CN115407157A - Complex multi-state electrical appliance load event distinguishing method and device - Google Patents

Complex multi-state electrical appliance load event distinguishing method and device Download PDF

Info

Publication number
CN115407157A
CN115407157A CN202211347743.7A CN202211347743A CN115407157A CN 115407157 A CN115407157 A CN 115407157A CN 202211347743 A CN202211347743 A CN 202211347743A CN 115407157 A CN115407157 A CN 115407157A
Authority
CN
China
Prior art keywords
load event
load
current waveform
current
state
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.)
Pending
Application number
CN202211347743.7A
Other languages
Chinese (zh)
Inventor
丁贵立
倪嘉豪
唐健耀
王宗耀
许志浩
康兵
王晓虎
颜高洋
程巧
施嘉兵
袁净帅
何登旋
刘文轩
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangxi Paiyuan Technology Co ltd
Nanchang Institute of Technology
Original Assignee
Jiangxi Paiyuan Technology Co ltd
Nanchang Institute of Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Jiangxi Paiyuan Technology Co ltd, Nanchang Institute of Technology filed Critical Jiangxi Paiyuan Technology Co ltd
Priority to CN202211347743.7A priority Critical patent/CN115407157A/en
Publication of CN115407157A publication Critical patent/CN115407157A/en
Pending legal-status Critical Current

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

Abstract

The invention belongs to the technical field of household appliance control, and relates to a complex multi-state electrical appliance load event distinguishing method and a complex multi-state electrical appliance load event distinguishing device, wherein the method obtains instantaneous current of an original circuit and calculates an effective value of the instantaneous current; calculating the change rate of the current waveform according to the effective value of the instantaneous current to determine the change rate as a time node of the occurrence of the load event; extracting a steady-state current waveform of one period or a plurality of periods from a time node of the occurrence of the load event; after Savitzky-Golay filtering and discrete wavelet transformation, calculating the correlation of current instantaneous values before and after a time node of a load event occurrence and after Savitzky-Golay filtering and discrete wavelet transformation by using a Pearson correlation coefficient, and judging whether the load event is an electric appliance switching load event or a multi-state electric appliance working mode switching load event according to the Pearson correlation coefficient. The invention realizes the judgment of the load event type through the correlation of the current instantaneous value.

Description

Complex multi-state electrical appliance load event distinguishing method and device
Technical Field
The invention belongs to the technical field of household appliance control, and particularly relates to a method and a device for judging a complex multi-state appliance load event.
Background
The multi-state electric appliance refers to an electric appliance of which the main working elements work indirectly after the electric appliance is put into operation or are restricted by a built-in circuit to execute different working strategies. The multi-state electric appliance has at least two working modes, different current waveforms can be generated when the multi-state electric appliance works in different modes, circuit characteristics can be changed by switching between the two working modes, and common multi-state household electric appliances comprise a washing machine, an air conditioner, a refrigerator, an electric cooker, an electric fan and the like.
The load event refers to the change of the running state of the electric equipment, is a key link of non-intrusive load monitoring, and specifically comprises two types of load switching events of the electric appliance and load switching events of the working mode of the multi-state electric appliance. The household load of residents has more multi-state electric appliances, such as air conditioners, washing machines, refrigerators and other electric appliances, the multi-state working mode characteristics often cause that switching load events of the electric appliances and conversion load events of the working modes of the multi-state electric appliances are difficult to identify, and the precision of an NILM (non-embedded identification) algorithm is reduced.
Relevant scholars at home and abroad carry out more researches on the problem of the identification precision of the NILM load event, but the existing documents are developed on the detection precision of the load event, but the researches are rarely related to the discrimination research of the switching load event of the electric appliance under the complex multi-state electric appliance scene and the conversion load event of the working mode of the multi-state electric appliance.
Disclosure of Invention
In order to solve the difficult problem of distinguishing switching load events of electric appliances and switching load events of working modes of multi-state electric appliances, the invention provides a method and a device for distinguishing the load events of complex multi-state household electric appliances, wherein the method uses a Savitzky-Golay algorithm to process and collect the steady-state current waveform of the electric appliances, and filters high-frequency noise components contained in the waveform to obtain an approximate current waveform; extracting an approximate current waveform as a main feature based on discrete wavelet transform; and calculating the correlation coefficient of the steady-state waveforms before and after the load event based on the Pearson correlation coefficient, and judging the type of the load event according to the magnitude of the correlation coefficient. In the operation monitoring of the household appliance, the method provided by the invention is added after the variable point detection process in the prior art, so that the identification precision of the NILM algorithm can be improved.
The invention is realized by the following technical scheme. A complex multi-state electric appliance load event distinguishing method comprises the following steps:
a method for distinguishing complex polymorphic electric appliance load events comprises the following steps:
step S1: acquiring instantaneous current of an original circuit, and calculating an effective value of the instantaneous current;
step S2: calculating the current waveform change rate according to the instantaneous current effective value, and appointing a time node of the load event when the current waveform change rate is greater than a threshold value m;
and step S3: calculating the derivative of the effective value of the instantaneous current before and after the occurrence of the load event when the derivative value is less than the threshold valueaWhen the load event occurs, the current waveform is determined to enter a steady state, and the steady-state current waveform of one period or a plurality of periods is extracted from the time node of the load eventShaping;
and step S4: filtering high-frequency components and clutter signals in a steady-state current waveform before and after a load event occurs by using a Savitzky-Golay filtering algorithm;
step S5: extracting approximate components of the steady-state current waveforms before and after the load event is filtered by using discrete wavelet transform to obtain current instantaneous values after Savitzky-Golay filtering and discrete wavelet transform before and after the load event is generated;
step S6: and calculating the correlation of the current instantaneous values before and after the time node of the load event occurrence through Savitzky-Golay filtering and discrete wavelet transformation by using a Pearson correlation coefficient, and judging whether the load event is an electric appliance switching load event or a multi-state electric appliance working mode switching load event according to the Pearson correlation coefficient.
Further preferably, the condition for assuming that the current waveform enters the steady state in step S3 is as follows:
Figure 416960DEST_PATH_IMAGE001
Figure 656312DEST_PATH_IMAGE002
time node for occurrence of load event
Figure 954569DEST_PATH_IMAGE003
Rear end
Figure 279371DEST_PATH_IMAGE004
The derivative of the current waveform at the time of day,
Figure 85391DEST_PATH_IMAGE005
time node for occurrence of load event
Figure 913670DEST_PATH_IMAGE006
Front side
Figure 507462DEST_PATH_IMAGE007
The derivative of the current waveform at time.
More preferably, in step S3, the steady-state current waveform is extracted as follows:
Figure 53981DEST_PATH_IMAGE008
in the formula (I), the compound is shown in the specification,
Figure 810496DEST_PATH_IMAGE009
is the steady state current waveform prior to the occurrence of a load event,
Figure 617915DEST_PATH_IMAGE010
for a steady state current waveform after a load event occurs,
Figure 664500DEST_PATH_IMAGE011
at the last point in time of the steady state current waveform before the load event occurs,
Figure 698315DEST_PATH_IMAGE012
and T is a power frequency period, and is the time point when the steady-state current waveform starts after the load event occurs.
Further preferably, the process of step S4 is: firstly, the width of the sliding window is selected to be 2k+1, sliding window will load the steady state current waveform before the event happens
Figure 846137DEST_PATH_IMAGE013
Is divided into 2k+1 data, wherein the time t takes the value
Figure 649008DEST_PATH_IMAGE014
，kIs a positive integer based on 2 in the sliding windowk+1 data, structurenA polynomial of order for calculating the steady state current waveform before the filtered load event
Figure 725549DEST_PATH_IMAGE015
：
Figure 105714DEST_PATH_IMAGE016
In the formula (I), the compound is shown in the specification,
Figure 932594DEST_PATH_IMAGE017
respectively 0,1,2, \8230,n-1，norder coefficients, all coefficients constituting a coefficient matrix
Figure 855550DEST_PATH_IMAGE018
；
And (3) calculating a fitting residual error by using a least square method to construct an objective function, wherein the equation of the residual error is as follows:
Figure 368571DEST_PATH_IMAGE020
Figure 111399DEST_PATH_IMAGE021
is the coefficient of the g order; equation for residual errorEDerivation is performed so that the partial derivative of the parameter is 0:
Figure 601024DEST_PATH_IMAGE022
solving a current coefficient matrix using known data in a sliding window
Figure 378488DEST_PATH_IMAGE023
When the sliding window is moved, the central point of the sliding window is taken as the steady-state current waveform before the occurrence of the load event
Figure 62410DEST_PATH_IMAGE024
When the sliding window passes through all the data to be smoothed, all the return values are the complete filtered steady-state current waveform before the load event occurs
Figure 292534DEST_PATH_IMAGE025
；
Steady state current waveform after filtered load event occurs
Figure 845570DEST_PATH_IMAGE026
And the filtered steady state current waveform before the load event occurs
Figure 477539DEST_PATH_IMAGE027
The same calculation process is performed.
More preferably, step S5 is to obtain the instantaneous current value after Savitzky-Golay filtering and discrete wavelet transform before the load event occurs by solving the following equation
Figure 801204DEST_PATH_IMAGE028
After a load event occurs, current instantaneous value is subjected to Savitzky-Golay filtering and discrete wavelet transformation
Figure 784204DEST_PATH_IMAGE029
：
Figure 615631DEST_PATH_IMAGE030
Figure 836528DEST_PATH_IMAGE031
Figure 862253DEST_PATH_IMAGE032
Figure 332549DEST_PATH_IMAGE033
In the formula:WT(a,t) In order to be the function of the inner product,
Figure 967667DEST_PATH_IMAGE034
is a function of the wavelet transform,ais a scale factor, x is the amount of time,DWT(q,p) The function is a function obtained by discretizing the scale parameters according to power series and uniformly discretizing the time,
Figure 43070DEST_PATH_IMAGE035
is a discrete wavelet transform function, t is time,q、prespectively, a scale factor and a shift factor of the discrete wavelet.
More preferably, in step S6, the pearson correlation coefficient is calculated as follows:
Figure 239697DEST_PATH_IMAGE036
in the formula (I), the compound is shown in the specification,
Figure 931709DEST_PATH_IMAGE037
represents the average value of the current after Savitzky-Golay filtering and discrete wavelet transformation before the load event happens in a discrete time period from 1 to n,
Figure 104939DEST_PATH_IMAGE038
represents the average value of the current after Savitzky-Golay filtering and discrete wavelet transformation after the load event happens in a discrete time period from 1 to n,
Figure 566008DEST_PATH_IMAGE039
is shown asiA time periodt i The instantaneous value of the current after Savitzky-Golay filtering and discrete wavelet transformation before the load event occurs,
Figure 933535DEST_PATH_IMAGE040
is shown asiA period of timet i And (3) current instantaneous values after Savitzky-Golay filtering and discrete wavelet transformation after the load event occurs.
The invention also provides a device for judging the complex multi-state electric appliance load event, and the device stores and executes the computer program instruction of the method for judging the complex multi-state electric appliance load event.
The invention has the beneficial effects that: firstly, filtering steady-state current waveforms before and after a load event by adopting a Savitzky-Golay filtering algorithm, and removing high-frequency components and noise components; secondly, decomposing the filtered current waveform through discrete wavelet transform to obtain an approximate component waveform; and finally, calculating the correlation coefficient of the waveform of the front approximate component and the rear approximate component of the load event based on the Pearson correlation coefficient, and judging the property of the load event according to the correlation coefficient. Through a large amount of current waveform tests of complex multi-state electrical appliances such as an air conditioner, a washing machine, a refrigerator and the like, experimental results show that: the Pearson correlation coefficient of the switching load event of the electrical appliance processed by the method is below 0.6, and the Pearson correlation coefficient of the mode switching load event of the multi-state electrical appliance is above 0.65, and the standard can be used as a main criterion for judging the type of the load event in a complex multi-state electrical appliance scene.
Drawings
Fig. 1 is a flow chart of a method for judging a complex multi-state electrical appliance load event according to the invention.
Fig. 2 is a schematic diagram of sliding window movement smoothing.
FIG. 3 is a graph of the results of filtering the operating conditions of the washing machine in a steady-state time sequence prior to a load event in the fast wash mode.
Fig. 4 is a filtering result of an operating state of the washing machine in a steady time sequence after a load event in a fast wash mode.
Fig. 5 is a result of filtering an operation state of the washing machine in a steady time sequence before a load event in a rinsing mode.
Fig. 6 is a result of filtering an operation state of the washing machine in a steady time sequence after a load event in a rinsing mode.
FIG. 7 is a discrete wavelet transform result of a steady state time sequence of a washing machine before a load event in a fast wash mode.
FIG. 8 is a result of discrete wavelet transform in a steady state time series after a load event for a washing machine in a fast wash mode.
Fig. 9 is a discrete wavelet transform result of a steady state time sequence of the washing machine before a load event in a rinsing mode.
Fig. 10 is a result of discrete wavelet transform in a steady state time series after a load event in a rinsing mode of a washing machine.
Detailed Description
The technical idea of the present invention is further explained in detail below with reference to the accompanying drawings.
Referring to fig. 1, a method for discriminating a complex multi-state electrical appliance load event includes the following steps:
step S1: obtaining instantaneous current of original circuiti(t) And calculating an instantaneous effective current valueI(t)：
Figure 706319DEST_PATH_IMAGE041
In the formula: t is time, T is power frequency period, 0.02s is taken,I(t 1 ), I(t 2 ),…, I(t n ) Are respectively ast 1 ,t 2 ,…,t n Instantaneous current effective value of 1,2, \8230, instantaneous current effective value of time,nis the sample point order.
S2, determining a time node of the occurrence of the load event: calculating the current waveform change rate according to the instantaneous current effective value
Figure 184705DEST_PATH_IMAGE042
When the change rate of the appointed current waveform is greater than the threshold value m, the appointed current waveform is the time node of the occurrence of the load event
Figure 270253DEST_PATH_IMAGE043
。
Namely:
Figure 543103DEST_PATH_IMAGE044
，
Figure 209708DEST_PATH_IMAGE045
time node for occurrence of load event
Figure 226205DEST_PATH_IMAGE043
Current waveform rate of change.
Step S3, extracting a steady-state current waveform from a time node of a load event: because of the current transient process of part of the electric appliances, the effective value of the instantaneous current before and after the occurrence of the load event needs to be calculated
Figure 160401DEST_PATH_IMAGE046
When the derivative value is smaller than the threshold value a, the current waveform is considered to enter a steady state;
Figure 604152DEST_PATH_IMAGE001
Figure 492473DEST_PATH_IMAGE047
time node for occurrence of load event
Figure 47082DEST_PATH_IMAGE043
Rear end
Figure 101364DEST_PATH_IMAGE048
The derivative of the current waveform at the time of day,
Figure 981595DEST_PATH_IMAGE049
time node for occurrence of load event
Figure 357213DEST_PATH_IMAGE043
Front side
Figure 715513DEST_PATH_IMAGE050
The derivative of the current waveform at time.
At this time, the steady state current waveform of one or several cycles is extracted from the time node where the load event occurs:
Figure 624301DEST_PATH_IMAGE008
in the formula (I), the compound is shown in the specification,
Figure 409854DEST_PATH_IMAGE051
is the steady state current waveform prior to the occurrence of a load event,
Figure 272768DEST_PATH_IMAGE052
for a steady state current waveform after a load event occurs,
Figure 434759DEST_PATH_IMAGE053
at the last point in time of the steady state current waveform before the load event occurs,
Figure 28132DEST_PATH_IMAGE054
is the point in time at which the steady state current waveform begins after the occurrence of a load event.
Step S4, savitzky-Golay filtering: and filtering high-frequency components and clutter signals in a steady-state current waveform before and after the load event occurs by using a Savitzky-Golay (S-G) filtering algorithm.
Compared with other similar filtering algorithms, the Savitzky-Golay algorithm can keep the distribution characteristics such as relative maximum, minimum and width, and can keep the main waveform and remove disturbance of the multi-state electric appliance in each working mode.
Basic process of Savitzky-Golay algorithm: referring to FIG. 2, first, a sliding window width of 2 is selectedk+1, sliding window converts the steady-state current waveform before the occurrence of a load event
Figure 984586DEST_PATH_IMAGE055
Is divided into 2k+1 data, wherein the time period t takes the value of
Figure 69217DEST_PATH_IMAGE056
，kIs a positive integer based on 2 in the sliding windowk+1 data, structurenOrder polynomial (general required order)nLess than the width of the sliding window), calculating the steady state current waveform after filtering before the occurrence of the load event
Figure 34899DEST_PATH_IMAGE057
：
Figure 387121DEST_PATH_IMAGE058
In the formula (I), the compound is shown in the specification,
Figure 780056DEST_PATH_IMAGE059
respectively 0,1,2, \8230,n-1，norder coefficient, allThe coefficients jointly form a coefficient matrix
Figure 617562DEST_PATH_IMAGE060
。
In order to solve the optimal coefficient, a least square method is used for calculating a fitting residual error to construct an objective function, and the equation of the residual error is as follows:
Figure 121356DEST_PATH_IMAGE061
Figure 954182DEST_PATH_IMAGE021
is the coefficient of the g order; to optimize the fitting polynomial, the residual equation needs to be minimized. Equation of the errorEThe derivation is such that its partial derivative to the parameter is 0.
Figure 157499DEST_PATH_IMAGE062
The current coefficient matrix can be solved by using the known data in the sliding window
Figure 341356DEST_PATH_IMAGE063
. When the sliding window is moved, the central point of the sliding window is taken as the steady-state current waveform before the occurrence of the load event
Figure 383261DEST_PATH_IMAGE064
The result of smoothing. When the sliding window passes through all the data to be smoothed, all the return values are the complete steady-state current waveform before the filtered load event occurs
Figure 477119DEST_PATH_IMAGE065
。
The sliding window is used for generating a steady-state current waveform after a load event
Figure 444813DEST_PATH_IMAGE066
Is divided into 2k+1 data based on 2 in a sliding windowk +1 data, structurenOrder polynomialFormula (number of general requirements)nLess than the width of the sliding window), calculating the steady state current waveform after the filtered load event occurs
Figure 256911DEST_PATH_IMAGE067
Is calculated by
Figure 102508DEST_PATH_IMAGE068
The calculation process is the same.
Step S5, discrete wavelet transform: and extracting approximate components of the steady-state current waveforms before and after the load event is filtered by Discrete Wavelet Transform (DWT) to obtain current instantaneous values after Savitzky-Golay filtering and discrete wavelet transform before and after the load event is generated.
The advantage of discrete wavelet transform is that both frequency and position information is taken into account. For multi-mode data, due to multi-harmonic and multi-pulse of the electric appliance, effective information and characteristics of the same waveform are just retained to the maximum extent by the characteristics of the multi-mode data, and comparison of steady-state waveforms before and after a load event is facilitated. A wavelet is a function that is characterized by a change over a certain period of time and has two properties: have a finite duration and abrupt frequency and amplitude; the mean value over the selected time is 0, as follows:
Figure 50872DEST_PATH_IMAGE069
in the formula:
Figure 726485DEST_PATH_IMAGE070
is the Fourier frequency domain
Figure 760300DEST_PATH_IMAGE071
The conjugate of the transformed light beam is obtained,
Figure 144008DEST_PATH_IMAGE072
is the frequency.
The wavelet transform is to shift the post-position of basic wavelet functiontAt different scales with the signal to be analyzed (filtered load event occurrence)Previous steady state current waveform
Figure 946879DEST_PATH_IMAGE073
) Inner product is made, x is the amount of time, i.e.:
Figure 521954DEST_PATH_IMAGE074
in the formula:WT(a,t) For a post-shift of the basic wavelet functiontThen, at different scalesaLower, steady state current waveform before occurrence of filtered load event
Figure 777486DEST_PATH_IMAGE075
The inner product function of (a) is,
Figure 964885DEST_PATH_IMAGE076
is a function of the wavelet transform,ais a scale factor, and is a function of,aand > 0, the scaling of the basic wavelet is controlled by a scale factor. The duration of the wavelet widens with increasing value at different scales, and the amplitude is equal to
Figure 887842DEST_PATH_IMAGE077
The inverse ratio decreases, but the waveform shape remains unchanged.
And the discrete wavelet transform is to scale factoraDiscretizing according to power series, and discretizing time to obtain uniform values (meeting the Nyquist sampling theorem).
Figure 633819DEST_PATH_IMAGE078
Similarly, the steady state current waveform after the filtered load event occurs
Figure 642226DEST_PATH_IMAGE079
Comprises the following steps:
Figure 633316DEST_PATH_IMAGE080
Figure 410779DEST_PATH_IMAGE081
in the formula:DWT(q,p) The function is a function obtained by discretizing the scale parameters according to power series and uniformly discretizing the time,
Figure 327657DEST_PATH_IMAGE082
is a discrete wavelet transform function, t is time,q、prespectively, a scale factor and a translation factor of the discrete wavelet, R is a discrete time set,
Figure 557781DEST_PATH_IMAGE083
for the instantaneous value of the current after Savitzky-Golay filtering and discrete wavelet transform before the occurrence of the load event,
Figure 352562DEST_PATH_IMAGE084
the current instantaneous value is the current instantaneous value after Savitzky-Golay filtering and discrete wavelet transformation after a load event occurs;
by passingDWT(q,p) Function can be obtained
Figure 718952DEST_PATH_IMAGE085
And
Figure 800872DEST_PATH_IMAGE086
。
s6, correlation calculation and load event judgment: and calculating the correlation of the current instantaneous values before and after the time node of the occurrence of the load event after Savitzky-Golay filtering and wavelet transformation by using a Pearson correlation coefficient, wherein the load event is switched by the electric appliance when the Pearson correlation coefficient is less than 0.6, and the load event is switched by the working mode of the multi-state electric appliance when the Pearson correlation coefficient is more than 0.6.
Figure 783872DEST_PATH_IMAGE087
In the formula (I), the compound is shown in the specification,
Figure 975819DEST_PATH_IMAGE088
represents the average value of the current after Savitzky-Golay filtering and discrete wavelet transformation before the load event happens in a discrete time period from 1 to n,
Figure 603240DEST_PATH_IMAGE089
represents the average value of the current after Savitzky-Golay filtering and discrete wavelet transformation after a load event happens in a discrete time period from 1 to n,
Figure 127500DEST_PATH_IMAGE090
denotes the firstiA period of timet i The instantaneous value of the current after Savitzky-Golay filtering and discrete wavelet transform before the load event occurs,
Figure 332217DEST_PATH_IMAGE091
is shown asiA period of timet i And (3) carrying out Savitzky-Golay filtering and discrete wavelet transform on the current instantaneous value after the load event occurs.
Taking a drum-type variable frequency washing machine as an example, the monitored current waveforms of the washing machine in the fast washing mode and the rinsing mode are different, the processing is performed according to steps S1-S6, and the washing machine load data is subjected to S-G filtering processing, wherein the filtering results of the operating states of the washing machine in different modes and different time periods are shown in fig. 3-6. The filtered data can greatly eliminate noise signals and ultrahigh frequency signals generated by frequency conversion power electronic devices and the like on the basis of ensuring that the waveform characteristics are unchanged. Approximate components of the waveform are further extracted as main features by using a discrete wavelet transform, and the processed waveform is shown in fig. 7-10.
To accurately measure the similarity of steady state waveforms before and after a load event, a pearson correlation coefficient is selected to describe the correlation. The comparison shows that when the multi-state electric appliance mode conversion load event occurs in the same electric appliance (namely the multi-state electric appliance working mode conversion load event), the correlation coefficient is obviously improved after processing, and the Pearson correlation coefficient is about 0.85.
The embodiment also provides a complex multi-state electrical appliance load event judging device, and the device stores computer program instructions for executing the complex multi-state electrical appliance load event judging method.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein. The scheme in the embodiment of the application can be implemented by adopting various computer languages, such as object-oriented programming language Java and transliterated scripting language JavaScript.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While the preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all alterations and modifications as fall within the scope of the application.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations as well.

Claims (8)

1. A method for judging a load event of a complex multi-state electric appliance is characterized by comprising the following steps:
step S1: acquiring instantaneous current of an original circuit, and calculating an effective value of the instantaneous current;
step S2: calculating the current waveform change rate according to the instantaneous current effective value, and appointing a time node of the occurrence of the load event when the current waveform change rate is greater than a threshold value m;
and step S3: calculating the derivative of the effective value of the instantaneous current before and after the occurrence of the load event, when the derivative value is less than the threshold valueaWhen the current waveform enters a steady state, extracting the steady state current waveform of one period or a plurality of periods from the time node of the occurrence of the load event;
and step S4: filtering high-frequency components and clutter signals in a steady-state current waveform before and after a load event occurs by using a Savitzky-Golay filtering algorithm;
step S5: extracting approximate components of steady-state current waveforms before and after the load event is filtered by using discrete wavelet transform to obtain current instantaneous values after Savitzky-Golay filtering and discrete wavelet transform before and after the load event is generated;
step S6: and calculating the correlation of the current instantaneous values before and after the time node of the load event occurrence through Savitzky-Golay filtering and discrete wavelet transformation by using a Pearson correlation coefficient, and judging whether the load event is an electric appliance switching load event or a multi-state electric appliance working mode switching load event according to the Pearson correlation coefficient.
2. The method for distinguishing the load events of the complex multi-state electric appliances according to the claim 1, wherein the condition that the current waveform enters the steady state in the step S3 is determined as follows:
Figure 220898DEST_PATH_IMAGE001
Figure 184044DEST_PATH_IMAGE002
time node for occurrence of load event
Figure 760518DEST_PATH_IMAGE003
Rear end
Figure 968777DEST_PATH_IMAGE004
The derivative of the current waveform at the time of day,
Figure 845466DEST_PATH_IMAGE005
time node for occurrence of load event
Figure 979513DEST_PATH_IMAGE006
Front side
Figure 308863DEST_PATH_IMAGE007
The derivative of the current waveform at time.
3. The method for distinguishing the load events of the complex multi-state electric appliances according to the claim 2, wherein in the step S3, the mode of extracting the steady-state current waveform is as follows:
Figure 320813DEST_PATH_IMAGE008
in the formula (I), the compound is shown in the specification,
Figure 52008DEST_PATH_IMAGE009
is the steady state current waveform prior to the occurrence of a load event,
Figure 356957DEST_PATH_IMAGE010
for a steady state current waveform after a load event occurs,
Figure 439182DEST_PATH_IMAGE011
at the last point in time of the steady state current waveform before the load event occurs,
Figure 989243DEST_PATH_IMAGE012
and T is a power frequency period, and is the time point when the steady-state current waveform starts after the load event occurs.
4. The method for distinguishing the load events of the complex multi-state electric appliances according to the claim 3, wherein the process of the step S4 is as follows: firstly, the width of the sliding window is selected to be 2k+1, sliding window converts the steady-state current waveform before the occurrence of a load event
Figure 574945DEST_PATH_IMAGE013
Is divided into 2k+1 data, wherein the time t takes the value
Figure 310515DEST_PATH_IMAGE014
，kIs a positive integer based on 2 in the sliding windowk+1 data, structurenAn order polynomial to calculate the steady state current waveform before the occurrence of the filtered load event
Figure 880036DEST_PATH_IMAGE015
：
Figure 873269DEST_PATH_IMAGE016
In the formula (I), the compound is shown in the specification,
Figure 329789DEST_PATH_IMAGE017
respectively 0,1,2, \8230,n-1，norder coefficients, all coefficients together forming a coefficient matrix
Figure 727273DEST_PATH_IMAGE018
；
And (3) calculating a fitting residual error by using a least square method to construct an objective function, wherein the equation of the residual error is as follows:
Figure 767779DEST_PATH_IMAGE020
Figure 908910DEST_PATH_IMAGE021
is the coefficient of the g order; equation for residual errorEDerivation, such that its partial derivative to parameter is 0:
Figure 78991DEST_PATH_IMAGE022
solving a current coefficient matrix using known data in a sliding window
Figure 162223DEST_PATH_IMAGE023
When the sliding window is moved, the central point of the sliding window is taken as the steady-state current waveform before the occurrence of the load event
Figure 175178DEST_PATH_IMAGE025
As a result of the smoothing, when the sliding window passesAll the data to be smoothed are filtered, and all the return values are the complete steady-state current waveform before the load event after filtering occurs
Figure 870733DEST_PATH_IMAGE027
；
Steady state current waveform after filtered load event occurs
Figure 285534DEST_PATH_IMAGE028
And the filtered steady state current waveform before the load event occurs
Figure 279946DEST_PATH_IMAGE029
The same calculation process is performed.
5. The method for distinguishing the load events of the complex multi-state electric appliances as claimed in claim 4, wherein the step S5 is to obtain the current instantaneous value after Savitzky-Golay filtering and discrete wavelet transform before the load events occur by solving the following equation
Figure 780198DEST_PATH_IMAGE030
After a load event occurs, current instantaneous value is subjected to Savitzky-Golay filtering and discrete wavelet transformation
Figure 279444DEST_PATH_IMAGE031
：
Figure 814330DEST_PATH_IMAGE032
Figure 239364DEST_PATH_IMAGE033
Figure 226912DEST_PATH_IMAGE034
Figure 264269DEST_PATH_IMAGE035
In the formula:WT(a,t) In order to be the function of the inner product,
Figure 388083DEST_PATH_IMAGE036
is a function of the wavelet transform,ais a scale factor, x is an amount of time,DWT(q,p) The function is a function obtained by discretizing the scale parameters according to power series and uniformly discretizing the time,
Figure 984018DEST_PATH_IMAGE037
is a discrete wavelet transform function, t is time,q、prespectively, a scale factor and a shift factor of the discrete wavelet.
6. The method for distinguishing the load events of the complex multi-state electric appliances according to the claim 5, wherein in the step S6, the Pearson correlation coefficient is calculated according to the following formula:
Figure 458862DEST_PATH_IMAGE038
in the formula (I), the compound is shown in the specification,
Figure 299910DEST_PATH_IMAGE039
represents the average value of the current after Savitzky-Golay filtering and discrete wavelet transformation before the load event happens in a discrete time period from 1 to n,
Figure 543809DEST_PATH_IMAGE040
represents the average value of the current after Savitzky-Golay filtering and discrete wavelet transformation after a load event happens in a discrete time period from 1 to n,
Figure 936745DEST_PATH_IMAGE041
is shown asiA time periodt i The instantaneous value of the current after Savitzky-Golay filtering and discrete wavelet transform before the load event occurs,
Figure 7206DEST_PATH_IMAGE042
is shown asiA period of timet i And (3) current instantaneous values after Savitzky-Golay filtering and discrete wavelet transformation after the load event occurs.
7. The method for distinguishing the load events of the complex multi-state electric appliances according to claim 6, wherein the load events of the switching of the electric appliances are determined when the Pearson correlation coefficient is less than 0.6, and the load events of the switching of the working modes of the multi-state electric appliances are determined when the Pearson correlation coefficient is more than 0.6.
8. A device for discriminating a complex polymorphic electric appliance load event, characterized in that the device stores computer program instructions for executing the method for discriminating a complex polymorphic electric appliance load event according to any one of claims 1 to 7.
CN202211347743.7A 2022-10-31 2022-10-31 Complex multi-state electrical appliance load event distinguishing method and device Pending CN115407157A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211347743.7A CN115407157A (en) 2022-10-31 2022-10-31 Complex multi-state electrical appliance load event distinguishing method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211347743.7A CN115407157A (en) 2022-10-31 2022-10-31 Complex multi-state electrical appliance load event distinguishing method and device

Publications (1)

Publication Number Publication Date
CN115407157A true CN115407157A (en) 2022-11-29

Family

ID=84168644

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211347743.7A Pending CN115407157A (en) 2022-10-31 2022-10-31 Complex multi-state electrical appliance load event distinguishing method and device

Country Status (1)

Country Link
CN (1) CN115407157A (en)

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103116098A (en) * 2013-01-25 2013-05-22 重庆大学 Household appliance operating state identification method based on cross correlation coefficient
CN106786534A (en) * 2016-12-28 2017-05-31 天津求实智源科技有限公司 A kind of non-intrusive electrical load transient process discrimination method and system
CN109165604A (en) * 2018-08-28 2019-01-08 四川大学 The recognition methods of non-intrusion type load and its test macro based on coorinated training
CN110018369A (en) * 2019-03-05 2019-07-16 天津工业大学 A kind of household electrical appliances intelligent recognition and monitoring method based on non-intrusion type load decomposition
CN110569877A (en) * 2019-08-07 2019-12-13 武汉中原电子信息有限公司 Non-invasive load identification method and device and computing equipment
CN110956220A (en) * 2019-12-11 2020-04-03 深圳市活力天汇科技股份有限公司 Non-invasive household appliance load identification method
CN111722028A (en) * 2019-03-19 2020-09-29 华北电力大学 Load identification method based on high-frequency data
CN113010985A (en) * 2021-03-05 2021-06-22 重庆邮电大学 Non-invasive load identification method based on parallel AANN
CN114707542A (en) * 2022-03-17 2022-07-05 上海梦象智能科技有限公司 Non-invasive electrical data identification method based on instantaneous load characteristics

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103116098A (en) * 2013-01-25 2013-05-22 重庆大学 Household appliance operating state identification method based on cross correlation coefficient
CN106786534A (en) * 2016-12-28 2017-05-31 天津求实智源科技有限公司 A kind of non-intrusive electrical load transient process discrimination method and system
CN109165604A (en) * 2018-08-28 2019-01-08 四川大学 The recognition methods of non-intrusion type load and its test macro based on coorinated training
CN110018369A (en) * 2019-03-05 2019-07-16 天津工业大学 A kind of household electrical appliances intelligent recognition and monitoring method based on non-intrusion type load decomposition
CN111722028A (en) * 2019-03-19 2020-09-29 华北电力大学 Load identification method based on high-frequency data
CN110569877A (en) * 2019-08-07 2019-12-13 武汉中原电子信息有限公司 Non-invasive load identification method and device and computing equipment
CN110956220A (en) * 2019-12-11 2020-04-03 深圳市活力天汇科技股份有限公司 Non-invasive household appliance load identification method
CN113010985A (en) * 2021-03-05 2021-06-22 重庆邮电大学 Non-invasive load identification method based on parallel AANN
CN114707542A (en) * 2022-03-17 2022-07-05 上海梦象智能科技有限公司 Non-invasive electrical data identification method based on instantaneous load characteristics

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
陶鹏等: "电力负荷运行状态在线检测系统研究与应用", 《中国电力教育》 *

Similar Documents

Publication Publication Date Title
CN110441641B (en) Low-current grounding line selection method and system based on zero-sequence direct-current component
EP3079226A1 (en) Device's state estimation device, device's power consumption estimation device, and program
CN111738128A (en) Series fault arc detection method based on morphological filtering and MMG
CN109447473B (en) Power load monitoring method, device, equipment and readable storage medium
Guo et al. 19 An Enhanced EWMA Controller for Processes Subject to Random Disturbances
CN111426905A (en) Power distribution network common bus transformation relation abnormity diagnosis method, device and system
CN112816789A (en) Conductor internal resistance abnormity identification method, device, equipment and computer storage medium
CN110456159B (en) System side harmonic impedance estimation method and system based on corrected independent random vector
CN109901003B (en) Inverter power fault detection method and system
CN111161097A (en) Method and device for detecting switch event based on event detection algorithm of hypothesis test
CN115407157A (en) Complex multi-state electrical appliance load event distinguishing method and device
CN111126780A (en) Non-invasive load monitoring method and storage medium
CN112742280B (en) Chaotic state detection method and system of hybrid system
JP5570929B2 (en) Power converter and power supply system
Renaux et al. Non-intrusive load monitoring: an architecture and its evaluation for power electronics loads
Schirmer et al. Improving energy disaggregation performance using appliance-driven sampling rates
CN113837895A (en) Power distribution network abnormal event identification method and system based on power disturbance data
CN116449081B (en) Data acquisition system, device and storage medium with self-adaptive regulation and control function
Garnier et al. The CONTSID toolbox for Matlab: extensions and latest developments
CN113704698B (en) Event detection method and system for non-intrusive load identification
CN115269241A (en) Method, device and storage medium for carrying out anomaly detection on periodic data
CN112688325B (en) Wind power plant subsynchronous oscillation monitoring method based on two-stage improved ITD algorithm
CN109684937A (en) A kind of signal antinoise method and device based on FFT and Mathematical Morphology method
Qing-wei et al. A practical approach of online control performance monitoring
CN108037350A (en) A kind of parameter identification method of voltage waveform, system

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
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20221129