribution G ij ; The contribution G is calculated by the following formula ij ,Right now

[0020]

[0021] The eigenvalue p ij The larger the value, the greater the correlation factor matrix c ij The larger the coverage length in the corresponding characteristic direction, the corresponding contribution G is obtained. ij The larger the value is, the further data c with contribution greater than the preset threshold is selected. ij Construct the fault feature set Y; that is, Y = [y1, y2, y3.., y z ], where 1≤z≤n;

[0022] Furthermore, the fault feature set Y is matched with each other and the following formula is used for relationship matching:

[0023]

[0024] Among them, W α Represents the eigenvector corresponding to the fault mode, 1≤α≤n, represents the smoothing factor;

[0025] The probability of the fault feature set corresponding to different types of fault modes after relationship matching is calculated by the following formula:

[0026]

[0027] Among them, f a represents the probability of the failure mode, ρ represents the dimension of the failure mode sample, represents the smoothing factor;

[0028] The final fault type T is determined by the probability density function, that is, T = argmax (f a ); the final fault type is obtained through the fault probability. Then, the training of the verification and evaluation model of the gateway electric energy meter is completed by continuously iterating the data;

[0029] Step 3: Obtain the operation data of the gateway electric energy meter in real time and transmit the data to the data processing center;

[0030] Step 4: inputting the operation data of the gateway electric energy meter into the trained verification and evaluation model of the gateway electric energy meter;

[0031] Step 5: Provide feedback and warning on the fault type, and send a warning prompt to the client through the cloud server;

[0032] The cloud server pushes maintenance suggestions from an expert database according to the fault type, and sends the maintenance suggestions to a handheld device of the maintenance personnel.

[0033] Furthermore, the gateway electric energy meter in the data acquisition device includes voltage and current transformers used in different types of electric energy meters, and their secondary circuit electric energy metering cabinets, wherein the electric energy meter is divided into direct access and access via transformer according to the access form.

[0034] Furthermore, when collecting historical electricity metering data, a timed automatic collection method is pre-set to trigger the collection of electric energy data and non-electricity related data, wherein the sampling interval is adjusted according to the load size during the current time period to collect the voltage and current on site at a high frequency.

[0035] Furthermore, the faulty operating conditions of the gateway electricity meter include secondary circuit miswiring, pressure loss fault, collection error, appearance damage, meter stoppage, error out of tolerance, clock abnormality, power abnormality, display fault, interface damage, and parameter error.

[0036] In addition, the present invention also proposes a device based on health evaluation and intelligent operation and maintenance of gateway electric energy meters, which includes the following modules:

[0037] A historical data collection module is used to use data collection equipment to regularly collect historical energy metering data of multiple gateway energy meters in the past period of time. Exemplary historical energy metering data include on-site voltage, current, on-site temperature and humidity data, as well as historical energy metering data information such as the number of active energy pulses, current active power, reactive power, power factor, etc. measured by the gateway energy meters under inspection;

[0038] The gateway electric energy meter calibration and evaluation model training module is used to remove outliers and noise data from the collected historical data, and then perform preprocessing, and establish the gateway electric energy meter calibration and evaluation model through multi-source data feature fusion and probability judgment;

[0039] In the process of establishing the verification and evaluation model of the gateway electric energy meter, the voltage, current, on-site temperature and humidity data in the multi-source historical data, as well as the number of active electric energy pulses, current active power, reactive power and power factor measured by the gateway electric energy meter under inspection are preprocessed to generate the corresponding standard feature matrix;

[0040] In the preprocessing process, the historical data collected by each gateway electric energy meter is linearly transformed to obtain the corresponding initial sample matrix X mn , that is, the initial sample matrix X is obtained by the following formula mn ;

[0041] X mn =(x ij ) mn

[0042] Among them, x ij Represents the initial sample matrix X mn The historical data of each gateway electric energy meter in, 1≤i≤m, 1≤j≤n, m, n represent the initial sample matrix X mn Dimensions;

[0043] The historical data x of each gateway electric energy meter ij Perform standard transformation to obtain the historical data of each gateway electric energy meter after standard transformation And according to Create a standard transformation matrix Among them, x is calculated by the following formula ij Perform standard transformation to obtain

[0044]

[0045] Among them, s j Denotes the initial sample matrix X mn Medium ij The standard value of the jth column of Represents the initial sample matrix X mn Medium ij The average value of the jth column of ;

[0046] By the following formula Create a standard transformation matrix Right now

[0047]

[0048] Among them, 1≤i≤m, 1≤j≤n, m and n represent the standard transformation matrix Dimensions;

[0049] The standard transformation matrix is ​​calculated by the following formula The historical data of each gateway electric energy meter after the standard conversion The correlation factor c ij , establish the correlation factor matrix C mn ;Right now

[0050]

[0051] in, Represents the standard transformation matrix Any two data in Represents the standard transformation matrix The average value of the i-th row, Represents the standard transformation matrix The mean of the jth column;

[0052] The correlation factor matrix C is calculated by the characteristic equation mn Perform eigendecomposition and obtain c ij The characteristic λ ij and the corresponding eigenvalue p ij , calculate the correlation factor matrix C according to the eigenvalue o mn Medium ij Contribution G ij ; The contribution G is calculated by the following formula ij ,Right now

[0053]

[0054] The eigenvalue p ij The larger the value, the greater the correlation factor matrix c ij The larger the coverage length in the corresponding characteristic direction, the corresponding contribution G is obtained. ij The larger the value is, the further data c with contribution greater than the preset threshold is selected. ij Construct the fault feature set Y; that is, Y = [y1, y2, y3.., y z ], where 1≤z≤n;

[0055] Furthermore, the fault feature set Y is matched with each other and the following formula is used for relationship matching:

[0056]

[0057] Among them, W α Represents the eigenvector corresponding to the fault mode, 1≤α≤n, represents the smoothing factor;

[0058] The probability of the fault feature set corresponding to different types of fault modes after relationship matching is calculated by the following formula:

[0059]

[0060] Among them, f a represents the probability of the failure mode, ρ represents the dimension of the failure mode sample, represents the smoothing factor;

[0061] The final fault type T is determined by the probability density function, that is, T = argmax (f a ); the final fault type is obtained through the fault probability. Then, the training of the verification and evaluation model of the gateway electric energy meter is completed by continuously iterating the data;

[0062] Real-time data acquisition module, used to obtain the operation data of the gateway electric energy meter in real time and transmit the data to the data processing center;

[0063] A data analysis module, used for inputting the operation data of the gateway electric energy meter into the trained verification and evaluation model of the gateway electric energy meter;

[0064] The early warning and operation and maintenance module provides feedback and early warning for the fault type, and sends early warning prompts to the client through the cloud server;

[0065] The cloud server pushes maintenance suggestions from an expert database according to the fault type, and sends the maintenance suggestions to a handheld device of the maintenance personnel.

[0066] Furthermore, the gateway electric energy meter in the data acquisition device includes voltage and current transformers used in different types of electric energy meters, and their secondary circuit electric energy metering cabinets, wherein the electric energy meter is divided into direct access and access via transformer according to the access form.

[0067] Furthermore, when collecting historical electricity metering data, a timed automatic collection method is pre-set to trigger the collection of electric energy data and non-electricity related data, wherein the sampling interval is adjusted according to the load size during the current time period to collect the voltage and current on site at a high frequency.

[0068] Furthermore, the faulty operating conditions of the gateway electricity meter include secondary circuit miswiring, pressure loss fault, collection error, appearance damage, meter stoppage, error out of tolerance, clock abnormality, power abnormality, display fault, interface damage, and parameter error.

[0069] The present invention also proposes an electronic device, which includes a processor and a memory: the memory is used to store program code and transmit the program code to the processor, and the processor is used to execute the above-mentioned method based on gateway electricity meter health evaluation and intelligent operation and maintenance according to the instructions in the program code.

[0070] The present invention also proposes a computer-readable storage medium, which is used to store program code and transmit the program code to a processor, and the processor is used to execute the above-mentioned method based on gateway electricity meter health evaluation and intelligent operation and maintenance according to the instructions in the program code.

[0071] Therefore, the present invention can remove abnormal values ​​and noise data from the operating data of the gateway electric energy meter, and then pre-process the denoised data to generate a corresponding standard transformation matrix; calculate the correlation factors between the various data of the gateway electric energy meters in the standard transformation matrix, and construct a correlation factor matrix; perform feature decomposition on the correlation factor matrix through the characteristic equation, and obtain the characteristics of the correlation factors and the corresponding eigenvalues, calculate the contribution of the correlation factors in the correlation factor matrix according to the eigenvalues, and determine the fault feature set through the contribution; perform relationship matching on the fault feature set, and calculate the probability of the fault feature set corresponding to different types of fault modes after relationship matching, and then determine the final fault type through the probability density function. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0073] Figure 1 It is a method flow chart based on health evaluation and intelligent operation and maintenance of gateway electric energy meters;

[0074] Figure 2 It is a module diagram based on health evaluation and intelligent operation and maintenance of gateway electricity meters. DETAILED DESCRIPTION

[0075] like Figure 1 As shown in the figure, a method for health evaluation and intelligent operation and maintenance of gateway electric energy meters is provided, and the specific steps are as follows:

[0076] Step 1: Use data acquisition equipment to regularly collect historical electricity metering data from multiple gateway electricity meters over a period of time. Exemplary historical electricity metering data include on-site voltage, current, on-site temperature and humidity data, as well as historical electricity metering data information such as the number of active energy pulses, current active power, reactive power, power factor, etc. measured by the gateway electricity meters under inspection.

[0077] Specifically, the gateway electric energy meter in the data acquisition device includes voltage and current transformers used in different types of electric energy meters, and their secondary circuit electric energy metering cabinets, wherein the electric energy meter is divided into direct access and access via transformer according to the access form.

[0078] Furthermore, when collecting historical electricity metering data, a scheduled automatic collection is pre-set, exemplarily by default once an hour, or continuous collection is performed at a fixed time, or the collection of electricity energy data and non-electricity related data is triggered according to the agreed load point method, which mainly includes high-frequency collection within the current time period, exemplarily by default the high-frequency collection time is 1 minute, and the sampling interval can also be adjusted according to the load size to collect the voltage and current on site. Similarly, the data collection equipment also needs to collect the number of electric energy pulses output by the inspected gateway electricity meter, as well as the surrounding environmental data such as temperature and humidity.

[0079] Furthermore, the voltage, current, on-site temperature and humidity data of each gateway electric energy meter in the collected historical data, as well as the number of active energy pulses, current active power, reactive power, and power factor measured by the detected gateway electric energy meter are analyzed for operation and maintenance data, and it is determined that the gateway electric energy meter is in a faulty operation state. Exemplarily, the faulty operation state of the gateway electric energy meter includes secondary circuit miswiring, pressure loss fault, collection error, appearance damage, meter stoppage, error out of tolerance, clock abnormality, power abnormality, display fault, interface damage, parameter error, etc.

[0080] Step 2: Remove outliers and noise data from the collected historical data, then perform preprocessing, and establish a verification and evaluation model for gateway electric energy meters through multi-source data feature fusion and probability judgment;

[0081] Specifically, in the process of establishing the verification and evaluation model of the gateway electric energy meter, the corresponding standard feature matrix is ​​generated by preprocessing the voltage, current, on-site temperature and humidity data in the multi-source historical data, as well as the number of active electric energy pulses, current active power, reactive power and power factor measured by the inspected gateway electric energy meter.

[0082] Specifically, in the preprocessing process, the historical data collected by each gateway electric energy meter is linearly transformed to obtain the corresponding initial sample matrix X mn , that is, the initial sample matrix X is obtained by the following formula mn ;

[0083] X mn =(x ij ) mn

[0084] Among them, x ij Represents the initial sample matrix X mn The historical data of each gateway electric energy meter in, 1≤i≤m, 1≤j≤n, m, n represent the initial sample matrix X mn Dimension.

[0085] Furthermore, each historical data x of each gateway electric energy meterij Perform standard transformation to obtain the historical data of each gateway electric energy meter after standard transformation And according to Create a standard transformation matrix Among them, x is calculated by the following formula ij Perform standard transformation to obtain

[0086]

[0087] Among them, s j Denotes the initial sample matrix X mn Medium ij The standard value of the jth column of Represents the initial sample matrix X mn Medium ij The average value of the jth column of ;

[0088] Further, through the following formula Create a standard transformation matrix Right now

[0089]

[0090] Among them, 1≤i≤m, 1≤j≤n, m and n represent the standard transformation matrix Dimensions;

[0091] Furthermore, the standard transformation matrix is ​​calculated by the following formula: The historical data of each gateway electric energy meter after the standard conversion The correlation factor c ij , establish the correlation factor matrix C mn ;Right now

[0092]

[0093] in, Represents the standard transformation matrix Any two data in Represents the standard transformation matrix The average value of the i-th row, Represents the standard transformation matrix The mean of the jth column;

[0094] Further, the correlation factor matrix C is transformed into mn Perform eigendecomposition and obtain c ij The characteristic λ ij and the corresponding eigenvalue p ij , calculate the correlation factor matrix C according to the eigenvalue p mn Mediumij Contribution G ij ; The contribution G is calculated by the following formula ij ,Right now

[0095]

[0096] Specifically, the eigenvalue p ij The larger the value, the greater the correlation factor matrix c ij The larger the coverage length in the corresponding characteristic direction, the corresponding contribution G is obtained. ij The larger the value is, the further data c with contribution greater than the preset threshold is selected. ij Construct the fault feature set Y; that is, Y = [y1, y2, y3.., y z ], where 1≤z≤n;

[0097] Furthermore, the fault feature set Y is matched with each other and the following formula is used for relationship matching:

[0098]

[0099] Among them, W α Represents the eigenvector corresponding to the fault mode, 1≤α≤n, represents the smoothing factor;

[0100] Furthermore, the probability of the fault feature set corresponding to different types of fault modes after relationship matching is calculated by the following formula:

[0101]

[0102] Among them, f a represents the probability of the failure mode, ρ represents the dimension of the failure mode sample, represents the smoothing factor;

[0103] Furthermore, the final fault type T is determined by the probability density function, that is, T = argmax(f a ); the final fault type is obtained through the fault probability. Then, the training of the verification and evaluation model of the gateway electric energy meter is completed by continuously iterating the data.

[0104] Step 3: Obtain the operation data of the gateway electric energy meter in real time and transmit the data to the data processing center;

[0105] Specifically, the real-time acquisition of the operation data of the gateway electric energy meter includes the voltage, current, on-site temperature and humidity data of each gateway electric energy meter, as well as the number of active electric energy pulses, current active power, reactive power, and power factor measured by the detected gateway electric energy meter, and the operation data of the gateway electric energy meter is stored in a preset database;

[0106] Step 4: inputting the operation data of the gateway electric energy meter into the trained verification and evaluation model of the gateway electric energy meter;

[0107] Specifically, outliers and noise data are removed from the operation data of the gateway electric energy meter, and then the denoised data is preprocessed; in the preprocessing process, the denoised data, i.e., the voltage, current, on-site temperature and humidity data, and the number of active electric energy pulses, current active power, reactive power, and power factor measured by the inspected gateway electric energy meter are preprocessed, and in the preprocessing process, the voltage, current, on-site temperature and humidity data, and the initial sample matrix composed of the number of active electric energy pulses, current active power, reactive power, and power factor measured by the inspected gateway electric energy meter are subjected to standard transformation to generate a corresponding standard transformation matrix;

[0108] Further, the correlation factors between the data of the electric energy meters at each gateway in the standard transformation matrix are calculated, and a correlation factor matrix is ​​constructed;

[0109] Further, the correlation factor matrix is ​​eigen-decomposed by a characteristic equation, and the characteristics and corresponding eigenvalues ​​of the correlation factors are obtained, the contribution of the correlation factors in the correlation factor matrix is ​​calculated according to the eigenvalues, and the fault feature set is determined by the contribution;

[0110] Further, relationship matching is performed on the fault feature set, and the probability of the fault feature set corresponding to different types of fault modes after relationship matching is calculated, and then the final fault type is determined by a probability density function;

[0111] Step 5: Provide feedback and warning on the fault type, and send a warning prompt to the client through the cloud server; the cloud server pushes maintenance suggestions from the expert database according to the fault type, and sends the maintenance suggestions to the maintenance personnel's handheld device.

[0112] Example 2

[0113] like Figure 2 As shown, a device based on health evaluation and intelligent operation and maintenance of gateway electric energy meters includes:

[0114] The historical data collection module is used to use data collection equipment to regularly collect historical electricity metering data of multiple gateway electricity meters in the past period of time. Exemplary historical electricity metering data include on-site voltage, current, on-site temperature and humidity data, as well as historical electricity metering data information such as the number of active energy pulses, current active power, reactive power, power factor, etc. measured by the inspected gateway electricity meters.

[0115] Specifically, the gateway electric energy meter in the data acquisition device includes voltage and current transformers used in different types of electric energy meters, and their secondary circuit electric energy metering cabinets, wherein the electric energy meter is divided into direct access and access via transformer according to the access form.

[0116] Furthermore, when collecting historical electricity metering data, a scheduled automatic collection is pre-set, exemplarily by default once an hour, or continuous collection is performed at a fixed time, or the collection of electricity energy data and non-electricity related data is triggered according to the agreed load point method, which mainly includes high-frequency collection within the current time period, exemplarily by default the high-frequency collection time is 1 minute, and the sampling interval can also be adjusted according to the load size to collect the voltage and current on site. Similarly, the data collection equipment also needs to collect the number of electric energy pulses output by the inspected gateway electricity meter, as well as the surrounding environmental data such as temperature and humidity.

[0117] Furthermore, the voltage, current, on-site temperature and humidity data of each gateway electric energy meter in the collected historical data, as well as the number of active energy pulses, current active power, reactive power, and power factor measured by the detected gateway electric energy meter are analyzed for operation and maintenance data, and it is determined that the gateway electric energy meter is in a faulty operation state. Exemplarily, the faulty operation state of the gateway electric energy meter includes secondary circuit miswiring, pressure loss fault, collection error, appearance damage, meter stoppage, error out of tolerance, clock abnormality, power abnormality, display fault, interface damage, parameter error, etc.

[0118] The verification and evaluation model training module of the gateway electric energy meter is used to remove outliers and noise data from the collected historical data, and then pre-process it, and establish the verification and evaluation model of the gateway electric energy meter through multi-source data feature fusion and probability judgment.

[0119] Specifically, in the process of establishing the verification and evaluation model of the gateway electric energy meter, the corresponding standard feature matrix is ​​generated by preprocessing the voltage, current, on-site temperature and humidity data in the multi-source historical data, as well as the number of active electric energy pulses, current active power, reactive power and power factor measured by the inspected gateway electric energy meter.

[0120] Specifically, in the preprocessing process, the historical data collected by each gateway electric energy meter is linearly transformed to obtain the corresponding initial sample matrix X mn , that is, the initial sample matrix X is obtained by the following formula mn ;

[0121] X mn =(x ij ) mn

[0122] Among them, x ijRepresents the initial sample matrix X mn The historical data of each gateway electric energy meter in, 1≤i≤m, 1≤j≤n, m, n represent the initial sample matrix X mn Dimension.

[0123] Furthermore, each historical data x of each gateway electric energy meter ij Perform standard transformation to obtain the historical data of each gateway electric energy meter after standard transformation And according to Create a standard transformation matrix Among them, x is calculated by the following formula ij Perform standard transformation to obtain

[0124]

[0125] Among them, s j Denotes the initial sample matrix X mn Medium ij The standard value of the jth column of Represents the initial sample matrix X mn Medium ij The average value of the jth column of ;

[0126] Further, through the following formula Create a standard transformation matrix Right now

[0127]

[0128] Among them, 1≤i≤m, 1≤j≤n, m and n represent the standard transformation matrix Dimensions;

[0129] Furthermore, the standard transformation matrix is ​​calculated by the following formula: The historical data of each gateway electric energy meter after the standard conversion The correlation factor c ij , establish the correlation factor matrix C mn ;Right now

[0130]

[0131] in, Represents the standard transformation matrix Any two data in Represents the standard transformation matrix The average value of the i-th row, Represents the standard transformation matrix The mean of the jth column;

[0132] Further, the correlation factor matrix C is calculated by the characteristic equation mn Perform eigendecomposition and obtain c ij The characteristic λ ij and the corresponding eigenvalue p ij , calculate the correlation factor matrix c according to the eigenvalue p mn Medium ij Contribution G ij ; The contribution G is calculated by the following formula ij ,Right now

[0133]

[0134] Specifically, the eigenvalue p ij The larger the value, the greater the correlation factor matrix c ij The larger the coverage length in the corresponding characteristic direction, the corresponding contribution G is obtained. ij The larger the value is, the further data c with contribution greater than the preset threshold is selected. ij Construct the fault feature set Y; that is, Y = [y1, y2, y3.., y z ], where 1≤z≤n;

[0135] Furthermore, the fault feature set Y is matched with each other and the following formula is used for relationship matching:

[0136]

[0137] Among them, W α Represents the eigenvector corresponding to the fault mode, 1≤α≤n, represents the smoothing factor;

[0138] Furthermore, the probability of the fault feature set corresponding to different types of fault modes after relationship matching is calculated by the following formula:

[0139]

[0140] Among them, f a represents the probability of the failure mode, ρ represents the dimension of the failure mode sample, represents the smoothing factor;

[0141] Furthermore, the final fault type T is determined by the probability density function, that is, T = argmax(f a ); the final fault type is obtained through the fault probability. Then, the training of the verification and evaluation model of the gateway electric energy meter is completed by continuously iterating the data.

[0142] The real-time data acquisition module is used to obtain the operating data of the gateway electric energy meter in real time and transmit the data to the data processing center.

[0143] Specifically, the real-time acquisition of the operation data of the gateway electric energy meter includes the voltage, current, on-site temperature and humidity data of each gateway electric energy meter, as well as the number of active electric energy pulses, current active power, reactive power, and power factor measured by the detected gateway electric energy meter, and the operation data of the gateway electric energy meter is stored in a preset database;

[0144] The data analysis module is used to input the operation data of the gateway electric energy meter into the verification and evaluation model of the trained gateway electric energy meter.

[0145] Specifically, outliers and noise data are removed from the operation data of the gateway electric energy meter, and then the denoised data is preprocessed; in the preprocessing process, the denoised data, i.e., the voltage, current, on-site temperature and humidity data, and the number of active electric energy pulses, current active power, reactive power, and power factor measured by the inspected gateway electric energy meter are preprocessed, and in the preprocessing process, the voltage, current, on-site temperature and humidity data, and the initial sample matrix composed of the number of active electric energy pulses, current active power, reactive power, and power factor measured by the inspected gateway electric energy meter are subjected to standard transformation to generate a corresponding standard transformation matrix;

[0146] Further, the correlation factors between the data of the electric energy meters at each gateway in the standard transformation matrix are calculated, and a correlation factor matrix is ​​constructed;

[0147] Further, the correlation factor matrix is ​​eigen-decomposed by a characteristic equation, and the characteristics and corresponding eigenvalues ​​of the correlation factors are obtained, the contribution of the correlation factors in the correlation factor matrix is ​​calculated according to the eigenvalues, and the fault feature set is determined by the contribution;

[0148] Furthermore, relationship matching is performed on the fault feature set, and the probability that the fault feature set corresponds to different types of fault modes after relationship matching is calculated, and then the final fault type is determined by a probability density function.

[0149] The early warning and operation and maintenance module is used to provide feedback and early warning on the fault type, and send early warning prompts to the client through the cloud server; the cloud server pushes maintenance suggestions from the expert database according to the fault type, and sends the maintenance suggestions to the maintenance personnel's handheld device.

[0150] The present invention also relates to an electronic device, characterized in that the electronic device includes a processor and a memory: the memory is used to store program code and transmit the program code to the processor, and the processor is used to execute a method based on gateway electricity meter health evaluation and intelligent operation and maintenance described in the above method embodiment according to instructions in the program code.

[0151] The present invention also relates to a computer-readable storage medium, characterized in that the computer-readable storage medium is used to store program code and transmit the program code to a processor, and the processor is used to execute a method based on gateway electricity meter health evaluation and intelligent operation and maintenance described in the above method embodiment according to instructions in the program code.

[0152] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other to form a technical solution.

Claims

1. A method for health evaluation and intelligent operation and maintenance of gateway electric energy meters, characterized in that The following steps are involved: Step 1: using data acquisition equipment to regularly collect historical energy metering data of multiple gateway energy meters, the historical energy metering data includes on-site voltage, current, on-site temperature and humidity data, and the number of active energy pulses, current active power, reactive power, and power factor historical energy metering data information measured by the gateway energy meters under inspection; Step 2: Remove outliers and noise data from the collected historical data, then perform preprocessing, and establish a verification and evaluation model for gateway electric energy meters through multi-source data feature fusion and probability judgment; In the process of establishing the verification and evaluation model of the gateway electric energy meter, the voltage, current, on-site temperature and humidity data in the multi-source historical data, as well as the number of active electric energy pulses, current active power, reactive power and power factor measured by the gateway electric energy meter under inspection are preprocessed to generate the corresponding standard feature matrix; In the preprocessing process, the historical data collected by each gateway electric energy meter is linearly transformed to obtain the corresponding initial sample matrix X mn , that is, the initial sample matrix X is obtained by the following formula mn ; X mn =(x ij ) mn Among them, x ij Represents the initial sample matrix X mn The historical data of each gateway electric energy meter in, 1≤i≤m, 1≤j≤n, m, n represent the initial sample matrix X mn Dimensions; The historical data of each gateway electric energy meter x ij Perform standard transformation to obtain the historical data of each gateway electric energy meter after standard transformation And according to Create a standard transformation matrix Among them, x is calculated by the following formula ij Perform standard transformation to obtain Among them, s j Represents the initial sample matrix X mn Medium ij The standard value of the jth column of Represents the initial sample matrix X mn Medium ij The average value of the jth column of ; By the following formula Create a standard transformation matrix Right now Among them, 1ψiψm, 1ψjψn, m and n represent the standard transformation matrix Dimensions; The standard transformation matrix is ​​calculated by the following formula The historical data of each gateway electric energy meter after the standard conversion The correlation factor c ij , establish the correlation factor matrix C mn ;Right now in, Represents the standard transformation matrix Any two data in Represents the standard transformation matrix The average value of the i-th row, Represents the standard transformation matrix The mean of the jth column; The correlation factor matrix C is calculated by the characteristic equation mn Perform eigendecomposition and obtain c ij The characteristic λ ij and the corresponding eigenvalue p ij , calculate the correlation factor matrix c according to the eigenvalue p mn Medium ij Contribution G ij ; The contribution G is calculated by the following formula ij ,Right now The eigenvalue p ij The larger the value, the greater the correlation factor matrix c ij The larger the coverage length in the corresponding characteristic direction, the corresponding contribution G is obtained. i The larger j is, the further data c with contribution greater than the preset threshold is selected. ij Construct the fault feature set Y; that is, Y = [y1, y2, y3.., y z ], where 1≤z≤n; Furthermore, the fault feature set Y is matched with each other and the following formula is used for relationship matching: Among them, W α Represents the eigenvector corresponding to the fault mode, 1≤α≤n, represents the smoothing factor; The probability of the fault feature set corresponding to different types of fault modes after relationship matching is calculated by the following formula: Among them, f a represents the probability of the failure mode, ρ represents the dimension of the failure mode sample, represents the smoothing factor; The final fault type T is determined by the probability density function, that is, T = argmax (f a ); thereby obtaining the final fault type through the fault probability, and then completing the training of the verification and evaluation model of the gateway electric energy meter by continuously iterating the data; Step 3: Obtain the operation data of the gateway electric energy meter in real time and transmit the data to the data processing center; Step 4: inputting the operation data of the gateway electric energy meter into the trained verification and evaluation model of the gateway electric energy meter; Step 5: Provide feedback and warning on the fault type, and send a warning prompt to the client through the cloud server; The cloud server pushes maintenance suggestions from an expert database according to the fault type, and sends the maintenance suggestions to a handheld device of the maintenance personnel.

2. A method for health assessment and intelligent operation and maintenance of gateway electric energy meters according to claim 1, characterized in that: The gateway electric energy meter in the data acquisition device includes voltage and current transformers used in different types of electric energy meters, and their secondary circuit electric energy metering cabinets, wherein the electric energy meter is divided into direct access and access via transformer according to the access form.

3. A method for health assessment and intelligent operation and maintenance of gateway electric energy meters according to claim 2, characterized in that: When collecting historical electricity metering data, a timed automatic collection method is pre-set to trigger the collection of electric energy data and non-electricity related data. Among them, the sampling interval is adjusted according to the load size during the current time period to collect the voltage and current on site at a high frequency.

4. The method for health assessment and intelligent operation and maintenance of gateway electric energy meters according to claim 1 is characterized in that: The faulty operating conditions of the gateway electricity meter include secondary circuit miswiring, pressure loss fault, collection error, appearance damage, meter stoppage, error out of tolerance, clock abnormality, power abnormality, display failure, interface damage, and parameter error.

5. A device based on health evaluation and intelligent operation and maintenance of gateway electric energy meters, characterized in that The device includes the following modules: The historical data collection module is used to collect historical energy metering data of multiple gateway energy meters in the past period of time by using data collection equipment. Exemplary historical energy metering data include on-site voltage, current, on-site temperature and humidity data, and active energy pulse number, current active power, reactive power, and power factor historical energy metering data information measured by the gateway energy meters under inspection; The gateway electric energy meter calibration and evaluation model training module is used to remove outliers and noise data from the collected historical data, and then perform preprocessing, and establish the gateway electric energy meter calibration and evaluation model through multi-source data feature fusion and probability judgment; In the process of establishing the verification and evaluation model of the gateway electric energy meter, the voltage, current, on-site temperature and humidity data in the multi-source historical data, as well as the number of active electric energy pulses, current active power, reactive power and power factor measured by the gateway electric energy meter under inspection are preprocessed to generate the corresponding standard feature matrix; In the preprocessing process, the historical data collected by each gateway electric energy meter is linearly transformed to obtain the corresponding initial sample matrix X mn , that is, the initial sample matrix X is obtained by the following formula mn ; X mn =(x ij ) mn Among them, x ij Represents the initial sample matrix X mn The historical data of each gateway electric energy meter in, 1≤i≤m, 1≤j≤n, m, n represent the initial sample matrix X mn Dimensions; The historical data of each gateway electric energy meter x ij Perform standard transformation to obtain the historical data of each gateway electric energy meter after standard transformation And according to Create a standard transformation matrix Among them, x is calculated by the following formula ij Perform standard transformation to obtain Among them, s j Represents the initial sample matrix X mn Medium ij The standard value of the jth column of Represents the initial sample matrix X mn Medium ij The average value of the jth column of ; By the following formula Create a standard transformation matrix Right now Among them, 1≤i≤m, 1≤j≤n, m and n represent the standard transformation matrix Dimensions; The standard transformation matrix is ​​calculated by the following formula The historical data of each gateway electric energy meter after the standard conversion The correlation factor c ij , establish the correlation factor matrix C mn ;Right now in, Represents the standard transformation matrix Any two data in Represents the standard transformation matrix The average value of the i-th row, Represents the standard transformation matrix The mean of the jth column; The correlation factor matrix C is calculated by the characteristic equation mn Perform eigendecomposition and obtain c ij The characteristic λ ij and the corresponding eigenvalue p ij , calculate the correlation factor matrix C according to the eigenvalue p mn Medium ij Contribution G ij ; The contribution G is calculated by the following formula ij ,Right now The eigenvalue p ij The larger the value, the greater the correlation factor matrix c ij The larger the coverage length in the corresponding characteristic direction, the corresponding contribution G is obtained. i The larger j is, the further data c with contribution greater than the preset threshold is selected. ij Construct the fault feature set Y; that is, Y = [y1, y2, y3.., y z ], where 1≤z≤n; Furthermore, the fault feature set Y is matched with each other and the following formula is used for relationship matching: Among them, W α Represents the eigenvector corresponding to the fault mode, 1≤α≤n, represents the smoothing factor; The probability of the fault feature set corresponding to different types of fault modes after relationship matching is calculated by the following formula: Among them, f a represents the probability of the failure mode, ρ represents the dimension of the failure mode sample, represents the smoothing factor; The final fault type T is determined by the probability density function, that is, T = argmax (f a ); thereby obtaining the final fault type through the fault probability, and then completing the training of the verification and evaluation model of the gateway electric energy meter by continuously iterating the data; Real-time data acquisition module, used to obtain the operation data of the gateway electric energy meter in real time and transmit the data to the data processing center; A data analysis module, used for inputting the operation data of the gateway electric energy meter into the trained verification and evaluation model of the gateway electric energy meter; The early warning and operation and maintenance module provides feedback and early warning for the fault type, and sends early warning prompts to the client through the cloud server; The cloud server pushes maintenance suggestions from an expert database according to the fault type, and sends the maintenance suggestions to a handheld device of the maintenance personnel.

6. The device based on health evaluation and intelligent operation and maintenance of gateway electric energy meters according to claim 5 is characterized in that: The gateway electric energy meter in the data acquisition device includes voltage and current transformers used in different types of electric energy meters, and their secondary circuit electric energy metering cabinets, wherein the electric energy meter is divided into direct access and access via transformer according to the access form.

7. The device based on health evaluation and intelligent operation and maintenance of gateway electric energy meters according to claim 6 is characterized in that: When collecting historical electricity metering data, a timed automatic collection method is pre-set to trigger the collection of electric energy data and non-electricity related data. Among them, the sampling interval is adjusted according to the load size during the current time period to collect the voltage and current on site at a high frequency.

8. The device based on health evaluation and intelligent operation and maintenance of gateway electric energy meters according to claim 5 is characterized in that: The faulty operating conditions of the gateway electricity meter include secondary circuit miswiring, pressure loss fault, collection error, appearance damage, meter stoppage, error out of tolerance, clock abnormality, power abnormality, display failure, interface damage, and parameter error.

9. An electronic device, characterized in that: The electronic device includes a processor and a memory: the memory is used to store program code and transmit the program code to the processor, and the processor is used to execute a method based on health evaluation and intelligent operation and maintenance of gateway electricity meters as described in any one of claims 1 to 4 according to instructions in the program code.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store program code and transmit the program code to a processor, and the processor is used to execute the method for health assessment and intelligent operation and maintenance of gateway electric energy meters according to any one of claims 1 to 4 according to the instructions in the program code.

Images