Electric energy meter state analysis method and system based on service driving

Abstract

The invention discloses an electric energy meter state analysis method and system based on service driving, and the method comprises the steps: obtaining indexes of all links of the whole life cycle of an electric energy meter, and carrying out the weight analysis of the obtained indexes; calculating an electric energy meter state comprehensive score based on the obtained index weight analysis result; based on the calculated electric energy meter state comprehensive score, electric energy meter state interval division is carried out according to the historical electric energy meter quality score, and an electric energy meter state analysis result is obtained. According to the method, automatic evaluation and state updating of the state of the electric energy meter can be realized, multi-dimensional fusion analysis can be carried out on massive metering data of the electric energy meter, the use condition and demand characteristics of the electric energy can be deeply known, and a more accurate power consumption prediction and load regulation and control strategy can be provided.

Description

Technical Field

[0001] The present invention relates to the field of full life cycle management and data analysis of electric energy meters, and in particular to a business-driven electric energy meter state analysis method and system. Background Art

[0002] In the new historical period, accelerating the construction of a national modern advanced measurement system is a major issue in the metrology work of the new era. With the vigorous development of the digital economy, digital technology has become a new development engine. Driving the development of the metrology industry with digitalization is both an urgent need and the direction of industrial development.

[0003] In order to promote the digital transformation of metering services, it is necessary to improve the quality control level of electric energy meters, improve the efficiency and accuracy of metering services, and provide reliable basic data support for energy management and billing. However, the scattered quality data of electric energy meters, insufficient application of data value mining, and insufficient intelligence, automation, and digitization of electric energy meter quality inspection and verification, operation status monitoring, fault analysis, quality analysis and prediction, make it difficult to meet the needs of the digital transformation of the power grid for massive and accurate metering data.

[0004] CN202410330575.3A discloses a method and system for managing the operation status of an electric energy meter; including: collecting comprehensive data and environmental data of the electric energy meter; establishing an operation status function according to the comprehensive data of the electric energy meter; obtaining environmental impact factors according to environmental data; constructing an electric energy meter state matrix; obtaining the state trend of the electric energy meter according to the operation status function, environmental impact factors and the electric energy meter state matrix; the state trend of the electric energy meter is divided into overall stability, slow deterioration, rapid deterioration, slow optimization and rapid optimization; when the state trend of the electric energy meter is slow deterioration or rapid deterioration, a reminder message is sent to the electric energy meter management terminal. The operation status function constructed by the invention based only on the comprehensive data and environmental data of the electric energy meter may affect the rigor of the state trend analysis and judgment.

[0005] CN202111530000.9A discloses a method for analyzing and evaluating the state of an electric energy meter based on a generalized regression neural network. The method comprises the following steps: S1: collecting data; S2: data preprocessing; S3: establishing a neural network model and training the model through data; S4: model optimization, determining the smoothing factor of the model; S5: model analysis and evaluation effect determination. The invention only uses the neural network model, which will affect the rigor of the state classification evaluation. It uses historical data in conjunction with the real-time data rolling learning method, and based on the state classification evaluation index system data and hyperparameter optimization dynamic evaluation mechanism, jointly builds a state classification evaluation model. Although the neural network model can handle complex nonlinear relationships, its internal mechanism is often difficult to explain intuitively, which may affect the understanding and trust of the model output. If there is not enough data or the regularization technology is inappropriate, the neural network model is prone to overfitting, resulting in poor generalization ability of the model on new data.

[0006] Therefore, it is necessary to carry out in-depth application of advanced technologies such as digital twins, big data, and artificial intelligence in the entire life cycle of electric energy meters, activate massive data resources, explore the value of data related to electric energy meter quality, realize comprehensive recording and real-time update of information, promote the digital transformation of the electric energy meter industry, and improve overall operational efficiency and quality control level. It is possible to predict the state trend of electric energy meters based on operating state functions, environmental factors, and electric energy meter state matrix. Summary of the invention

[0007] In order to solve the deficiencies in the prior art, the present invention provides a business-driven multi-factor intelligent association analysis method, including factor association analysis, weight fusion index system construction, index scoring analysis and weight analysis, and electric energy meter state prediction.

[0008] In order to achieve the above-mentioned invention object, the present invention specifically adopts the following technical solutions:

[0009] A method for analyzing electric energy meter status based on business drive, characterized in that the method comprises:

[0010] Step 1: Extract the common factors of the initial indicators of each link in the whole life cycle of the electric energy meter as the status indicators of each link in the whole life cycle of the electric energy meter;

[0011] Step 2: Perform adaptive weight analysis based on the obtained state indicators to obtain the fusion weights of each link and each state indicator under each link;

[0012] Step 3: Based on the status indicators of each link in the whole life cycle of the electric energy meter and the fusion weight of each status indicator, the comprehensive score of the electric energy meter status is calculated;

[0013] Step 4: Based on the calculation result of the electric energy meter quality score in step 3, the electric energy meter status interval is divided according to the historical electric energy meter quality score; and the status analysis result of the electric energy meter is obtained.

[0014] More preferably,

[0015] In step 1, each link of the whole life cycle of the electric energy meter includes procurement and arrival, equipment acceptance, verification and testing, warehousing and distribution, equipment installation, equipment operation, equipment removal, equipment scrapping and the environment in which it is located;

[0016] The initial indicators of each link in the whole life cycle of the electric energy meter specifically include:

[0017] Initial indicators for purchased goods include the pre-production full performance test pass rate and pre-production sample comparison pass rate;

[0018] The initial indicators for equipment acceptance include the return and exchange ratio, pre-delivery sample comparison pass rate, pre-delivery full performance test pass rate, post-arrival sample comparison pass rate, and post-arrival random inspection test pass rate;

[0019] The initial indicators of the verification and testing phase include the full inspection acceptance rate, the electric energy meter re-inspection pass rate, and the proportion of verification and testing fault types;

[0020] Initial indicators for the warehousing and distribution phase include new product inventory age, finished product inventory age, and distribution failure rate; initial indicators for the equipment installation phase include the failure rate of smart energy meters during installation and the proportion of quality failure types of smart energy meters during installation;

[0021] The initial indicators of equipment operation include the one-time data collection success rate of low-voltage users, the daily data collection success rate of low-voltage users, the one-time data collection success rate of special transformer users, the daily data collection success rate of special transformer users, the qualified rate of random inspection of operation tables, the qualified rate of on-site verification, and the qualified rate of supervision random inspection;

[0022] Initial indicators for equipment dismantling include the recycling rate of dismantled assets and the qualified rate of dismantled measurement assets;

[0023] Initial indicators for equipment scrapping include the scrapping rate of measured assets due to quality factors;

[0024] The initial indicators of the environment include maximum temperature and minimum temperature.

[0025] More preferably,

[0026] In step 1, the common factors of the initial indicators of each link in the whole life cycle of the electric energy meter are extracted as the status indicators of each link in the whole life cycle of the electric energy meter, specifically:

[0027] Calculate the correlation coefficient for every two initial indicator samples in each link. If the correlation between an initial indicator sample and any other initial indicator sample in the same link is weakly correlated, then the initial indicator is judged as a weakly correlated indicator, otherwise it is a strongly correlated indicator;

[0028] For weak correlation indicators, they are directly used as common factors;

[0029] For the strong correlation indicators, the Bartlett sphericity test and KMO test were performed, and the common factors were extracted for the strong correlation indicators that met the test criteria, otherwise they were directly used as common factors;

[0030] The proposed common factors are used as status indicators of each link in the entire life cycle of the electricity meter.

[0031] More preferably,

[0032] The test criterion is as follows:

[0033]

[0034] Among them, P is the Bartlett sphericity test value, and γ is the KMO test value.

[0035] More preferably,

[0036] For the strong correlation index that meets the test criteria as a factor, calculate the characteristic value, variance contribution rate and cumulative variance contribution rate of each factor sample data in each link. When the following conditions are met, the factor is extracted as a common factor:

[0037] For the strong correlation indicators that meet the test criteria, the characteristic values ​​of the sample data of each factor in each link are calculated, and the variance contribution rate of each characteristic value and the cumulative variance contribution rate of each factor are calculated. When the cumulative variance contribution rate of the factor is greater than or equal to 88%, these factors are extracted as common factors and other factors are discarded;

[0038] λ1+λ2+...+λ r = total variance;

[0039]

[0040] Cumulative variance contribution f = Variance contribution rate 1 + Variance contribution rate 2 + ... + Variance contribution rate f

[0041] Among them, λ1, λ2, λ f , r Respectively represent the eigenvalues ​​of the 1st, 2nd, fth and rth factor sample data of a certain link, and r represents the total number of factors in the link; variance contribution ratef represents the variance contribution rate of the fth factor;

[0042] When the current f cumulative variance contribution rates are greater than or equal to 88%, 1-f factors are extracted as the common factors of this link.

[0043] More preferably,

[0044] In step 2, the following are included:

[0045] 2.1 Calculate the initial weights of each link and each state indicator under different models;

[0046] 2.2 Calculate the initial weight mean of each link calculated under different models and the initial weight mean of each state indicator calculated under different models;

[0047]

[0048]

[0049] in, represents the initial weight mean of the jth link, represents the initial weight mean of the j-th state indicator, α ij represents the initial weight of the jth link under the i-th model, β ik represents the initial weight of the kth state indicator under the i-th model, and m is the number of models;

[0050] 2.3 Calculate the absolute deviation of the initial weights of each link under different models and the absolute deviation of the initial weights of each state indicator under different models;

[0051]

[0052]

[0053] Among them, S ij represents the absolute deviation of the j-th link weight under the i-th model, S i ' k represents the absolute deviation of the kth state indicator weight under the i-th model;

[0054] 2.4 Based on the absolute deviation, calculate the fusion weight of each link and the fusion weight of each state indicator.

[0055] More preferably,

[0056] The different models at least include a hierarchical analysis model and an entropy weight method model;

[0057] In each model, the sum of the initial weights of each link is 1, and the sum of the initial weights of each state indicator in each link is 1.

[0058] More preferably,

[0059] The fusion weight of each link and the fusion weight of each status indicator are calculated as follows:

[0060]

[0061]

[0062] Among them, α j represents the fusion weight of the jth link, β k Represents the fusion weight of the k-th state indicator.

[0063] More preferably,

[0064] In step 3, the comprehensive score of the electric energy meter status is calculated based on the status indicators of each link in the whole life cycle of the electric energy meter and the fusion weight of each status indicator, which is:

[0065] 3.1 Collect the values ​​of each status indicator of the electric energy meter and calculate the score of the status indicator;

[0066] 3.2 The status indicator score of each link is multiplied by the fusion weight of the status indicator corresponding to each status indicator, and then the status score value of each link is obtained;

[0067] 3.3 The status score of each link is multiplied by the corresponding fusion weight of each link and then summarized to obtain the comprehensive score of the electricity meter status.

[0068] More preferably,

[0069] In 3.1, the status indicator score includes a positive indicator score or a negative indicator score. The positive or negative score is determined according to the nature of the status indicator. The higher the value of the positive indicator, the more qualified the equipment is. The lower the value of the negative indicator, the more qualified the equipment is.

[0070] Positive indicator score C i_score Calculated using the following formula:

[0071]

[0072] Negative indicator score C i_score Calculated using the following formula:

[0073]

[0074] Among them, C i is the state indicator value, (C i ) min , (C i ) maxThey are the minimum and maximum values ​​of the historical values ​​of the status indicator respectively.

[0075] More preferably,

[0076] In step 4, based on the electric energy meter quality score calculation result in step 3, the electric energy meter state interval is divided according to the historical electric energy meter quality score, specifically including:

[0077] 4.1 Collect the historical quality score data of the electric energy meter and the corresponding quality status of the electric energy meter;

[0078] 4.2 Set the quality status classification level of the electric energy meter, and use the number of classification levels as the number of categories;

[0079] 4.3 Randomly select the historical scoring data of the electric energy meter as the initial analog center;

[0080] 4.3 Calculate the probability calculation result of each electric energy meter that needs to be divided into state intervals belonging to each category, and update the center of the category according to the probability calculation result until the category center no longer changes or the change is less than the set threshold.

[0081] More preferably,

[0082] The probability calculation result μ of the energy meter belonging to each category sw Calculate as follows:

[0083]

[0084] Among them, u sw represents the possibility that the sth electric energy meter belongs to the jth category, x s represents the comprehensive score of the sth electric energy meter status, v w represents the center score data of the w-th category, v q represents the cluster center score data of all categories, || represents the distance calculation, t represents the number of categories, and p is the calculation index;

[0085]

[0086] x is the number of times the category center is updated.

[0087] On the other hand, the present invention discloses an electric energy meter state analysis system based on the above-mentioned electric energy meter state analysis method, including a calculation and confirmation module for each link and state index of the whole life cycle of the electric energy meter, a fusion weight calculation module for each link and each state index under each link, an electric energy meter state comprehensive score module and an electric energy meter state analysis module; characterized in that:

[0088] The calculation and confirmation module of each link and state index of the whole life cycle of the electric energy meter extracts the common factors of the initial index of each link of the whole life cycle of the electric energy meter as the state index of each link of the whole life cycle of the electric energy meter;

[0089] The fusion weight calculation module of each link and each state indicator under each link performs adaptive weight analysis based on the obtained state indicators to obtain the fusion weight of each link and each state indicator under each link;

[0090] The comprehensive score module of the electric energy meter status is based on the status indicators of each link in the whole life cycle of the electric energy meter, and the fusion weight calculation of each status indicator to obtain the comprehensive score of the electric energy meter status;

[0091] The electric energy meter status analysis module divides the electric energy meter status intervals based on the electric energy meter quality score calculation results and the historical electric energy meter quality scores to obtain the electric energy meter status analysis results.

[0092] The beneficial effects of the present invention are as follows:

[0093] (1) The present invention is based on business drive, determines each link of the whole life cycle of the electric energy meter, explores the influence degree of various influencing factors on the state of the electric energy meter according to the influence relationship of the influencing factors on the state of the electric energy meter, and can effectively mine the influencing factors with a relatively large correlation with the state of the electric energy meter, laying a foundation for carrying out automated comprehensive evaluation of the whole life cycle of the electric energy meter.

[0094] (2) Through the weight impact analysis of the electric energy meter indicator system constructed based on multimodal correlation factors, the comprehensive score of the electric energy meter status can be calculated in real time according to the weight performance and data conditions of each indicator dimension, and the latest status of the electric energy meter can be intuitively displayed.

[0095] (3) Promote the digital transformation of metering: Apply advanced technologies such as data feature dimensionality reduction, machine learning, and big data analysis to form a complete digital management system. Through its application throughout the life cycle of electricity meters, it can promote the digital transformation of power metering and improve the technical and management levels of electricity meters.

[0096] (4) Help improve the level of electricity service: By improving the management efficiency and power quality of electricity meters throughout their life cycle, we can better provide electricity services and optimize the analysis and application of electricity metering data, thereby improving the company's operational efficiency and service quality and meeting the growing electricity demand. BRIEF DESCRIPTION OF THE DRAWINGS

[0097] Figure 1 It is a flowchart of the method for analyzing the state of an electric energy meter based on business drive; DETAILED DESCRIPTION

[0098] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only embodiments of a part of the present invention, rather than all embodiments. Based on the spirit of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the protection scope of the present invention.

[0099] like Figure 1 As shown, the present invention discloses a method for analyzing the state of an electric energy meter based on business drive, and the method specifically comprises the following steps:

[0100] Step 1: Extract the common factors of the initial indicators of each link in the whole life cycle of the electric energy meter as the status indicators of each link in the whole life cycle of the electric energy meter;

[0101] The various links of the whole life cycle of the electric energy meter include procurement and arrival, equipment acceptance, verification and testing, warehousing and distribution, equipment installation, equipment operation, equipment removal, equipment scrapping and the environment in which it is located;

[0102] The initial indicators of each link in the whole life cycle of the electric energy meter specifically include:

[0103] Initial indicators for purchased goods include the pre-production full performance test pass rate and pre-production sample comparison pass rate;

[0104] The initial indicators for equipment acceptance include the return and exchange ratio, pre-delivery sample comparison pass rate, pre-delivery full performance test pass rate, post-arrival sample comparison pass rate, and post-arrival random inspection test pass rate;

[0105] The initial indicators of the verification and testing phase include the full inspection acceptance rate, the electric energy meter re-inspection pass rate, and the proportion of verification and testing fault types;

[0106] Initial indicators for the warehousing and distribution phase include new product inventory age, finished product inventory age, and distribution failure rate; initial indicators for the equipment installation phase include the failure rate of smart energy meters during installation and the proportion of quality failure types of smart energy meters during installation;

[0107] The initial indicators of equipment operation include the one-time data collection success rate of low-voltage users, the daily data collection success rate of low-voltage users, the one-time data collection success rate of special transformer users, the daily data collection success rate of special transformer users, the qualified rate of random inspection of operation tables, the qualified rate of on-site verification, and the qualified rate of supervision random inspection;

[0108] Initial indicators for equipment dismantling include the recycling rate of dismantled assets and the qualified rate of dismantled measurement assets;

[0109] Initial indicators for equipment scrapping include the scrapping rate of measured assets due to quality factors;

[0110] The initial indicators of the environment include maximum temperature and minimum temperature.

[0111] The common factors of the initial indicators of each link in the whole life cycle of the electric energy meter are extracted as the status indicators of each link in the whole life cycle of the electric energy meter, specifically:

[0112] Calculate the correlation coefficient for every two initial indicator samples in each link. If the correlation between an initial indicator sample and any other initial indicator sample in the same link is weakly correlated, then the initial indicator is judged as a weakly correlated indicator, otherwise it is a strongly correlated indicator;

[0113] For weak correlation indicators, they are directly used as common factors;

[0114] For the strong correlation indicators, the Bartlett sphericity test and KMO test were performed, and the common factors were extracted for the strong correlation indicators that met the test criteria, otherwise they were directly used as common factors;

[0115] The test criterion is as follows:

[0116]

[0117] Among them, P is the Bartlett sphericity test value, and γ is the KMO test value.

[0118] For the strong correlation indicators that meet the test criteria, the characteristic values ​​of the sample data of each factor in each link are calculated, and the variance contribution rate of each characteristic value and the cumulative variance contribution rate of each factor are calculated. When the cumulative variance contribution rate of the factor is greater than or equal to 88%, these factors are extracted as common factors and other factors are discarded;

[0119] λ1+λ2+...+λ r = total variance;

[0120]

[0121] Cumulative variance contribution f = Variance contribution rate 1 + Variance contribution rate 2 + ... + Variance contribution rate f

[0122] Among them, λ1, λ2, λ f , r Respectively represent the eigenvalues ​​of the 1st, 2nd, fth and rth factor sample data of a certain link, and r represents the total number of factors in the link; variance contribution rate f represents the variance contribution rate of the fth factor;

[0123] When the current f cumulative variance contribution rates are greater than or equal to 88%, 1-f factors are extracted as the common factors of this link.

[0124] The proposed common factors are used as status indicators of each link in the entire life cycle of the electricity meter.

[0125] Step 2: Perform adaptive weight analysis based on the obtained state indicators to obtain the fusion weights of each link and each state indicator under each link;

[0126] Specifically include the following:

[0127] 2.1 Calculate the initial weights of each link and each state indicator under different models;

[0128] The different models at least include a hierarchical analysis model and an entropy weight method model;

[0129] In each model, the sum of the initial weights of each link is 1, and the sum of the initial weights of each state indicator in each link is 1.

[0130] 2.2 Calculate the initial weight mean of each link calculated under different models and the initial weight mean of each state indicator calculated under different models;

[0131]

[0132]

[0133] Among them, α j represents the initial weight mean of the jth link, represents the initial weight mean of the j-th state indicator, α ij represents the initial weight of the jth link under the i-th model, β ik represents the initial weight of the kth state indicator under the i-th model, and m is the number of models;

[0134] 2.3 Calculate the absolute deviation of the initial weights of each link under different models and the absolute deviation of the initial weights of each state indicator under different models;

[0135]

[0136]

[0137] Among them, S ij represents the absolute deviation of the j-th link weight under the i-th model, S i ' k represents the absolute deviation of the kth state indicator weight under the i-th model;

[0138] 2.4 Based on the absolute deviation, calculate the fusion weight of each link and the fusion weight of each state indicator.

[0139] The fusion weight of each link and the fusion weight of each status indicator are calculated as follows:

[0140]

[0141]

[0142] Among them, α j represents the fusion weight of the jth link, β k Represents the fusion weight of the k-th state indicator.

[0143] Step 3: Based on the status indicators of each link in the whole life cycle of the electric energy meter and the fusion weight of each status indicator, the comprehensive score of the electric energy meter status is calculated;

[0144] Based on the status indicators of each link in the whole life cycle of the electric energy meter and the fusion weight calculation of each status indicator, the comprehensive score of the electric energy meter status is obtained, which is:

[0145] 3.1 Collect the values ​​of each state indicator of the electric energy meter and calculate the state indicator score; the state indicator score includes a positive indicator score or a negative indicator score, and the positive or negative score is determined according to the nature of the state indicator. Among them, the higher the value of the positive indicator, the more qualified the equipment is, and the lower the value of the negative indicator, the more qualified the equipment is;

[0146] Positive indicator score C i_score Calculated using the following formula:

[0147]

[0148] Negative indicator score C i_score Calculated using the following formula:

[0149]

[0150] Among them, C i is the state indicator value, (C i ) min , (C i ) max They are the minimum and maximum values ​​of the historical values ​​of the status indicator respectively;

[0151] 3.2 The status indicator score of each link is multiplied by the fusion weight of the status indicator corresponding to each status indicator, and then the status score value of each link is obtained;

[0152] 3.3 The status score of each link is multiplied by the corresponding fusion weight of each link and then summarized to obtain the comprehensive score of the electricity meter status.

[0153] Step 4: Based on the calculation result of the electric energy meter quality score in step 3, the electric energy meter status interval is divided according to the historical electric energy meter quality score; and the status analysis result of the electric energy meter is obtained.

[0154] Based on the calculation result of the electric energy meter quality score in step 3, the electric energy meter status interval is divided according to the historical electric energy meter quality score, specifically including:

[0155] 4.1 Collect the historical quality score data of the electric energy meter and the corresponding quality status of the electric energy meter;

[0156] 4.2 Set the quality status classification level of the electric energy meter, and use the number of classification levels as the number of categories;

[0157] 4.3 Randomly select the historical scoring data of the electric energy meter as the initial analog center;

[0158] 4.3 Calculate the probability calculation result of each electric energy meter that needs to be divided into state intervals belonging to each category, and update the center of the category according to the probability calculation result until the category center no longer changes or the change is less than the set threshold.

[0159] In a preferred embodiment of the present invention,

[0160] The probability calculation result μ of the energy meter belonging to each category sw Calculate as follows:

[0161]

[0162] Among them, u sw represents the possibility that the sth electric energy meter belongs to the jth category, x s represents the comprehensive score of the sth electric energy meter status, v w represents the center score data of the wth category, t represents the number of categories, and v q represents the cluster center score data of all categories, || represents the distance calculation, and p is the calculation index;

[0163]

[0164] x is the number of times the category center is updated.

[0165] In this application, the final quality status level of the electric energy meter is divided into three levels, A, B, and C, based on business experience. Then t is equal to 3 at this time.

[0166] The update of the category center and the cluster center update formula are preferably performed in the following manner:

[0167]

[0168] Among them, v w Represents the center of the w-th category, which needs to be updated in each iteration; This is the p-th power of the probability that the s-th meter's comprehensive score belongs to the w-th category; xs Represents the comprehensive score of the sth electric energy meter; is the comprehensive score of all electric energy meters x s The result of weighted summation, the weight is is the score of all electric energy meters x s The sum of the probabilities of belonging to the corresponding category centers.

[0169] The present application also discloses an electric energy meter state analysis system based on the above-mentioned electric energy meter state analysis method, including a calculation and confirmation module for each link and state index of the whole life cycle of the electric energy meter, a fusion weight calculation module for each link and each state index under each link, an electric energy meter state comprehensive score module and an electric energy meter state analysis module;

[0170] The calculation and confirmation module of each link and state index of the whole life cycle of the electric energy meter extracts the common factors of the initial index of each link of the whole life cycle of the electric energy meter as the state index of each link of the whole life cycle of the electric energy meter;

[0171] The fusion weight calculation module of each link and each state indicator under each link performs adaptive weight analysis based on the obtained state indicators to obtain the fusion weight of each link and each state indicator under each link;

[0172] The comprehensive score module of the electric energy meter status is based on the status indicators of each link in the whole life cycle of the electric energy meter, and the fusion weight calculation of each status indicator to obtain the comprehensive score of the electric energy meter status;

[0173] The electric energy meter status analysis module divides the electric energy meter status intervals based on the electric energy meter quality score calculation results and the historical electric energy meter quality scores to obtain the electric energy meter status analysis results.

[0174] Embodiment 1, the following specific embodiments further illustrate the technical solution of the present invention in detail.

[0175] Step 1: Determine the indicators of each link in the life cycle of the energy meter by factor extraction

[0176] 1.1 Selection of common factor extraction method

[0177] By studying typical common factor extraction techniques, the present invention finds that the singular value decomposition algorithm is suitable for decomposing directly on the original matrix; the PCA principal component analysis algorithm aims to maximize the variance, and has weak interpretability; the factor analysis aims to analyze the structure behind the data, and has strong interpretability. The present invention combines the characteristics of large data volume, multiple data dimensions, complex data sources, and diverse data structures (structured and unstructured) throughout the life cycle of batch electric energy meters, based on the scope of application and application scenarios of several mainstream dimensionality reduction algorithms, in order to better analyze the relationship between the full-cycle life cycle data of batch electric energy meters and extract the practical value in the data, the factor analysis technology is used to conduct a dimensionality reduction study on the full-life cycle characteristics of batch electric energy meters.

[0178] 1.2 Extraction of common factors of various dimensions of batch electric energy meters

[0179] Based on factor analysis technology, the present invention carries out batch electric energy meter common factor extraction from various dimensions such as procurement and arrival, equipment acceptance, verification and testing, warehousing and distribution, equipment installation, equipment operation, equipment removal, asset scrapping, and environment. The common factor extraction process mainly includes data inspection of each indicator and common factor generation. Among them, the data inspection of each indicator includes KMO test analysis (an indicator used to compare the simple correlation coefficient and partial correlation coefficient between variables. The closer the KMO value is to 1, the stronger the correlation between variables is, and the more suitable the original variables are for factor analysis; the closer the KMO value is to 0, the weaker the correlation between variables is, and the less suitable the original variables are for factor analysis), Bartlett test analysis (used to test whether the variances between multiple samples are equal), and the common factor generation is mainly analyzed by variance contribution rate, scree plot, etc.

[0180] 1.2.1 Extraction of common factors of procurement arrival dimension

[0181] Carry out factor analysis of the procurement arrival dimension, and use factor analysis technology to extract common factors for the pre-production full performance test pass rate and pre-production sample comparison pass rate. Before factor analysis, perform Bartlett test and KMO test analysis on the indicators.

[0182] The Bartlett test is shown in the following table:

[0183] Table 1-1 Bartlett inspection values ​​for purchased goods

[0184]

[0185] As can be seen from the table, the p-value of the Bartlett sphericity test is less than 0.05, indicating that there is a correlation between the original variables and factor analysis can be performed.

[0186] The KMO test is shown in the following table:

[0187] Table 1-2 KMO inspection value of purchased goods

[0188]

[0189] As can be seen from the table, the KMO test value is greater than 0.6, which also shows that there is a correlation between the variables and can be analyzed.

[0190] Calculate the eigenvalues ​​and eigenvectors of the factors and their contributions. Since there are only two indicators in total, when extracting common factors, only the variance contribution rate of one common factor can be considered. The results are shown in the following table:

[0191] Table 1-3 Purchase arrival characteristic value and variance contribution table

[0192] Eigenvalue Variance Contribution Cumulative variance contribution 0.762381 0.873406 0.873406

[0193] It can be seen that when the eigenvalue is 1, the variance contribution rate and the cumulative variance contribution rate are both 0.873406, and the explainability of the two indicators of procurement and arrival is 87%, and a common factor is extracted. Since the factor analysis method cannot generate factor names, combined with business understanding, the common factor is named the pre-production full performance test pass rate. This common factor integrates 87% of the information of the two indicators of procurement and arrival. The comparison relationship between the indicators and the common factors is shown in the following table:

[0194] Table 1-4 Comparison table of procurement arrival dimension indicators and common factors

[0195]

[0196] 1.2.2 Extraction of common factors of equipment acceptance dimensions

[0197] Conduct factor analysis on equipment acceptance dimensions, and use factor analysis techniques to extract common factors for the return and exchange ratio, pre-delivery sample comparison pass rate, pre-delivery full performance test pass rate, post-arrival sample comparison pass rate, and post-arrival random inspection test pass rate. Before factor analysis, perform Bartlett test and KMO test analysis on the indicators.

[0198] The Bartlett test is shown in the following table:

[0199] Table 1-5 Bartlett test values ​​for equipment acceptance

[0200]

[0201] As can be seen from the table, the p-value of the Bartlett sphericity test is less than 0.05, indicating that there is a correlation between the original variables and factor analysis can be performed.

[0202] The KMO test is shown in the following table:

[0203] Table 1-6 Equipment acceptance KMO inspection values

[0204]

[0205] As can be seen from the table, the KMO test value is greater than 0.6, which also shows that there is a correlation between the variables and can be analyzed.

[0206] The slope between the number of common factors 3 and 4 becomes gentler, indicating that the information that can be explained by the subsequent 4th common factor has been reduced even more, and the first three common factors can already cover most of the information, so the number of common factors can be selected as 3.

[0207] Table 1-7 Equipment acceptance characteristic values ​​and variance contribution

[0208] Eigenvalue Variance Contribution Cumulative variance contribution 2.345213 0.326708 0.326708 1.789065 0.222195 0.548903 0.917625 0.333438 0.882341 0.650908 0.038423 0.920764 0.441457 0.079236 1

[0209] The present invention analyzes that when the eigenvalue is 0.917625, the cumulative variance contribution rate reaches 0.882341, the explainability of the five indicators of the equipment acceptance dimension reaches 88%, and three common factors are extracted.

[0210] The common factors are named as the proportion of returned goods, the qualified rate of full performance test before delivery, and the qualified rate of random inspection test after arrival. They integrate 88% of the information of the five indicators of equipment acceptance dimension, as shown in the following table:

[0211] Table 1-8 Comparison table of equipment acceptance dimension indicators and common factors

[0212]

[0213] 1.2.3 Extraction of common factors of test dimensions

[0214] The present invention conducts factor analysis of the verification and detection dimension, and extracts common factors from the full inspection acceptance rate, the electric energy meter re-inspection pass rate (inventory), and the proportion of verification and detection fault types using factor analysis technology. Before performing factor analysis, the indicators are first analyzed by Bartlett test and KMO test.

[0215] The Bartlett test is shown in the following table:

[0216] Table 1-9 Bartlett test values ​​for testing

[0217]

[0218] As can be seen from the table, the p-value of the Bartlett sphericity test is less than 0.05, indicating that there is a correlation between the original variables and factor analysis can be performed.

[0219] The KMO test is shown in the following table:

[0220] Table 1-10 KMO test values ​​for calibration and testing

[0221]

[0222] As can be seen from the table, the KMO test value is greater than 0.6, which also shows that there is a correlation between the variables and factor analysis can be performed.

[0223] The slope between the number of common factors 2 and 3 becomes gentler, indicating that the information that can be explained by the subsequent third common factor has been reduced even more, and the first two common factors can already cover most of the information, so the number of common factors can be selected as 2.

[0224] Table 1-11 Test characteristic values ​​and variance contributions

[0225] Eigenvalue Variance Contribution Cumulative variance contribution 1.587902 0.326708 0.388903 0.879657 0.474568 0.863471 0.671529 0.136529 1

[0226] It can be seen that when the eigenvalue is 0.879657, the cumulative variance contribution rate reaches 0.863471, the explainability of the three indicators of the test dimension reaches 86%, and three common factors are extracted.

[0227] The common factors named the full inspection acceptance rate and the proportion of inspection and detection fault types, which integrate 86% of the information of the three indicators of the inspection and detection dimensions, are shown in the following table:

[0228] Table 1-12 Comparison table of the relationship between the inspection dimension indicators and common factors

[0229]

[0230] 1.2.4 Common Factor Extraction of Warehousing and Distribution Dimensions

[0231] The present invention conducts factor analysis of the warehousing and distribution dimension, and extracts common factors from the new product inventory age, finished product inventory age, and distribution failure rate using factor analysis technology. Before factor analysis, the indicators are first analyzed by Bartlett test and KMO test.

[0232] The Bartlett test is shown in the following table:

[0233] Table 1-13 Bartlett test value for warehousing and distribution

[0234]

[0235] As can be seen from the table, the p-value of the Bartlett sphericity test is less than 0.05, indicating that there is a correlation between the original variables and factor analysis can be performed.

[0236] The KMO test is shown in the following table:

[0237] Table 1-14 KMO inspection values ​​for warehousing and distribution

[0238]

[0239] As can be seen from the table, the KMO test value is greater than 0.6, which also shows that there is a correlation between the variables and can be analyzed.

[0240] The slope between the number of common factors 2 and 3 becomes gentler, indicating that the information that can be explained by the subsequent third common factor has been reduced even more, and the first two common factors can already cover most of the information, so the number of common factors can be selected as 2.

[0241] Table 1-15 Warehousing and distribution characteristic values ​​and variance contributions

[0242] Eigenvalue Variance Contribution Cumulative variance contribution 1.657035 0.427801 0.427801 0.776012 0.485671 0.913472 0.654387 0.086528 1

[0243] It can be seen that when the eigenvalue is 0.776012, the cumulative variance contribution rate reaches 0.913472, the explainability of the three indicators of the warehousing and distribution dimension reaches 91%, and two common factors are extracted.

[0244] The common factor naming includes the proportion of returned goods, the qualified rate of full performance test before delivery, and the qualified rate of random inspection test after arrival, which integrates 91% of the information of the three indicators of warehousing and distribution, as shown in the following table:

[0245] Table 1-16 Comparison table of warehousing and distribution dimension indicators and common factors

[0246]

[0247] 1.2.5 Extraction of common factors of equipment installation dimensions

[0248] After analysis, the present invention finds that the correlation between the two indicators "failure rate of smart electric energy meters during installation" and "proportion of quality failure types of smart electric energy meters during installation" is very low, and is not suitable for feature dimensionality reduction to extract common factors. The two indicators are retained as common factors of the equipment installation dimension.

[0249] 1.2.6 Extraction of common factors of equipment operation dimensions

[0250] The present invention carries out factor analysis of equipment operation dimension, and extracts common factors from the one-time collection success rate of new low-voltage users, the daily collection success rate of low-voltage households, the one-time collection success rate of special transformer users, the daily collection success rate of special transformer users, the collection success rate, the accuracy rate of remote fee control, the unqualified rate of operation meter inspection, the qualified rate of operation meter sampling inspection, the qualified rate of on-site verification, the qualified rate of supervision sampling inspection, the proportion of (quality factor) fault types, the failure rate of operation meter, (actual operation) meter age, operation time, the number of phase failures of electric energy meters, the number of voltage over-limit times, the number of current over-limit times, the number of abnormal opening of electric energy meter cover, the number of power failures, the number of over-tolerance times, the number of flying away of electric energy meters, the number of stopping of electric energy meters, the number of opening cover of electric energy meters at the beginning of electric energy meters, the number of electric energy meter degree drops, the number of over-tolerance times of electric energy meter clock, the number of voltage loss times, the number of power outages of electric energy meters, and the number of abnormal current loops of electric energy meters using factor analysis technology. Before factor analysis, the indicators are first analyzed by Bartlett test and KMO test.

[0251] The Bartlett test is shown in the following table:

[0252] Table 1-17 Bartlett test values ​​for equipment operation

[0253]

[0254] As can be seen from the table, the p-value of the Bartlett sphericity test is less than 0.05, indicating that there is a correlation between the original variables and factor analysis can be performed.

[0255] The KMO test is shown in the following table:

[0256] Table 1-18 Equipment operation KMO test values

[0257]

[0258] As can be seen from the table, the KMO test value is greater than 0.6, which also shows that there is a correlation between the variables and can be analyzed.

[0259] Table 1-19 Equipment operation characteristic values ​​and variance contribution

[0260]

[0261] It can be seen that when the eigenvalue is 0.799801, the cumulative variance contribution rate reaches 0.883809, the explainability of 27 indicators of equipment operation dimension reaches 88%, and 6 common factors are extracted.

[0262] The common factors are named operation time, energy meter inspection pass rate, energy meter failure rate, remote fee control accuracy rate, collection success rate, and energy meter abnormal event times. They integrate 88% of the information of 27 indicators of equipment operation, as shown in the following table:

[0263] Table 1-20 Comparison table of equipment operation dimension indicators and common factors

[0264]

[0265] 1.2.7 Extraction of common factors of equipment dismantling dimensions

[0266] Through analysis, the present invention finds that the correlation degree of the indicators "recycling rate of dismantled assets" and "qualified rate of inspection of dismantled measurement assets" is very low, and it is not suitable to perform feature dimensionality reduction to extract common factors. The two indicators are retained as common factors of the equipment dismantling dimension.

[0267] 1.2.8 Extraction of common factors of equipment scrapping dimensions

[0268] The only indicator of equipment scrapping dimension is the scrapping rate of measured assets caused by quality factors. It is no longer possible to perform feature dimensionality reduction to extract common factors. This indicator is retained as the common factor of equipment scrapping dimension.

[0269] 1.2.9 Common Factor Extraction of Environmental Dimensions

[0270] The two indicators of maximum temperature and minimum temperature in the environmental dimension have a low degree of correlation, and it is not suitable to use the feature dimensionality reduction method to extract common factors and derive the average temperature as the common factor of the environmental dimension.

[0271] in:

[0272] Average temperature = (maximum temperature + minimum temperature) / 2

[0273] 1.3 Common factor result table of each dimension of batch electric energy meter

[0274] Table 1-21 Common factor results of batch electric energy meters

[0275]

[0276] Step 2: Perform adaptive weight analysis based on the obtained state indicators to obtain the fusion weights of each link and each state indicator under each link;

[0277] The present invention studies a variety of conventional weight algorithms on the market, and combines the characteristics of electric energy meter quality analysis to select basic weight algorithms (hierarchical analysis method, entropy weight method) and fusion weight algorithms (grey correlation method) to dynamically and adaptively generate batch electric energy meter index weight results. The hierarchical analysis method and entropy weight method are not described in detail in the present invention. The grey correlation method is used to associate the hierarchical analysis method and entropy weight method in the basic fusion algorithm to form the final weight of the batch electric energy meter index. The main steps are as follows:

[0278] (1) Generate weight matrix w(m×n)

[0279]

[0280] Where: w ij Represents the weight of indicator j under the i-th model.

[0281] (2) Calculate the weighted average of each indicator

[0282]

[0283] (3) Calculate the absolute deviation of indicator weights under different models

[0284]

[0285] Note: when When ij Use 1 to replace it uniformly.

[0286] (4) Calculate the comprehensive weight of indicators W(n)

[0287]

[0288] Where W j satisfy

[0289] The present invention firstly externalizes the expert judgment matrix of batch electric energy meter indicators based on the hierarchical analysis method. When the weight changes, the judgment matrix can be modified at any time. The model program will automatically read the value of the judgment matrix to calculate and generate the weight of the electric energy meter indicators; secondly, the entropy weight method will read the batch electric energy meter indicator data every month for analysis, and adaptively update the entropy weight method weight of the batch electric energy meter indicators for that month according to the data changes. Finally, the program will dynamically and adaptively update the weights of the two weights every month to generate the batch electric energy meter indicator weights for that month.

[0290] When the data indicators of the next month change, the weight program will adaptively generate entropy weights based on the changed data. Similarly, the judgment matrix of the hierarchical analysis will also be regenerated, and then dynamic adaptive weights will be formed on a monthly rolling basis.

[0291] The present invention conducts weight analysis on each dimension and the indicators within the dimension based on the hierarchical analysis method. Three business experts with rich experience in measurement business are invited to judge the weight score of the matrix each time. In order to avoid large differences in the weight score of the matrix, the average value of the judgment matrix of the three business experts is taken as the final judgment matrix weight score. Finally, the weight result is calculated based on the judgment matrix weight score through steps such as product and vector normalization.

[0292] 1) Hierarchical analysis and judgment matrix of indicators in each dimension

[0293] A hierarchical analysis and judgment matrix is ​​constructed for dimensions such as procurement arrival, equipment acceptance, calibration and testing, warehousing and distribution, equipment installation, equipment preparation, equipment dismantling, equipment scrapping, and environment.

[0294] The average score of each indicator is calculated for 20 indicators, including the pass rate of full performance inspection before production, the pass rate of full performance inspection before delivery, the pass rate of random inspection after arrival, the proportion of returned and exchanged goods, the pass rate of full inspection and acceptance, the proportion of calibration and inspection failure types, inventory age, distribution failure rate, failure rate of smart electricity meters during installation, the proportion of quality failure types of smart electricity meters during installation, operating time, electricity meter inspection pass rate, electricity meter failure rate, remote fee control accuracy rate, collection success rate, number of abnormal events of electricity meters, dismantling and reuse rate, calibration pass rate of dismantled metering assets, scrapping rate of metering assets due to quality factors, and average temperature.

[0295] 2) Weight of hierarchical analysis of indicators in each dimension

[0296] The principle and formula of the hierarchical analysis method are not repeated here. The present invention combines the results of the judgment matrix of each dimension and each indicator to calculate the weight value.

[0297] It can be seen that equipment operation has the highest weight among the nine dimensions, reaching 27%, followed by verification and inspection, with a weight of 20.54%, and warehousing and distribution has the lowest weight of 3.72%. The weights of equipment scrapping, equipment dismantling, procurement arrival, and environment are relatively close.

[0298] Table 2-1 Results of weight analysis of indicators in each dimension

[0299]

[0300]

[0301] The principle and formula of the entropy weight method will not be repeated here. The present invention combines the data of each dimension and each indicator to calculate the weight value. Since there is no data in the secondary dimension, only the weight value of the third-level indicator is calculated when the entropy weight method is used to calculate the weight. The results are as follows:

[0302] Table 2-2 Results of entropy weight method for each dimension indicator

[0303]

[0304] The subjective hierarchical method weight result value and the objective entropy weight result value in the basic algorithm are integrated, and the grey correlation is used to calculate the comprehensive weight of each dimension and index of the batch electric energy meter. Among them, the secondary dimension weight only has the analytic hierarchy process weight value, so the analytic hierarchy process weight is used as the final weight of the secondary dimension; the tertiary index adopts the grey correlation method to integrate the analytic hierarchy process weight and the entropy weight method weight results to form the final weight, as shown in the following figure 2-3.

[0305] Table 2-3 Results of different algorithm weights for batch electric energy meters

[0306]

[0307]

[0308] Step 3: Based on the indicators of each link in the whole life cycle of the electric energy meter determined in step 1, the weight of each indicator in each link obtained in step 2, and the comprehensive score of the electric energy meter status calculated by the weight of each link, perform power quality scoring analysis;

[0309] 3.1 Confirmation of the calculation logic of each indicator score

[0310] The present invention integrates the full cycle process of the electric energy meter, analyzes the historical maximum value (Ci) max and the minimum score (Ci) min of the index by statistically analyzing the existing historical data, and non-dimensionalizes the existing index Ci to calculate the score of each index.

[0311] Positive indicators:

[0312]

[0313] Among them C i Represents indicators in batch electric energy meters, such as the full performance test pass rate before production (%), the full performance test pass rate before delivery (%), and the random inspection pass rate after arrival (%).

[0314] Reverse indicators:

[0315]

[0316] Among them C i Indicates indicators in batch electric energy meters, such as electric energy meter failure rate and operating time.

[0317] 3.2 Calculate the score of indicators under each dimension

[0318] Based on the aforementioned index scoring calculation logic, the present invention combines the index data of the batch electric energy meter to calculate and generate the index scoring results of the batch electric energy meter, as shown in the following table (taking the arrival batch 31240506****1001 as an example):

[0319]

[0320] 3.4 Calculation of total quality score

[0321] Based on the above-mentioned scoring of each dimension and total quality score calculation logic, the present invention calculates the evaluation scores of each dimension and the total quality score of the 31240506****1001 batch of electric energy meters.

[0322] Step 4: Based on the calculation result of the electric energy meter quality score in step 3, the electric energy meter status interval is divided according to the historical electric energy meter quality score to obtain the analysis result of the electric energy meter status;

[0323] Then, based on the historical status score distribution, the present invention uses the Gath-Geva clustering algorithm to form the situation awareness level interval of batch electric energy meters, as shown in the following table:

[0324] Table 4-1 Gath-Geva cluster interval distribution table of batch electric energy meters

[0325] Serial number Minimum Maximum Situational Awareness Level Level Description 1 -- 77.95 C Focus 2 77.95 90.07 B General concerns 3 90.07 -- A Mild concern

[0326] Ultimately, timely perception of the current status of the electricity meter is achieved.

[0327] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0328] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples of computer-readable storage media (a non-exhaustive list) include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the foregoing. As used herein, a computer-readable storage medium is not to be interpreted as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0329] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0330] The computer program instructions for performing the operation of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed completely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be customized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present disclosure.

[0331] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A business-driven electric energy meter status analysis method, characterized in that: The method comprises: Step 1: Extract the common factors of the initial indicators of each link in the whole life cycle of the electric energy meter as the status indicators of each link in the whole life cycle of the electric energy meter; Step 2: Perform adaptive weight analysis based on the obtained state indicators to obtain the fusion weights of each link and each state indicator under each link; Step 3: Based on the status indicators of each link in the whole life cycle of the electric energy meter and the fusion weight of each status indicator, the comprehensive score of the electric energy meter status is calculated; Step 4: Based on the electric energy meter quality score calculation result in step 3, the electric energy meter status interval is divided according to the historical electric energy meter quality score to obtain the electric energy meter status analysis result.

2. The method for analyzing the status of an electric energy meter based on business drive according to claim 1, characterized in that: In step 1, each link of the whole life cycle of the electric energy meter includes procurement and arrival, equipment acceptance, verification and testing, warehousing and distribution, equipment installation, equipment operation, equipment removal, equipment scrapping and the environment in which it is located; The initial indicators of each link in the whole life cycle of the electric energy meter specifically include: Initial indicators for purchased goods include the pre-production full performance test pass rate and pre-production sample comparison pass rate; The initial indicators for equipment acceptance include the return and exchange ratio, pre-delivery sample comparison pass rate, pre-delivery full performance test pass rate, post-arrival sample comparison pass rate, and post-arrival random inspection test pass rate; The initial indicators of the verification and testing phase include the full inspection and acceptance pass rate, the electric energy meter re-inspection pass rate, and the proportion of verification and testing fault types; Initial indicators for the warehousing and distribution phase include new product inventory age, finished product inventory age, and distribution failure rate; initial indicators for the equipment installation phase include the failure rate of smart energy meters during installation and the proportion of quality failure types of smart energy meters during installation; The initial indicators of equipment operation include the one-time data collection success rate of low-voltage users, the daily data collection success rate of low-voltage users, the one-time data collection success rate of special transformer users, the daily data collection success rate of special transformer users, the qualified rate of random inspection of operation tables, the qualified rate of on-site verification, and the qualified rate of supervision random inspection; Initial indicators for equipment dismantling include the recycling rate of dismantled assets and the qualified rate of dismantled measurement assets; Initial indicators for equipment scrapping include the scrapping rate of measured assets due to quality factors; The initial indicators of the environment include maximum temperature and minimum temperature.

3. A business-driven electric energy meter status analysis method according to claim 1 or 2, characterized in that: In step 1, the common factors of the initial indicators of each link in the whole life cycle of the electric energy meter are extracted as the status indicators of each link in the whole life cycle of the electric energy meter, specifically: Calculate the correlation coefficient for every two initial indicator samples in each link. If the correlation between an initial indicator sample and any other initial indicator sample in the same link is weakly correlated, then the initial indicator is judged as a weakly correlated indicator, otherwise it is a strongly correlated indicator; For weak correlation indicators, they are directly used as common factors; For the strong correlation indicators, the Bartlett sphericity test and KMO test were performed, and the common factors were extracted for the strong correlation indicators that met the test criteria, otherwise they were directly used as common factors; The proposed common factors are used as status indicators of each link in the entire life cycle of the electricity meter.

4. The method for analyzing the status of an electric energy meter based on business drive according to claim 3 is characterized in that: The test criterion is as follows: Among them, P is the Bartlett sphericity test value, and γ is the KMO test value.

5. The method for analyzing the status of an electric energy meter based on business drive according to claim 3 is characterized in that: For the strong correlation indicators that meet the test criteria, the characteristic values ​​of the sample data of each factor in each link are calculated, and the variance contribution rate of each characteristic value and the cumulative variance contribution rate of each factor are calculated. When the cumulative variance contribution rate of the factor is greater than or equal to 88%, these factors are extracted as common factors and other factors are discarded; λ1+λ2+...+λ r = total variance; Cumulative variance contribution f = Variance contribution rate 1 + Variance contribution rate 2 + ... + Variance contribution rate f Among them, λ1, λ2, λ f , r Respectively represent the eigenvalues ​​of the 1st, 2nd, fth and rth factor sample data of a certain link, and r represents the total number of factors in the link; variance contribution rate f represents the variance contribution rate of the fth factor; When the current f cumulative variance contribution rates are greater than or equal to 88%, 1-f factors are extracted as the common factors of this link.

6. The method for analyzing the status of an electric energy meter based on business drive according to claim 1, characterized in that: In step 2, the following are included: 2.1 Calculate the initial weights of each link and each state indicator under different models; 2.2 Calculate the initial weight mean of each link calculated under different models and the initial weight mean of each state indicator calculated under different models; in, represents the initial weight mean of the jth link, represents the initial weight mean of the j-th state indicator, α ij represents the initial weight of the jth link under the i-th model, β ik represents the initial weight of the kth state indicator under the i-th model, and m is the number of models; 2.3 Calculate the absolute deviation of the initial weights of each link under different models and the absolute deviation of the initial weights of each state indicator under different models; Among them, S ij represents the absolute deviation of the j-th link weight under the i-th model, S i ' k represents the absolute deviation of the kth state indicator weight under the i-th model; 2.4 Based on the absolute deviation, calculate the fusion weight of each link and the fusion weight of each state indicator.

7. The method for analyzing the status of an electric energy meter based on business drive according to claim 6, characterized in that: The different models at least include a hierarchical analysis model and an entropy weight method model; In each model, the sum of the initial weights of each link is 1, and the sum of the initial weights of each state indicator in each link is 1.

8. The method for analyzing the status of an electric energy meter based on business drive according to claim 6, characterized in that: The fusion weight of each link and the fusion weight of each status indicator are calculated as follows: Among them, α j represents the fusion weight of the jth link, β k Represents the fusion weight of the k-th state indicator.

9. The method for analyzing the status of an electric energy meter based on business drive according to claim 1, characterized in that: In step 3, the comprehensive score of the electric energy meter status is calculated based on the status indicators of each link in the whole life cycle of the electric energy meter and the fusion weight of each status indicator, which is: 3.1 Collect the values ​​of each status indicator of the electric energy meter and calculate the score of the status indicator; 3.2 The status indicator score of each link is multiplied by the fusion weight of the status indicator corresponding to each status indicator, and then the status score value of each link is obtained; 3.3 The status score of each link is multiplied by the corresponding fusion weight of each link and then summarized to obtain the comprehensive score of the electricity meter status.

10. The method for analyzing the status of an electric energy meter based on business drive according to claim 9, characterized in that: In 3.1, the status indicator score includes a positive indicator score or a negative indicator score. The positive or negative score is determined according to the nature of the status indicator. The higher the value of the positive indicator, the more qualified the equipment is. The lower the value of the negative indicator, the more qualified the equipment is. Positive indicator score C i_score Calculated using the following formula: Negative indicator score C i_score Calculated using the following formula: Among them, C i is the state indicator value, (C i ) min , (C i ) max They are the minimum and maximum values ​​of the historical values ​​of the status indicator respectively.

11. The method for analyzing electric energy meter status based on business drive according to claim 1, characterized in that: In step 4, based on the electric energy meter quality score calculation result in step 3, the electric energy meter state interval is divided according to the historical electric energy meter quality score, specifically including: 4.1 Collect the historical quality score data of the electric energy meter and the corresponding quality status of the electric energy meter; 4.2 Set the quality status classification level of the electric energy meter, and use the number of classification levels as the number of categories; 4.3 Randomly select the historical scoring data of the electric energy meter as the initial analog center; 4.3 Calculate the probability calculation result of each electric energy meter that needs to be divided into state intervals belonging to each category, and update the center of the category according to the probability calculation result until the category center no longer changes or the change is less than the set threshold.

12. The method for analyzing the status of an electric energy meter based on business drive according to claim 11, characterized in that: The probability calculation result μ of the energy meter belonging to each category sw Calculate as follows: Among them, u sw represents the possibility that the sth electric energy meter belongs to the jth category, x s represents the comprehensive score of the sth electric energy meter status, v w represents the center score data of the w-th category, v q represents the cluster center score data of all categories, | | represents the distance calculation, t represents the number of categories, and p is the calculation index; x is the number of times the category center is updated.

13. An electric energy meter state analysis system based on the electric energy meter state analysis method according to any one of claims 1 to 12, comprising a calculation and confirmation module for each link and state index of the whole life cycle of the electric energy meter, a fusion weight calculation module for each link and each state index under each link, an electric energy meter state comprehensive score module and an electric energy meter state analysis module; characterized in that: The calculation and confirmation module of each link and state index of the whole life cycle of the electric energy meter extracts the common factors of the initial index of each link of the whole life cycle of the electric energy meter as the state index of each link of the whole life cycle of the electric energy meter; The fusion weight calculation module of each link and each state indicator under each link performs adaptive weight analysis based on the obtained state indicators to obtain the fusion weight of each link and each state indicator under each link; The comprehensive score module of the electric energy meter status is based on the status indicators of each link in the whole life cycle of the electric energy meter, and the fusion weight calculation of each status indicator to obtain the comprehensive score of the electric energy meter status; The electric energy meter status analysis module divides the electric energy meter status intervals based on the electric energy meter quality score calculation results and the historical electric energy meter quality scores to obtain the electric energy meter status analysis results.

Images