CN109040002A - Real-time reconstruction method of smart meter data based on compressed sensing - Google Patents
Real-time reconstruction method of smart meter data based on compressed sensing Download PDFInfo
- Publication number
- CN109040002A CN109040002A CN201810520192.7A CN201810520192A CN109040002A CN 109040002 A CN109040002 A CN 109040002A CN 201810520192 A CN201810520192 A CN 201810520192A CN 109040002 A CN109040002 A CN 109040002A
- Authority
- CN
- China
- Prior art keywords
- layer
- compression
- reconstruction
- measured value
- signal
- 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.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
- H04L69/04—Protocols for data compression, e.g. ROHC
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
- H03M7/00—Conversion of a code where information is represented by a given sequence or number of digits to a code where the same, similar or subset of information is represented by a different sequence or number of digits
- H03M7/30—Compression; Expansion; Suppression of unnecessary data, e.g. redundancy reduction
Abstract
The invention discloses a real-time reconstruction method of intelligent electric meter data based on compressive sensing, which comprises the following steps: s1: the transmission model of the electric meter data based on real-time compressed sensing mainly comprises three parts: smart meter, access point and control center S2: the compressed sensor-auto-encoder proposed herein mainly consists of two parts: compression encoding and reconstruction encoding, S3: in the compression coding part of the signal, data is mainly compressed, and circuit communication links are reduced. S4: when the signal is reconstructed and coded. The time consumed by the intelligent electric meter is very short, and the intelligent electric meter data transmission method is very suitable for transmission of intelligent electric meter data with high real-time requirements.
Description
Technical field
The present invention relates to smart grid field, the Real-time Reconstruction of specifically a kind of intelligent electric meter data based on compression sensing
Method.
Background technique
With the construction of smart grid, more and more intelligent electric meters are mounted.It is reported that intelligence all over the world in 2022
Energy ammeter data is estimated to reach 1,100,000,000.Under normal conditions, intelligent electric meter every 30 minutes numbers of acquisition or with higher frequency
Rate acquires data, so, a large amount of intelligent electric meter will generate a large amount of data.But intelligent electric meter data pass through power line
Communication link is transmitted.And power line communications link is a kind of communication link of Bandwidth-Constrained.Therefore, a large amount of intelligence of transmission
Ammeter data will lead to the problem of the loss of propagation delay time and data packet.This gives the real-time Transmission of smart grid, and monitoring is brought
Huge challenge.
In recent years, Donoho, Candes and Tao et al. propose compression sensing (Compressive Sensing, CS) simultaneously
It demonstrates when signal is sparse signal or when some sparse transformation base can be converted into sparse vector, signal can be lower than
It is reconstructed in the case where Nyquist rate.Compression sensing breach Nyquist-Shaimon sampling thheorem constraint thus by
Extensive concern.
The rarefaction representation of signal, designs calculation matrix and restructing algorithm is three big core contents of compression sensing.One
Good sparse transformation base can make signal have very high sparsity on transform domain, and a good calculation matrix can make
The compression ratio of signal is lower, and a good restructing algorithm makes signal very low or the case where measured value is mixed into noise in compression ratio
Under still can accurately reconstruct signal.Therefore, sparse transformation base is designed, calculation matrix and restructing algorithm are to ammeter data
Reconstruction accuracy it is most important.
Although compression sensing can reduce the transmission pressure and carrying cost of communication network.But people are much studied at present
It is used Magnetic resonance imaging, the fields such as recognition of face by member, but is compressed and sensed when signal is reconstructed in these fields still
There are this biggish time delays, not too much meet the high engineering of requirement of real-time.Therefore, more and more scholars begin to focus on research
The real-time for compressing sensing, has widened the application field of CS.Currently the Research Literature about Real Time Compression sensing has:
[1]S.Tang,Y.Xu,and X.Tang,“Real-time reconstruction of multi-area
power system signals based on compressed sensing”2017International Electrical
and Energy Conference(CIEEC2017),pp.389-394,2017.
[2]Y.Xu,Z.Y,J.Zhang,Z.Fei and W.Liu,“Real-time compressive sensing
based control strategy for a multi-area power system,”IEEE Trans.Smart Grid,
2017.
[3]H.Palangi,R.Ward,and L.Deng,“Distributed compressive sensing:a
deep learning approach,”IEEE Trans.Signal Processing,vol.64,no.17,pp.4504-
4518,Sep.2016.
[4]W.Yu,C.Chen,T.He,B.Yang and X.Guan,“Adaptive compressive engine
for real-time electrocardiogram monitoring under unreliable wireless
channels,”IET Commun.,vol.10,no.6,pp.607-615,2016.
[5]B.Sun,and H.Feng,“Efficient compressed sensing for wireless neural
recording:a deep learning approach,”IEEE Signal Processing Letters,vol.24,
no.6,pp.863-867,2017.
[6]Z.Zhang,T.P.Jung,S.Makeig,Z.Pi,and B.D.Rao,“Spatiotemporal sparse
Bayesian learning with applications to compressed sensing of multichannel
physiological signals.”IEEE Trans.on Neural System and Rehabilitation
Engineering,vol.22,no.6,pp.1186-1197,2014.
Wherein, the training that document [1] will propose a kind of more regional dictionary learning algorithm and be used for more regional electric power signals, together
When electric power signal is reconstructed using management loading restructing algorithm, the experimental results showed that this method use electric power signal
Real-time Reconstruction.Real Time Compression Application in Sensing has been arrived more regional electric control systems by document [2], enables system with less
Signal carries out transmission and restores original signal by suitable restructing algorithm.Document [3] is proposed using a kind of deep learning
Method solves the problems, such as multitask measurement vector reconstruction sparse vector in compression sensing.Shot and long term memory network has been used to go in text
Capture reconstructed in signal reconstruction using these characteristic informations after the characteristic information between the sparse vector of different channels it is sparse
Vector, and then reconstruction signal.Document [4] proposes a kind of self-adapting compressing sensing, and main method is being transmitted across by data
Measurement square to be used is needed when Loss Rate sparse error and output error in journey go adaptive adjusting to transmit data next time
The dimension size of battle array, reduces the reconstructed error of signal.Document [5] proposes a kind of method of two-value autocoder to nerve
Tracer signal is reconstructed, and simulation result shows that this method reconstructed velocity will be faster than other methods and signal-to-noise ratio and classification accuracy
Also higher.Document [6] proposes that a kind of space-time management loading algorithm solves the reconstruction of multitask measurement vector.To
Management loading has been widened in the application of compression sensing.
Summary of the invention
The purpose of the present invention is to provide a kind of Real-time Reconstruction methods of intelligent electric meter data based on compression sensing, with solution
Certainly the problems mentioned above in the background art.
To achieve the above object The technical solution adopted by the invention is as follows:
A kind of transmission method of the intelligent electric meter data based on Real Time Compression sensing, it is main including the following steps:
S1: the mode of the ammeter data based on Real Time Compression sensing mainly consists of three parts: intelligent electric meter, access
Point (AP) and control centre.Intelligent electric meter will compress after data collection, then will be measured by power line communications link
Value is transferred to AP, and the measured value being collected into is transferred to control centre using cordless communication network by AP, finally in control centre pair
Measured value is reconstructed.
S2: the compression of the signal based on autocoder principle mainly consists of two parts with reconstruct: compressed encoding and again
Structure coding, compressed encoding are mainly made of input layer and compression compression layer.Reconstruct coding is mainly made of reconstruction of layer and output layer.
S3: in the compressed encoding part of signal: the compression encoding process for compressing sensing mainly utilizes Φ pairs of calculation matrix
Signal carries out compressed encoding.Its input layer is original data x ∈ R firstN×1, compression layer is measured value y ∈ RM×1, wherein pressing
The measured value of contracting layer can be obtained by following formula by input layer:
Y=Φ x (1)
Wherein Φ is calculation matrix and Φ ∈ RM×N(M<N).Current common calculation matrix has: random gaussian matrix, at random
Bernoulli Jacob's matrix and simple two values matrix etc..
In order to preferably compress to ammeter data, obtain lower compression ratio, the present invention to calculation matrix done with
Under improvement:
It is available that singular value decomposition is carried out to calculation matrix Φ:
Wherein U and V is unitary matrice,By formula (15) both sides
The left side all multiplied byWhereinTherefore formula (15) are converted
At:
T=Ψ x (3)
Wherein:
S4: signal reconstruction coding: after obtaining measured value t, measured value by communication network transmission to control centre, I
Need that measured value is reverted to original intelligent electric meter data in control centre.Its problem can sum up are as follows:
Wherein v is sparse vector, ACS=Ψ D is perception matrix.D is the sparse transformation base of ammeter data x.
Since formula (18) is a np problem, it is difficult to solve in actual engineering.Therefore, a kind of solution is by l0Norm
It is transformed into l1On norm solves, such as orthogonal matching pursuit (OMP), sparse Adaptive matching tracking (SAMP) and regularization are just
It hands over match tracing (ROMP).Other solutions have management loading algorithm, combinational algorithm etc..
In order to preferably reconstruct ammeter data, the invention proposes a kind of compression sensing _ autocoder restructing algorithm,
Its way is as follows:
(1) is normalized measured value t, i.e., before reconstruction of layer of entering
Wherein tminFor the minimum value of measured value t, tmaxFor the maximum value of measured value t.
(2) assumes that reconstruction of layer is L, and the dimension of every layer of reconstruction of layer isWhereinGive the power of every layer of reconstruction of layer
Matrix WlWith bias vector bl.So lthThe activation primitive of reconstruction of layer can be expressed by following formula:
zl=f (Wlzl-1+bl) (6)
Wherein z0=t.F (α) is the activation primitive of α.Under normal conditions, the common activation primitive of f (α) has sigmoid letter
Number, tanh function, ReLU function etc., the activation primitive that the present invention uses are sigmoid, are defined as follows:
(3) output layer is the ammeter data of reconstruct, expression formula are as follows:
Wherein xminFor the minimum value of ammeter data x, xmaxFor the maximum value of ammeter data x.
(4) is needed to weight matrix to preferably reconstruct ammeter data for preferably capturing ammeter data feature
WlWith bias vector blIt is finely adjusted.Use direction propagation algorithm of the present invention is finely adjusted parameter.It is specific as follows:
Δbl=η δl (12)
δl-1=(Wl Tδl)⊙f'(Wl-1zl-2+bl-1) (13)
Wl=Wl-ΔWl (14)
bl=bl-Δbl (15)
Wherein η is learning rate, and ⊙ indicates to press element multiplication.
S4: the performance indicator of compression sensing is measured in this invention in terms of following three:
(1) compression ratio:
Wherein, M indicates that the dimension using measured value after compression, N indicate the dimension of signal itself.
(2) signal-to-noise ratio:
Wherein,xIndicate original signal,Indicate reconstruction signal.
(3) relative error:
Compared with prior art, the beneficial effects of the present invention are: time for reconstructing every time of setting of the invention is very short, very
Suitable for the transmission of the intelligent electric meter data high to requirement of real-time, and the dimension of transmission data is greatly reduced, reduced
Transmission cost.
Detailed description of the invention
Fig. 1 is the real time communication illustraton of model of the intelligent electric meter data sensed the present invention is based on Real Time Compression.
Fig. 2 is the data compression coding figure of intelligent electric meter of the present invention.
Fig. 3 is one week datagram that intelligent electric meter of the present invention is collected.
Fig. 4 is present invention compression sensing _ autocoder schematic diagram.
Fig. 5 is the original intelligent electric meter data in compression ratio ρ=12/48 and reconstruct number when measured value of the present invention is not mixed into noise
According to figure.
Fig. 6 is the original intelligent electric meter data in compression ratio ρ=12/48 and reconstruct data when measured value of the present invention is mixed into noise
Figure.
Fig. 7 is the compression ratio of intelligent electric meter data and the relational graph of signal-to-noise ratio when measured value of the present invention is not mixed into noise.
Fig. 8 is the compression ratio of intelligent electric meter data and the relational graph of signal-to-noise ratio when measured value of the present invention is mixed into noise.
Specific embodiment
It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the present invention can phase
Mutually combination.
In the description of the present invention, it is to be understood that, term " center ", " longitudinal direction ", " transverse direction ", "upper", "lower",
The orientation or positional relationship of the instructions such as "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outside" is
It is based on the orientation or positional relationship shown in the drawings, is merely for convenience of description of the present invention and simplification of the description, rather than instruction or dark
Show that signified device or element must have a particular orientation, be constructed and operated in a specific orientation, therefore should not be understood as pair
Limitation of the invention.In addition, term " first ", " second " etc. are used for description purposes only, it is not understood to indicate or imply phase
To importance or implicitly indicate the quantity of indicated technical characteristic.The feature for defining " first ", " second " etc. as a result, can
To explicitly or implicitly include one or more of the features.In the description of the present invention, unless otherwise indicated, " multiple "
It is meant that two or more.
In the description of the present invention, it should be noted that unless otherwise clearly defined and limited, term " installation ", " phase
Even ", " connection " shall be understood in a broad sense, for example, it may be being fixedly connected, may be a detachable connection, or be integrally connected;It can
To be mechanical connection, it is also possible to be electrically connected;It can be directly connected, can also can be indirectly connected through an intermediary
Connection inside two elements.For the ordinary skill in the art, above-mentioned term can be understood by concrete condition
Concrete meaning in the present invention.
The present invention will be described in detail below with reference to the accompanying drawings and embodiments.
Refering to fig. 1~8, in the embodiment of the present invention, a kind of Real-time Reconstruction side of the intelligent electric meter data based on compression sensing
Method, including the following steps:
S1: the mode of the ammeter data based on Real Time Compression sensing mainly consists of three parts: intelligent electric meter, access
Point (AP) and control centre.Intelligent electric meter will compress after data collection, then will be measured by power line communications link
Value is transferred to AP, and the measured value being collected into is transferred to control centre using cordless communication network by AP, finally in control centre pair
Measured value is reconstructed.
S2: the compression of the signal based on autocoder principle mainly consists of two parts with reconstruct: compressed encoding and again
Structure coding, compressed encoding are mainly made of input layer and compression compression layer.Reconstruct coding is mainly made of reconstruction of layer and output layer.
S3: in the compressed encoding part of signal: the compression encoding process for compressing sensing mainly utilizes Φ pairs of calculation matrix
Signal carries out compressed encoding.Its input layer is original data x ∈ R firstN×1, compression layer is measured value y ∈ RM×1, wherein pressing
The measured value of contracting layer can be obtained by following formula by input layer:
Y=Φ x (19)
Wherein Φ is calculation matrix and Φ ∈ RM×N(M<N).Current common calculation matrix has: random gaussian matrix, at random
Bernoulli Jacob's matrix and simple two values matrix etc..
In order to preferably compress to ammeter data, obtain lower compression ratio, the present invention to calculation matrix done with
Under improvement:
It is available that singular value decomposition is carried out to calculation matrix Φ:
Wherein U and V is unitary matrice,By formula (15) both sides
The left side all multiplied byWhereinTherefore formula (15) are converted
At:
T=Ψ x (21)
Wherein:
S4: signal reconstruction coding: after obtaining measured value t, measured value by communication network transmission to control centre, I
Need that measured value is reverted to original intelligent electric meter data in control centre.Its problem can sum up are as follows:
Wherein v is sparse vector, ACS=Ψ D is perception matrix.D is the sparse transformation base of ammeter data x.
Since formula (18) is a np problem, it is difficult to solve in actual engineering.Therefore, a kind of solution is by l0Norm
It is transformed into l1On norm solves, such as orthogonal matching pursuit (OMP), sparse Adaptive matching tracking (SAMP) and regularization are just
It hands over match tracing (ROMP).Other solutions have management loading algorithm, combinational algorithm etc..
In order to preferably reconstruct ammeter data, the invention proposes a kind of compression sensing _ autocoder restructing algorithm,
Its way is as follows:
(1) is normalized measured value t, i.e., before reconstruction of layer of entering
Wherein tminFor the minimum value of measured value t, tmaxFor the maximum value of measured value t.
(2) assumes that reconstruction of layer is L, and the dimension of every layer of reconstruction of layer isWhereinGive the power of every layer of reconstruction of layer
Matrix WlWith bias vector bl.So lthThe activation primitive of reconstruction of layer can be expressed by following formula:
zl=f (Wlzl-1+bl) (24)
Wherein z0=t.F (α) is the activation primitive of α.Under normal conditions, the common activation primitive of f (α) has sigmoid letter
Number, tanh function, ReLU function etc., the activation primitive that the present invention uses are sigmoid, are defined as follows:
(3) output layer is the ammeter data of reconstruct, expression formula are as follows:
Wherein xminFor the minimum value of ammeter data x, xmaxFor the maximum value of ammeter data x.
(4) is needed to weight matrix to preferably reconstruct ammeter data for preferably capturing ammeter data feature
WlWith bias vector blIt is finely adjusted.Use direction propagation algorithm of the present invention is finely adjusted parameter.It is specific as follows:
Δbl=η δl (30)
δl-1=(Wl Tδl)⊙f'(Wl-1zl-2+bl-1) (31)
Wl=Wl-ΔWl (32)
bl=bl-Δbl (33)
Wherein η is learning rate, and ⊙ indicates to press element multiplication.
S4: the performance indicator of compression sensing is measured in this invention in terms of following three:
(1) compression ratio:
Wherein, M indicates that the dimension using measured value after compression, N indicate the dimension of signal itself.
(2) signal-to-noise ratio:
Wherein,xIndicate original signal,Indicate reconstruction signal.
(3) relative error:
The time that setting of the invention reconstructs every time is very short, is readily applicable to the intelligent electric meter data high to requirement of real-time
Transmission, and greatly reduce transmission data dimension, reduce transmission cost.
Fig. 3 show true intelligent electric meter data, the data of acquisition in intelligent electric meter every 30 minutes.Intelligent electric meter data
Show that the electricity of consumption mainly concentrates 8:00 to 6:00 in afternoon in morning.
Step 1: setting compression ratio ρ=12/48, the i.e. dimension of measured value t are 12, calculation matrix Φ ∈ RM×N, design one
A only 5 elements are " 1 ", the calculation matrix that remaining element is " 0 " and their positions are random.Then calculation matrix passes through formula
(16) it is converted, obtains new calculation matrix Ψ.
Step 2: dividing compression sensing _ autocoder test set { xtestAnd training set { xtrain}.Then setting solution
The code process number of plies is L=3, learning rate η=0.1, the number of iterations 5000, reconstruction of layer multipleInitialize weight matrix WlWith it is inclined
Set vector bl, Wl=rand-0.5, bl=rand-0.5.
Step 3: according to formula (21) and formula (23)-(33) training set { xtrainTrain compression sensing _ autocoder
Algorithm.It is as shown in Figure 4 to compress sensing _ autocoder algorithm flow.
Step 4: using measurement collection { xtestTested.Point several situations are discussed in simulation process, measured value without
It is mixed into noise and measured value is mixed into noise.Point situation is successively discussed below:
1) measured value is without being mixed into noise.First test set compress using calculation matrix Ψ designed by the first step
To ttest=Ψ xtest, then it is made after being normalized using formula (23), is finally passed using the trained compression of third step
Sense _ autocoder algorithm reconstructs to obtain reconstruct dataIntelligent electric meter initial data and reconstruct data result are as shown in Figure 5.
2) measured value is mixed into noise, and noise is e=0.1 × randn (M, 1).The measurement square mentioned first using the first step
Battle array Ψ is compressed to obtain to test setThen it is made to use compression sensing _ automatic after being normalized
Encoder is reconstructed to obtain reconstruct dataIntelligent electric meter initial data and reconstruct data result are as shown in Figure 6.
Step 5: Fig. 7-8 is the relationship of compression ratio and signal-to-noise ratio, the lower dimension for illustrating measured value of compression ratio is smaller.From
Figure can be seen that no matter whether measured value is mixed into noise, as the dimension of the raising of compression ratio, i.e. measured value t is bigger, signal-to-noise ratio
It is higher.In addition, signal-to-noise ratio will receive the influence of noise.
Step 6: simulated environment of the invention and machine configuration are as follows:
(1), processor: Intel (R) Core (TM) i7-6700CPU@3.40GHz3.41GHz
(2), memory: 16.0GB is installed
(3), system type: Windows 10 (64 bit manipulation system)
(4), simulated conditions: MATLAB R2016b
The relative error and reconstitution time comparison of the different compression ratios of table 1
Table 1 shows different compression ratios, relative error, the comparison of reconstitution time.With the increase of compression ratio, now to accidentally
Difference gradually decreases, and reconstitution time gradually increases.But the time reconstructed every time is very short, is readily applicable to the intelligence high to requirement of real-time
The transmission of energy ammeter data, and the dimension of transmission data is greatly reduced, reduce transmission cost.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Within mind and principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (4)
1. a kind of Real-time Reconstruction method of the intelligent electric meter data based on compression sensing, which is characterized in that including following step
It is rapid:
S1: the mode of the ammeter data based on Real Time Compression sensing consists of three parts: intelligent electric meter, access point and control
Center processed, intelligent electric meter will compress after data collection, and measured value is then transferred to access by power line communications link
Point, the measured value being collected into is transferred to control centre using cordless communication network by access point, finally in control centre to measurement
Value is reconstructed;
S2: the compression of the signal based on autocoder principle consists of two parts with reconstruct: compressed encoding and reconstruct coding, pressure
It reduces the staff code to be made of input layer and compression compression layer, reconstruct coding is made of reconstruction of layer and output layer;
S3: in the compressed encoding part of signal: the compression encoding process for compressing sensing mainly utilizes calculation matrix Φ to signal
Carry out compressed encoding.Its input layer is original data x ∈ R firstN×1, compression layer is measured value y ∈ RM×1, wherein compression layer
Measured value can be obtained by following formula by input layer:
Y=Φ x (1)
Wherein Φ is calculation matrix and Φ ∈ RM×N(M<N).Current common calculation matrix has: random gaussian matrix, random uncle exert
Sharp matrix and simple two values matrix etc..
2. the transmission method of the intelligent electric meter data according to claim 1 based on Real Time Compression sensing, which is characterized in that
In step S3, ammeter data is compressed using calculation matrix is improved.The specific improved method of calculation matrix is as follows:
It is available that singular value decomposition is carried out to calculation matrix Φ:
Wherein U and V is unitary matrice,By the left side on formula (1) both sides
Side all multiplied byWhereinTherefore formula (1) is converted into:
T=Ψ x (3)
Wherein:
S4: signal reconstruction coding: after obtaining measured value t, measured value passes through communication network transmission to control centre, control centre
Measured value is reverted into original intelligent electric meter data, is obtained by following formula:
Wherein v is sparse vector, ACS=Ψ D is perception matrix, and D is the sparse transformation base of ammeter data x.
3. the Real-time Reconstruction method of the intelligent electric meter data according to claim 2 based on compression sensing, which is characterized in that
In step S3, the reconstruct of ammeter data specifically includes the following steps:
(1) is normalized measured value t, i.e., before reconstruction of layer of entering
Wherein tminFor the minimum value of measured value t, tmaxFor the maximum value of measured value t.
(2) assumes that reconstruction of layer is L, and the dimension of every layer of reconstruction of layer isWhereinGive the weight matrix of every layer of reconstruction of layer
WlWith bias vector bl.So lthThe activation primitive of reconstruction of layer can be expressed by following formula:
zl=f (Wlzl-1+bl) (5)
Wherein z0=t.F (α) is the activation primitive of α.Under normal conditions, the common activation primitive of f (α) has sigmoid function,
Tanh function, ReLU function etc., the activation primitive that the present invention uses are sigmoid, are defined as follows:
(3) output layer is the ammeter data of reconstruct, expression formula are as follows:
Wherein xminFor the minimum value of ammeter data x, xmaxFor the maximum value of ammeter data x.
(4) is needed to weight matrix W to preferably reconstruct ammeter data for preferably capturing ammeter data featurelWith it is inclined
Set vector blIt is finely adjusted.Use direction propagation algorithm of the present invention is finely adjusted parameter.It is specific as follows:
Δbl=η δl (11)
Wl=Wl-ΔWl (13)
bl=bl-Δbl (14)
Wherein η is learning rate, and ⊙ indicates to press element multiplication.Hold power matrix WlWith bias vector blAfter update, so that it may pass through
Formula (7) obtains reconstruction signal.
4. the transmission method of the intelligent electric meter data according to claim 1 based on Real Time Compression sensing, which is characterized in that
The performance indicator of compression sensing is measured in terms of following three:
(1) compression ratio:
Wherein, M indicates that the dimension using measured value after compression, N indicate the dimension of signal itself.
(2) signal-to-noise ratio:
Wherein, x indicates original signal,Indicate reconstruction signal.
(3) relative error:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810520192.7A CN109040002B (en) | 2018-05-28 | 2018-05-28 | Real-time reconstruction method of smart meter data based on compressed sensing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810520192.7A CN109040002B (en) | 2018-05-28 | 2018-05-28 | Real-time reconstruction method of smart meter data based on compressed sensing |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109040002A true CN109040002A (en) | 2018-12-18 |
CN109040002B CN109040002B (en) | 2022-05-13 |
Family
ID=64611292
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810520192.7A Active CN109040002B (en) | 2018-05-28 | 2018-05-28 | Real-time reconstruction method of smart meter data based on compressed sensing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109040002B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114024575A (en) * | 2021-09-29 | 2022-02-08 | 广东电网有限责任公司电力调度控制中心 | Data compression transmission method suitable for low-voltage power line carrier communication |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103607292A (en) * | 2013-10-28 | 2014-02-26 | 国家电网公司 | Fast distributed monitoring method for electric-power communication network services |
US20140195200A1 (en) * | 2011-08-18 | 2014-07-10 | University Of Antwerp | Smart Data Sampling and Data Reconstruction |
CN105827250A (en) * | 2016-03-16 | 2016-08-03 | 江苏大学 | Electric-energy quality data compression and reconstruction method based on self-adaptive dictionary learning |
WO2017007546A1 (en) * | 2015-07-03 | 2017-01-12 | Intel Corporation | Apparatus and method for data compression in a wearable device |
CN107689795A (en) * | 2017-07-10 | 2018-02-13 | 广东顺德中山大学卡内基梅隆大学国际联合研究院 | A kind of how regional electrical control method perceived based on Real Time Compression |
-
2018
- 2018-05-28 CN CN201810520192.7A patent/CN109040002B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140195200A1 (en) * | 2011-08-18 | 2014-07-10 | University Of Antwerp | Smart Data Sampling and Data Reconstruction |
CN103607292A (en) * | 2013-10-28 | 2014-02-26 | 国家电网公司 | Fast distributed monitoring method for electric-power communication network services |
WO2017007546A1 (en) * | 2015-07-03 | 2017-01-12 | Intel Corporation | Apparatus and method for data compression in a wearable device |
CN105827250A (en) * | 2016-03-16 | 2016-08-03 | 江苏大学 | Electric-energy quality data compression and reconstruction method based on self-adaptive dictionary learning |
CN107689795A (en) * | 2017-07-10 | 2018-02-13 | 广东顺德中山大学卡内基梅隆大学国际联合研究院 | A kind of how regional electrical control method perceived based on Real Time Compression |
Non-Patent Citations (1)
Title |
---|
SUIGU TANG等: "Real-time reconstruction of multi-area power system signals based on compressed sensing", 《PROCEEDINGS OF 2017 INTERNATIONAL ELECTRICAL AND ENERGY CONFERENCE (CIEEC 2017)》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114024575A (en) * | 2021-09-29 | 2022-02-08 | 广东电网有限责任公司电力调度控制中心 | Data compression transmission method suitable for low-voltage power line carrier communication |
Also Published As
Publication number | Publication date |
---|---|
CN109040002B (en) | 2022-05-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106405339B (en) | Based on the associated transmission line malfunction reason discrimination method of low-and high-frequency wavelet character | |
CN109685314A (en) | A kind of non-intruding load decomposition method and system based on shot and long term memory network | |
CN111710150A (en) | Abnormal electricity consumption data detection method based on countermeasure self-coding network | |
CN113283175B (en) | Photovoltaic power station joint fault diagnosis method based on asynchronous decentralized federal learning | |
CN105827250A (en) | Electric-energy quality data compression and reconstruction method based on self-adaptive dictionary learning | |
CN111753968A (en) | Non-invasive load monitoring intelligent ammeter and electricity decomposition method | |
CN113113030B (en) | High-dimensional damaged data wireless transmission method based on noise reduction self-encoder | |
CN102664705B (en) | Aircraft data link adaptive encoding and modulation method | |
CN107689795A (en) | A kind of how regional electrical control method perceived based on Real Time Compression | |
CN110446173B (en) | Efficient and energy-saving satellite-borne wireless sensor network data compression method | |
CN105811993A (en) | Data collection method based on compression dictionary learning in wireless sensor network | |
CN109040002A (en) | Real-time reconstruction method of smart meter data based on compressed sensing | |
Lokumarambage et al. | Wireless end-to-end image transmission system using semantic communications | |
CN109672885B (en) | Video image coding and decoding method for intelligent monitoring of mine | |
Wang | Application of Wireless Sensor Network based on LoRa in City Gas Meter Reading. | |
CN115473771A (en) | Model evolution-based environment sensing method | |
CN107094043A (en) | MMSE method for detecting low complexity signal after improvement based on block iteration method | |
CN105242237B (en) | A kind of electromagnetic vector array parameter method of estimation based on compressed sensing | |
CN110415403A (en) | A kind of control method of the intelligent lock system based on edge calculations | |
CN113938254A (en) | Attention mechanism-based layered source-channel joint coding transmission system and transmission method thereof | |
CN105472657A (en) | Data reconstruction method for wireless sensor network based on low-rank tensor | |
CN104794571A (en) | Method of acquiring mass battery data of storage power station | |
CN114630207B (en) | Multi-sensing-node sensing data collection method based on noise reduction self-encoder | |
CN109525598A (en) | A kind of fault-tolerant compression method of wireless sense network depth and system based on variation mixing | |
CN104901704A (en) | Body sensing network signal reconstruction method with spatial-temporal correlation characteristics |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |