CN109307798B - Power signal filtering method for switch event detection - Google Patents
Power signal filtering method for switch event detection Download PDFInfo
- Publication number
- CN109307798B CN109307798B CN201810995998.1A CN201810995998A CN109307798B CN 109307798 B CN109307798 B CN 109307798B CN 201810995998 A CN201810995998 A CN 201810995998A CN 109307798 B CN109307798 B CN 109307798B
- Authority
- CN
- China
- Prior art keywords
- power signal
- matrix
- filtering
- function
- calculating
- 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.)
- Expired - Fee Related
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R21/00—Arrangements for measuring electric power or power factor
- G01R21/001—Measuring real or reactive component; Measuring apparent energy
- G01R21/002—Measuring real component
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Complex Calculations (AREA)
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
Abstract
The invention provides a power signal filtering method for detecting a switching event, which is used for solving the problems that the background noise with obvious non-stationarity and non-Gaussian characteristic cannot be effectively filtered for a power signal and the switching event is difficult to be accurately detected in the prior art. The method comprises the following steps: and summing all scales of the obtained filtered power signals to obtain a filtered power matrix. According to the invention, through carrying out multi-scale summation after filtering the power signals, noise with non-stationarity and non-Gaussian characteristic is effectively filtered, the power signals are effectively filtered, accurate switching event detection is carried out, the filtering method is simple, and the calculation speed is high, so that accurate decomposition of the household electrical load is completed.
Description
Technical Field
The invention belongs to the field of electric power, and particularly relates to a power signal filtering method for switch event detection.
Background
With the development of smart grids, the analysis of household electrical loads becomes more and more important. Through the analysis of the power load, a family user can obtain the power consumption information of each electric appliance and a refined list of the power charge in time; the power department can obtain more detailed user power utilization information, can improve the accuracy of power utilization load prediction, and provides a basis for overall planning for the power department. The current electric load decomposition is mainly divided into an invasive load decomposition method and a non-invasive load decomposition method. The non-invasive load decomposition method does not need to install monitoring equipment on the internal electric equipment of the load, can obtain the load information of each electric equipment only according to the total information of the electric load, has less investment and convenient use, and is suitable for the decomposition of household load electricity. In the non-invasive load decomposition algorithm, the detection of the switching event of the electrical equipment is the most important link. The operating state of any one of the electric devices changes, the consumed power value of any one of the electric devices also changes inevitably, and the change is reflected in the total power consumed by all the electric devices, and the commonly used switch event detection takes the change value Δ P of the active power P as the judgment basis of the event detection. Fig. 1 is a schematic diagram showing a simulation of energy decomposition of a household electrical load in the prior art.
This method requires setting a reasonable threshold for the power variation value, and has the following problems: the instantaneous power value at the starting time of some electric appliances has larger peak, namely pulse noise (for example, the starting current of a motor is far larger than the rated current), which causes the change value of the steady-state power of the electric appliances to be inaccurate, thereby influencing the judgment of a switching event; the transient processes of different household appliances are either long or short (the duration and the frequency of occurrence of impulse noise are very different), so that the determination of the power variation value becomes difficult; due to the fact that the active power changes suddenly when the quality of the electric energy changes (such as voltage drop), misjudgment is likely to happen. Fig. 2 shows a power signal (also referred to as a power data sequence) measured by a switching event detection device and a conventional detection algorithm in the prior art, and the distribution of noise (including impulse noise) in the signal can be seen, where the intensity of the (impulse) noise in the power sequence is very large, and 3 switching events are detected although only one true switching event exists. Therefore, filtering the power signal is an important step in the switching event detection process.
Noise cancellation devices commonly used in the prior art are low pass filters and median filters. The low-pass filter has a good effect of eliminating stationarity and noise conforming to the Gaussian law, but has a poor effect of eliminating noise with non-stationarity and non-Gaussian characteristics. In daily life, nonlinear electrical appliances such as a juicer, a coffee maker, a soymilk maker and the like using a motor are increasingly applied, and more impulse noise appears in background noise. The noise has high instantaneous power and presents more obvious non-stationarity and non-Gaussian characteristics, and the common noise elimination equipment in the prior art cannot effectively filter power signals and is difficult to accurately detect switching events.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a power signal filtering method for detecting a switching event, aiming at non-stationary and non-Gaussian power signals in the prior art, the method can also effectively filter signals and accurately detect the switching event, and is simple in filtering method and high in calculation speed, so that the household electrical load is accurately decomposed.
To solve the above technical problem, an embodiment of the present invention provides a power signal filtering method for switching event detection, where the method includes: and summing all the scales to obtain a filtered power signal matrix.
Wherein the method further comprises the steps of:
before solving the filtered power signal matrix:
step S1, inputting measured power signals;
step S2, segmenting the actually measured power signal, and rearranging the segmented data into a power signal matrix;
step S3, calculating the discrete Fourier transform function of the basis function;
step S4, calculating the discrete value of the basis function according to the Fourier transform function of the basis function;
step S5, calculating the discrete value of the mother function according to the discrete value of the basis function;
step S6, calculating the projection of the power signal under the mother function according to the discrete value of the mother function to obtain a transformed power signal matrix;
step S7, filtering the transformed power signal matrix;
after obtaining the filtered power signal matrix, the method further includes:
and step S9, converting the filtered power signal matrix into a data sequence, rearranging the data, and obtaining the power signal with the noise filtered.
Wherein the step S2 further includes:
step S21, setting the length of the input actual measurement power signal as N, and dividing all data into NRA segment;
step S22, segmenting according to data sequence, wherein each segment contains NCThe number of the data is one,
step S23, if N < NR×NCZero-filling the insufficient part;
step S24, rearranges the segmented data into a matrix form.
Wherein, the step S3 is to calculate the basis functionFourier transform function ofThe method specifically comprises the following steps:
wherein:
Wherein, in the step S4, discrete values of the basis functions are obtainedComprises the following steps:
wherein, the discrete value ψ of the mother function is obtained in the step S5lmijComprises the following steps:
l=1,2,L,10
m=1,2,L,10
i=1,2,L,NR
j=1,2,L,NC
wherein, the step S6 of obtaining the projection of the power signal under the mother function further includes:
step S61, dividing the discrete value psi of the mother functionlmijConversion to a matrix of mother functions:
wherein psilmijIs a matrixlmRow i and column j elements of (1);
step S62, the power signal matrix and the mother function matrix are subjected to point multiplication to obtain the transformed power signal
Number matrix:
wherein, L is 1,2, L, 10; m is 1,2, L, 10.
Wherein, the filtering the transformed power signal matrix in step S7 further includes:
Wherein, mean [ psi ]lmij]Indicating that the variables l and m are fixed, psi in the case where i, j take all valueslmijThe median value of (d);
step S72, according to the threshold lambdalmThe value of each element of the new matrix is determined.
Wherein the content of the first and second substances,
in the step S71, by
Calculating a threshold lambdalm;
Wherein, mean [ psi ]lmij]Indicating that the variables l and m are fixed, psi in the case where i, j take all valueslmijThe median value of (d);
the step S72 further includes:
if it is notIf any element in (b) is smaller than the threshold λ, making this element 0; after the processing, a new matrix is obtainedThe ith row and jth column elements of the matrix are denoted as
Step S8, where the summation of all the scales is performed to obtain a filtered power signal matrix, and specifically, the filtered power signal matrix is:
P1=[Pij](6)
wherein the content of the first and second substances,i, j represents a scale, and all i, j are summed, i.e., all scales are summed.
The technical scheme of the invention has the following beneficial effects:
according to the invention, through carrying out multi-scale summation after filtering the power signal, noise with non-stationarity and non-Gaussian characteristic is effectively filtered, and the power signal is effectively filtered. The signal-to-noise ratio of the signal after noise filtering is improved by about 5dB, and a good foundation is laid for subsequent switch event detection. The filtering method is simple and the calculation speed is high.
Drawings
In order to more clearly illustrate the embodiments of the present invention and the prior art, the following technical scheme description figures of the present invention are briefly introduced, and it is obvious that other figures can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a schematic diagram of a prior art energy-resolved simulation of a domestic electrical load;
FIG. 2 is a graph of power signals measured by a prior art switching event detection device and a conventional detection algorithm;
FIG. 3 is a flow chart illustrating a filtering method for detecting a switching event according to an embodiment of the present invention;
fig. 4 is a schematic diagram of data segmentation and matrix arrangement.
Detailed Description
In order to make the technical problems, technical solutions and advantages of the present invention more apparent, the following detailed description is given with reference to the accompanying drawings and specific embodiments.
The invention provides a novel filtering method aiming at the noise characteristic in a power signal, aiming at the problems that the common noise elimination equipment in the prior art can not effectively filter the noise with non-stationarity and non-Gaussian characteristic and is difficult to accurately detect the switching event.
The present invention will be described in further detail below with reference to specific embodiments and drawings.
Examples
The present embodiment provides a power signal filtering method for detecting a switching event, and fig. 3 illustrates the power signal filtering method according to the present embodiment. As shown in fig. 3, the power signal filtering method for switching event detection includes the steps of:
in step S1, the measured power signal is input, and the signal length is set to N.
Specifically, the measured power signal sequences P (1), P (2), L, P (N) are input, and N is the length of the power signal sequence.
Step S2, segmenting the measured power signal and rearranging the segmented data into a power signal matrix P.
Specifically, the method comprises the following steps:
step S21, dividing all data into NRAnd (4) section.
Preferably, said N isR256 or 512 or 1024.
Step S22, segmenting according to data sequence, wherein each segment contains NCAnd (4) data.
Wherein the content of the first and second substances,symbolMeaning rounding up. For exampleThe purpose of this is that all data is involved in the operation and not discarded.
Step S23, if N < NR×NCThe insufficient portion is zero-padded.
Step S24, rearranges the segmented data into a matrix form.
Fig. 4 is a schematic diagram of data segmentation and matrix arrangement. As shown in FIG. 4, a segment of data is a row, and thus, the power signal matrix P has N in totalRLine, NCAnd (4) columns.
Specifically, the transformed function is:
wherein:
Step S4, calculating discrete value of basis function according to Fourier transform of basis function
Wherein, i is 1,2, L, NR;j=1,2,L,NC
Step S5, discrete value according to basis functionCalculating the discrete value psi of the mother functionlmij:
Wherein, L is 1,2, L, 10; m is 1,2, L, 10; i is 1,2, L, NR;j=1,2,L,NC;
Step S6, discrete value psi according to mother functionlmijAnd solving the projection of the power signal under the mother function to obtain a transformed power signal matrix.
Further, the method specifically comprises the following steps:
step S61, dividing the discrete value psi of the mother functionlmijConversion to a matrix of mother functions:
phi in formula (3)lmijRepresentation matrixlmThe ith row and the jth column of the matrix, the matrix having N in totalRLine, NCAnd (4) columns.
Step S62, the power signal matrix and the mother function matrix are subjected to point multiplication to obtain the transformed power signal
Number matrix:
wherein, L is 1,2, L, 10; m is 1,2, L, 10.
In step S7, the transformed power signal matrix is filtered.
Specifically, the filtering process includes the following steps:
step S71, calculating the threshold lambda by the following equationlm:
Wherein, mean [ psi ]lmij]Indicating that the variables l and m are fixed, psi in the case where i, j take all valueslmijThe median value of (d);
in step S72, the value of each element of the new matrix is determined based on the threshold value.
In particular, ifIs smaller than a threshold value lambdalmLet this element be 0; after the processing, a new matrix is obtainedThe ith row and jth column elements of the matrix are denoted as
Step S8, summing all scales, and solving a filtered power signal matrix:
P1=[Pij](6)
in this step, i, j represents a scale, and the sum of all i, j is the sum of all scales.
And step S9, converting the filtered power signal matrix into a data sequence, rearranging the data, and obtaining the power signal with the noise filtered.
Specifically, the method comprises the following steps: will matrix P1The first line of data is used as the first section, the 2 nd line of data is used as the second section, and so on, the last line of data is used as the last section, the sections are connected in sequence, and before interception is carried outThe N data of the surface form a data sequence, which is the power signal after filtering out the background noise (especially the strong impulse noise).
According to the technical scheme, the power signal filtering method for detecting the switching event effectively filters noise with non-stationarity and non-Gaussian characteristics through multi-scale filtering, effectively filters the power signal, accurately detects the switching event, is simple in filtering method and high in calculation speed, and accordingly achieves accurate decomposition of the household power load.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (9)
1. A method of filtering a power signal for switching event detection, the method comprising:
step S1, inputting measured power signals;
step S2, segmenting the actually measured power signal, and rearranging the segmented data into a power signal matrix;
step S3, calculating the discrete Fourier transform function of the basis function;
step S4, calculating the discrete value of the basis function according to the Fourier transform function of the basis function;
step S5, calculating the discrete value of the mother function according to the discrete value of the basis function;
step S6, calculating the projection of the power signal under the mother function according to the discrete value of the mother function to obtain a transformed power signal matrix;
step S7, filtering the transformed power signal matrix;
step S8, summing all scales, and solving a filtered power signal matrix;
and step S9, converting the filtered power signal matrix into a data sequence, rearranging the data, and obtaining the power signal with the noise filtered.
2. The power signal filtering method according to claim 1, wherein the step S2 further comprises:
step S21, setting the length of the input actual measurement power signal as N, and dividing all data into NRA segment;
step S22, segmenting according to data sequence, wherein each segment contains NCThe number of the data is one,
step S23, if N < NR×NCZero-filling the insufficient part;
step S24, rearranges the segmented data into a matrix form.
6. the method for filtering a power signal according to claim 1, wherein the step S6 of calculating the projection of the power signal under the mother function further comprises:
step S61, dividing the discrete value psi of the mother functionlmijConversion to a matrix of mother functions:
wherein psilmijIs a matrixlmRow i and column j elements of (1);
step S62, performing dot product operation on the power signal matrix and the mother function matrix to obtain a transformed power signal matrix:
wherein, l is 1,2, …, 10; m is 1,2, …, 10.
7. The method for filtering a power signal according to claim 1, wherein the step S7 of matrix filtering the transformed power signal further comprises:
Wherein the content of the first and second substances,indicating that the variables l and m are fixed, psi in the case where i, j take all valueslmijThe median value of (d);
step S72, according to the threshold lambdalmThe value of each element of the new matrix is determined.
8. The power signal filtering method according to claim 7,
in the step S71, by
Calculating a threshold lambdalm;
Wherein, mean [ psi ]lmij]Indicating that the variables l and m are fixed, psi in the case where i, j take all valueslmijThe median value of (d);
the step S72 further includes:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810995998.1A CN109307798B (en) | 2018-08-29 | 2018-08-29 | Power signal filtering method for switch event detection |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810995998.1A CN109307798B (en) | 2018-08-29 | 2018-08-29 | Power signal filtering method for switch event detection |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109307798A CN109307798A (en) | 2019-02-05 |
CN109307798B true CN109307798B (en) | 2020-08-28 |
Family
ID=65223912
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810995998.1A Expired - Fee Related CN109307798B (en) | 2018-08-29 | 2018-08-29 | Power signal filtering method for switch event detection |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109307798B (en) |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109902961A (en) * | 2019-03-04 | 2019-06-18 | 广东石油化工学院 | The multi-scale filtering method and system filtered for power signal in Energy Decomposition |
CN110196354B (en) * | 2019-04-23 | 2021-09-17 | 广东石油化工学院 | Method and device for detecting switching event of load |
CN110221121A (en) * | 2019-06-28 | 2019-09-10 | 广东石油化工学院 | A kind of on-load switch event detecting method and system based on Signal separator |
CN110531149B (en) * | 2019-08-31 | 2021-06-18 | 广东石油化工学院 | Power signal filtering method and system based on waveform regularization |
CN110542855B (en) * | 2019-09-08 | 2021-09-21 | 广东石油化工学院 | Load switch event detection method and system based on discrete cosine transform |
CN110726870A (en) * | 2019-10-20 | 2020-01-24 | 广东石油化工学院 | Load switch event detection method and system based on data purity |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1610144A2 (en) * | 2004-06-23 | 2005-12-28 | Whirlpool Corporation | Method for calibrating offset from DC bus |
JP4765180B2 (en) * | 2001-03-07 | 2011-09-07 | パナソニック株式会社 | Method for measuring output power of inductive load of electric powered vehicle |
CN102288285A (en) * | 2011-05-24 | 2011-12-21 | 南京航空航天大学 | Blind source separation method for single-channel vibration signals |
CN102841341A (en) * | 2012-09-03 | 2012-12-26 | 深圳先进技术研究院 | Pulse radar moving-target detection method |
CN106908657A (en) * | 2017-03-06 | 2017-06-30 | 中国海洋大学 | A kind of load switch detection means and method based on power |
-
2018
- 2018-08-29 CN CN201810995998.1A patent/CN109307798B/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4765180B2 (en) * | 2001-03-07 | 2011-09-07 | パナソニック株式会社 | Method for measuring output power of inductive load of electric powered vehicle |
EP1610144A2 (en) * | 2004-06-23 | 2005-12-28 | Whirlpool Corporation | Method for calibrating offset from DC bus |
CN102288285A (en) * | 2011-05-24 | 2011-12-21 | 南京航空航天大学 | Blind source separation method for single-channel vibration signals |
CN102841341A (en) * | 2012-09-03 | 2012-12-26 | 深圳先进技术研究院 | Pulse radar moving-target detection method |
CN106908657A (en) * | 2017-03-06 | 2017-06-30 | 中国海洋大学 | A kind of load switch detection means and method based on power |
Non-Patent Citations (3)
Title |
---|
Research of Transmission Parameters of Four-Conductor Cables for Power Line Communication;Mengcong Tang 等;《2008 International Conference on Computer Science and Software Engineering》;20081214;全文 * |
基于稀疏矩阵变换的电网故障电流快速算法;陈志光 等;《广东电力》;20151009;第28卷(第9期);全文 * |
矩阵稀疏化中的最优子波滤波器组设计;彭志威 等;《西安电子科技大学学报》;19990220(第1期);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN109307798A (en) | 2019-02-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109307798B (en) | Power signal filtering method for switch event detection | |
CN109188069B (en) | Pulse noise filtering method for load switch event detection | |
CN108918932B (en) | Adaptive filtering method for power signal in load decomposition | |
CN108918931B (en) | Power signal self-adaptive filtering method in load decomposition | |
CN111680590A (en) | Power signal filtering method and system by using contraction gradient | |
CN108918927B (en) | Power signal filtering method and system in energy decomposition | |
CN108918929B (en) | Power signal self-adaptive filtering method in load decomposition | |
CN109241874B (en) | Power signal filtering method in energy decomposition | |
CN111666870A (en) | Power signal reconstruction method and system by utilizing quadratic constraint | |
CN110187167B (en) | Method and device for detecting load switch event based on manifold classification | |
CN110196354B (en) | Method and device for detecting switching event of load | |
CN110542855B (en) | Load switch event detection method and system based on discrete cosine transform | |
CN109740582B (en) | Power signal noise filtering method and system for energy decomposition | |
CN112434567B (en) | Power signal filtering method and system by using noise jitter property | |
CN111830405A (en) | Load switch event detection method and system by using frequency difference | |
CN111832474A (en) | Power signal filtering method and system by using energy scale | |
CN111585544A (en) | Method and filter for filtering power signal impulse noise | |
CN110221119B (en) | Load switch event detection method and system based on power and akie fusion information | |
CN109194367B (en) | Power signal reconstruction method in energy decomposition | |
CN111639606A (en) | Power signal filtering method and system utilizing Dantzig total gradient minimization | |
CN112307986B (en) | Load switch event detection method and system by utilizing Gaussian gradient | |
CN111737645A (en) | Power signal reconstruction method and system by using prediction matrix | |
CN111400649A (en) | Harmonic responsibility quantification method and device, computer equipment and storage medium | |
CN110514884B (en) | Power signal filtering method and system based on delay vector | |
CN110705426B (en) | Power signal filtering method and system by using deblurring operator |
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 | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200828 Termination date: 20210829 |