CN108663202A - GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm and system - Google Patents
GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm and system Download PDFInfo
- Publication number
- CN108663202A CN108663202A CN201810414184.4A CN201810414184A CN108663202A CN 108663202 A CN108663202 A CN 108663202A CN 201810414184 A CN201810414184 A CN 201810414184A CN 108663202 A CN108663202 A CN 108663202A
- Authority
- CN
- China
- Prior art keywords
- gis
- chaos
- signal
- bird
- value
- 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
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Feedback Control In General (AREA)
Abstract
The invention discloses a kind of GIS mechanical failure diagnostic methods and system based on chaos cuckoo algorithm, two kinds of common GIS mechanical breakdowns are vibrated for loosened screw and metal particle, chaos cuckoo algorithm optimization VMD parameters are used first, then VMD decomposition is carried out to GIS normal vibrations signal and fault-signal, extract the feature vector of different faults type signal, training sample set of eigenvectors is clustered finally by the K means clustering algorithms of linear decrease weight PSO optimizations, obtain different cluster centres, recycle minimum Eustachian distance principle that test sample set of eigenvectors is identified, realize the diagnosis of GIS mechanical breakdowns.
Description
Technical field
The present invention relates to a kind of GIS mechanical failure diagnostic methods and system based on chaos cuckoo algorithm.
Background technology
In recent years, along with the continuous improvement of voltage class, to the reliability of power quality, more stringent requirements are proposed.
While further increasing, power grid accident takes place frequently for national electricity consumption, and GIS is as important link, once breaking down may
Large-scale power outage is caused, therefore ensures that the safe operation of GIS is benefited the nation and the people.
Currently, although the GIS coefficients of stability are high in electric system, still there are equipment operation a period of time or
Just occurs the precedent of accident when starting shipment.In addition, in the process of running, since voltage class is very high, presence is very strong inside GIS
Magnetic field, be easy that tiny flaw existing for equipment is made gradually to spread, thus damage equipment, bring huge economic losses.
Therefore, the mechanical performance of GIS is monitored on-line, its likelihood of failure is predicted simultaneously by monitoring vibration signal
Differentiate its fault type, it is significant to safe operation of power system, and have a good application prospect and promotional value.
Invention content
The present invention is to solve the above-mentioned problems, it is proposed that a kind of GIS mechanical fault diagnosis based on chaos cuckoo algorithm
Method and system.
First, the present invention provides a kind of GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm.This method
With chaos cuckoo algorithm, mode decomposition is carried out to GIS abnormal transient vibration signals, to find and judge the failure of GIS device,
It avoids that more serious mechanical breakdown occurs.
Secondly, the present invention provides a kind of GIS Diagnosis system of mechanical failure based on chaos cuckoo algorithm, this system
Present Research for current GIS machine performances live detection and fault diagnosis technology and there are the problem of, based on vibration letter
Number processing carry out GIS mechanical fault detections.
To achieve the goals above, the present invention adopts the following technical scheme that:
A kind of GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm, specifically include following steps:
According to typical fault Simulated GlS electric fault and mechanical breakdown, GIS normal vibrations signal and fault-signal are obtained;
With chaos cuckoo algorithm optimization variation mode decomposition parameter, cuckoo population is improved using chaotic maps
Population number;
It carries out variation mode decomposition (VMD) respectively to GIS normal vibrations signal and fault-signal, extracts different faults class
The feature vector of type signal;
Training sample set of eigenvectors is clustered by clustering algorithm, obtains different cluster centres;
Test sample set of eigenvectors is identified using minimum Eustachian distance principle, realizes examining for GIS mechanical breakdowns
It is disconnected.
Further, typical fault Simulated GlS electric fault and mechanical breakdown data in historical data are collected, GIS is formed
Normal vibration signal set and fault-signal set.
Preferably, GIS simulation test platforms, the free particle vibration of Simulated GlS and the common machine of two class of loosened screw are built
Tool failure, to extract the characteristic information of two class failures.
Further, with chaos cuckoo algorithm optimization variation mode decomposition parameter, cuckoo is improved using chaotic maps
After the detailed process of the population number of bird population is included in cuckoo algorithm change Bird's Nest position, chaos optimization search is carried out, is used
Chaotic maps obtain chaos sequence, change Bird's Nest position according to fitness function.
Specifically, including the following steps:
1) each parameter of initialization CCS algorithms and determining fitness function;
2) initialization probability parameter randomly generates n Bird's Nest, in affecting parameters precognition mode number K and secondary punishment
The combination [K, α] of item α is used as Bird's Nest position, randomly generates initial position of a certain number of affecting parameters combinations as Bird's Nest;
3) variation mode decomposition operation is done to signal under different Bird's Nest locality conditions, it is corresponding calculates each Bird's Nest position
Fitness value;
4) retain the most suitable Bird's Nest position of prior-generation, update Bird's Nest using Levy flight algorithms, compare fitness value size;
5) judge whether fitness function changes in parameter setting iterations, if there is variation, be transferred in next step;
If there is no variation, to current optimal Bird's Nest chaotic maps;
6) step 4) change Bird's Nest position is utilized, is compared with random number and probability parameter, if random number is joined more than probability
Number then changes Bird's Nest position at random, on the contrary then constant, finally retains one group of best Bird's Nest position;
7) loop iteration goes to step 3), until iterations reach after maximum set value output optimal adaptation angle value and
Bird's Nest position.
Further, the variation mode decomposition of GIS vibration signal x (t) obtains n Intrinsic mode function BIMF, accordingly
It can be calculated and correspond to respective energy value E1, E2 ... En;Do not consider the influence of participation component, the energy summation of n BIMF
Certain total energy value for being equal to original vibration signal;Due to each BIMF components BIMF1, BIMF2 ..., BIMFn include to differ
The frequency constituent of sample corresponds to different energy matrix E={ E1, E2 ... En }, constitutes GIS vibrations letter in frequency domain
Number Energy distribution.
Further, by the K-means clustering algorithms of linear decrease weight PSO optimizations to training sample feature vector
Collection is clustered, and different cluster centres is obtained.
Further, the PSO algorithm calculating process after optimization includes:
A) initialization population scale, the speed of particle and location parameter;
B) fitness for evaluating each particle makes a body position and adaptive value in particle with adaptive optimal control value preserve
In global extremum gbest;
C) particle position and speed are updated;
D) according to current iteration step number, weighted value is updated:
E) fitness value of each particle is made comparisons with individual extreme value pbest, global extremum gbest, if being better than
Pbest, gbest are then substituted, then relatively more current all pbest, gbest, update global extremum gbest;
If f) meeting the corresponding conditions of end condition, iteration if not then return to step c) are terminated.
Further, linear decrease weight PSO algorithms are introduced into K-means clusters, are encoded using cluster centre
Mode, the Euclidean distance function of data sample point and barycenter is designed as to the fitness value function of linear decrease weight PSO.
Further, K-means algorithmic procedures include:
Randomly select k cluster center of mass point;
Class is broken up:The distance of the data and cluster centre in sample is calculated, foundation is apart from nearest criterion, sample data
Distribute to corresponding cluster center;
Recalculate center of mass point:Calculate it is all kinds of in all sample datas mean value, and this is defined as in new cluster
The heart;
The step of repeating class differentiation and recalculating center of mass point, until final algorithmic stability, cluster terminate.
Further, linear decrease weight PSO algorithms are introduced into K-means clusters by the present invention, using in cluster
The Euclidean distance function of data sample point and barycenter, is designed as the fitness value of linear decrease weight PSO by the mode of heart coding
Function.
Further, according to the definition of Energy-Entropy and root-mean-square value, the entropy and root-mean-square value of all samples are calculated, profit
Root-mean-square value and VMD entropy are subjected to two-dimentional cluster with K-means clustering methods are improved, obtain more initial cluster centers, respectively table
Levy the vibration signal under GIS different conditions;
To remaining sample, the signal under different conditions is found according to the Euclidean distance iteration at sample and known cluster center
Cluster centre, and then classify to the vibration signal of sample.
A kind of GIS Diagnosis system of mechanical failure based on chaos cuckoo algorithm, runs on processor or can storage medium
On, it is configured as executing to give an order:
According to typical fault Simulated GlS electric fault and mechanical breakdown, GIS normal vibrations signal and fault-signal are obtained;
With chaos cuckoo algorithm optimization variation mode decomposition parameter, cuckoo population is improved using chaotic maps
Population number;
Variation mode decomposition is carried out to GIS normal vibrations signal and fault-signal, extracts the spy of different faults type signal
Sign vector;
Training sample set of eigenvectors is clustered by clustering algorithm, obtains different cluster centres;
Test sample set of eigenvectors is identified using minimum Eustachian distance principle, realizes examining for GIS mechanical breakdowns
It is disconnected.
Compared with prior art, beneficial effects of the present invention are:
1, the present invention provides a kind of mechanical failure diagnostic method based on chaos cuckoo algorithm optimization VMD parameters, uses
Chaos cuckoo algorithm carries out mode decomposition to GIS abnormal transient vibration signals and is avoided with finding and judging the failure of GIS device
More serious mechanical breakdown occurs.
2, the problem of being directed to signal characteristic abstraction algorithm variation mode decomposition (VMD) parameter uncertainty, proposes based on mixed
The intelligent optimizing of the VMD parameters of ignorant cuckoo algorithm, effectively increases signal decomposition rate and accuracy;
3, chaos cuckoo optimization algorithm is introduced into the optimization of VMD parameters, solve VMD algorithm parameters and be difficult to determine
The problem of.
Description of the drawings
The accompanying drawings which form a part of this application are used for providing further understanding of the present application, the application's
Illustrative embodiments and their description do not constitute the improper restriction to the application for explaining the application.
Fig. 1 is vibration signal acquisition system figure;
Fig. 2 is CS algorithm flow schematic diagrames;
Fig. 3 is CCS algorithm flow schematic diagrames;
Fig. 4 is chaos cuckoo algorithm optimization VMD parameter algorithm flow diagrams;
Fig. 5 is particle cluster algorithm basic framework schematic diagram;
Fig. 6 is linear decrease weight PSO-K-means clustering algorithm flow charts;
Fig. 7 is GIS mechanical fault diagnosis flow charts.
Specific implementation mode:
The invention will be further described with embodiment below in conjunction with the accompanying drawings.
It is noted that following detailed description is all illustrative, it is intended to provide further instruction to the application.Unless
Otherwise indicated, all technical and scientific terms used herein has and the application person of an ordinary skill in the technical field
Normally understood identical meanings.
It should be noted that term used herein above is merely to describe specific implementation mode, and be not intended to restricted root
According to the illustrative embodiments of the application.As used herein, unless the context clearly indicates otherwise, otherwise singular shape
Formula is also intended to include plural form, additionally, it should be understood that, when in the present specification use term "comprising" and/or
When " comprising ", existing characteristics, step, operation, device, component and/or combination thereof are indicated.
In the present invention, term for example "upper", "lower", "left", "right", "front", "rear", "vertical", "horizontal", " side ",
The orientation or positional relationship of the instructions such as "bottom" is to be based on the orientation or positional relationship shown in the drawings, only to facilitate narration is originally
The relative for inventing each component or component structure relationship and determination, not refers in particular to either component or element in the present invention, Bu Nengli
Solution is limitation of the present invention.
In the present invention, term such as " affixed ", " connected ", " connection " shall be understood in a broad sense, and expression can be fixed company
It connects, can also be to be integrally connected or be detachably connected;It can be directly connected, it can also be indirectly connected through an intermediary.
Related scientific research for this field or technical staff can determine above-mentioned term in the present invention specific as the case may be
Meaning is not considered as limiting the invention.
The purpose of the present invention is for mechanical breakdown common GIS, provide a kind of based on chaos cuckoo algorithm optimization
The mechanical failure diagnostic method of VMD parameters, and develop the GIS mechanical breakdown software and hardwares system analyzed based on abnormal transient vibration signal
System.
Two kinds of common GIS mechanical breakdowns are vibrated for loosened screw and metal particle, use chaos cuckoo to calculate first
Method optimizes VMD parameters, then carries out VMD decomposition, extraction different faults type letter to GIS normal vibrations signal and fault-signal
Number feature vector, finally by the K-means clustering algorithms of linear decrease weight PSO optimization to training sample feature vector
Collection is clustered, and different cluster centres is obtained, and minimum Eustachian distance principle is recycled to carry out test sample set of eigenvectors
The diagnosis of GIS mechanical breakdowns is realized in identification.
As shown in fig. 7, a kind of mechanical failure diagnostic method based on chaos cuckoo algorithm optimization VMD parameters, with mixed
Ignorant cuckoo algorithm, includes the following steps:
(1) typical fault Simulated GlS electric fault and mechanical breakdown, such as the vibration of loosened screw, metal particle are set,
To obtain GIS normal vibrations signal and fault-signal;
As shown in Figure 1, perceiving the vibration physical signal of GIS using vibrating sensor, will be vibrated using signal condition equipment
The electric signal that sensor obtains is improved, and using the data after data collecting card acquisition conditioning, allusion quotation is obtained using computer
Type fault simulation GIS electric faults and mechanical fault signals.
(2) chaos cuckoo algorithm optimization VMD parameters are used, the population number of cuckoo population is improved using chaotic maps,
Utmostly improve algorithm the convergence speed and accuracy rate;
(3) VMD decomposition is carried out to GIS normal vibrations signal and fault-signal, extracts the feature of different faults type signal
Vector;
(4) training sample set of eigenvectors is carried out by the K-means clustering algorithms of linear decrease weight PSO optimizations
Cluster, obtains different cluster centres;
(5) test sample set of eigenvectors is identified using minimum Eustachian distance principle, realizes GIS mechanical breakdowns
Diagnosis.
Wherein, second step has following steps:
Before the best parameter group [K, α] of chaos cuckoo algorithm search VMD, clear fitness function is needed, and kurtosis
As for characterizing the parameter of waveform spike degree size, may be defined as:
In formula:μ is the mean value of signal x;σ is the standard deviation of signal x.In the present invention, kurtosis function is selected as chaos cuckoo
The fitness function of bird algorithm, since each Bird's Nest position corresponds to a parameter combination [K, α], and kurtosis functional value
Maximum Bird's Nest position includes most fault message, so, very big kurtosis valueIt is the ultimate aim of search.Such as figure
Shown in 4, detailed process is as follows:
1) each parameter of initialization CCS algorithms and determining fitness function;
2) initialization probability parameter PaIt is 0.25, randomly generates n Bird's Nest, Bird's Nest is used as with affecting parameters combination [K, α]
Position randomly generates initial position of a certain number of affecting parameters combinations as Bird's Nest;
3) VMD operations are done to signal under different Bird's Nest locality conditions, calculates each corresponding fitness value in Bird's Nest position
4) retain the most suitable Bird's Nest position of prior-generationIt is flown using Levy and updates Bird's Nest, compare fitness value size;
5) judge whether fitness function changes in parameter setting iterations, if there is variation, be transferred in next step;
If there is no variation, to current optimal Bird's Nest chaotic maps;
6) step 4) change Bird's Nest position is utilized, with random number r ∈ [0,1] and PaComparison, if r>Pa, then random to changeIt is on the contrary then constant, finally retain one group of best bird's nest position;
7) loop iteration goes to step 3), until iterations reach after maximum set value output optimal adaptation angle value and
Bird's Nest position.
Levy is distributed:
Wherein,
Therefore, as can be seen from the above equation, characteristic function and Levy steady-state distributions are determined by tetra- real parameters of α, β, μ, σ.Its
In, characteristic index α ∈ [0,2], offset parameter β ∈ [- 1,1], μ characterize displacement parameter, scale σ>0, k is initial parameter.
VMD specific implementation process is as follows, and the arrow in formula indicates assignment:
1) parameters are initialized;
Specifically:λ1, the initial value of n is 0,
{μk}:={ μ1,…,μkIndicate the abbreviation of each set of modes;
{ωk}:={ ω1,…,ωkIndicate the abbreviation of each mode top frequency;
λ1For Lagrange multiplier.
2) iterations are updated, update iterations n is n+1;
3) in each iteration cycle process, to all centre frequency ω >=0,
Update
α is secondary penalty term;
Update ωk:
4) to all ω >=0, dual promotion is carried out:
5) 2) -4 are repeated), until meeting following iterative constrained condition:
{μk}:={ μ1,…,μkIndicate the abbreviation of each set of modes;
{ωk}:={ ω1,…,ωkIndicate the abbreviation of each mode top frequency;
λ:Lagrange multiplier;ξ>0 is discrimination precision.
ωkOr ω is centre frequency;
RespectivelyThe corresponding Fourier transformation of f (t), λ (t).
VMD Energy-Entropies:
By the way that n Intrinsic mode function BIMF can be obtained to the VMD decomposition of laboratory simulation GIS vibration signal x (t),
It can be calculated accordingly corresponds to respective energy value E1, E2 ... En.Do not consider the influence of participation component, the energy of n BIMF
Measure the total energy value that summation is centainly equal to original vibration signal.Due to each BIMF components BIMF1, BIMF2 ..., BIMFn packet
Containing different frequency constituent, different energy matrix E={ E1, E2 ... En } is corresponded to, constitutes GIS in frequency domain
The Energy distribution of vibration signal.According to the above analysis, the definition of VMD Energy-Entropies can be obtained:
In formula:pi=Ei/ E be i-th limit band Intrinsic mode function BIMFi energy account for gross energy number.
Root-mean-square value:
Root-mean-square value (RMS) is a kind of semaphore scale being commonly used, and the characterization signal vibration that it can be reliable becomes
Change situation.The mean-square value of signal x (t) is a kind of measure of mean power, and expression formula is:
That square root of corresponding positive number is exactly root-mean-square value xrms, i.e.,:
It is as follows to optimize later PSO algorithm calculating process:
1) initialization population scale, the speed of particle and position etc.;
2) fitness for evaluating each particle makes a body position and adaptive value in particle with adaptive optimal control value preserve
In global extremum gbest;
3) particle position and speed are updated according to the following formula:
Wherein, population position xi=(xi1,xi2,…,xid);
Speed vi=(vi1,vi2,…,vid);
History optimal location pi=(pi1,pi2,…,pid);
Aceleration pulse c1And c2。
xi,j(t+1)=xi,j(t)+vi,j(t+1), j=1,2 ..., d
vi,j(t+1)=wvi,j(t)+c1r1[pi,j-xi,j(t)]+c2r2[pg,j-xi,j(t)]
4) weighted value is updated according to the following formula:
In formula:wmaxWeight coefficient maximum value;
wminWeight coefficient minimum value;
T- current iteration step numbers.
5) fitness value of each particle is made comparisons with individual extreme value pbest, global extremum gbest, if being better than
Pbest, gbest are then substituted.Then relatively more current all pbest, gbest, update gbest.
If 6) meet the corresponding conditions of end condition, algorithm just terminates if not then return to step 3).
As shown in fig. 6, linear decrease weight PSO-K-means clustering algorithms:
K-means algorithms are estimating as similarity Euclidean distance, and it is accurate that clustering criteria function chooses error sum of squares
Then function.The basic calculating process of the algorithm is as follows:
1) it initializes:Randomly select k cluster center of mass point;
2) class is broken up:The distance of the data and cluster centre in sample is calculated, foundation is apart from nearest criterion, sample number
According to distributing to corresponding cluster center;
3) center of mass point is recalculated:Calculate it is all kinds of in all sample datas mean value, and this is defined as in new cluster
The heart;
4) step 2) is repeated and 3) until final algorithmic stability, cluster terminate.
The algorithm is more sensitive to the selection of initial clustering center of mass point, is susceptible to local optimum, and therefore, the present invention will
Linear decrease weight PSO algorithms are introduced into K-means clusters, by the way of cluster centre coding, by data sample point
The fitness value function of linear decrease weight PSO is designed as with the Euclidean distance function of barycenter.
Linear decrease weight PSO-K-means clustering algorithms are as shown in Figure 6.
By taking the vibration of free particle and loosened screw failure the two GIS failures as an example, the VMD entropy of fault-signal compares
It is small, and root-mean-square value is larger, can utilize accordingly improved K-means clusterings by root-mean-square value and VMD entropy the two
Dimension is combined, and carrys out the otherness of characterization failure type.
Acquisition normal vibration signal, loosened screw signal and each 80 groups of samples of particle vibration signal respectively, appoint and take three types
It is used as training sample for each 60 groups in number, remaining 20 groups are regarded test sample;Appoint and take each 20 groups of above-mentioned three classes signal, uses chaos
Cuckoo optimization VMD parameter combinations [K, α] obtain average value 20 times.
According to the definition of Energy-Entropy and root-mean-square value, the entropy and root-mean-square value of all samples are calculated, utilizes improvement K-
Root-mean-square value and VMD entropy are carried out two-dimentional cluster by means clustering methods, are obtained three initial cluster centers, are characterized GIS respectively
Normal vibration signal, GIS loosened screws signal and GIS free metal particle vibration signals.To remaining sample, according to sample and
The Euclidean distance iteration at known cluster center finds cluster centre.The three cluster centres difference acquired by improved clustering
For C1 (c11, c12), C2 (c21,c22), C3 (c31,c32)。
Analysis result is as follows:
Test sample is calculated at a distance from three cluster centres using Euclidean distance, i.e.,:
In formula, (cj1,cj2) what is indicated is C1, the coordinate of C2, C3;d1Indicate test sample and loosened screw signal center
The distance of C1;d2Test sample is indicated at a distance from free metal particle Vibration Fault Signal center C2, d3Indicate test sample
At a distance from GIS normal vibrations signal center C3, j=1,2,3.
To remaining each 20 groups of test samples, using minimum Eustachian distance principle to micro- according to normal condition, free metal
60 groups of test samples that grain vibration fault, loosened screw failure sequence form are judged.The minimum Eustachian distance d of definitionj:
Solve n-th of sample djValue, corresponding sample belong to jth class state, j=1, and 2,3.
Laboratory GIS simulation test platforms can also be built, the free particle vibration of Simulated GlS and two class of loosened screw are normal
The mechanical breakdown seen effectively extracts the characteristic information of two class failures, and linear decrease weight PSO-K-means clusters is combined to calculate
Method identifies GIS mechanical breakdowns, tentatively establishes GIS mechanical fault diagnosis identification library.
The foregoing is merely the preferred embodiments of the application, are not intended to limit this application, for the skill of this field
For art personnel, the application can have various modifications and variations.Within the spirit and principles of this application, any made by
Modification, equivalent replacement, improvement etc., should be included within the protection domain of the application.
Above-mentioned, although the foregoing specific embodiments of the present invention is described with reference to the accompanying drawings, is not protected to the present invention
The limitation of range, those skilled in the art should understand that, based on the technical solutions of the present invention, people in the art
Member need not make the creative labor the various modifications or changes that can be made still within protection scope of the present invention.
Claims (10)
1. a kind of GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm, it is characterized in that:Include the following steps:
According to typical fault Simulated GlS electric fault and mechanical breakdown, GIS normal vibrations signal and fault-signal are obtained;
With chaos cuckoo algorithm optimization variation mode decomposition parameter, the population of cuckoo population is improved using chaotic maps
Number;
Variation mode decomposition is carried out respectively to GIS normal vibrations signal and fault-signal, extracts the spy of different faults type signal
Sign vector;
Training sample set of eigenvectors is clustered by clustering algorithm, obtains different cluster centres;
Test sample set of eigenvectors is identified using minimum Eustachian distance principle, realizes the diagnosis of GIS mechanical breakdowns.
2. a kind of GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm as described in claim 1, it is characterized in that:
Collect historical data in typical fault Simulated GlS electric fault and mechanical breakdown data, formed GIS normal vibrations signal set and
Fault-signal set.
3. a kind of GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm as described in claim 1, it is characterized in that:
With chaos cuckoo algorithm optimization variation mode decomposition parameter, the tool of the population number of cuckoo population is improved using chaotic maps
After body process is included in cuckoo algorithm change Bird's Nest position, chaos optimization search is carried out, chaos sequence is obtained using chaotic maps
Row change Bird's Nest position according to fitness function.
4. a kind of GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm as claimed in claim 3, it is characterized in that:
Include the following steps:
1) each parameter of initialization CCS algorithms and determining fitness function;
2) initialization probability parameter randomly generates n Bird's Nest, with the group of precognition mode number and secondary penalty term in affecting parameters
Cooperation is Bird's Nest position, randomly generates initial position of a certain number of affecting parameters combinations as Bird's Nest;
3) variation mode decomposition operation is done to signal under different Bird's Nest locality conditions, calculates each corresponding adaptation in Bird's Nest position
Angle value;
4) retain the most suitable Bird's Nest position of prior-generation, update Bird's Nest using Levy flight algorithms, compare fitness value size;
5) judge whether fitness function changes in parameter setting iterations, if there is variation, be transferred in next step;If not depositing
Changing, to current optimal Bird's Nest chaotic maps;
6) step 4) change Bird's Nest position is utilized, is compared with random number and probability parameter, if random number is more than probability parameter, with
Machine changes Bird's Nest position, on the contrary then constant, finally retains one group of best bird's nest position;
7) loop iteration goes to step 3), until iterations reach output optimal adaptation angle value and Bird's Nest after maximum set value
Position.
5. a kind of GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm as described in claim 1, it is characterized in that:
The variation mode decomposition of GIS vibration signal x (t) obtains n Intrinsic mode function BIMF, can calculate it accordingly and correspond to respectively
Energy value E1, E2 ... En;Do not consider that the influence of participation component, the energy summation of n BIMF are centainly equal to original vibration signal
Total energy value;It is corresponding since each BIMF components BIMF1, BIMF2 ..., BIMFn include different frequency constituent
Different energy matrix E={ E1, E2 ... En }, the Energy distribution of GIS vibration signals in frequency domain is constituted.
6. a kind of GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm as described in claim 1, it is characterized in that:
Training sample set of eigenvectors is clustered by the K-means clustering algorithms of linear decrease weight PSO optimizations, obtains difference
Cluster centre.
7. a kind of GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm as claimed in claim 6, it is characterized in that:
PSO algorithm calculating process after optimization includes:
A) initialization population scale, the speed of particle and location parameter;
B) fitness for evaluating each particle makes a body position and adaptive value in particle with adaptive optimal control value be stored in the overall situation
In extreme value gbest;
C) particle position and speed are updated;
D) according to current iteration step number, weighted value is updated:
E) fitness value of each particle is made comparisons with individual extreme value pbest, global extremum gbest, if better than pbest,
Gbest is then substituted, and then relatively more current all pbest, gbest, update gbest;
If f) meeting the corresponding conditions of end condition, iteration if not then return to step c) are terminated.
8. a kind of GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm as described in claim 1, it is characterized in that:
Linear decrease weight PSO algorithms are introduced into K-means clusters, by the way of cluster centre coding, by data sample
Point is designed as the fitness value function of linear decrease weight PSO with the Euclidean distance function of barycenter.
9. a kind of GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm as claimed in claim 8, it is characterized in that:
Linear decrease weight PSO-K-means cluster process includes:
Randomly select k cluster center of mass point;
Class is broken up:The distance for calculating the data and cluster centre in sample distributes to sample data according to apart from nearest criterion
Corresponding cluster center;
Recalculate center of mass point:Calculate it is all kinds of in all sample datas mean value, and this is defined as to new cluster centre;
The step of repeating class differentiation and recalculating center of mass point, until final algorithmic stability, cluster terminate.
The algorithm is more sensitive to the selection of initial clustering center of mass point, is susceptible to local optimum, and therefore, the present invention will be passed linearly
Subtract weight PSO algorithms and be introduced into K-means clusters, by the way of cluster centre coding, by data sample point and barycenter
Euclidean distance function be designed as the fitness value function of linear decrease weight PSO.
10. a kind of GIS Diagnosis system of mechanical failure based on chaos cuckoo algorithm, it is characterized in that:Run on processor or can
On storage medium, it is configured as executing to give an order:
According to typical fault Simulated GlS electric fault and mechanical breakdown, GIS normal vibrations signal and fault-signal are obtained;
With chaos cuckoo algorithm optimization variation mode decomposition parameter, the population of cuckoo population is improved using chaotic maps
Number;
Variation mode decomposition is carried out to GIS normal vibrations signal and fault-signal, extract the feature of different faults type signal to
Amount;
Training sample set of eigenvectors is clustered by clustering algorithm, obtains different cluster centres;
Test sample set of eigenvectors is identified using minimum Eustachian distance principle, realizes the diagnosis of GIS mechanical breakdowns.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810414184.4A CN108663202A (en) | 2018-05-03 | 2018-05-03 | GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810414184.4A CN108663202A (en) | 2018-05-03 | 2018-05-03 | GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108663202A true CN108663202A (en) | 2018-10-16 |
Family
ID=63781795
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810414184.4A Pending CN108663202A (en) | 2018-05-03 | 2018-05-03 | GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108663202A (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109376652A (en) * | 2018-10-24 | 2019-02-22 | 国网江苏省电力有限公司检修分公司 | Paralleling reactor of extra-high voltage method for diagnosing faults, apparatus and system based on artificial immune particle swarm clustering algorithm |
CN109856530A (en) * | 2018-12-25 | 2019-06-07 | 国网江苏省电力有限公司南京供电分公司 | A kind of load ratio bridging switch on-line monitoring method for diagnosing faults |
CN110059437A (en) * | 2019-04-28 | 2019-07-26 | 国网四川省电力公司南充供电公司 | A kind of GIS vibr ation signals extracting method based on variation mode decomposition |
CN110082082A (en) * | 2019-04-28 | 2019-08-02 | 国网四川省电力公司南充供电公司 | A kind of GIS state identification method based on vibration signal Principal Component Analysis |
CN110161125A (en) * | 2019-06-17 | 2019-08-23 | 哈尔滨工业大学 | The Aeroengine Smart monitoring method combined based on acceleration with sound emission cognition technology |
CN110688219A (en) * | 2019-09-05 | 2020-01-14 | 浙江理工大学 | Adaptive weight load balancing algorithm based on reverse chaotic cuckoo search |
CN110849626A (en) * | 2019-11-18 | 2020-02-28 | 东南大学 | Self-adaptive sparse compression self-coding rolling bearing fault diagnosis system |
CN113219824A (en) * | 2021-02-19 | 2021-08-06 | 哈尔滨工业大学 | Dynamic system control method based on variational modal decomposition |
CN114113984A (en) * | 2021-11-29 | 2022-03-01 | 平安壹账通云科技(深圳)有限公司 | Fault drilling method, device, terminal equipment and medium based on chaotic engineering |
US20230062548A1 (en) * | 2021-08-27 | 2023-03-02 | Hamilton Sundstrand Corporation | Online health monitoring and fault detection for high voltage dc distribution networks |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104318329A (en) * | 2014-10-20 | 2015-01-28 | 国家电网公司 | Power load forecasting method of cuckoo search algorithm improved support vector machine |
CN106443259A (en) * | 2016-09-29 | 2017-02-22 | 国网山东省电力公司电力科学研究院 | Transformer fault diagnosis new method based on Euclidean clustering and SPO-SVM |
-
2018
- 2018-05-03 CN CN201810414184.4A patent/CN108663202A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104318329A (en) * | 2014-10-20 | 2015-01-28 | 国家电网公司 | Power load forecasting method of cuckoo search algorithm improved support vector machine |
CN106443259A (en) * | 2016-09-29 | 2017-02-22 | 国网山东省电力公司电力科学研究院 | Transformer fault diagnosis new method based on Euclidean clustering and SPO-SVM |
Non-Patent Citations (1)
Title |
---|
侯焰: "基于异常振动分析的GIS机械故障诊断技术研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109376652A (en) * | 2018-10-24 | 2019-02-22 | 国网江苏省电力有限公司检修分公司 | Paralleling reactor of extra-high voltage method for diagnosing faults, apparatus and system based on artificial immune particle swarm clustering algorithm |
CN109856530B (en) * | 2018-12-25 | 2021-11-02 | 国网江苏省电力有限公司南京供电分公司 | On-load tap-changer on-line monitoring fault diagnosis method |
CN109856530A (en) * | 2018-12-25 | 2019-06-07 | 国网江苏省电力有限公司南京供电分公司 | A kind of load ratio bridging switch on-line monitoring method for diagnosing faults |
CN110059437A (en) * | 2019-04-28 | 2019-07-26 | 国网四川省电力公司南充供电公司 | A kind of GIS vibr ation signals extracting method based on variation mode decomposition |
CN110082082A (en) * | 2019-04-28 | 2019-08-02 | 国网四川省电力公司南充供电公司 | A kind of GIS state identification method based on vibration signal Principal Component Analysis |
CN110059437B (en) * | 2019-04-28 | 2023-12-01 | 国网四川省电力公司南充供电公司 | GIS vibration signal characteristic quantity extraction method based on variation modal decomposition |
CN110082082B (en) * | 2019-04-28 | 2020-11-17 | 国网四川省电力公司南充供电公司 | GIS state identification method based on vibration signal principal component analysis method |
CN110161125A (en) * | 2019-06-17 | 2019-08-23 | 哈尔滨工业大学 | The Aeroengine Smart monitoring method combined based on acceleration with sound emission cognition technology |
CN110688219A (en) * | 2019-09-05 | 2020-01-14 | 浙江理工大学 | Adaptive weight load balancing algorithm based on reverse chaotic cuckoo search |
CN110849626B (en) * | 2019-11-18 | 2022-06-07 | 东南大学 | Self-adaptive sparse compression self-coding rolling bearing fault diagnosis system |
CN110849626A (en) * | 2019-11-18 | 2020-02-28 | 东南大学 | Self-adaptive sparse compression self-coding rolling bearing fault diagnosis system |
CN113219824B (en) * | 2021-02-19 | 2022-01-07 | 哈尔滨工业大学 | Dynamic system control method based on variational modal decomposition |
CN113219824A (en) * | 2021-02-19 | 2021-08-06 | 哈尔滨工业大学 | Dynamic system control method based on variational modal decomposition |
US20230062548A1 (en) * | 2021-08-27 | 2023-03-02 | Hamilton Sundstrand Corporation | Online health monitoring and fault detection for high voltage dc distribution networks |
US11874318B2 (en) * | 2021-08-27 | 2024-01-16 | Hamilton Sundstrand Corporation | Online health monitoring and fault detection for high voltage DC distribution networks |
CN114113984A (en) * | 2021-11-29 | 2022-03-01 | 平安壹账通云科技(深圳)有限公司 | Fault drilling method, device, terminal equipment and medium based on chaotic engineering |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108663202A (en) | GIS mechanical failure diagnostic methods based on chaos cuckoo algorithm and system | |
CN112699913B (en) | Method and device for diagnosing abnormal relationship of household transformer in transformer area | |
CN110503245B (en) | Prediction method for large-area delay risk of airport flight | |
CN107132310B (en) | Transformer equipment health status method of discrimination based on gauss hybrid models | |
CN108399163A (en) | Bluebeard compound polymerize the text similarity measure with word combination semantic feature | |
CN109191922A (en) | A kind of extensive four-dimensional dynamic track prediction method and device | |
CN114021799A (en) | Day-ahead wind power prediction method and system for wind power plant | |
CN109684673B (en) | Feature extraction and cluster analysis method for transient stability result of power system | |
CN107579846B (en) | Cloud computing fault data detection method and system | |
CN106896219B (en) | The identification of transformer sub-health state and average remaining lifetime estimation method based on Gases Dissolved in Transformer Oil data | |
CN109273096A (en) | A kind of risk management grading evaluation method based on machine learning | |
CN108681660A (en) | A kind of non-coding RNA based on association rule mining and disease relationship prediction technique | |
CN108919067A (en) | A kind of recognition methods for GIS partial discharge mode | |
CN110535149A (en) | A kind of area Gong Biantai electric load three-phase imbalance prediction technique | |
CN109239669A (en) | A kind of self-evolution Radar Targets'Detection algorithm based on deep learning | |
CN111680875A (en) | Unmanned aerial vehicle state risk fuzzy comprehensive evaluation method based on probability baseline model | |
CN108810799B (en) | Multi-floor indoor positioning method and system based on linear discriminant analysis | |
CN115618261A (en) | Photovoltaic access power distribution network electric leakage identification method based on NCA and SSA-KELM | |
CN109921938A (en) | Fault detection method under a kind of cloud computing environment | |
CN103020733B (en) | Method and system for predicting single flight noise of airport based on weight | |
CN111863135B (en) | False positive structure variation filtering method, storage medium and computing device | |
CN108090623B (en) | Risk assessment method for power grid power failure accident | |
CN106251861A (en) | A kind of abnormal sound in public places detection method based on scene modeling | |
CN105046106A (en) | Protein subcellular localization and prediction method realized by using nearest-neighbor retrieval | |
Rousseeuw et al. | Monitoring system of phytoplankton blooms by using unsupervised classifier and time modeling |
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 |
Application publication date: 20181016 |
|
RJ01 | Rejection of invention patent application after publication |