CN113759271A - Redundant power supply current sharing state identification method based on frequency spectrum and LSTM network - Google Patents
Redundant power supply current sharing state identification method based on frequency spectrum and LSTM network Download PDFInfo
- Publication number
- CN113759271A CN113759271A CN202111324442.8A CN202111324442A CN113759271A CN 113759271 A CN113759271 A CN 113759271A CN 202111324442 A CN202111324442 A CN 202111324442A CN 113759271 A CN113759271 A CN 113759271A
- Authority
- CN
- China
- Prior art keywords
- power supply
- spectrum
- switching power
- frequency
- frequency spectrum
- 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
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/40—Testing power supplies
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R23/00—Arrangements for measuring frequencies; Arrangements for analysing frequency spectra
- G01R23/16—Spectrum analysis; Fourier analysis
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention provides a redundant power supply current sharing state identification method based on frequency spectrums and an LSTM network, which belongs to the technical field of power supply state identification, and comprises the steps of firstly obtaining output power-central frequency curves of all switching power supplies in a redundant power supply sample, then respectively obtaining spectrum characteristic training sample data of all switching power supplies, constructing and training the LSTM network, then dividing a ripple signal of a load end of a to-be-identified redundant power supply into frequency spectrum intervals, sequentially inputting the frequency spectrum intervals into the LSTM network in sequence to obtain corresponding labels, and screening out the frequency spectrum intervals corresponding to all the switching power supplies; and determining the output power of each switching power supply according to the central frequency of the frequency spectrum interval and an output power-central frequency curve, calculating to obtain output current, and further determining the current sharing state of the redundant power supplies. The invention accurately identifies the ripple component in the frequency spectrum by using the LSTM network based on the frequency spectrum characteristics of the switching power supply, thereby realizing the accurate identification of the current-sharing state of the redundant power supply.
Description
Technical Field
The invention belongs to the technical field of power supply state identification, and particularly relates to a redundant power supply current sharing state identification method based on a frequency spectrum and an LSTM network.
Background
In the fields of electric power, petroleum, natural gas, medical treatment and the like, the reliability of an electronic system depends on the stability of a redundant power supply, and the current sharing state of the power supply is an important factor influencing the service life and the reliability of the redundant power supply. Under the condition of uneven current, the working temperature of the heavy-load power supply is higher than that of the light-load power supply by more than 10 ℃, so that the average service life of the heavy-load power supply is relatively reduced by about 50%. Meanwhile, the non-current-sharing state of the parallel power supply also influences the power supply parameters of the power supply input end, so that the harmonic wave of the whole system is enhanced, and the performance of the system is reduced. Therefore, monitoring the current sharing state of the parallel redundant power supply has great significance for improving the reliability of the system and delaying the service life of the system.
The traditional method for detecting the current-sharing state of the redundant power supply is based on current detection sensors on all power supply branches, and the current-sharing state of the redundant power supply is judged by comparing the currents of all the branches. Although this method is accurate and reliable, it has many limitations in practical applications. For example, current detection requires sensors to be connected in series in the current branch, which increases the risk of system failure; for a distributed system, the current detection of each branch needs to solve the problems of power supply, communication and the like through a special cable, and the problems of high cost, poor maintainability and the like exist.
In recent years, a novel redundant power source current sharing state detection method is realized based on the frequency spectrum characteristics of ripple waves at a load end. The method has the characteristics of single-point type and non-invasive type, and has great advantages in the aspects of cost and maintainability because only the ripple voltage at the load end needs to be measured. However, such methods can only determine which current branch components exist in the system at present, but cannot determine the specific current branch corresponding to each current component, and it is difficult to implement uniform current control.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a redundant power supply current sharing state identification method based on a frequency spectrum and an LSTM network, frequency spectrum intervals are divided based on the frequency spectrum characteristics of a switching power supply, and the frequency spectrum intervals corresponding to the switching power supplies of all branches are accurately identified and positioned by using the LSTM (Long Short-Term Memory) network and a sliding window, so that the current sharing state identification of the redundant power supply is realized.
The specific technical scheme of the invention is as follows:
the redundant power supply current sharing state identification method based on the frequency spectrum and the LSTM network is characterized by comprising the following steps of:
s1: respectively acquiring an output power-center frequency curve of the output power of each switching power supply changing along with the center frequency for a redundant power supply sample consisting of the switching power supplies of the D branches; wherein D is more than or equal to 2;
s2: order toi=1；
S3: acquisition switching power supplyiThe specific acquisition process of the spectrum characteristic training sample data is as follows:
s3.1: adjustable switch power supplyiThe actual current of each branch circuit is changed, the ripple signal of the load end is collected for P times in a preset period T, and the corresponding original ripple frequency spectrum signal is obtained through processing
, j=1, …, P; simultaneously acquiring output current of each switching power supply at each acquisition moment of the ripple signal; wherein P is more than or equal to 1;
s3.2: for the original ripple wave frequency spectrum signal obtained in S3.1
, j=1, …, P divides the spectrum interval, specifically:
intercept frequencyfIn [ (1-a)f m , (1+a)f m ]Raw ripple spectrum signal in interval
, j=1,…,PFor a ripple spectrum signal
, j=1, …, P, in ripple spectrum signal
, j=1, …, P constructed with a width ofNWindow of in preset step lengthN step Sliding M-1 times to obtain a width ofNOf the frequency spectrum interval
, j=1,…,P, m=1,2, …, M; wherein the content of the first and second substances,ais a preset constant, and 0<a<1；f m The frequency of the maximum harmonic component of the internal switching signal of each switching power supply in the ripple signal is obtained; presetting step lengthN step Not more thanN；
S3.3: obtaining a spectrum sample of each switching power supply, specifically:
from the original ripple spectral signal
, j=1, …, P, the output power of each switching power supply is calculated from the output current of each switching power supply, the corresponding center frequency is determined from the output power-center frequency curve of each switching power supply, and then the center frequency is used as the widthNCenter-truncated ripple spectrum signal
, j=1, …, P, the obtained spectrum interval being the spectrum sample of the corresponding switching power supply
, j=1,…,P, d=1,2,…,D；
S3.4: for the spectral interval obtained at S3.2
, j=1,…,P, m=1,2, …, M is labeled, specifically:
calculating spectral intervals
, j=1,…,P, m=1,2, …, M and spectrum samples corresponding to each switching power supply
, j=1,…,P, dOverlap ratio between =1,2, …, D
, j=1,…,P, m=1,2,…,M, d=1,2, …, D, screening out the overlap ratio with each switching power supply
Maximum value of (max) ((
) If the maximum value max (
) If the frequency spectrum interval is larger than the preset overlap rate threshold value, the frequency spectrum interval is divided
, j=1,…,P, mLabel for =1,2, …, M is maximum max (max: = tag of label of M with labels of tag of M, of label for M, of label for 1,2, …, of M, and/1, 2, …
) Number corresponding to switching power supplyd(ii) a Otherwise, the frequency spectrum interval is divided
, j=1,…,P, mLabel of =1,2, …, M is labeled "unknown class";
all frequency spectrum intervals
, j=1,…,P, m=1,2, …, M andcorresponding label as switch power supplyiTraining sample data of the spectrum characteristics;
s4: judgment ofiIf D is equal to D, if yes, go to S5; otherwise, it ordersi=i+1, go back to S3;
s5: constructing an LSTM network, training based on the spectrum characteristic training sample data of each switching power supply, and inputting the spectrum characteristic training sample data into the spectrum interval corresponding to each switching power supply
, j=1,…,P, m=1,2, …, M, the training target is the corresponding label, and the LSTM network after training is obtained;
s6: collecting ripple signals of a load end of a redundant power supply to be identified, and processing to obtain corresponding original ripple frequency spectrum signals
Dividing the frequency spectrum interval by adopting the same method as S3.2; sequentially inputting the frequency spectrum intervals into the trained LSTM network according to the window sliding sequence to obtain labels corresponding to the frequency spectrum intervals, and screening the frequency spectrum intervals corresponding to the switching power supplies; and determining the output power of each switching power supply according to the center frequency of the frequency spectrum interval and an output power-center frequency curve, calculating to obtain output current, and finally determining the current sharing state of the redundant power supply based on the output current of each switching power supply.
Further, the preset period T in S3.1 is determined by the sampling rate of the ripple signalf s Sum frequencyf m The formula is determined as follows:
T=γf m /f s
wherein gamma is a preset integer and gamma is greater than 2.
Further, the collected ripple signals are processed through fast fourier transform and normalization in S3.1 and S6.
Further, the spectral interval in S3.2
Number of (3) M = Round (2T)f m /N step )。
Further, width in S3.2NThe method is specifically as follows, wherein the effective signal spectrum width is larger than the effective signal spectrum width under the condition that the ripple wave spectrum in all the switching power supplies is widest: to the switching power supplyiDefining the original ripple spectrum signal
, j=1, …, the spectrum region formed by continuous frequency points with normalization values all greater than 0.01 in P is the spectrum width, and the switching power supply is obtainediMaximum spectral width at different output powers, and then defining the widthNGreater than the maximum of the corresponding maximum spectral widths of all switching power supplies.
Further, the overlap ratio is calculated in S3.4
The formula of (1) is as follows:
wherein the content of the first and second substances,
is a frequency spectrum interval
, j=1,…,P, m=1,2, …, M and switching power supplydOf the spectrum sample
, j=1,…,P, dWidth of overlapping spectrum of =1,2, …, D;
further, the predetermined threshold of the overlap ratio in S3.4 is greater than or equal to 50% + (N step /2N)。
Further, the LSTM network constructed in S5 further includes a convolutional neural network.
The invention has the beneficial effects that:
the invention provides a redundant power supply current-sharing state identification method based on frequency spectrums and an LSTM network.
Drawings
Fig. 1 is a flowchart of a redundant power supply current sharing state identification method based on spectrum and LSTM network according to embodiment 1 of the present invention;
FIG. 2 is a structural framework of a redundant power supply sample employed in embodiment 1 of the present invention;
fig. 3 is a flowchart of acquiring spectrum characteristic training sample data of the switching power supply in embodiment 1 of the present invention;
fig. 4 is a diagram showing an example of division of spectrum space and spectrum samples in embodiment 1 of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described with reference to the following embodiments and the accompanying drawings.
The following non-limiting examples are presented to enable those of ordinary skill in the art to more fully understand the present invention and are not intended to limit the invention in any way.
Example 1
The embodiment provides a redundant power supply current sharing state identification method based on a frequency spectrum and an LSTM network, and the flow is shown in fig. 1, and the method comprises the following steps:
s1: by adopting a redundant power supply sample which is formed by the switching power supplies with the D branches and has the structure shown in figure 2, each switching power supply is respectively connected in parallel through a diode and output to a load end, and ripple wave acquisition modules are connected in parallel at two ends of the load end to acquire ripple wave signals and transmit the ripple wave signals to a signal processing unit for processing;
respectively acquiring an output power-center frequency curve of the output power of each switching power supply changing along with the center frequency according to the historical use data of each switching power supply; wherein D is more than or equal to 2;
s2: order toi=1；
S3: acquisition switching power supplyiThe process of the spectrum feature training sample data is shown in fig. 3, and the specific acquisition process is as follows:
s3.1: adjustable switch power supplyiThe ripple collecting module collects ripple signals at the load end for P times in a preset period T, the signal processing unit carries out fast Fourier transform and normalization processing on the ripple signals to obtain corresponding original ripple frequency spectrum signals,j=1, …, P; simultaneously acquiring output current of each switching power supply at each acquisition moment of the ripple signal; wherein P is more than or equal to 1; the preset period T is determined by the sampling rate of the ripple signalf s Sum frequencyf m The formula is determined as follows:
T=γf m /f s
wherein gamma is a preset integer and gamma is greater than 2;
s3.2: for the original ripple wave frequency spectrum signal obtained in S3.1
, j=1, …, P divides the spectrum interval, specifically:
intercept frequencyfIn [ (1-a)f m , (1+a)f m ]Raw ripple spectrum signal in interval
, j=1, …, P is ripple spectrum signal
, j=1, …, P, in ripple spectrum signal
, j=1, …, P constructed with a width ofNWindow of in preset step lengthN step Sliding M-1 times to obtain a width ofNOf the frequency spectrum interval
, j=1,…,P, m=1,2, …, M, see the area enclosed by the black dashed line as shown in fig. 4; wherein the content of the first and second substances,ais a preset constant, and 0<a<1；f m The frequency of the maximum harmonic component of the internal switching signal of each switching power supply in the ripple signal is obtained; presetting step lengthN step Not more thanN；M=Round(2Tf m /N step ) (ii) a To the switching power supplyiDefining the original ripple spectrum signal
, j=1, …, the spectrum region formed by continuous frequency points with normalization values all greater than 0.01 in P is the spectrum width, and the switching power supply is obtainediMaximum spectral width at different output powers, and then defining the widthNGreater than the maximum value of the maximum spectrum widths corresponding to all switching power supplies;
s3.3: obtaining a spectrum sample of each switching power supply, specifically:
from the original ripple spectral signal
, j=1, …, P, calculating the output power of each switching power supply, determining the corresponding center frequency according to the output power-center frequency curve of each switching power supply, and then taking the center frequency as the widthNCenter of (2), intercept the original ripple spectrum signal
, j=1, …, P, see the area enclosed by the solid black line shown in fig. 4, and the obtained spectrum interval is the spectrum sample of the corresponding switching power supply
, j=1,…,P, d=1,2,…,D；
S3.4: for the spectral interval obtained at S3.2
, j=1,…,P, m=1,2, …, M is labeled, specifically:
calculating spectral intervals
, j=1,…,P, m=1,2, …, M and spectrum samples corresponding to each switching power supply
, j=1,…,P, dOverlap ratio between =1,2, …, D
：
Wherein the content of the first and second substances,
is a frequency spectrum interval
, j=1,…,P, m=1,2, …, M and switching power supplydOf the spectrum sample
, j=1,…,P, dWidth of overlapping spectrum of =1,2, …, D, see the overlapping area of the area enclosed by the solid black line and the area enclosed by the dashed black line as shown in fig. 4;
for frequency spectrum interval
, j=1,…,P, m=1,2, …, M, screening out the overlap ratio with each switching power supply
Maximum value of (max) ((
) If the maximum value max (
) If the frequency spectrum interval is larger than the preset overlap rate threshold value, the frequency spectrum interval is divided
, j=1,…,P, mLabel for =1,2, …, M is maximum max (max: = tag of label of M with labels of tag of M, of label for M, of label for 1,2, …, of M, and/1, 2, …
) Number corresponding to switching power supplyd(ii) a Otherwise, the frequency spectrum interval is divided
, j=1,…,P, mLabel of =1,2, …, M is labeled "unknown class"; wherein the predetermined overlap ratio threshold is greater than or equal to 50% + (N step /2N)；
All frequency spectrum intervals
, j=1,…,P, m=1,2, …, M and corresponding label as switching power supplyiTraining sample data of the spectrum characteristics;
s4: judgment ofiIf D is equal to D, if yes, go to S5; otherwise, it ordersi=i+1, go back to S3;
s5: constructing an LSTM network, training based on the spectrum characteristic training sample data of each switching power supply, and inputting the spectrum characteristic training sample data into the spectrum interval corresponding to each switching power supply
, j=1,…,P, m=1,2, …, M, the training target is the corresponding label, and the LSTM network after training is obtained;
s6: collecting ripple signals of a load end of a redundant power supply to be identified, and obtaining corresponding signals after fast Fourier transform and normalization processingOf the original ripple spectrum signal
Dividing the frequency spectrum interval by adopting the same method as S3.2; sequentially inputting the frequency spectrum intervals into the trained LSTM network according to the window sliding sequence to obtain labels corresponding to the frequency spectrum intervals, and screening the frequency spectrum intervals corresponding to the switching power supplies; and determining the output power of each switching power supply according to the center frequency of the frequency spectrum interval and an output power-center frequency curve, calculating to obtain output current, and finally determining the current sharing state of the redundant power supply based on the output current of each switching power supply.
Although illustrative embodiments of the present invention have been described above to facilitate the understanding of the present invention by those skilled in the art, it should be understood that the present invention is not limited to the scope of the embodiments, and various changes may be made apparent to those skilled in the art as long as they are within the spirit and scope of the present invention as defined and defined by the appended claims, and all matters of the invention which utilize the inventive concepts are protected.
Claims (8)
1. The redundant power supply current sharing state identification method based on the frequency spectrum and the LSTM network is characterized by comprising the following steps of:
s1: respectively acquiring an output power-center frequency curve of the output power of each switching power supply changing along with the center frequency for a redundant power supply sample consisting of the switching power supplies of the D branches; wherein D is more than or equal to 2;
s2: order toi=1；
S3: acquisition switching power supplyiThe specific process of the spectrum characteristic training sample data is as follows:
s3.1: adjustable switch power supplyiThe ripple signals of the load end are collected for P times in a preset period T, and original ripple frequency spectrum signals corresponding to the ripple signals are obtained after processing
, j=1, …, P; at the same time obtainThe output current of each switching power supply at each collection time of the ripple signal is taken; wherein P is more than or equal to 1;
s3.2: for the original ripple wave frequency spectrum signal obtained in S3.1
Dividing a frequency spectrum interval, specifically:
intercept frequencyfIn [ (1-a)f m , (1+a)f m ]Raw ripple spectrum signal in interval
For a ripple spectrum signal
, j=1, …, P, in ripple spectrum signal
In a width ofNWindow of in preset step lengthN step Sliding M-1 times to obtain a width ofNOf the frequency spectrum interval
, j=1,…,P, m=1,2, …, M; wherein the content of the first and second substances,ais a preset constant, and 0<a<1；f m The frequency of the maximum harmonic component of the internal switching signal of each switching power supply in the ripple signal is obtained; presetting step lengthN step Not more thanN；
S3.3: obtaining a spectrum sample of each switching power supply, specifically:
from the original ripple spectral signal
The output power of each switching power supply is obtained by calculation according to the corresponding output current of each switching power supply, and the corresponding middle power supply is determined according to the output power-center frequency curve of each switching power supplyHeart frequency and then width by center frequencyNCenter-truncated ripple spectrum signal
The obtained spectrum interval is a spectrum sample of the corresponding switching power supply
, j=1,…,P, d=1,2,…,D；
S3.4: for the spectral interval obtained at S3.2
Marking, specifically:
calculating spectral intervals
With spectral samples corresponding to each switching power supply
Overlap ratio between
, j=1,…,P, m=1,2,…,M, d=1,2, …, D, screening out the overlap ratio
Maximum value of (max) ((
) If the maximum value max (
) If the frequency spectrum interval is larger than the preset overlap rate threshold value, the frequency spectrum interval is divided
Is marked with a maximum value max (
) Number corresponding to switching power supplyd(ii) a Otherwise, the frequency spectrum interval is divided
The label of (1) is labeled "unknown class";
all frequency spectrum intervals
And the corresponding label is used as a switch power supplyiTraining sample data of the spectrum characteristics;
s4: judgment ofiIf D is equal to D, if yes, go to S5; otherwise, it ordersi=i+1, go back to S3;
s5: constructing an LSTM network, training based on the spectrum characteristic training sample data of each switching power supply, and inputting the spectrum characteristic training sample data into the spectrum interval corresponding to each switching power supply
The training target is a corresponding label to obtain a trained LSTM network;
s6: collecting ripple signals of a load end of a redundant power supply to be identified, and processing to obtain corresponding original ripple frequency spectrum signals
Dividing the frequency spectrum interval by adopting the same method as S3.2; sequentially inputting the frequency spectrum intervals into the trained LSTM network according to the window sliding sequence to obtain labels corresponding to the frequency spectrum intervals, and screening the frequency spectrum intervals corresponding to the switching power supplies; and determining the output power of each switching power supply according to the center frequency of the frequency spectrum interval and an output power-center frequency curve, calculating to obtain output current, and finally determining the current sharing state of the redundant power supply based on the output current of each switching power supply.
2. Spectrum and LS based on claim 1The method for identifying the current sharing state of the redundant power supply of the TM network is characterized in that the preset period T in S3.1 is determined by the sampling rate of ripple signalsf s Sum frequencyf m The formula is determined as follows:
T=γf m /f s
wherein gamma is a preset integer and gamma is greater than 2.
3. The method for identifying the current sharing status of the redundant power supplies based on the frequency spectrum and the LSTM network according to claim 1, wherein the collected ripple signals are processed through fast Fourier transform and normalization in S3.1 and S6.
4. The method for identifying current sharing status of redundant power supplies based on spectrum and LSTM network as claimed in claim 1, wherein the spectrum interval in S3.2
Number of (3) M = Round (2T)f m /N step )。
5. The method for identifying current sharing status of redundant power supplies based on spectrum and LSTM network as claimed in claim 1, wherein width in S3.2NThe value taking process is as follows: to the switching power supplyiDefining the original ripple spectrum signal
, j=1, …, the spectrum region formed by continuous frequency points with normalization values all greater than 0.01 in P is the spectrum width, and the switching power supply is obtainediMaximum spectral width at different output powers, and thus the widthNGreater than the maximum of the corresponding maximum spectral widths of all switching power supplies.
6. The method for identifying current sharing status of redundant power supplies based on spectrum and LSTM network as claimed in claim 1, wherein the calculation weight in S3.4Rate of stack
The formula of (1) is as follows:
wherein the content of the first and second substances,
is a frequency spectrum interval
And a switching power supplydOf the spectrum sample
The width of the overlapping spectrum.
7. The method for identifying current sharing status of redundant power supplies based on spectrum and LSTM network as claimed in claim 1, wherein the predetermined threshold value of overlap ratio in S3.4 is greater than or equal to 50% + (c) ((r))N step /2N)。
8. The method for identifying the current sharing status of the redundant power supplies based on the spectrum and the LSTM network according to claim 1, wherein the LSTM network constructed in S5 further comprises a convolutional neural network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111324442.8A CN113759271B (en) | 2021-11-10 | 2021-11-10 | Redundant power supply current sharing state identification method based on frequency spectrum and LSTM network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111324442.8A CN113759271B (en) | 2021-11-10 | 2021-11-10 | Redundant power supply current sharing state identification method based on frequency spectrum and LSTM network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113759271A true CN113759271A (en) | 2021-12-07 |
| CN113759271B CN113759271B (en) | 2022-01-25 |
Family
ID=78784815
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111324442.8A Active CN113759271B (en) | 2021-11-10 | 2021-11-10 | Redundant power supply current sharing state identification method based on frequency spectrum and LSTM network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113759271B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116520182A (en) * | 2023-05-11 | 2023-08-01 | 珠海中瑞电力科技有限公司 | DCS device power supply module abnormality early diagnosis method |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4276590A (en) * | 1979-04-30 | 1981-06-30 | The Perkin-Elmer Corporation | Current sharing modular power system |
| JPH06276750A (en) * | 1993-03-19 | 1994-09-30 | Fuji Electric Co Ltd | Control method of inverter |
| US6628106B1 (en) * | 2001-07-30 | 2003-09-30 | University Of Central Florida | Control method and circuit to provide voltage and current regulation for multiphase DC/DC converters |
| TW200421687A (en) * | 2003-04-09 | 2004-10-16 | Phoenixtec Power Co Ltd | Alternating current output parallel power system and share current control method thereof |
| US20100013306A1 (en) * | 2008-07-18 | 2010-01-21 | Heineman Douglas E | Intelligent Management of Current Sharing Group |
| CN102594157A (en) * | 2012-04-05 | 2012-07-18 | 重庆安谐新能源技术有限公司 | Signal acquisition device and processing method for alternating current power current-sharing output system |
| CN102621373A (en) * | 2012-04-05 | 2012-08-01 | 重庆安谐新能源技术有限公司 | Signal collection device and signal processing method of current-sharing input system of alternating-current power supply |
| CN105958469A (en) * | 2016-06-15 | 2016-09-21 | 厦门科灿信息技术有限公司 | Synchronization and current-sharing methods for programmable multi-machine parallel power supply system |
| CN113131604A (en) * | 2021-04-29 | 2021-07-16 | 电子科技大学 | Automatic control system for current-sharing output of redundant power supply |
| CN113189513A (en) * | 2021-04-29 | 2021-07-30 | 电子科技大学 | Ripple-based redundant power supply current sharing state identification method |
-
2021
- 2021-11-10 CN CN202111324442.8A patent/CN113759271B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4276590A (en) * | 1979-04-30 | 1981-06-30 | The Perkin-Elmer Corporation | Current sharing modular power system |
| JPH06276750A (en) * | 1993-03-19 | 1994-09-30 | Fuji Electric Co Ltd | Control method of inverter |
| US6628106B1 (en) * | 2001-07-30 | 2003-09-30 | University Of Central Florida | Control method and circuit to provide voltage and current regulation for multiphase DC/DC converters |
| TW200421687A (en) * | 2003-04-09 | 2004-10-16 | Phoenixtec Power Co Ltd | Alternating current output parallel power system and share current control method thereof |
| US20100013306A1 (en) * | 2008-07-18 | 2010-01-21 | Heineman Douglas E | Intelligent Management of Current Sharing Group |
| CN102594157A (en) * | 2012-04-05 | 2012-07-18 | 重庆安谐新能源技术有限公司 | Signal acquisition device and processing method for alternating current power current-sharing output system |
| CN102621373A (en) * | 2012-04-05 | 2012-08-01 | 重庆安谐新能源技术有限公司 | Signal collection device and signal processing method of current-sharing input system of alternating-current power supply |
| CN105958469A (en) * | 2016-06-15 | 2016-09-21 | 厦门科灿信息技术有限公司 | Synchronization and current-sharing methods for programmable multi-machine parallel power supply system |
| CN113131604A (en) * | 2021-04-29 | 2021-07-16 | 电子科技大学 | Automatic control system for current-sharing output of redundant power supply |
| CN113189513A (en) * | 2021-04-29 | 2021-07-30 | 电子科技大学 | Ripple-based redundant power supply current sharing state identification method |
Non-Patent Citations (2)
| Title |
|---|
| D.J.PERREAULT ET AL.: "Switching-ripple-based current sharing for paralleled power converters", 《IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS I:FUNDAMENTAL THEORY AND APPLICATIONS》 * |
| 林燎源: "多逆变器并联均流及功率灵活分配若干关键问题研究", 《中国优秀博硕士学位论文全文数据库（博士） 工程科技II辑》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116520182A (en) * | 2023-05-11 | 2023-08-01 | 珠海中瑞电力科技有限公司 | DCS device power supply module abnormality early diagnosis method |
| CN116520182B (en) * | 2023-05-11 | 2023-11-14 | 珠海中瑞电力科技有限公司 | DCS device power supply module abnormality early diagnosis method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113759271B (en) | 2022-01-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110336534B (en) | Fault diagnosis method based on photovoltaic array electrical parameter time series feature extraction | |
| CN108664901B (en) | Improved LMD-based micro-grid power quality disturbance signal detection method | |
| CN109546659B (en) | Power distribution network reactive power optimization method based on random matrix and intelligent scene matching | |
| CN113759271B (en) | Redundant power supply current sharing state identification method based on frequency spectrum and LSTM network | |
| CN112684401B (en) | Method and system for processing operation error data of electric energy meter in light-load station area | |
| CN103323672B (en) | A kind of online resonance prior-warning device of the parallel power condenser and method | |
| CN116183058A (en) | Monitoring method of intelligent capacitor | |
| CN116307944B (en) | Distribution box remote monitoring system based on artificial intelligence and Internet of things | |
| CN116628616A (en) | Data processing method and system for high-power charging energy | |
| CN111856209A (en) | Power transmission line fault classification method and device | |
| CN112540220A (en) | Voltage sag detection circuit and device | |
| CN109767353B (en) | Photovoltaic power generation power prediction method based on probability distribution function | |
| KR101835377B1 (en) | Method for estimating state-of-charge of battery pack based on multivariate wavelet transform and apparatus thereof | |
| CN113189513B (en) | Ripple-based redundant power supply current sharing state identification method | |
| CN108415880B (en) | Line loss characteristic analysis method based on sample entropy and wavelet transformation | |
| CN112182499B (en) | Low-voltage distribution network topological structure identification method based on time sequence electric quantity data | |
| CN116626575B (en) | Fitting correction method and device for data acquisition equipment | |
| CN107576848B (en) | A kind of template setting and template detection method based on spectrum analysis | |
| CN116187968A (en) | Substation equipment state evaluation method and system, related equipment and medium | |
| CN112710898A (en) | New energy rapid frequency response test system and method | |
| CN112327216B (en) | Cable health state assessment device and method based on edge calculation | |
| CN110658385A (en) | Power consumption classification metering method based on complex trapezoid method | |
| CN116510223B (en) | Self-identification fire pump electrical parameter inspection monitoring system based on Internet of things | |
| CN117315470B (en) | Water quality parameter inversion system based on ground-air full-spectrum hyperspectral data | |
| CN117233629B (en) | Method, system, equipment and medium for testing electrical performance of lithium ion battery |
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 |