CN113626646A - Intelligent electric meter data compression method and device and electronic equipment - Google Patents
Intelligent electric meter data compression method and device and electronic equipment Download PDFInfo
- Publication number
- CN113626646A CN113626646A CN202110813537.XA CN202110813537A CN113626646A CN 113626646 A CN113626646 A CN 113626646A CN 202110813537 A CN202110813537 A CN 202110813537A CN 113626646 A CN113626646 A CN 113626646A
- Authority
- CN
- China
- Prior art keywords
- vector
- feature
- power consumption
- data
- data compression
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/901—Indexing; Data structures therefor; Storage structures
- G06F16/9024—Graphs; Linked lists
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/906—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/047—Probabilistic or stochastic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Probability & Statistics with Applications (AREA)
- Algebra (AREA)
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
Abstract
The application relates to the field of smart power grids, in particular to a data compression method and device for a smart meter and electronic equipment. And because data compression is carried out at one end of the intelligent electric meter, the data volume transmitted to the control center by the intelligent electric meter can be reduced, and the burden of the intelligent power grid system is reduced.
Description
Technical Field
The application relates to the field of smart power grids, in particular to a smart meter data compression method, a smart meter data compression device and electronic equipment.
Background
The smart grid is the intellectualization of the grid, and is established on the basis of an integrated, high-speed two-way communication network, and is applied through advanced sensing and measuring technology, advanced equipment technology, advanced control method and advanced decision support system technology.
In the smart grid, the smart meter can collect detailed information in the power utilization process of a user, including various analog information and digital information, the former includes, for example, a real-time current waveform in the power utilization process of the user, and the latter includes power utilization quantity, rated power and the like of the user at various time points, so that the control center can perform various controls such as load balancing and the like based on the detailed information.
However, since the power transmission line does not explicitly distinguish between the uplink and downlink channels, when the control center needs to collect the detailed information from the smart meter, an additional channel connection, such as a wireless channel, must be established, which obviously increases the system burden, such as the communication burden and the computational resource burden.
Therefore, an optimized communication scheme for a smart grid is desired.
Disclosure of Invention
Accordingly, in order to solve the above technical problems: since the power transmission line does not explicitly distinguish between the uplink and downlink channels, when the control center needs to collect the detailed information from the smart meter, an additional channel connection, such as a wireless channel, must be established, which obviously increases the system burden, such as the communication burden and the computational resource burden. The technical idea of the application is as follows: the data volume transmitted to the control center by the intelligent electric meter is reduced by data compression at one end of the intelligent electric meter, and the data compression proportion is set to be adaptively adjusted at one end of the intelligent electric meter based on the absolute electricity consumption and electricity consumption change data of a user in the electricity utilization process, so that the communication load of the intelligent power grid system is reduced.
Specifically, the solution of the application is as follows: the method comprises the steps of firstly, obtaining absolute power consumption data and power consumption change data of each intelligent electric meter in a preset time period, respectively constructing an absolute power consumption vector and a power consumption change vector, and then respectively multiplying the absolute power consumption vector and the power consumption change vector by transposes thereof to obtain a first matrix representing absolute power consumption correlation and a second matrix representing power consumption change correlation. And then, respectively passing the first matrix and the second matrix through a convolutional neural network to obtain a first characteristic diagram and a second characteristic diagram so as to express deep high-dimensional state correlation information of absolute power consumption and power consumption change among the intelligent electric meters.
Then, a default data compression ratio of each smart meter, such as a data compression ratio calculated in a previous time period, may be considered to be a prior probability under the condition represented by the first feature map and the second feature map, so as to obtain a prior probability vector. And respectively multiplying the prior probability vector serving as a query vector by a first feature map and a second feature map to respectively obtain a first feature vector and a second feature vector, and then according to a Bayesian probability formula, multiplying the feature value of each position of the prior probability vector by the feature value of each position of the first feature vector and then dividing the feature value of each position of the second feature vector to obtain a posterior probability value under the known absolute value feature distribution and the known variation value feature distribution, so that the posterior probability value forms a coding feature vector.
Then, by decoding the encoded eigenvector using a decoder including a plurality of fully-connected layers to obtain the length of the encoded eigenvector, the data compression ratio to be currently set can be obtained based on the eigenvalue of each position of the decoded eigenvector.
Based on the technical scheme, the following technical effects can be achieved: firstly, in the technical scheme of the application, data compression is carried out at one end of the intelligent electric meter to reduce the data volume transmitted to the control center by the intelligent electric meter, so that the system load is reduced.
Secondly, considering that the real-time current waveform in the power utilization process of the user contains abundant information, and such analog information can be converted into digital information for transmission through fourier transform, while the number of frequency components selected in the fourier transform will significantly affect the overall data volume, by setting a data compression method based on the number of frequency components selected in the fourier transform, the data compression ratio can be increased while ensuring effective information as much as possible when data compression is performed.
Thirdly, because the selected frequency component data is obviously related to the information quantity required to be transmitted by each intelligent electric meter, and the information quantity is related to the absolute power consumption and the power consumption change of each intelligent electric meter, namely, the larger the absolute power consumption or the larger the power consumption change is, the larger the information quantity required to be transmitted is, the absolute power consumption and the power consumption change data of each intelligent electric meter are selected as the data base in the scheme of the application.
Fourthly, deep high-dimensional state correlation information of absolute power consumption and power consumption change data among the intelligent electric meters can be sufficiently mined by using the convolutional neural network, so that a subsequently set data compression ratio can be strongly correlated with the absolute power consumption and power consumption change data among the intelligent electric meters, and the accuracy and the adaptability of data compression ratio setting are improved in such a way.
Fifthly, the data compression ratio of each intelligent electric meter is determined by fully utilizing the absolute power consumption and the characteristic distribution of power consumption change data of each intelligent electric meter based on the Bayesian idea, so that the finally set data compression ratio can fully consider the power consumption information of each intelligent electric meter, and the set data compression ratio can be more adaptive to the communication load of the whole intelligent power grid.
According to one aspect of the application, a method for compressing data of a smart meter is provided, which comprises the following steps:
acquiring absolute power consumption data and power consumption change data of each intelligent electric meter in a preset time period;
constructing the absolute power consumption data and the power consumption change data of each intelligent electric meter into an absolute power consumption vector and a power consumption change vector respectively;
multiplying the absolute power consumption vector and the power consumption change vector by transposes thereof respectively to obtain a first matrix for representing absolute power consumption correlation and a second matrix for representing power consumption transformation correlation;
obtaining a first feature map and a second feature map from the first matrix and the second matrix respectively by using a convolutional neural network;
acquiring a default data compression ratio of each intelligent electric meter and constructing the default data compression ratio of each intelligent electric meter as a prior probability vector;
taking the prior probability vector as a query vector to perform matrix multiplication with the first feature map and the second feature map respectively to obtain a first feature vector and a second feature vector;
calculating posterior probability values corresponding to the feature values of each position in the prior probability vectors according to a Bayesian formula based on the feature values of each position in the first feature vector and the second feature vector to obtain coding feature vectors formed by the posterior probability values of each position;
decoding the encoded eigenvector using a decoder to obtain a decoded eigenvector of equal length to the encoded eigenvector; and
and determining the data compression ratio to be currently set by each intelligent electric meter based on the characteristic value of each position of the decoding characteristic vector.
According to another aspect of the present application, there is also provided a smart meter data compression apparatus, including:
the electric quantity data unit is used for acquiring absolute electric quantity data and electric quantity change data of each intelligent electric meter in a preset time period;
the vector construction unit is used for constructing the absolute power consumption data and the power consumption change data of each intelligent ammeter into an absolute power consumption vector and a power consumption change vector respectively;
a matrix construction unit, configured to multiply the absolute power consumption vector and the power consumption change vector by transposes thereof, respectively, to obtain a first matrix representing an absolute power consumption correlation and a second matrix representing a power consumption transformation correlation;
a feature map extraction unit, configured to obtain a first feature map and a second feature map from the first matrix and the second matrix, respectively, using a convolutional neural network;
the prior probability vector construction unit is used for acquiring the default data compression proportion of each intelligent electric meter and constructing the default data compression proportion of each intelligent electric meter into a prior probability vector;
the vector query unit is used for taking the prior probability vector as a query vector to carry out matrix multiplication with the first feature map and the second feature map respectively so as to obtain a first feature vector and a second feature vector;
the encoding feature vector generating unit is used for calculating posterior probability values corresponding to the feature values of each position in the prior probability vectors according to a Bayesian formula on the basis of the feature values of each position in the first feature vector and the second feature vector so as to obtain encoding feature vectors formed by the posterior probability values of each position;
a decoding unit configured to decode the encoded eigenvector using a decoder to obtain a decoded eigenvector of equal length to the encoded eigenvector; and
and the data compression ratio setting unit is used for determining the data compression ratio to be set currently by each intelligent electric meter based on the characteristic value of each position of the decoding characteristic vector.
According to still another aspect of the present application, there is provided an electronic apparatus including: a processor; and a memory in which computer program instructions are stored, which, when executed by the processor, cause the processor to execute the smart meter data compression method as described above.
According to yet another aspect of the present application, there is provided a computer readable medium having stored thereon computer program instructions, which when executed by a processor, cause the processor to execute the smart meter data compression method as described above.
Drawings
The above and other objects, features and advantages of the present application will become more apparent by describing in more detail embodiments of the present application with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of the embodiments of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings, like reference numbers generally represent like parts or steps.
Fig. 1 illustrates an application scenario of a data compression method for a smart meter according to an embodiment of the application.
Fig. 2 illustrates a flowchart of a data compression method of a smart meter according to an embodiment of the present application.
Fig. 3 is a schematic diagram illustrating a system architecture of a data compression method for a smart meter according to an embodiment of the present application.
Fig. 4 illustrates a flowchart of acquiring absolute power consumption data and power consumption change data of each smart meter in a predetermined time period in the smart meter data compression method according to the embodiment of the application.
Fig. 5 is a flowchart illustrating that in the method for compressing data of a smart meter according to the embodiment of the present application, a data compression ratio to be currently set for each smart meter is determined based on the feature value of each position of the decoded feature vector.
Fig. 6 illustrates a block diagram of a smart meter data compression apparatus according to an embodiment of the present application.
Fig. 7 illustrates a block diagram of a power data unit of the smart meter data compression apparatus according to an embodiment of the present application.
Fig. 8 illustrates a block diagram of a data compression ratio setting unit of the smart meter data compression apparatus according to an embodiment of the present application.
FIG. 9 illustrates a block diagram of an electronic device in accordance with an embodiment of the present application.
Detailed Description
Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be understood that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and that the present application is not limited by the example embodiments described herein.
Overview of a scene
As described above, in the smart grid, the smart meter can collect detailed information of the user during power utilization, including various analog information and digital information, the former includes, for example, a real-time current waveform of the user during power utilization, and the latter includes the power utilization amount, rated power, etc. of the user at each time point, so that the control center can perform various controls such as load balancing based on the detailed information.
However, since the power transmission line does not explicitly distinguish between the uplink and downlink channels, when the control center needs to collect the detailed information from the smart meter, an additional channel connection, such as a wireless channel, must be established, which obviously increases the system burden, such as the communication burden and the computational resource burden.
In order to solve the technical problem, according to the technical scheme, data compression is performed at one end of the intelligent electric meter to reduce the data volume transmitted to the control center by the intelligent electric meter, and a data compression ratio is set to be adaptively adjusted at one end of the intelligent electric meter based on the absolute power consumption and power consumption change data of a user in the power utilization process, so that the communication load of the intelligent power grid system is reduced.
Based on this, the application provides a smart meter data compression method, which includes: acquiring absolute power consumption data and power consumption change data of each intelligent electric meter in a preset time period; constructing the absolute power consumption data and the power consumption change data of each intelligent electric meter into an absolute power consumption vector and a power consumption change vector respectively; multiplying the absolute power consumption vector and the power consumption change vector by transposes thereof respectively to obtain a first matrix for representing absolute power consumption correlation and a second matrix for representing power consumption transformation correlation; obtaining a first feature map and a second feature map from the first matrix and the second matrix respectively by using a convolutional neural network; acquiring a default data compression ratio of each intelligent electric meter and constructing the default data compression ratio of each intelligent electric meter as a prior probability vector; taking the prior probability vector as a query vector to perform matrix multiplication with the first feature map and the second feature map respectively to obtain a first feature vector and a second feature vector; calculating posterior probability values corresponding to the feature values of each position in the prior probability vectors according to a Bayesian formula based on the feature values of each position in the first feature vector and the second feature vector to obtain coding feature vectors formed by the posterior probability values of each position; decoding the encoded eigenvector using a decoder to obtain a decoded eigenvector of equal length to the encoded eigenvector; and determining the data compression ratio to be set currently by each intelligent electric meter based on the characteristic value of each position of the decoding characteristic vector.
Further, as shown in fig. 1, in an application scenario of the method for compressing data of a smart meter, a communication link exists between each smart meter (e.g., M as illustrated in fig. 1) and a control center (e.g., S as illustrated in fig. 1), for example, the smart meter communicates in a wired or wireless manner, and the smart meter compresses data before transmitting power consumption information to reduce the amount of data transmitted by the smart meter to the control center. In addition, in the application scenario of the application, the smart meter adaptively adjusts the data compression ratio of the smart meter through a smart meter data compression algorithm to meet the load condition of the smart grid.
Specifically, the method acquires absolute power consumption data and power consumption change data of each intelligent electric meter in a preset time period at one end of the intelligent electric meter. And then, processing the absolute power consumption data and the power consumption change data by using a smart meter data compression algorithm to determine the data compression ratio to be set currently for each smart meter, so that the finally set data compression ratio can be more adaptive to the communication load of the whole smart power grid.
Having described the general principles of the present application, various non-limiting embodiments of the present application will now be described with reference to the accompanying drawings.
Exemplary method
Fig. 2 illustrates a flowchart of a data compression method of a smart meter according to an embodiment of the present application. As shown in fig. 2, the method for compressing data of a smart meter according to the embodiment of the application includes the steps of: s110, acquiring absolute electricity consumption data and electricity consumption change data of each intelligent electric meter in a preset time period; s120, constructing the absolute power consumption data and the power consumption change data of each intelligent ammeter into an absolute power consumption vector and a power consumption change vector respectively; s130, multiplying the absolute power consumption vector and the power consumption change vector by the transpose of the absolute power consumption vector and the power consumption change vector respectively to obtain a first matrix for representing absolute power consumption correlation and a second matrix for representing power consumption conversion correlation; s140, obtaining a first characteristic diagram and a second characteristic diagram from the first matrix and the second matrix respectively by using a convolutional neural network; s150, acquiring the default data compression ratio of each intelligent electric meter and constructing the default data compression ratio of each intelligent electric meter into a prior probability vector; s160, taking the prior probability vector as a query vector to perform matrix multiplication with the first feature map and the second feature map respectively to obtain a first feature vector and a second feature vector; s170, calculating posterior probability values corresponding to the feature values of each position in the prior probability vectors according to a Bayesian formula based on the feature values of each position in the first feature vector and the second feature vector to obtain coding feature vectors formed by the posterior probability values of each position; s180, decoding the coded feature vector by using a decoder to obtain a decoded feature vector with the length equal to that of the coded feature vector; and S190, determining the data compression ratio to be set currently by each intelligent electric meter based on the characteristic value of each position of the decoding characteristic vector.
Fig. 3 is a schematic diagram illustrating a system architecture of a data compression method for a smart meter according to an embodiment of the present application. As shown IN fig. 3, IN the network architecture of the data compression method for the smart meters, firstly, the obtained absolute power consumption data (for example, IN0 as illustrated IN fig. 3) and power consumption change data (for example, IN1 as illustrated IN fig. 3) of each smart meter IN a predetermined period are respectively constructed into an absolute power consumption vector (for example, V1 as illustrated IN fig. 3) and a power consumption change vector (for example, V2 as illustrated IN fig. 3), and particularly, IN the embodiment of the present application, the absolute power consumption data and the power consumption change data of each smart meter IN the predetermined period are generated by an analog-to-digital converter based on a real-time current waveform transmitted to the smart meter by a user IN a power utilization process. Then, the absolute power consumption vector and the power consumption change vector are multiplied by their transposes, respectively, to obtain a first matrix (e.g., M1 as illustrated in fig. 3) for representing absolute power consumption correlations and a second matrix (e.g., M2 as illustrated in fig. 3) for representing power consumption transformation correlations. Then, a convolutional neural network (e.g., CNN as illustrated in fig. 3) is used to obtain a first signature (e.g., F1 as illustrated in fig. 3) and a second signature (e.g., F2 as illustrated in fig. 3) from the first matrix and the second matrix, respectively. Then, the default data compression ratio for obtaining each smart meter is constructed into a prior probability vector (for example, Vp as illustrated in fig. 3). Then, the prior probability vector is taken as a query vector to be subjected to matrix multiplication with the first feature map and the second feature map respectively to obtain a first feature vector (for example, Vf1 as illustrated in fig. 3) and a second feature vector (for example, Vf2 as illustrated in fig. 3). Then, a posterior probability value corresponding to the feature value of each position in the prior probability vector is calculated based on the feature value of each position in the first feature vector and the second feature vector according to a bayesian formula to obtain a coded feature vector (for example, Ven as illustrated in fig. 3) composed of the posterior probability values of the positions. The encoded feature vector is then decoded using a Decoder (e.g., a Decoder as illustrated in fig. 3) to obtain a decoded feature vector (e.g., Vde as illustrated in fig. 3) of equal length to the encoded feature vector. Finally, a data compression ratio (for example, R1 to Rn as illustrated in fig. 3) to be currently set by each of the smartmeters is determined based on the eigenvalues of each position of the decoded eigenvector.
In step S110, the absolute power consumption data and power consumption change data of each smart meter in a predetermined period of time are acquired. Particularly, in the technical scheme of the application, the data volume transmitted to the control center by the intelligent electric meter is reduced by data compression at one end of the intelligent electric meter. In order to determine the data compression ratio, first, absolute power consumption data and power consumption change data of each smart meter in a predetermined time period are acquired.
Correspondingly, in the embodiment of the present application, the process of acquiring the absolute power consumption data and the power consumption change data of each smart meter in a predetermined time period includes: firstly, acquiring real-time current waveforms transmitted to each intelligent electric meter during the electricity utilization process of the user. That is, in the technical scheme of this application, every smart electric meter's absolute power consumption data and power consumption change data are based on the real-time current waveform generation that the user transmitted to smart electric meter in the power consumption process.
It should be understood that the real-time current waveform during the power consumption process of the user is an analog signal, and needs to be converted into a digital signal for propagation. Accordingly, in the embodiment of the present application, fourier transform is performed on each of the real-time current waveforms to obtain the current value of each of the real-time current waveforms at the respective harmonic frequencies, that is, analog-to-digital conversion is performed on the real-time current waveforms by fourier transform. Thus, the current value of the real-time current waveform at each harmonic frequency can be determined as the absolute power consumption data, and the difference between the current values of the real-time current waveform at adjacent harmonic frequencies can be determined as the power consumption change data.
In particular, in the embodiment of the present application, it is considered that a real-time current waveform in a power utilization process of a user contains abundant information, and such analog information can be converted into digital information by fourier transform for transmission, and the number of frequency components selected in the fourier transform will significantly affect the overall data amount, so that, in some examples of the present application, by providing a data compression method based on the number of frequency components selected in the fourier transform, it is possible to increase a data compression ratio while securing effective information as much as possible in data compression. That is, in some examples of the present application, the process of acquiring the absolute power consumption data and the power consumption change data of each smart meter in a predetermined time period further includes: and intercepting the numerical values of the first N harmonic frequencies from the current value of each real-time current waveform at the harmonic frequency.
Here, since the selected frequency component data is obviously related to the information amount required to be transmitted by each smart meter, and the information amount is related to the absolute power consumption and the power consumption change of each smart meter, that is, the larger the absolute power consumption or the larger the power consumption change is, the larger the information amount required to be transmitted is, the absolute power consumption and the power consumption change data of each smart meter are selected as the data base in the solution of the present application.
Fig. 4 illustrates a flowchart of acquiring absolute power consumption data and power consumption change data of each smart meter in a predetermined time period in the smart meter data compression method according to the embodiment of the application. As shown in fig. 4, in the embodiment of the present application, acquiring absolute power consumption data and power consumption change data of each smart meter in a predetermined time period includes the steps of: s210, acquiring a real-time current waveform of a user in the electricity utilization process transmitted to each intelligent electricity meter; s220, performing Fourier transform on each real-time current waveform to obtain a current value of each real-time current waveform under each harmonic frequency; s230, intercepting numerical values of the first N harmonic frequencies from current values of each real-time current waveform at each harmonic frequency, and S240, determining the current values of the real-time current waveforms at each harmonic frequency as absolute power consumption data; and S250, determining the difference of current values of the real-time current waveform under adjacent harmonic frequencies as the power consumption change data.
In steps S120 and S130, the absolute power consumption data and the power consumption change data of each smart meter are respectively constructed as an absolute power consumption vector and a power consumption change vector, and the absolute power consumption vector and the power consumption change vector are respectively multiplied by transposes thereof to obtain a first matrix for representing absolute power consumption correlation and a second matrix for representing power consumption transformation correlation. Here, the characteristic values of the respective positions in the first matrix represent the correlation between the absolute power consumption of the respective smart meters, for example, the characteristic value of the 1 st row and the 3 rd column of the first matrix represents the correlation between the absolute power consumption of the third smart meter and the absolute power consumption of the first smart meter; similarly, the characteristic values of the positions in the second matrix represent the correlation between the electricity consumption change quantities of the intelligent electric meters.
In step S140, a first feature map and a second feature map are obtained from the first matrix and the second matrix, respectively, using a convolutional neural network. Namely, the first matrix and the second matrix are respectively passed through a convolutional neural network to obtain a first characteristic diagram and a second characteristic diagram so as to express deep high-dimensional state correlation information of absolute power consumption and power consumption change among the intelligent electric meters.
It should be understood that the deep high-dimensional state correlation information of the absolute power consumption and power consumption change data among the intelligent electric meters can be sufficiently mined by using the convolutional neural network, so that the subsequently set data compression ratio can be strongly correlated with the absolute power consumption and power consumption change data among the intelligent electric meters, and the accuracy and the adaptability of the data compression ratio setting are improved.
Those of ordinary skill in the art will appreciate that convolutional neural networks have superior performance in extracting local features. In particular, in the embodiment of the present application, the convolutional neural network obtains a first feature map and a second feature map from the first matrix and the second matrix respectively according to the following formulas;
wherein the formula is: f. ofi＝sigmoid(Ni×fi-1+Bi)
Wherein f isi-1Is the input of the i-th convolutional neural network, fiIs the output of the ith convolutional neural network, NiIs the convolution kernel of the ith convolutional neural network, and BiAnd (4) for the bias vector of the ith layer of convolutional neural network, sigmoid indicates that the activation function is a sigmoid activation function.
In this way, the finally obtained feature values of the first feature map and the second feature map at the respective positions are within the interval of 0 to 1, that is, the first feature map and the second feature map are projected into a probability space through a sigmoid activation function, so as to facilitate the application and calculation of the subsequent bayesian formula.
Of course, in other examples of the present application, the convolutional neural network may have other structures, for example, only the activation function of the last layer of the convolutional neural network is set as a sigmoid activation function, and the types of the activation functions of other layers of the convolutional neural network are not limited, so that the first feature map and the second feature map can be projected into a probability space to facilitate the application and calculation of the subsequent bayesian formula, and the selection of different activation functions can improve and optimize the capability of the convolutional neural network in terms of nonlinear expression.
That is, in other examples of the present application, in obtaining a first feature map and a second feature map from the first matrix and the second matrix, respectively, using a convolutional neural network, a last layer of the convolutional neural network is activated with a Sigmoid function to convert a feature value of each position in the first feature map and the second feature map into an interval of 0 to 1.
It should be noted that, in other examples of the present application, feature values of respective positions in the first feature map and the second feature map may also be projected into a probability space in other manners. For example, in other examples of the present application, the network structure of the convolutional neural network may not be changed, and the first feature map and the second feature map may be input into a Softmax-like classification function to transform the first feature map and the second feature map into a probability space after the first feature map and the second feature map are generated.
Here, the Softmax-like classification function is yi ═ exp (xi)/Σ exp (xi), where xi is a feature value of each position in the first feature map or the second feature map. Based on the functional characteristics of the Softmax-like classification function, the sum of the feature values of the positions in the first feature map converted into the probability space is 1, and the sum of the feature values of the positions in the second feature map converted into the probability space is 1, so that the feature values of the positions in the first feature map and the second feature map further represent the distribution probability of each position relative to the global state, and the calculation form of conditional probability-total probability of the Bayesian formula is more satisfied.
In step S150, a default data compression ratio of each smart meter is obtained and is configured as a prior probability vector. In a specific implementation, the data compression ratio calculated in the previous time period can be used as the default data compression of each smart meter, and then the default data compression ratio of each smart meter is arranged into a vector according to the dimension of the smart meter. It should be understood that the historical data compression ratio of each smart meter can be regarded as a prior probability under the condition represented by the first feature map and the second feature map, and therefore, the obtained vector is a prior probability vector.
In step S160, the prior probability vector is used as a query vector to perform matrix multiplication with the first feature map and the second feature map respectively to obtain a first feature vector and a second feature vector. That is, the prior probability vector is multiplied by a first feature map and a second feature map as a query vector to fuse information in the first feature map and information in the second feature map into the prior probability vector to obtain the first feature vector and the second feature vector, respectively.
In step S170, a posterior probability value corresponding to the feature value of each position in the prior probability vector is calculated based on the feature value of each position in the first feature vector and the second feature vector according to a bayesian formula, so as to obtain a coded feature vector consisting of the posterior probability values of the positions. Specifically, in the embodiment of the present application, based on the feature value of each position in the first feature vector and the second feature vector, a posterior probability value corresponding to the feature value of each position in the prior probability vector is calculated according to the following formula, so as to obtain a coded feature vector composed of posterior probability values of respective positions;
wherein the formula is: yi ═ Σxi∈J，ai∈L，bj∈Kxi ai/bi, wherein xi represents a characteristic value of each position in the prior probability vector; ai represents the eigenvalue of each position in the first eigenvector and bi represents the eigenvalue of each position in the second eigenvector
That is, after the first feature vector and the second feature vector are obtained, the feature value of each position of the prior probability vector is multiplied by the feature value of each position of the first feature vector and then divided by the feature value of each position of the second feature vector according to a bayesian probability formula to obtain posterior probability values under the known absolute value feature distribution and the known variation value feature distribution, so that the posterior probability values form the coding feature vector.
It should be understood that the characteristic distribution of the absolute power consumption and power consumption change data of each smart meter is fully utilized to determine the data compression ratio of each smart meter based on the bayesian idea, so that the finally set data compression ratio can fully consider the power consumption information of each smart meter, and the set data compression ratio can be more adaptive to the communication load of the whole smart grid.
In step S180, the encoded feature vector is decoded using a decoder to obtain a decoded feature vector having a length equal to the encoded feature vector. In a specific example, the decoder is a decoder composed of a plurality of fully-connected layers, and the information of each position in the coded feature vector can be fully utilized in the decoding process to improve the decoding precision.
In step S190, a data compression ratio to be currently set for each of the smartmeters is determined based on the feature values of each position of the decoded feature vector. In this embodiment of the application, the process of determining the data compression ratio to be currently set for each smart meter based on the feature value of each position of the decoded feature vector includes: firstly, the decoding feature vector is converted into a probability space by a Softmax-like classification function, so that the feature value of each position in the decoding feature vector is within an interval of 0 to 1. In this way, the eigenvalue of each position in the decoded eigenvector is used to represent the bandwidth ratio that should be allocated to the corresponding smart meter.
Then, based on the total data transmission bandwidth and the decoded feature vector converted into probability space, the allocated data bandwidth of each smart meter is determined. Finally, calculating the data compression ratio of each intelligent electric meter based on the allocated data bandwidth and the current data quantity of each intelligent electric meter, namely dividing the current data quantity of each intelligent electric meter by the allocated data bandwidth and determining the quotient value of the current data quantity of each intelligent electric meter as the data compression ratio of each intelligent electric meter.
Fig. 5 is a flowchart illustrating that in the method for compressing data of a smart meter according to the embodiment of the present application, a data compression ratio to be currently set for each smart meter is determined based on the feature value of each position of the decoded feature vector. As shown in fig. 5, in the embodiment of the present application, determining the data compression ratio to be currently set by each smart meter based on the feature value of each position of the decoded feature vector includes the steps of: s310, converting the decoding feature vector into a probability space by a Softmax-like classification function so that the feature value of each position in the decoding feature vector is within a range from 0 to 1; s320, determining the distribution data bandwidth of each intelligent electric meter based on the total data transmission bandwidth and the decoding feature vector converted into the probability space; and S330, calculating a data compression ratio of each intelligent electric meter based on the allocated data bandwidth and the current data amount of each intelligent electric meter.
In summary, the data compression method for the smart meter according to the embodiment of the present application is illustrated, which reduces the data amount transmitted to the control center by the smart meter by performing data compression at one end of the smart meter, and sets a data compression ratio to be adaptively adjusted based on the absolute power consumption and power consumption change data of the user during the power consumption process at one end of the smart meter, so as to reduce the communication load of the smart grid system.
Exemplary devices
Fig. 6 illustrates a block diagram of a smart meter data compression apparatus according to an embodiment of the present application.
As shown in fig. 6, the smart meter data compression apparatus 600 according to the embodiment of the present application includes: the electric quantity data unit 610 is used for acquiring absolute electricity consumption data and electricity consumption change data of each intelligent electric meter in a preset time period; the vector construction unit 620 is configured to respectively construct the absolute power consumption data and the power consumption change data of each smart meter into an absolute power consumption vector and a power consumption change vector; a matrix construction unit 630, configured to multiply the absolute power consumption vector and the power consumption change vector by transposes thereof, respectively, to obtain a first matrix representing an absolute power consumption correlation and a second matrix representing a power consumption transformation correlation; a feature map extracting unit 640, configured to obtain a first feature map and a second feature map from the first matrix and the second matrix, respectively, using a convolutional neural network; a priori probability vector construction unit 650, configured to obtain a default data compression ratio of each smart meter and construct the default data compression ratio of each smart meter as a priori probability vector; a vector querying unit 660, configured to perform matrix multiplication on the prior probability vector as a query vector and the first feature map and the second feature map respectively to obtain a first feature vector and a second feature vector; a coding feature vector generating unit 670, configured to calculate a posterior probability value corresponding to the feature value of each position in the prior probability vector based on the feature value of each position in the first feature vector and the second feature vector according to a bayesian formula, so as to obtain a coding feature vector formed by the posterior probability values of each position; a decoding unit 680 configured to decode the encoded eigenvector using a decoder to obtain a decoded eigenvector with a length equal to the encoded eigenvector; and a data compression ratio setting unit 690 for determining a data compression ratio to be currently set for each of the smartmeters based on the eigenvalues of each position of the decoded eigenvector.
In one example, in the above-mentioned smart meter data compression apparatus 600, as shown in fig. 7, the electric quantity data unit 610 includes: a real-time current waveform obtaining subunit 611, configured to obtain a real-time current waveform of a user during power consumption, which is transmitted to each of the smart meters; an analog-to-digital conversion unit 612, configured to perform fourier transform on each real-time current waveform to obtain a current value of each real-time current waveform at each harmonic frequency; a frequency clipping sub-unit 613, configured to clip values at the first N harmonic frequencies from current values of each of the real-time current waveforms at the respective harmonic frequencies; an absolute power consumption data determining subunit 614, configured to determine a current value of the real-time current waveform at each harmonic frequency as the absolute power consumption data; and a power consumption change data determining subunit 615 configured to determine a difference between current values of the real-time current waveform at adjacent harmonic frequencies as the power consumption change data.
In one example, in the above-mentioned smart meter data compression apparatus 600, the last layer of the convolutional neural network is activated by a Sigmoid function to convert the feature value of each position in the first feature map and the second feature map into an interval of 0 to 1.
In one example, in the above-mentioned smart meter data compression apparatus 600, the convolutional neural network obtains a first characteristic map and a second characteristic map from the first matrix and the second matrix respectively according to the following formulas;
wherein the formula is: f. ofi＝sigmoid(Ni×fi-1+Bi)
Wherein f isi-1Is the input of the i-th convolutional neural network, fiIs the output of the ith convolutional neural network, NiIs the convolution kernel of the ith convolutional neural network, and BiAnd (4) for the bias vector of the ith layer of convolutional neural network, sigmoid indicates that the activation function is a sigmoid activation function.
In an example, in the above smart meter data compression apparatus 600, the feature map extraction unit 640 is further configured to input the first feature map and the second feature map into a Softmax classification function to convert the first feature map and the second feature map into a probability space, and a sum of feature values of each position in the first feature map converted into the probability space is 1, and a sum of feature values of each position in the second feature map converted into the probability space is 1. Wherein the Softmax-like classification function is: yi ═ exp (xi)/Σ exp (xi), where xi is a feature value of each position in the first feature map or the second feature map.
In an example, in the foregoing smart meter data compression apparatus 600, the encoding feature vector generating unit 670 is further configured to: calculating a posterior probability value corresponding to the feature value of each position in the prior probability vector according to the following formula based on the feature value of each position in the first feature vector and the second feature vector so as to obtain a coding feature vector consisting of the posterior probability values of all the positions;
wherein the formula is: yi ═ Σxi∈J，ai∈L，bj∈Kxi ai/bi, wherein xi represents a characteristic value of each position in the prior probability vector; ai represents the eigenvalue of each position in the first eigenvector and bi represents the eigenvalue of each position in the second eigenvector.
In one example, in the above-mentioned smart meter data compression apparatus 600, as shown in fig. 8, the data compression ratio setting unit 690 includes: a probability transforming unit 691 configured to transform the decoded feature vector into a probability space with a Softmax-like classification function such that feature values of respective positions in the decoded feature vector are within an interval of 0 to 1; an allocated data bandwidth subunit 692 for determining an allocated data bandwidth for each of the smart meters based on a total data transmission bandwidth and the decoded feature vector translated to a probability space; and a data compression ratio calculation subunit 693, configured to calculate a data compression ratio of each of the smart meters based on the allocated data bandwidth and the current data size of each of the smart meters.
Here, it will be understood by those skilled in the art that the detailed functions and operations of the respective units and modules in the above-described smart meter data compression apparatus 600 have been described in detail in the above description of the smart meter data compression method with reference to fig. 1 to 5, and thus, a repetitive description thereof will be omitted.
As described above, the smart meter data compression apparatus 600 according to the embodiment of the present application may be implemented in various terminal devices, such as a smart meter and the like. In one example, the smart meter data compression apparatus 600 according to the embodiment of the present application may be integrated into a terminal device as a software module and/or a hardware module. For example, the smart meter data compression apparatus 600 may be a software module in the operation apparatus of the terminal device, or may be an application developed for the terminal device; of course, the smart meter data compression apparatus 600 may also be one of many hardware modules of the terminal device.
Alternatively, in another example, the smart meter data compression apparatus 600 and the terminal device may be separate devices, and the smart meter data compression apparatus 600 may be connected to the terminal device through a wired and/or wireless network and transmit the interaction information according to an agreed data format.
Exemplary electronic device
Next, an electronic apparatus according to an embodiment of the present application is described with reference to fig. 9.
FIG. 9 illustrates a block diagram of an electronic device in accordance with an embodiment of the present application.
As shown in fig. 9, the electronic device 10 includes one or more processors 11 and memory 12.
The processor 11 may be a Central Processing Unit (CPU) or other form of processing unit having data processing capabilities and/or instruction execution capabilities, and may control other components in the electronic device 10 to perform desired functions.
In one example, the electronic device 10 may further include: an input device 13 and an output device 14, which are interconnected by a bus device and/or other form of connection mechanism (not shown).
The input device 13 may include, for example, a keyboard, a mouse, and the like.
The output device 14 can output various information including a data compression ratio and the like to the outside. The output devices 14 may include, for example, a display, speakers, a printer, and a communication network and its connected remote output devices, among others.
Of course, for simplicity, only some of the components of the electronic device 10 relevant to the present application are shown in fig. 9, and components such as buses, input/output interfaces, and the like are omitted. In addition, the electronic device 10 may include any other suitable components depending on the particular application.
Exemplary computer program product and computer-readable storage Medium
In addition to the above-described methods and apparatuses, embodiments of the present application may also be a computer program product comprising computer program instructions that, when executed by a processor, cause the processor to perform the steps of the method for compressing data of a smart meter according to various embodiments of the present application described in the above-mentioned "exemplary methods" section of this specification.
The computer program product may be written with program code for performing the operations of embodiments of the present application in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server.
Furthermore, embodiments of the present application may also be a computer-readable storage medium having stored thereon computer program instructions, which, when executed by a processor, cause the processor to perform the steps in the smart meter data compression method according to various embodiments of the present application described in the "exemplary methods" section above in this specification.
The computer-readable storage medium may take any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may include, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor device, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
Claims (10)
1. A data compression method for a smart meter is characterized by comprising the following steps:
acquiring absolute power consumption data and power consumption change data of each intelligent electric meter in a preset time period;
constructing the absolute power consumption data and the power consumption change data of each intelligent electric meter into an absolute power consumption vector and a power consumption change vector respectively;
multiplying the absolute power consumption vector and the power consumption change vector by transposes thereof respectively to obtain a first matrix for representing absolute power consumption correlation and a second matrix for representing power consumption transformation correlation;
obtaining a first feature map and a second feature map from the first matrix and the second matrix respectively by using a convolutional neural network;
acquiring a default data compression ratio of each intelligent electric meter and constructing the default data compression ratio of each intelligent electric meter as a prior probability vector;
taking the prior probability vector as a query vector to perform matrix multiplication with the first feature map and the second feature map respectively to obtain a first feature vector and a second feature vector;
calculating posterior probability values corresponding to the feature values of each position in the prior probability vectors according to a Bayesian formula based on the feature values of each position in the first feature vector and the second feature vector to obtain coding feature vectors formed by the posterior probability values of each position;
decoding the encoded eigenvector using a decoder to obtain a decoded eigenvector of equal length to the encoded eigenvector; and
and determining the data compression ratio to be currently set by each intelligent electric meter based on the characteristic value of each position of the decoding characteristic vector.
2. The smart meter data compression method according to claim 1, wherein the acquiring of the absolute power consumption data and the power consumption change data of each smart meter in a predetermined time period comprises:
acquiring a real-time current waveform of a user in the electricity utilization process transmitted to each intelligent electricity meter;
performing Fourier transform on each real-time current waveform to obtain a current value of each real-time current waveform at each harmonic frequency;
determining the current value of the real-time current waveform under each harmonic frequency as the absolute power consumption data; and
and determining the difference of the current values of the real-time current waveform under the adjacent harmonic frequencies as the power consumption change data.
3. The smart meter data compression method of claim 2, wherein after fourier transforming each of the real-time current waveforms to obtain a current value of each of the real-time current waveforms at a respective harmonic frequency, further comprising:
and intercepting the numerical values of the first N harmonic frequencies from the current value of each real-time current waveform at the harmonic frequency.
4. The smart meter data compression method according to claim 1, wherein in obtaining the first and second feature maps from the first and second matrices, respectively, using a convolutional neural network, a last layer of the convolutional neural network is activated with a Sigmoid function to convert the feature value of each position in the first and second feature maps to an interval of 0 to 1.
5. The smart meter data compression method of claim 1, wherein obtaining a first signature graph and a second signature graph from the first matrix and the second matrix, respectively, using a convolutional neural network comprises:
the convolutional neural network obtains a first characteristic diagram and a second characteristic diagram from the first matrix and the second matrix respectively according to the following formula;
wherein the formula is: f. ofi＝sigmoid(Ni×fi-1+Bi)
Wherein f isi-1Is the input of the i-th convolutional neural network, fiIs the output of the ith convolutional neural network, NiIs the convolution kernel of the ith convolutional neural network, and BiAnd (4) for the bias vector of the ith layer of convolutional neural network, sigmoid indicates that the activation function is a sigmoid activation function.
6. The method of claim 1, wherein the obtaining the first and second signature graphs from the first and second matrices using a convolutional neural network comprises:
inputting the first feature map and the second feature map into a Softmax-like classification function to convert the first feature map and the second feature map into a probability space, wherein the sum of feature values of each position in the first feature map converted into the probability space is 1, and the sum of feature values of each position in the second feature map converted into the probability space is 1;
wherein the Softmax-like classification function is: yi ═ exp (xi)/Σ exp (xi), where xi is a feature value of each position in the first feature map or the second feature map.
7. The method of claim 1, wherein calculating posterior probability values corresponding to the eigenvalues of each position in the prior probability vector based on the eigenvalues of each position in the first eigenvector and the second eigenvector according to a Bayesian formula to obtain the encoded eigenvector consisting of the posterior probability values of each position comprises:
calculating a posterior probability value corresponding to the feature value of each position in the prior probability vector according to the following formula based on the feature value of each position in the first feature vector and the second feature vector so as to obtain a coding feature vector consisting of the posterior probability values of all the positions;
wherein the formula is: yi ═ Σxi∈J，ai∈L，bj∈Kxi ai/bi, wherein xi represents a characteristic value of each position in the prior probability vector; ai represents the eigenvalue of each position in the first eigenvector and bi represents the eigenvalue of each position in the second eigenvector.
8. The method of claim 1, wherein determining a data compression ratio to be currently set for each smart meter based on the eigenvalue of each position of the decoded eigenvector comprises:
transforming the decoded feature vector into a probability space with a Softmax-like classification function such that feature values of respective positions in the decoded feature vector are within an interval of 0 to 1;
determining an allocated data bandwidth for each of the smart meters based on a total data transmission bandwidth and the decoded feature vector translated to a probability space; and
and calculating the data compression ratio of each intelligent electric meter based on the allocated data bandwidth and the current data quantity of each intelligent electric meter.
9. The utility model provides a smart electric meter data compression device which characterized in that includes:
the electric quantity data unit is used for acquiring absolute electric quantity data and electric quantity change data of each intelligent electric meter in a preset time period;
the vector construction unit is used for constructing the absolute power consumption data and the power consumption change data of each intelligent ammeter into an absolute power consumption vector and a power consumption change vector respectively;
a matrix construction unit, configured to multiply the absolute power consumption vector and the power consumption change vector by transposes thereof, respectively, to obtain a first matrix representing an absolute power consumption correlation and a second matrix representing a power consumption transformation correlation;
a feature map extraction unit, configured to obtain a first feature map and a second feature map from the first matrix and the second matrix, respectively, using a convolutional neural network;
the prior probability vector construction unit is used for acquiring the default data compression proportion of each intelligent electric meter and constructing the default data compression proportion of each intelligent electric meter into a prior probability vector;
the vector query unit is used for taking the prior probability vector as a query vector to carry out matrix multiplication with the first feature map and the second feature map respectively so as to obtain a first feature vector and a second feature vector;
the encoding feature vector generating unit is used for calculating posterior probability values corresponding to the feature values of each position in the prior probability vectors according to a Bayesian formula on the basis of the feature values of each position in the first feature vector and the second feature vector so as to obtain encoding feature vectors formed by the posterior probability values of each position;
a decoding unit configured to decode the encoded eigenvector using a decoder to obtain a decoded eigenvector of equal length to the encoded eigenvector; and
and the data compression ratio setting unit is used for determining the data compression ratio to be set currently by each intelligent electric meter based on the characteristic value of each position of the decoding characteristic vector.
10. An electronic device, comprising:
a processor; and
memory in which are stored computer program instructions which, when executed by the processor, cause the processor to carry out the smart meter data compression method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110813537.XA CN113626646A (en) | 2021-07-19 | 2021-07-19 | Intelligent electric meter data compression method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110813537.XA CN113626646A (en) | 2021-07-19 | 2021-07-19 | Intelligent electric meter data compression method and device and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113626646A true CN113626646A (en) | 2021-11-09 |
Family
ID=78380148
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110813537.XA Withdrawn CN113626646A (en) | 2021-07-19 | 2021-07-19 | Intelligent electric meter data compression method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113626646A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114870416A (en) * | 2022-04-28 | 2022-08-09 | 福建德尔科技股份有限公司 | Rectification control system and rectification control method for preparing electronic-grade monofluoromethane |
-
2021
- 2021-07-19 CN CN202110813537.XA patent/CN113626646A/en not_active Withdrawn
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114870416A (en) * | 2022-04-28 | 2022-08-09 | 福建德尔科技股份有限公司 | Rectification control system and rectification control method for preparing electronic-grade monofluoromethane |
| CN114870416B (en) * | 2022-04-28 | 2023-01-24 | 福建德尔科技股份有限公司 | Rectification control system and rectification control method for preparing electronic-grade monofluoromethane |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP4087239A1 (en) | Image compression method and apparatus | |
| CN114724386B (en) | Short-time traffic flow prediction method and system under intelligent traffic and electronic equipment | |
| CN114900126B (en) | Grounding test equipment and grounding test method for solar cell module | |
| CN115373813A (en) | Scheduling method and system based on GPU virtualization in cloud computing environment and electronic equipment | |
| CN115118675B (en) | Data stream transmission acceleration method and system based on intelligent network card equipment | |
| CN110909870B (en) | Training device and method | |
| WO2022168604A1 (en) | Softmax function approximation calculation device, approximation calculation method, and approximation calculation program | |
| CN114418121A (en) | Model training method, object processing method and device, electronic device and medium | |
| CN113626646A (en) | Intelligent electric meter data compression method and device and electronic equipment | |
| CN115878330A (en) | Thread operation control method and system | |
| CN110635833B (en) | Power distribution method and device based on deep learning | |
| CN115866253B (en) | Inter-channel conversion method, device, terminal and medium based on self-modulation | |
| JP2021033994A (en) | Text processing method, apparatus, device and computer readable storage medium | |
| CN112686365B (en) | Method, device and computer equipment for operating neural network model | |
| CN115617636A (en) | Distributed performance test system | |
| CN113902114A (en) | Quantization method, device and system of neural network model, electronic device and storage medium | |
| CN111382854B (en) | Convolutional neural network processing method, device, equipment and storage medium | |
| CN113554149A (en) | Neural network processing unit NPU, neural network processing method and device | |
| CN116567669B (en) | Method for processing terminal equipment information in wireless cloud access controller distributed manner | |
| CN116561696B (en) | Multi-dimensional user adjustable load rapid aggregation method and system thereof | |
| CN112561050A (en) | Neural network model training method and device | |
| CN114615147A (en) | Bandwidth control method and system for router and electronic equipment | |
| CN114501602A (en) | Method and system for determining transmission power of router and electronic equipment | |
| CN114692892B (en) | Method for processing numerical characteristics, model training method and device | |
| CN114237962B (en) | Alarm root cause judging method, model training method, device, equipment and medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20211109 |