CN108038497A - The working status decision model method for building up and equipment of a kind of alternating-current installation/AC installation - Google Patents
The working status decision model method for building up and equipment of a kind of alternating-current installation/AC installation Download PDFInfo
- Publication number
- CN108038497A CN108038497A CN201711267337.9A CN201711267337A CN108038497A CN 108038497 A CN108038497 A CN 108038497A CN 201711267337 A CN201711267337 A CN 201711267337A CN 108038497 A CN108038497 A CN 108038497A
- Authority
- CN
- China
- Prior art keywords
- installation
- alternating
- cluster
- working status
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/211—Selection of the most significant subset of features
- G06F18/2113—Selection of the most significant subset of features by ranking or filtering the set of features, e.g. using a measure of variance or of feature cross-correlation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Biology (AREA)
- Economics (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Public Health (AREA)
- Probability & Statistics with Applications (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses the working status decision method and equipment of a kind of alternating-current installation/AC installation, for establishing the condition adjudgement model of general alternating-current installation/AC installation and the state of alternating-current installation/AC installation being judged according to the condition adjudgement model.This method includes:Obtain the data sample set of alternating-current installation/AC installation;Each parameters of electric power obtains disturbance degree and is more than the parameters of electric power of default disturbance degree threshold value as the characteristic parameter in data sample set to the disturbance degree of the working status of alternating-current installation/AC installation in determining data sample set;The data of characteristic parameter described in each data sample in the data sample set are extracted, obtain the feature samples set of the alternating-current installation/AC installation;L cluster is divided into according to feature samples set, and obtains the barycenter of each cluster;The working status of the barycenter of L cluster and alternating-current installation/AC installation is associated, to obtain the working status decision model of alternating-current installation/AC installation.
Description
Technical field
The present invention relates to machine learning field, the working status decision model foundation side of more particularly to a kind of alternating-current installation/AC installation
Method and equipment.
Background technology
At present, the method judged the state of alternating-current installation/AC installation, typically to the alternating-current installation/AC installation of different power consumption into
Row monitoring, is modeled according to the data of monitoring collection, and finally can achieve the effect that is different conditions for alternating-current installation/AC installation
Corresponding threshold interval is set, and then according to the parameter of collection, such as can be which threshold interval power falls at, and then judge
Go out the state of alternating-current installation/AC installation.
But it is generally only to be monitored and model by single parameter at present, then if being due to that other specification goes out
When now fluctuating, it is more likely that the result inaccuracy of condition adjudgement is allowed for, and due to there was only the phase homotype of same type at present
Number the model of alternating-current installation/AC installation can be general, and the alternating-current installation/AC installation of different model cannot due to the difference of component, model
Share, therefore the reusability of current model is too low, applicability is not high.
The content of the invention
The embodiment of the present invention provides a kind of the working status decision model method for building up and equipment of alternating-current installation/AC installation, for building
The condition adjudgement model of vertical general alternating-current installation/AC installation.
First aspect, there is provided a kind of working status decision model method for building up of alternating-current installation/AC installation, this method include:
Obtain the data sample set of alternating-current installation/AC installation;Wherein, the data sample set includes the exchange of same type
Data sample of the electric equipment under different working condition, a data sample include the data of P parameters of electric power, and P is positive integer;
Determine the disturbance degree of each parameters of electric power in the data sample set to the working status of the alternating-current installation/AC installation,
Obtain disturbance degree and be more than the parameters of electric power of default disturbance degree threshold value as the characteristic parameter in the data sample set;
The data of characteristic parameter described in each data sample in the data sample set are extracted, obtain the alternating current
The feature samples set of equipment;
L cluster is divided into according to the feature samples set, and obtains the barycenter of each cluster;Wherein, a cluster includes institute
The partial data sample in data sample set is stated, the data sample that different clusters include is different, and the center that the barycenter is cluster,
The alternating-current installation/AC installation corresponds to a variety of working statuses, and the quantity L of cluster is identical with the quantity of the working status, and L is positive integer;
The working status of the barycenter and the alternating-current installation/AC installation of the L cluster is associated, to obtain the alternating current
The working status decision model of equipment;Wherein, a barycenter corresponds to a kind of working status.
Optionally, each influence of the parameters of electric power to the state of the alternating-current installation/AC installation in the data sample set is determined
Degree, obtains disturbance degree and is more than the parameters of electric power of default disturbance degree threshold value as the characteristic parameter in the data sample set, bag
Include:
Vertical data sample matrix is built jointly according to the set of data samples of the alternating-current installation/AC installation;Wherein, the data sample square
The data of different parameters of electric power in the corresponding data sample of row in battle array, the row correspondence in the data sample matrix are different
The data of same parameters of electric power in data sample;
Establish the covariance matrix of the data sample matrix;
It is total to calculate characteristic value shared by each characteristic value in the characteristic value of the covariance matrix, and the calculating characteristic value
The ratio of sum, and characteristic value corresponding parameters of electric power of the ratio more than or equal to preset ratio threshold value is determined as the spy
Levy parameter;Wherein, the corresponding characteristic value of a parameters of electric power is used to characterize one parameters of electric power to the alternating-current installation/AC installation
State disturbance degree.
Optionally, it is described that the feature samples set of the alternating-current installation/AC installation is divided into L cluster, including:
The feature samples set is repeatedly divided using K-means algorithms, until obtaining setting with the alternating current
The identical cluster of standby working status quantity;Wherein, for dividing each time, the mistake of each cluster obtained after current division is calculated
Poor quadratic sum SSE, and select the cluster of the value maximum of SSE in current cluster to be used as and be divided object;One cluster includes the feature
Partial Feature sample in sample set, the SSE of a cluster are used for the extent of polymerization for characterizing a cluster.
Optionally, one of which parameters of electric power is active power in the P parameters of electric power, then described by the L cluster
The working status of barycenter and the alternating-current installation/AC installation be associated, including:
According to the active power of the barycenter of the L cluster by the work of the barycenter of the L cluster and the alternating-current installation/AC installation
State is associated.
Optionally, after the working status decision model of the alternating-current installation/AC installation is obtained, the method further includes:
Obtain the working status decision model;
Obtain the data of the characteristic parameter of alternating-current installation/AC installation to be determined;The characteristic parameter is to the alternating-current installation/AC installation
The disturbance degree of state exceed the parameters of electric power of default disturbance degree threshold value;
The data of the characteristic parameter are inputted the working status decision model to be judged as a result, described judge that knot is used
In the working status for characterizing the alternating-current installation/AC installation.
Optionally, the data by the characteristic parameter input the working status decision model judged as a result,
It is described to judge that result is used for the working status for characterizing the alternating-current installation/AC installation, including:
Calculate the data and the Euclidean distance of all barycenter in the working status decision model of the characteristic parameter;Its
In, a barycenter corresponds to a kind of working status;
The working status of the centroid linkage of Euclidean distance minimum is determined as the alternating-current installation/AC installation to be determined
Working status.
Second aspect, there is provided a kind of working status decision model of alternating-current installation/AC installation establishes equipment, which includes:
First acquisition unit, for obtaining the data sample set of alternating-current installation/AC installation;Wherein, the data sample set bag
Data sample of the alternating-current installation/AC installation of same type under different working condition is included, a data sample includes P parameters of electric power
Data, P is positive integer;
Determination unit, for determining each work of the parameters of electric power to the alternating-current installation/AC installation in the data sample set
The disturbance degree of state, obtains disturbance degree and is more than the parameters of electric power of default disturbance degree threshold value as the spy in the data sample set
Levy parameter;
Extraction unit, for extracting the data of characteristic parameter described in each data sample in the data sample set,
Obtain the feature samples set of the alternating-current installation/AC installation;
Division unit, for being divided into L cluster according to the feature samples set, and obtains the barycenter of each cluster;Wherein,
One cluster includes the partial data sample in the data sample set, and the data sample that different clusters include is different, and the matter
The heart is the center of cluster, and the alternating-current installation/AC installation corresponds to a variety of working statuses, the quantity L of cluster and the quantity phase of the working status
Together, L is positive integer;
Associative cell, for the working status of the barycenter and the alternating-current installation/AC installation of the L cluster to be associated, with
To the working status decision model of the alternating-current installation/AC installation;Wherein, a barycenter corresponds to a kind of working status.
Optionally,
The determination unit, is additionally operable to build vertical data sample matrix jointly according to the set of data samples of the alternating-current installation/AC installation;
Wherein, the data of different parameters of electric power, the data sample in the corresponding data sample of row in the data sample matrix
The data of same parameters of electric power in the corresponding different data sample of row in matrix;Establish the covariance of the data sample matrix
Matrix;The characteristic value of the covariance matrix is calculated, and calculates characteristic value summation shared by each characteristic value in the characteristic value
Ratio, and by the ratio be greater than or equal to preset ratio threshold value the corresponding parameters of electric power of characteristic value be determined as the feature
Parameter;Wherein, the corresponding characteristic value of a parameters of electric power is used to characterize one parameters of electric power to the alternating-current installation/AC installation
The disturbance degree of state.
Optionally, the division unit, is additionally operable to using K-means algorithms repeatedly be drawn the feature samples set
Point, until obtaining the cluster identical with the working status quantity of the alternating-current installation/AC installation;Wherein, for dividing each time, calculate and work as
The error sum of squares SSE of each cluster obtained after preceding division, and select the cluster of the value maximum of SSE in current cluster to be used as and drawn
Divide object;One cluster includes the Partial Feature sample in the feature samples set, and the SSE of a cluster is used for one cluster of characterization
Extent of polymerization.
Optionally, the associative cell is closed the working status of the barycenter and the alternating-current installation/AC installation of the L cluster
Connection, including:
The associative cell exchanges the barycenter of the L cluster with described according to the active power of the barycenter of the L cluster
The working status of electric equipment is associated.
Optionally, the equipment further includes second acquisition unit and judging unit;
The second acquisition unit, for obtaining working status decision model, the working status decision model is used to sentence
Determine the working status of alternating-current installation/AC installation;And obtain the data of the characteristic parameter of alternating-current installation/AC installation to be determined;The feature ginseng
Number is the parameters of electric power for exceeding default disturbance degree threshold value to the disturbance degree of the state of the alternating-current installation/AC installation;
The judging unit, is judged for the data of the characteristic parameter to be inputted the working status decision model
As a result, described judge that knot is used for the working status for characterizing the alternating-current installation/AC installation.
Optionally, the judging unit, the data for being additionally operable to calculate the characteristic parameter judge mould with the working status
The Euclidean distance of all barycenter in type;Wherein, a barycenter corresponds to a kind of working status;By the matter of Euclidean distance minimum
The associated working status of the heart is determined as the working status of the alternating-current installation/AC installation to be determined.
The third aspect, there is provided a kind of computer installation, described device include at least one processor, and the processor is used for
Realize that the working status of the alternating-current installation/AC installation provided such as first aspect judges mould when performing the computer program stored in memory
The step of type method for building up.
Fourth aspect, there is provided a kind of computer-readable recording medium, is stored thereon with computer program, the computer journey
The step of the working status decision model method for building up of the alternating-current installation/AC installation provided such as first aspect is provided when sequence is executed by processor
Suddenly.
In embodiments of the present invention, can be divided according to the data of multiple parameters of electric power of multiple alternating-current installation/AC installations of acquisition
Multiple clusters, and the barycenter of each cluster is obtained, and be further associated the working status of barycenter and alternating-current installation/AC installation, and then
Obtain working status decision model.Since the working status decision model by mass data is handled to obtain, and
The data of multiple parameters of electric power of multiple equipment are combined, the working status of alternating-current installation/AC installation is judged from multiple dimensions,
So that judge that result can be more accurate.
Brief description of the drawings
In order to illustrate the technical solution of the embodiments of the present invention more clearly, it will make below to required in the embodiment of the present invention
Attached drawing is briefly described, it should be apparent that, attached drawing described below is only some embodiments of the present invention, for
For those of ordinary skill in the art, without creative efforts, other can also be obtained according to these attached drawings
Attached drawing.
Fig. 1 is the flow diagram of working status decision model method for building up provided in an embodiment of the present invention;
Fig. 2 is the flow diagram provided in an embodiment of the present invention that characteristic parameter is determined by PCA algorithms;
Fig. 3 is a kind of division result schematic diagram of cluster provided in an embodiment of the present invention;
Fig. 4 is the flow diagram of working status decision method provided in an embodiment of the present invention;
Fig. 5 is a kind of structure diagram that working status decision model provided in an embodiment of the present invention establishes equipment;
Fig. 6 is a kind of structure diagram of computer installation provided in an embodiment of the present invention.
Embodiment
To make the purpose, technical scheme and advantage of the embodiment of the present invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, the technical solution in the embodiment of the present invention is clearly and completely described.
The technical background of the embodiment of the present invention is described below.
At present, the method judged the working status of alternating-current installation/AC installation, is generally only to be carried out by single parameter
Detection and modeling, then if be due to that other specification fluctuates, it is more likely that allow for the result of working status judgement not
Accurately, and due to only having the model of the alternating-current installation/AC installation of the same model of same type can be general at present, and different model
Alternating-current installation/AC installation cannot be shared due to the difference of component, model, therefore the reusability of current model is too low, and applicability is not
It is high.
In consideration of it, the embodiment of the present invention provides a kind of working status decision model method for building up, in the method, Neng Gougen
Multiple clusters are divided according to the data of multiple parameters of electric power of multiple alternating-current installation/AC installations of acquisition, and obtain the barycenter of each cluster, by matter
The working status of the heart and alternating-current installation/AC installation is associated, and then obtains working status decision model.Since the working status judges
Model is handled to obtain by mass data, and combines the data of multiple parameters of electric power of multiple equipment, from more
A dimension judges the working status of alternating-current installation/AC installation, so that judging that result can be more accurate.
Technical solution provided in an embodiment of the present invention is introduced below in conjunction with the accompanying drawings.
Fig. 1 is referred to, one embodiment of the invention provides a kind of working status decision model method for building up of alternating-current installation/AC installation.
This method includes:
Step 101:Obtain the data sample set of alternating-current installation/AC installation;Wherein, data sample set includes same type
Data sample of the alternating-current installation/AC installation under different working condition, a data sample include the data of P parameters of electric power, and P is just
Integer;
Step 102:Each parameters of electric power is to the disturbance degree of the working status of alternating-current installation/AC installation in determining data sample set,
Obtain disturbance degree and be more than the parameters of electric power of default disturbance degree threshold value as the characteristic parameter in data sample set;
Step 103:The data of characteristic parameter in each data sample in data sample set are extracted, obtain alternating-current installation/AC installation
Feature samples set;
Step 104:L cluster is divided into according to data sample set, and obtains the barycenter of each cluster;Wherein, a cluster bag
The partial data sample in data sample set is included, the data sample that different clusters include is different, and the center that barycenter is cluster, exchange
Electric equipment corresponds to a variety of working statuses, and the quantity L of cluster is identical with the quantity of working status, and L is positive integer;
Step 105:The working status of the barycenter of L cluster and alternating-current installation/AC installation is associated, to obtain alternating-current installation/AC installation
Working status decision model;Wherein, a barycenter corresponds to a kind of working status.
In the embodiment of the present invention, the foundation of working status decision model needs to be based on substantial amounts of data, therefore firstly the need of
Obtain the data sample set for being used for establishing working status decision model.Specifically, due to different types of alternating-current installation/AC installation
Working status differs greatly, if will be used to model while different type, judges that result may be inaccurate, therefore usually building
During mould, the set of data samples of the alternating-current installation/AC installation of generally use same type builds the working status decision model of vertical the type jointly,
I.e. a type of alternating-current installation/AC installation corresponds to a kind of model.It is corresponding, when carrying out data acquisition, then can obtain mutually similar
The data sample set of the alternating-current installation/AC installation of type, the alternating-current installation/AC installation of multiple same types can be included in the data sample set
Data sample, and each alternating-current installation/AC installation can also correspond to multiple data samples.Wherein, each data sample can wrap
The data of P parameters of electric power are included, P is positive integer.It is more accurate that the judgement of working status is carried out for the model finally established, number
The quantity of the data sample included according to sample set is The more the better, such as can be 10 power exponent of the quantity of parameters of electric power
Again, i.e., the power exponent times that the order of magnitude of the data of each parameters of electric power is 10.
The collection of data sample can be used for adopting for data acquisition by connecting one in the power supply serial ports of alternating-current installation/AC installation
Collect equipment.The collecting device can gather the data of the instantaneous parameters of electric power of alternating-current installation/AC installation, and can also be incited somebody to action by network
The data of the parameters of electric power of collection upload onto the server or database is preserved.Wherein, parameters of electric power can be wattful power
Rate, reactive power, power factor, phase angle, voltage effective value, current effective value, electric voltage frequency, active energy, reactive energy,
Reverse active power mark etc., it is of course also possible to which including other possible parameters of electric power, the embodiment of the present invention does not limit this
System.
Collecting device by gather obtain data sample upload onto the server or database after, then can be from service
Device or database obtain data sample set.
In the embodiment of the present invention, since the data sample set of acquisition includes the data of multiple parameters of electric power, and wherein certain
A little parameters of electric power are to the disturbance degree very little of the working status of alternating-current installation/AC installation, even without disturbance degree, on the contrary in follow-up modeling
The complexity of calculating can be increased during calculating, therefore in order to reduce the complexity subsequently calculated, and also to extract to exchange
The working status of electric equipment has the parameters of electric power of considerable influence, after data sample is obtained, can also pass through principal component point
(each parameters of electric power is to alternating-current installation/AC installation in Principal Component Analysis, PCA algorithm determining data samples for analysis
Working status disturbance degree, obtain disturbance degree and be more than the parameters of electric power of default disturbance degree threshold value as the feature in data sample
Parameter.Wherein, the disturbance degree of a parameters of electric power refers to when the parameters of electric power fluctuates, can be to the work of alternating-current installation/AC installation
The influence degree that state produces.It is, of course, also possible to determine characteristic parameter by other dimension-reduction algorithms, such as can be by linear
Discrimination Analysis Algorithm (Linear Discriminant Analysis, LDA) or Local Liner Prediction (Locally
Linear embedding, LLE) determine characteristic parameter, or other possible algorithms carry out determining for characteristic parameters, this hair
Bright embodiment is without limitation.Wherein, presetting disturbance degree threshold value can rule of thumb be configured, alternatively, being surveyed according to experiment
Test result is configured, it is, of course, also possible to be configured according to other modes, the embodiment of the present invention is without limitation.Under
Face is described by taking PCA algorithms as an example to obtaining characteristic parameter.
Fig. 2 is referred to, determines that each parameters of electric power sets the alternating current in the data sample set by PCA algorithms
The disturbance degree of standby state, obtains disturbance degree and is more than the parameters of electric power of default disturbance degree threshold value as in the data sample set
Characteristic parameter detailed process it is as follows:
Step 201:Vertical data sample matrix is built jointly according to the set of data samples of alternating-current installation/AC installation;Wherein, data sample square
The data of different parameters of electric power in the corresponding data sample of row in battle array, the corresponding different data of row in data sample matrix
The data of same parameters of electric power in sample;
Step 202:Establish the covariance matrix of data sample matrix;
Step 203:It is total to calculate characteristic value shared by each characteristic value in the characteristic value of covariance matrix, and calculating characteristic value
The ratio of sum, and characteristic value corresponding parameters of electric power of the ratio more than or equal to preset ratio threshold value is determined as characteristic parameter;
Wherein, the corresponding characteristic value of a parameters of electric power is used to characterize disturbance degree of the parameters of electric power to the state of alternating-current installation/AC installation.
Specifically, the quantity for assuming the data sample that the data sample set obtained includes is M, each data sample includes
Each data sample, then can be converted into a vector in N-dimensional space by the data of N number of parameters of electric powerThe data sample bag
The data of each parameters of electric power included are the component in each dimension space in the vector, are x1, x2, x3...xn, then it is final
The data sample matrix of a M*N can be obtained by the data sample of acquisition, which can be expressed as:
It can be seen that the row in data sample matrix corresponds to the data of different parameters of electric power in a data sample, data
The data of same parameters of electric power in the corresponding different data sample of row in sample matrix.For example, x11, x12, x13...x1nRepresent
Each data of same data sample, x11, x21, x31...xm1Then represent the data of same parameters of electric power.
After data sample matrix is obtained, then the average of each parameters of electric power, that is, data sample square can be calculated
The average value of each row in battle array, the mean value computation formula of each of which row are:
Wherein, n be row sequence number, i.e. 0 < n≤N;M be row sequence number, i.e. 0 < m≤M.
After the average value of each row is tried to achieve, then covariance matrix can be established according to above-mentioned data sample matrix, assisted
Variance can be with the dispersion degree of characterize data, so as to know the data variance of each parameters of electric power, and by all electric power
Parameter is compared two-by-two, and the direction of the parameters of electric power of data variance maximum is first direction, the electricity with otherness maximum
The big direction of otherness time is second direction on the orthogonal direction in the direction of force parameter.Specifically, in covariance matrix, it is each
The association side of the data of the data of wherein one row in behavioral data sample matrix and wherein one row itself and the data of other row
Difference, covariance calculation formula are:
Wherein, cov (X, Y) represents the covariance of the X and Y row in data sample matrix;XiAnd YiIt is data sample matrix
In i-th of data in any one row, and XiAnd YiCan be the data of same column.
For clearer description covariance matrix, carried out below by taking the M data sample including 3 parameters of electric power as an example
Description.When only including 3 parameters of electric power, data sample matrix can be expressed as:
Wherein, X row, Y row, Z row correspond to the data row of three different parameters of electric power respectively.
Then its corresponding covariance matrix C, which can be obtained, according to the data sample matrix is:
Wherein, cov (X, X) is the covariance of X row and X row, and cov (X, Y) is the covariance of X row and Y row, and cov (X, Z) is
X arranges the covariance with Z row, other and so on.
After covariance matrix is obtained, then the characteristic value of the covariance matrix can be calculated.Specifically, for a square
Battle array A, there are equation AV=λ V, the i.e. product of the feature vector V of matrix A and matrix A, equal to matrix A eigenvalue λ and feature to
The product of V is measured, V can be orthogonal matrix.So, matrix A can be the covariance matrix of the embodiment of the present invention, and V is then the association
The feature vector of variance matrix, λ are the characteristic value of the covariance matrix, therefore can then calculate covariance by the equation
The characteristic value of matrix, it is identical with the quantity of parameters of electric power to calculate the quantity of the characteristic value of gained, and characteristic value and an electricity
Force parameter is corresponding.Since this calculating process belongs to the category of the prior art, details are not described herein.
After the characteristic value for obtaining covariance matrix is calculated, then characteristic value summation shared by each characteristic value can be calculated
Ratio, wherein, the corresponding characteristic value of a parameters of electric power is used to characterize working status of the parameters of electric power to alternating-current installation/AC installation
Disturbance degree.
Table 1 is referred to, is the reference example of characteristic value result of calculation provided in an embodiment of the present invention.Due to the data of collection
The quantity of sample is different, and the data that the data sample gathered every time includes also can be different, therefore the result in table 1 is only
For reference example, other result of calculations are not intended to limit.
Parameters of electric power title | Characteristic value | Account for total characteristic ratio |
Active power | 5.349519e+07 | 62.07% |
Reactive power | 2.179667e+07 | 25.29% |
Voltage effective value | 8.248376e+06 | 9.57% |
Current effective value | 1.315944e+06 | 1.52% |
Electric voltage frequency | 2.905635e+05 | 0.33% |
Power factor | 2.821686e+05 | 0.32% |
Phase angle | 2.310558e+05 | 0.26% |
Active energy | 2.056138e+05 | 0.23% |
Reactive energy | 1.609988e+05 | 0.18% |
Reverse active mark | 1.565486e+05 | 0.18% |
Table 1
Wherein, parameters of electric power title is classified as the title of the parameters of electric power of collection, and characteristic value is classified as in each parameters of electric power
The characteristic value of dimension, accounts for the ratio that total characteristic value ratio is total characteristic value shared by the corresponding characteristic value of each parameters of electric power.
After characteristic value summation shared by acquisition characteristic value is calculated, then ratio can be more than to the feature of preset ratio threshold value
It is worth corresponding parameters of electric power and is determined as characteristic parameter.For example, in the result of calculation shown in table 1, active power, reactive power,
Voltage effective value, the corresponding characteristic value of current effective value is larger, and the characteristic value of this four parameters of electric power and to account for characteristic value total
The 99% of sum, it is possible to which, by active power, reactive power, voltage effective value, current effective value is as characteristic parameter.
After characteristic parameter is determined, then the corresponding data of characteristic parameter in each data sample can be extracted, and removed
The data of other parameters of electric power outside characteristic parameter can then delete deletion, to obtain feature samples, in feature samples only
Including the corresponding data of characteristic parameter, to reduce the redundancy of data, and the complexity and calculation amount subsequently calculated is reduced.
It is identical in identical working status due to the difference of the model of alternating-current installation/AC installation in the embodiment of the present invention
The numerical value of the data of parameters of electric power can not possibly be identical, but numerical value can fluctuate within the specific limits, i.e., in identical work shape
Feature samples during state are similar, then then can be by clustering algorithm by sample clustering similar in feature samples to one
Rise, and then the working status of alternating-current installation/AC installation is judged according to cluster result.
In the embodiment of the present invention, feature samples set can repeatedly be divided using K- averages (means) algorithm, directly
To L cluster is obtained, L is positive integer.Wherein, the quantity L of cluster and the working status of the alternating-current installation/AC installation for the type to be judged
Quantity is identical.Wherein, for dividing each time, the error sum of squares SSE of each cluster obtained after current division is calculated, and is selected
The cluster conduct for selecting the value maximum of SSE in current cluster is divided object;One cluster includes the Partial Feature in feature samples set
Sample, the SSE of a cluster are used for the extent of polymerization for characterizing a cluster.
Specifically, after feature samples set is obtained, feature samples set can be drawn by K-means algorithms first
It is divided into two clusters.Wherein, K=2 in K-means algorithms, and for dividing each time, select error sum of squares in current cluster
The cluster of the value maximum of SSE is used as and is divided object.K-means algorithms are used to sample set being divided into K cluster, implement in the present invention
K is 2 in example, and finds the barycenter (Centroid) of each cluster, and will in the cluster completed by K-means algorithm partitions
Ask the barycenter of each cluster and the cluster as far as possible close.Certainly, the numerical value of K can also be other numerical value, such as 3 or 4 etc., the present invention
Embodiment example is without limitation.All illustrated below by taking K=2 as an example.
Wherein, first can be in the data field of the dimension where each characteristic parameter in the partition process of K-means algorithms
K data of interior random selection, are combined to obtain K barycenter.Specifically, using characteristic parameter as active power, reactive power,
Exemplified by voltage effective value, current effective value, then 2 data can be randomly choosed in this corresponding data of four parameters, then group
Two barycenter are synthesized, a barycenter includes one group of active power, reactive power, voltage effective value, current effective value and corresponds to respectively
Data.
After two barycenter are selected, then the Euclidean distance of each feature samples and the two barycenter is calculated.Wherein, Europe
Actual distance of family name's distance between two points in hyperspace.Wherein, Euclidean distance calculation formula is
Wherein, aiFor the data of the ith feature parameter of one of feature samples, the numbering of characteristic parameter is 1~n;bi
For the value of the ith feature parameter of one of barycenter.
After Euclidean distance of each feature samples to barycenter is calculated, compare this feature sample to the Euclidean of each barycenter
Distance, and this feature sample is divided into the cluster of the minimum corresponding barycenter of Euclidean distance, so that two clusters are obtained, and it is again true
The barycenter of the two fixed clusters.Specifically, the average of all data of dimension where each characteristic parameter in each cluster is calculated, and will
The average of each characteristic parameter is combined into the new barycenter of the cluster.
After the new barycenter of two clusters is obtained, pass through the European of each feature samples of new centroid calculation and the two barycenter
Distance, and the Euclidean distance based on calculating re-starts the division of cluster, and redefine new barycenter.After obtaining new barycenter,
Continue to repeat above-mentioned steps, untill the division result of the cluster of all feature samples no longer changes.
In the embodiment of the present invention, when cluster division result no longer changes, two clusters being divided, and calculate respectively this two
The error sum of squares (Sum of Squared Erroe, SSE) of a cluster, and retain the cluster of SSE minimums in the two clusters, and it is right
The cluster of SSE maximums continues with K-means algorithms and carries out two points in the two clusters.Wherein, the corresponding SSE of a cluster is used to characterize
The extent of polymerization of one cluster.
In to the two clusters the cluster of SSE maximums continue with K-means algorithms carry out two/after, also retain this time two
/ after two clusters in SSE minimums cluster.
Carry out second divide after, the total quantity of cluster is 3, select SSE maximums in this three clusters cluster continue into
Row division, untill the quantity of all clusters is identical with the quantity of the working status of alternating-current installation/AC installation.
Fig. 3 is referred to, is the division result schematic diagram of cluster provided in an embodiment of the present invention.Wherein, the division knot shown in Fig. 3
Fruit is carried out based on two characteristic parameters, i.e., is presented in two-dimensional space, and the quantity of the cluster after being divided is 3
It is a, i.e. cluster 1, cluster 2, cluster 3.Certainly, the division result shown in Fig. 3 is only a kind of division result, is not used to the explanation present invention
The unique division result of embodiment.
In the embodiment of the present invention, after the division of cluster is completed, the barycenter of each cluster is obtained, and barycenter and alternating current are set
Standby working status is associated, to obtain working status decision model.Wherein, a barycenter corresponds to a kind of working status.
When being associated, it can be associated according to the data of the actual working state of alternating-current installation/AC installation, such as can be according to barycenter
Barycenter and working status are associated by corresponding active power.For example, alternating-current installation/AC installation is in the standby state, power is mainly used
The element energy exchange such as capacitance and inductance in holding circuit, its active power is minimum in all working state, so taking
Holding state of the barycenter of active power minimum as equipment.
In the embodiment of the present invention, working status decision model can be stored in server or database equipment, with
Working status decision model is obtained easy to convenient when needing to carry out working status judgement.
Fig. 4 is referred to, after working status decision model establishes completion, then can be applied to the judgement of working status,
The flow that working status judges is as follows:
Step 401:Working status decision model is obtained, working status decision model is used for the work for judging alternating-current installation/AC installation
State;
Step 402:Obtain the data of the characteristic parameter of alternating-current installation/AC installation to be determined;Characteristic parameter is to alternating-current installation/AC installation
The disturbance degree of state exceed the parameters of electric power of default disturbance degree threshold value;
Step 403:The data input service state decision model of characteristic parameter is judged as a result, judging that knot is used for table
Levy the working status of alternating-current installation/AC installation.
In the embodiment of the present invention, working status decision model can be obtained from server or database, for subsequently handing over
The judgement of galvanic electricity equipment working state.Furthermore it is also possible to the friendship from server or the standby acquisition working status to be determined of database
The data of the characteristic parameter of galvanic electricity equipment.Specifically, the collection of data can be connected by the power supply serial ports in alternating-current installation/AC installation
One collecting device for being used for data acquisition.The collecting device can gather the data of the instantaneous parameters of electric power of alternating-current installation/AC installation,
And the data of the parameters of electric power of collection can also be uploaded onto the server by network or database is preserved.Collection is set
The data that the data of standby collection are either stored in server or database include the data of a variety of parameters of electric power, therefore, from
Collecting device is after either server or database obtain data, since the data of only characteristics of needs parameter are judged, because
This can also go out the corresponding data of characteristic parameter, the number of other parameters of electric power that will be in addition to characteristic parameter from extracting data
According to filtering out.Wherein, step 401 and step 402 do not have substantial sequencing, i.e., in embodiments of the present invention, step
401 and step 402 can perform at the same time or step 401 first carries out, or step 402 first carries out, and the present invention is implemented
Example is without limitation.
After the characteristic parameter of alternating-current installation/AC installation of working status decision model and working status to be determined is obtained, then may be used
To carry out the judgement of working status.Specifically, calculate the barycenter that the characteristic parameter obtained and working status decision model include
Euclidean distance, wherein it is possible to which the data of this feature parameter to be attributed to the cluster where the barycenter of Euclidean distance minimum, that is to say, that
The working status of the corresponding centroid linkage of Euclidean distance minimum is the work shape of the alternating-current installation/AC installation of working status to be determined
State.
In conclusion multiple clusters can be divided according to the data of multiple parameters of electric power of multiple alternating-current installation/AC installations of acquisition,
And the barycenter of each cluster is obtained, and be further associated the working status of barycenter and alternating-current installation/AC installation, and then obtain work
Make state decision model.Further, since the working status decision model by mass data is handled to obtain, and
And the data of multiple parameters of electric power of multiple equipment are combined, the working status of alternating-current installation/AC installation is sentenced from multiple dimensions
It is fixed, so that judging that result can be more accurate.Also, the embodiment of the present invention is also determined to the shape that works based on PCA algorithms
The higher characteristic parameter of state disturbance degree, avoids influence of the less dimension of other influences degree to judgement result.Meanwhile by
Dichotomy is utilized on the basis of K-means algorithms so that the cluster that cluster obtains every time is all global optimum, improves K-
The accuracy of means algorithms cluster.
Alternating current working status decision method provided in an embodiment of the present invention can be applied to all alternating-current installation/AC installations.Separately
Outside, the current working status of alternating-current installation/AC installation is judged and then according to sentencing by above-mentioned alternating current working status decision method
The disconnected time, then can count the history service condition of the alternating-current installation/AC installation, correspondingly, also can just be set for different alternating currents
Standby utilization rate is counted, so as to based on the obtained utilization rate of statistics with reference to the buying carried out for alternating-current installation/AC installation and
Allotment etc..
Fig. 5 is referred to, the same inventive concept based on the embodiment shown in Fig. 1, Fig. 2 and Fig. 4, one embodiment of the invention
A kind of working status decision model is provided and establishes equipment 50, which for example can be above-mentioned server or database, again
Either personal computer ((personal computer, PC), laptop etc..The equipment includes:
First acquisition unit 501, for obtaining the data sample set of alternating-current installation/AC installation;Wherein, data sample set bag
Data sample of the alternating-current installation/AC installation of same type under different working condition is included, a data sample includes P parameters of electric power
Data, P is positive integer;
Determination unit 502, for each parameters of electric power in determining data sample set to the working status of alternating-current installation/AC installation
Disturbance degree, obtain disturbance degree and be more than the parameters of electric power of default disturbance degree threshold value as the characteristic parameter in data sample set;
Extraction unit 503, for extracting the number of characteristic parameter described in each data sample in the data sample set
According to obtaining the feature samples set of the alternating-current installation/AC installation;
Division unit 504, for being divided into L cluster according to data sample set, and obtains the barycenter of each cluster;Wherein,
One cluster includes the partial data sample in data sample set, and the data sample that different clusters include is different, and barycenter is cluster
Center, alternating-current installation/AC installation correspond to a variety of working statuses, and the quantity L of cluster is identical with the quantity of working status, and L is positive integer;
Associative cell 505, for the working status of the barycenter of L cluster and alternating-current installation/AC installation to be associated, to be handed over
The working status decision model of galvanic electricity equipment;Wherein, a barycenter corresponds to a kind of working status.
Optionally,
Determination unit 502, specifically for building vertical data sample matrix jointly according to the set of data samples of alternating-current installation/AC installation;Its
In, the data of different parameters of electric power in the corresponding data sample of row in data sample matrix, the row in data sample matrix
The data of same parameters of electric power in corresponding different data sample;Establish the covariance matrix of data sample matrix;Calculate association side
The characteristic value of poor matrix, and calculate the ratio of characteristic value summation shared by each characteristic value in characteristic value, and ratio is more than or
The corresponding parameters of electric power of characteristic value equal to preset ratio threshold value is determined as characteristic parameter;Wherein, a parameters of electric power is corresponding
Characteristic value is used to characterize disturbance degree of the parameters of electric power to the state of alternating-current installation/AC installation.
Optionally, division unit 504, specifically for feature samples set is repeatedly divided using K-means algorithms,
Until obtaining the cluster identical with the working status quantity of alternating-current installation/AC installation;Wherein, for dividing each time, after calculating current division
The error sum of squares SSE of each obtained cluster, and select the cluster of the value maximum of SSE in current cluster to be used as and be divided object;
One cluster includes the Partial Feature sample in feature samples set, and the SSE of a cluster is used for the extent of polymerization for characterizing a cluster.
Optionally, the working status of the barycenter of L cluster and alternating-current installation/AC installation is associated by associative cell 505, including:
Associative cell 505 is according to the active power of the barycenter of L cluster by the barycenter of L cluster and the work shape of alternating-current installation/AC installation
State is associated.
Optionally, the equipment further includes second acquisition unit 406 and judging unit 407;
Second acquisition unit 506, for obtaining working status decision model, working status decision model is used to judge to exchange
The working status of electric equipment;And obtain the data of the characteristic parameter of alternating-current installation/AC installation to be determined;Characteristic parameter is to exchange
The disturbance degree of the state of electric equipment exceedes the parameters of electric power of default disturbance degree threshold value;
Judging unit 507, for the data input service state decision model of characteristic parameter to be judged as a result, judging
Knot is used for the working status for characterizing alternating-current installation/AC installation.
Optionally, judging unit 507, are additionally operable to calculate the data of characteristic parameter and owning in working status decision model
The Euclidean distance of barycenter;Wherein, a barycenter corresponds to a kind of working status;By the work shape of the centroid linkage of Euclidean distance minimum
State is determined as the working status of alternating-current installation/AC installation to be determined.
The equipment can be used for performing the method that the embodiment shown in Fig. 1, Fig. 2 and Fig. 4 is provided, therefore, for this
Function that each function module of equipment can be realized etc. refers to the description of the embodiment shown in Fig. 1, Fig. 2 and Fig. 4, seldom
Repeat.Wherein, second acquisition unit 506 and judging unit 507 are shown in the lump in Figure 5, but it is understood that second obtains
Unit 506 and judging unit 507 are not essential function modules, therefore shown in broken lines in Figure 5.
Fig. 6 is referred to, one embodiment of the invention also provides a kind of computer installation, which includes at least one
Processor 601, at least one processor 601 are used to realize the embodiment of the present invention when performing the computer program stored in memory
The step of working status decision model foundation of the alternating-current installation/AC installation of offer or working status decision method.
Optionally, at least one processor 601 can be specifically central processing unit, application-specific integrated circuit (English:
Application Specific Integrated Circuit, referred to as:ASIC), can be that one or more is used to control journey
The integrated circuit that sequence performs, can be use site programmable gate array (English:Field Programmable Gate
Array, referred to as:FPGA) the hardware circuit of exploitation, can be baseband processor.
Optionally, at least one processor 601 can include at least one processing core.
Optionally, which further includes memory 602, and memory 602 can include read-only storage (English:
Read Only Memory, referred to as:ROM), random access memory (English:Random Access Memory, referred to as:RAM)
And magnetic disk storage.Memory 602 is used to store data required when at least one processor 601 is run.The number of memory 602
Measure as one or more.Wherein, memory 602 is shown in the lump in figure 6, but it is understood that memory 602 is not essential
Function module, it is therefore shown in broken lines in figure 6.
One embodiment of the invention also provides a kind of computer-readable recording medium, is stored thereon with computer program, described
Realize that the working status decision model of alternating-current installation/AC installation provided in an embodiment of the present invention is built when computer program is executed by processor
The step of vertical or working status decision method.
In embodiments of the present invention, it should be understood that disclosed apparatus and method, can be real by another way
It is existing.For example, apparatus embodiments described above are only schematical, for example, the division of the unit or unit, is only
A kind of division of logic function, can there is an other dividing mode when actually realizing, for example, multiple units or component can combine or
Person is desirably integrated into another system, or some features can be ignored, or does not perform.Another, shown or discussed is mutual
Between coupling, direct-coupling or communication connection can be INDIRECT COUPLING or communication link by some interfaces, equipment or unit
Connect, can be electrical or other forms.
Each functional unit in embodiments of the present invention can be integrated in a processing unit, or unit also may be used
To be independent physical module.
If the integrated unit is realized in the form of SFU software functional unit and is used as independent production marketing or use
When, it can be stored in a computer read/write memory medium.Based on such understanding, the technical solution of the embodiment of the present invention
All or part can be embodied in the form of software product, which is stored in a storage medium
In, including some instructions are used so that a computer equipment, such as can be that personal computer, server, or network are set
It is standby etc., or all or part of step of each embodiment the method for processor (processor) the execution present invention.It is and foregoing
Storage medium includes:General serial bus USB (Universal Serial Bus flash drive), mobile hard disk, only
Read memory (read-only memory, ROM), random access memory (random access memory, RAM), magnetic disc
Or CD etc. is various can be with the medium of store program codes.
The above, above example are implemented above only the technical solution of the application to be described in detail
The explanation of example is only intended to help the method for understanding the embodiment of the present invention, should not be construed as the limitation to the embodiment of the present invention.This
The change or replacement that those skilled in the art can readily occur in, should all cover the embodiment of the present invention protection domain it
It is interior.
Claims (14)
- A kind of 1. working status decision model method for building up of alternating-current installation/AC installation, it is characterised in that including:Obtain the data sample set of alternating-current installation/AC installation;Wherein, alternating current of the data sample set including same type is set The standby data sample under different working condition, a data sample include the data of P parameters of electric power, and P is positive integer;Determine that each parameters of electric power to the disturbance degree of the working status of the alternating-current installation/AC installation, obtains in the data sample set Disturbance degree is more than the parameters of electric power of default disturbance degree threshold value as the characteristic parameter in the data sample set;The data of characteristic parameter described in each data sample in the data sample set are extracted, obtain the alternating-current installation/AC installation Feature samples set;L cluster is divided into according to the feature samples set, and obtains the barycenter of each cluster;Wherein, a cluster includes the number According to the partial data sample in sample set, the data sample that different clusters include is different, and the center that the barycenter is cluster, described Alternating-current installation/AC installation corresponds to a variety of working statuses, and the quantity L of cluster is identical with the quantity of the working status, and L is positive integer;The working status of the barycenter and the alternating-current installation/AC installation of the L cluster is associated, to obtain the alternating-current installation/AC installation Working status decision model;Wherein, a barycenter corresponds to a kind of working status.
- 2. the method as described in claim 1, it is characterised in that determine that each parameters of electric power is to institute in the data sample set The disturbance degree of the state of alternating-current installation/AC installation is stated, disturbance degree is obtained and is more than the parameters of electric power of default disturbance degree threshold value as the data Characteristic parameter in sample set, including:Vertical data sample matrix is built jointly according to the set of data samples of the alternating-current installation/AC installation;Wherein, in the data sample matrix The corresponding data sample of row in different parameters of electric power data, the corresponding different data of row in the data sample matrix The data of same parameters of electric power in sample;Establish the covariance matrix of the data sample matrix;The characteristic value of the covariance matrix is calculated, and calculates characteristic value summation shared by each characteristic value in the characteristic value Ratio, and the ratio is determined as the feature more than or equal to the corresponding parameters of electric power of characteristic value of preset ratio threshold value and is joined Number;Wherein, the corresponding characteristic value of a parameters of electric power is used to characterize shape of one parameters of electric power to the alternating-current installation/AC installation The disturbance degree of state.
- 3. the method as described in claim 1, it is characterised in that the feature samples set by the alternating-current installation/AC installation divides Into L cluster, including:The feature samples set is repeatedly divided using K-means algorithms, until obtaining and the alternating-current installation/AC installation The identical cluster of working status quantity;Wherein, for dividing each time, the error for calculating each cluster obtained after current division is put down Side and SSE, and select the cluster of the value maximum of SSE in current cluster to be used as and be divided object;One cluster includes the feature samples Partial Feature sample in set, the SSE of a cluster are used for the extent of polymerization for characterizing a cluster.
- 4. the method as described in claims 1 to 3 is any, it is characterised in that one of which electric power is joined in the P parameters of electric power Number is active power, then the working status of the barycenter and the alternating-current installation/AC installation by the L cluster is associated, including:According to the active power of the barycenter of the L cluster by the working status of the barycenter of the L cluster and the alternating-current installation/AC installation It is associated.
- 5. method as claimed in claim 4, it is characterised in that obtaining the working status decision model of the alternating-current installation/AC installation Afterwards, the method further includes:Obtain the working status decision model;Obtain the data of the characteristic parameter of alternating-current installation/AC installation to be determined;The characteristic parameter is the shape to the alternating-current installation/AC installation The disturbance degree of state exceedes the parameters of electric power of default disturbance degree threshold value;The data of the characteristic parameter are inputted the working status decision model to be judged as a result, the judgement result is used for Characterize the working status of the alternating-current installation/AC installation.
- 6. method as claimed in claim 5, it is characterised in that the data by the characteristic parameter input the work shape State decision model is judged as a result, described judge that result is used to characterize the working status of the alternating-current installation/AC installation, including:Calculate the data and the Euclidean distance of all barycenter in the working status decision model of the characteristic parameter;Wherein, One barycenter corresponds to a kind of working status;The working status of the centroid linkage of Euclidean distance minimum is determined as to the work of the alternating-current installation/AC installation to be determined State.
- 7. a kind of working status judgment models of alternating-current installation/AC installation establish equipment, it is characterised in that including:First acquisition unit, for obtaining the data sample set of alternating-current installation/AC installation;Wherein, the data sample set includes phase Data sample of the alternating-current installation/AC installation of same type under different working condition, a data sample include the number of P parameters of electric power According to P is positive integer;Determination unit, for determining that each parameters of electric power is to the working status of the alternating-current installation/AC installation in the data sample set Disturbance degree, obtain disturbance degree and be more than the parameters of electric power of default disturbance degree threshold value as the feature ginseng in the data sample set Number;Extraction unit, for extracting the data of characteristic parameter described in each data sample in the data sample set, obtains The feature samples set of the alternating-current installation/AC installation;Division unit, for being divided into L cluster according to the feature samples set, and obtains the barycenter of each cluster;Wherein, one Cluster includes the partial data sample in the data sample set, and the data sample that different clusters include is different, and the barycenter is The center of cluster, the alternating-current installation/AC installation correspond to a variety of working statuses, and the quantity L of cluster is identical with the quantity of the working status, and L is Positive integer;Associative cell, for the working status of the barycenter and the alternating-current installation/AC installation of the L cluster to be associated, to obtain State the working status decision model of alternating-current installation/AC installation;Wherein, a barycenter corresponds to a kind of working status.
- 8. equipment as claimed in claim 7, it is characterised in thatThe determination unit, specifically for building vertical data sample matrix jointly according to the set of data samples of the alternating-current installation/AC installation;Its In, the data of different parameters of electric power, the data sample square in the corresponding data sample of row in the data sample matrix The data of same parameters of electric power in the corresponding different data sample of row in battle array;Establish the covariance square of the data sample matrix Battle array;The characteristic value of the covariance matrix is calculated, and calculates characteristic value summation shared by each characteristic value in the characteristic value Ratio, and the ratio is determined as the feature more than or equal to the corresponding parameters of electric power of characteristic value of preset ratio threshold value and is joined Number;Wherein, the corresponding characteristic value of a parameters of electric power is used to characterize shape of one parameters of electric power to the alternating-current installation/AC installation The disturbance degree of state.
- 9. equipment as claimed in claim 7, it is characterised in that the division unit, specifically for by the feature samples collection Conjunction is repeatedly divided using K-means algorithms, until obtaining the cluster identical with the working status quantity of the alternating-current installation/AC installation; Wherein, for dividing each time, the error sum of squares SSE of each cluster obtained after current division is calculated, and is selected currently The cluster of the value maximum of SSE is used as and is divided object in cluster;One cluster includes the Partial Feature sample in the feature samples set, The SSE of one cluster is used for the extent of polymerization for characterizing a cluster.
- 10. the equipment as described in claim 7~9 is any, it is characterised in that the associative cell is by the barycenter of the L cluster It is associated with the working status of the alternating-current installation/AC installation, including:The associative cell sets the barycenter of the L cluster with the alternating current according to the active power of the barycenter of the L cluster Standby working status is associated.
- 11. equipment as claimed in claim 10, it is characterised in that the equipment further includes second acquisition unit and judges single Member;The second acquisition unit, for obtaining working status decision model, the working status decision model is used to judge to hand over The working status of galvanic electricity equipment;And obtain the data of the characteristic parameter of alternating-current installation/AC installation to be determined;The characteristic parameter is The parameters of electric power of default disturbance degree threshold value is exceeded to the disturbance degree of the state of the alternating-current installation/AC installation;The judging unit, obtains judging knot for the data of the characteristic parameter to be inputted the working status decision model Fruit, it is described to judge that knot is used for the working status for characterizing the alternating-current installation/AC installation.
- 12. equipment as claimed in claim 11, it is characterised in that the judging unit, joins specifically for calculating the feature Several data and the Euclidean distance of all barycenter in the working status decision model;Wherein, a barycenter corresponds to a kind of work Make state;The working status of the centroid linkage of Euclidean distance minimum is determined as to the work of the alternating-current installation/AC installation to be determined Make state.
- 13. a kind of computer installation, it is characterised in that described device includes processor, and the processor is used to perform memory Realized during the computer program of middle storage as any one of claim 1-6 the step of method.
- 14. a kind of computer-readable recording medium, is stored thereon with computer program, it is characterised in that:The computer program Realized when being executed by processor as any one of claim 1-6 the step of method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711267337.9A CN108038497B (en) | 2017-12-05 | 2017-12-05 | Method and equipment for establishing working state judgment model of alternating current equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711267337.9A CN108038497B (en) | 2017-12-05 | 2017-12-05 | Method and equipment for establishing working state judgment model of alternating current equipment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108038497A true CN108038497A (en) | 2018-05-15 |
CN108038497B CN108038497B (en) | 2020-08-25 |
Family
ID=62095043
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711267337.9A Active CN108038497B (en) | 2017-12-05 | 2017-12-05 | Method and equipment for establishing working state judgment model of alternating current equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108038497B (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108734603A (en) * | 2018-05-24 | 2018-11-02 | 河南工业大学 | Big data phase recognition methods based on cluster |
CN109145180A (en) * | 2018-06-27 | 2019-01-04 | 东华大学 | A kind of enterprise hot spots event method for digging based on increment cluster |
CN109615273A (en) * | 2019-01-16 | 2019-04-12 | 国网浙江省电力有限公司电力科学研究院 | A kind of electric car electrically-charging equipment method for evaluating state and system |
CN111488947A (en) * | 2020-04-28 | 2020-08-04 | 深圳力维智联技术有限公司 | Fault detection method and device for power system equipment |
CN111695884A (en) * | 2020-08-17 | 2020-09-22 | 广东新视野信息科技股份有限公司 | Internet of things big data visualization method and system based on smart construction site |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070010906A1 (en) * | 2005-07-11 | 2007-01-11 | Tokyo Electron Limited | Apparatus and system for monitoring a substrate processing, program for monitoring the processing and storage medium storing same |
CN106017955A (en) * | 2016-05-13 | 2016-10-12 | 上海数道信息科技有限公司 | Method and system for identifying mechanical equipment state types |
CN106156485A (en) * | 2016-06-16 | 2016-11-23 | 广州供电局有限公司 | Method for diagnosing fault of power transformer and device |
CN106204324A (en) * | 2016-07-07 | 2016-12-07 | 西安西热电站信息技术有限公司 | A kind of method determining that power plant's complex device key monitoring parameter and each parameters weighting distribute |
-
2017
- 2017-12-05 CN CN201711267337.9A patent/CN108038497B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070010906A1 (en) * | 2005-07-11 | 2007-01-11 | Tokyo Electron Limited | Apparatus and system for monitoring a substrate processing, program for monitoring the processing and storage medium storing same |
CN106017955A (en) * | 2016-05-13 | 2016-10-12 | 上海数道信息科技有限公司 | Method and system for identifying mechanical equipment state types |
CN106156485A (en) * | 2016-06-16 | 2016-11-23 | 广州供电局有限公司 | Method for diagnosing fault of power transformer and device |
CN106204324A (en) * | 2016-07-07 | 2016-12-07 | 西安西热电站信息技术有限公司 | A kind of method determining that power plant's complex device key monitoring parameter and each parameters weighting distribute |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108734603A (en) * | 2018-05-24 | 2018-11-02 | 河南工业大学 | Big data phase recognition methods based on cluster |
CN109145180A (en) * | 2018-06-27 | 2019-01-04 | 东华大学 | A kind of enterprise hot spots event method for digging based on increment cluster |
CN109145180B (en) * | 2018-06-27 | 2021-07-20 | 东华大学 | Enterprise hot event mining method based on incremental clustering |
CN109615273A (en) * | 2019-01-16 | 2019-04-12 | 国网浙江省电力有限公司电力科学研究院 | A kind of electric car electrically-charging equipment method for evaluating state and system |
CN111488947A (en) * | 2020-04-28 | 2020-08-04 | 深圳力维智联技术有限公司 | Fault detection method and device for power system equipment |
CN111488947B (en) * | 2020-04-28 | 2024-02-02 | 深圳力维智联技术有限公司 | Fault detection method and device for power system equipment |
CN111695884A (en) * | 2020-08-17 | 2020-09-22 | 广东新视野信息科技股份有限公司 | Internet of things big data visualization method and system based on smart construction site |
CN111695884B (en) * | 2020-08-17 | 2020-11-20 | 广东新视野信息科技股份有限公司 | Internet of things big data visualization method and system based on smart construction site |
Also Published As
Publication number | Publication date |
---|---|
CN108038497B (en) | 2020-08-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108038497A (en) | The working status decision model method for building up and equipment of a kind of alternating-current installation/AC installation | |
CN108429254B (en) | non-invasive electric load identification method | |
CN112699913A (en) | Transformer area household variable relation abnormity diagnosis method and device | |
Chicco et al. | Load pattern-based classification of electricity customers | |
CN108199404B (en) | Spectral clustering cluster division method of high-permeability distributed energy system | |
CN110070282B (en) | Low-voltage transformer area line loss influence factor analysis method based on comprehensive relevance | |
CN112149873B (en) | Low-voltage station line loss reasonable interval prediction method based on deep learning | |
CN104505894B (en) | Power management system and state estimation method based on mining lithium ion batteries | |
CN110097297A (en) | A kind of various dimensions stealing situation Intellisense method, system, equipment and medium | |
CN102074955A (en) | Method based on knowledge discovery technology for stability assessment and control of electric system | |
CN107422266B (en) | Fault diagnosis method and device for high-capacity battery energy storage system | |
CN110609200B (en) | Power distribution network earth fault protection method based on fuzzy metric fusion criterion | |
CN104809658B (en) | A kind of rapid analysis method of low-voltage distribution network taiwan area line loss | |
CN105807231B (en) | A kind of method and system for remaining battery capacity detection | |
CN103177188A (en) | Electric system load dynamic characteristic classifying method based on characteristic mapping | |
CN110110907B (en) | Method for extracting characteristic parameters of low-voltage transformer area | |
CN111525587B (en) | Reactive load situation-based power grid reactive voltage control method and system | |
CN106355209B (en) | Stealing diagnostic system and method based on decision Tree algorithms | |
CN109740962A (en) | Voltage stabilization probability evaluation method of failure based on scene subregion and cumulant | |
CN109298225A (en) | A kind of voltage metric data abnormality automatic identification model and method | |
CN106908752A (en) | The electric energy metrical abnormality diagnostic method and system of a kind of feature based packet | |
CN110738232A (en) | grid voltage out-of-limit cause diagnosis method based on data mining technology | |
CN112990627B (en) | Power quality evaluation method | |
CN111856209A (en) | Power transmission line fault classification method and device | |
CN113673480A (en) | Non-intrusive load identification method and device |
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 |