CN112307950B - Detail feature quantity extraction and intelligent analysis method for GIS vibration defect identification - Google Patents
Detail feature quantity extraction and intelligent analysis method for GIS vibration defect identification Download PDFInfo
- Publication number
- CN112307950B CN112307950B CN202011181967.6A CN202011181967A CN112307950B CN 112307950 B CN112307950 B CN 112307950B CN 202011181967 A CN202011181967 A CN 202011181967A CN 112307950 B CN112307950 B CN 112307950B
- Authority
- CN
- China
- Prior art keywords
- vibration
- gis
- sample
- identification
- defect
- 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.)
- Active
Links
- 230000007547 defect Effects 0.000 title claims abstract description 85
- 238000000605 extraction Methods 0.000 title claims abstract description 50
- 238000004458 analytical method Methods 0.000 title claims abstract description 31
- 238000000034 method Methods 0.000 claims abstract description 33
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 26
- 230000002159 abnormal effect Effects 0.000 claims abstract description 25
- 238000004891 communication Methods 0.000 claims abstract description 20
- 230000005484 gravity Effects 0.000 claims abstract description 9
- 230000006870 function Effects 0.000 claims description 47
- 239000011159 matrix material Substances 0.000 claims description 31
- 238000012549 training Methods 0.000 claims description 24
- 238000004422 calculation algorithm Methods 0.000 claims description 21
- 238000003745 diagnosis Methods 0.000 claims description 17
- 239000013598 vector Substances 0.000 claims description 16
- 238000001514 detection method Methods 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 14
- 239000004020 conductor Substances 0.000 claims description 13
- 238000012423 maintenance Methods 0.000 claims description 12
- 238000001228 spectrum Methods 0.000 claims description 11
- 238000005070 sampling Methods 0.000 claims description 10
- 238000010276 construction Methods 0.000 claims description 8
- 239000003463 adsorbent Substances 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000005315 distribution function Methods 0.000 claims description 6
- 239000002808 molecular sieve Substances 0.000 claims description 6
- URGAHOPLAPQHLN-UHFFFAOYSA-N sodium aluminosilicate Chemical compound [Na+].[Al+3].[O-][Si]([O-])=O.[O-][Si]([O-])=O URGAHOPLAPQHLN-UHFFFAOYSA-N 0.000 claims description 6
- 230000004044 response Effects 0.000 claims description 5
- 238000011161 development Methods 0.000 claims description 4
- 238000005065 mining Methods 0.000 claims description 4
- 230000005540 biological transmission Effects 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000007621 cluster analysis Methods 0.000 claims description 3
- 230000001186 cumulative effect Effects 0.000 claims description 3
- 230000010354 integration Effects 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 238000003032 molecular docking Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 230000010363 phase shift Effects 0.000 claims description 3
- 230000001172 regenerating effect Effects 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000004088 simulation Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 230000035945 sensitivity Effects 0.000 claims description 2
- 230000004069 differentiation Effects 0.000 abstract description 3
- 230000008929 regeneration Effects 0.000 abstract description 2
- 238000011069 regeneration method Methods 0.000 abstract description 2
- 230000008901 benefit Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 6
- 238000011156 evaluation Methods 0.000 description 4
- 230000018109 developmental process Effects 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000009413 insulation Methods 0.000 description 2
- 230000002452 interceptive effect Effects 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 239000012212 insulator Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012847 principal component analysis method Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/12—Fingerprints or palmprints
- G06V40/1347—Preprocessing; Feature extraction
-
- 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
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Software Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Multimedia (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Medical Informatics (AREA)
- Human Computer Interaction (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention relates to a detail feature quantity extraction and intelligent analysis method for GIS vibration defect identification, and belongs to the field of informatization. The method comprises the following steps: s1: constructing a data acquisition and display control unit; s2: constructing a combined detail feature extraction unit; s3: constructing a GIS equipment mechanical state identification unit; s4: constructing an identification result output unit; s5: constructing a wireless communication and cloud service unit; s6: and constructing a database standard sample of the GIS equipment abnormal sound vibration mechanical defect intelligent identification system. The invention adopts the regeneration phase movement auxiliary empirical mode decomposition to obtain the intrinsic mode function of the vibration signal, then obtains the independent mode sequence of the vibration signal based on the frequency domain windowing method and the amplitude and kurtosis criterion, further solves the detail combination characteristics of singular value, kurtosis, variation coefficient, amplitude, gravity center frequency and the like of each mode, and can increase the differentiation degree of the vibration signals in different states.
Description
Technical Field
The invention belongs to the field of informatization, and relates to a detail feature quantity extraction and intelligent analysis method for GIS vibration defect identification.
Background
The gas-insulated switchgear (gas insulated switchgear, GIS) has the advantages of small occupied area, low maintenance cost and high reliability, and is widely applied to the power grid. It is counted that latent insulation defects or mechanical defects generated in the manufacturing, installation and operation processes of the GIS device are main causes of faults. In recent years, the phenomenon that GIS equipment generates abnormal sound vibration in the operation process is gradually increased, and particularly the GIS equipment with the voltage level of 220kV and above is adopted. The damage of mechanical abnormal sound vibration is very large, which can lead to bolt loosening, gas leakage, gas pressure drop, insulator damage and the like, and can cause insulation accidents in serious cases.
The vibration detection method has the advantages of no invasiveness, interference resistance, high sensitivity and the like. The vibration signal development detection, feature extraction and defect state identification aiming at the mechanical abnormal sound defect of the GIS equipment are critical to guaranteeing the safe and stable operation of the GIS equipment and even the power grid. The invention provides a GIS equipment mechanical defect diagnosis system based on vibration detection and a method thereof, which mainly adopts a characteristic comparison mode to detect states. The invention provides a GIS (geographic information system) mechanical defect diagnosis method based on strong correlation frequency points between every two samples. The invention discloses a GIS state identification method based on a vibration signal principal component analysis method, which is a method for carrying out compression reduction and decision recognition on composite characteristics of energy of time domain, frequency domain skewness, average value and the like and an inherent mode function, and mainly analyzes the whole signal. The invention patent provides a GIS equipment state evaluation method based on support vector description and K nearest neighbor algorithm, and provides a method for performing state evaluation based on vibration signal root mean square frequency and GIS internal air pressure characteristics. In summary, the following problems still exist in the aspect of the defect diagnosis and research of the mechanical vibration abnormal sound of GIS equipment at present:
1) In terms of feature extraction, characterization of detailed information of the high-frequency response mode of the defect vibration signal is lacking. The existing feature extraction method is mostly based on the simple frequency spectrum, energy and other information of the vibration signal, or the integral signal is subjected to empirical mode decomposition to obtain the intrinsic mode function of the signal, so that feature extraction is performed, and mining of more local and detailed information of the signal is lacking. On the other hand, the empirical mode decomposition algorithm based on noise assistance has the problems of residual noise, false modes and the like, and can greatly influence the recognition accuracy of the mechanical state of GIS equipment.
2) The system is lack of a software and hardware platform and an identification analysis system which integrate vibration signal detection, vibration state characteristic quantity extraction and mechanical defect intelligent diagnosis.
3) Existing analysis methods and models do not take into account that the operating load conditions of the actual test equipment are changing, and the vibration characteristics of the test sample also change with the change of the load. Therefore, the diagnosis model with similar voltage and current conditions is matched for identification according to the load condition corresponding to the detection time point of the actual GIS equipment, and the identification accuracy of the system is improved. On the other hand, the traditional analysis method is limited to evaluating vibration data of a certain type of defects, and does not have the capability of continuously learning and updating a sample database and an identification model of a diagnosis system.
Aiming at the defects, the invention provides a combined detail characteristic quantity extraction and intelligent analysis system for identifying abnormal sound vibration defects of GIS equipment. Compared with the existing vibration state feature extraction and vibration defect diagnosis method, the method has three advantages, namely, the method overcomes the problems of residual noise and false modes of a noise auxiliary empirical mode decomposition method, performs refined feature extraction and combined analysis on each independent vibration mode, and can increase the differentiation degree of vibration signals; and secondly, a software platform integrating vibration signal detection, vibration state characteristic quantity extraction and mechanical defect intelligent diagnosis is established, and interactive information input and visual result display can be carried out with a user. And thirdly, the matching of the identification model can be intelligently carried out according to the load voltage and current information, meanwhile, the diagnosis system is learned and updated based on the GIS equipment abnormal sound vibration defect identification cloud server, the database and the network communication interface, the generalization capability of the system is improved, the defect identification accuracy is further improved, and the engineering value is remarkable.
Disclosure of Invention
In view of the above, the present invention aims to provide a detailed feature quantity extraction and intelligent analysis method for identifying a vibration defect of a GIS. The system mainly comprises a data acquisition and waveform display unit, a combined detail characteristic extraction unit, a mechanical state identification unit, an evaluation result output unit and a wireless communication and cloud service unit. The device and the system can realize high-sensitivity acquisition of the vibration signals of the GIS equipment, extraction of the detailed combination characteristics of the vibration mode response, intelligent diagnosis of the mechanical defect state and result display, and have the functions of self-learning and cloud updating of the defect identification model and the database. The GIS equipment abnormal sound vibration defect identification combined detail characteristic quantity extraction and intelligent analysis system is based on an integrated software and hardware platform, has the advantages of high efficiency and accuracy in identification and strong engineering applicability.
In order to achieve the above purpose, the present invention provides the following technical solutions:
A detail feature quantity extraction and intelligent analysis method for GIS vibration defect identification comprises the following steps:
s1: constructing a data acquisition and display control unit;
S2: constructing a combined detail feature extraction unit;
s3: constructing a GIS equipment mechanical state identification unit;
s4: constructing an identification result output unit;
s5: constructing a wireless communication and cloud service unit;
s6: and constructing a database standard sample of the GIS equipment abnormal sound vibration mechanical defect intelligent identification system.
Optionally, the S1 specifically is:
displaying the time domain and the frequency domain of the vibration signals collected on site and the collected historical data;
Recording GIS equipment operation load information of the acquired signals, wherein the GIS equipment operation load information comprises voltage and current information;
And carrying out instruction input and control on the acquisition signals, wherein the instruction input and control comprises whether sampling data stores an instruction, whether the instruction is continuously acquired, an instruction for starting and stopping an acquisition process and setting of sampling time and sampling rate.
Optionally, the S2 specifically is:
In the model training process, extracting features of a GIS equipment mechanical state database built in the system, and establishing a fingerprint feature database; extracting a combined detail feature matrix from a GIS equipment vibration sample to be evaluated;
The combined detail feature extraction unit acquires an intrinsic mode function IMFs of an acquired vibration signal sample by using a regenerative phase shift auxiliary empirical mode decomposition algorithm RPSEMD, and further acquires a training sample set source domain feature matrix by using an independent mode combined feature construction method of frequency domain windowing, and the method comprises the following steps of:
(1) Regenerated phase-shifted sine wave assisted empirical mode decomposition
Firstly, adopting an automatically generated sine wave as an analysis signal, as shown in a formula (1); wherein k corresponds to the decomposition extracted kth IMF modality component c k (t); and c k,fk and θ k are amplitude, frequency and phase, respectively;
sk(t|ak,fk,θk)＝ak cos(2πfkt+θk) (1)
Then, designing s k (t) based on IM cluster analysis and modal aliasing criteria;
Finally, changing the position of the extreme points by shifting s k (t) by θ k not only helps to preserve more detail of the independent IMs, but also ensures that the auxiliary signal s k (t) is completely cancelled out in the final result;
(2) Independent mode combination feature extraction based on frequency domain windowing
According to the independent mode combination characteristic extraction method of frequency domain windowing, RPSEMD decomposition is adopted to perform mode decomposition on a vibration signal x (t) to generate a mode component IMFs, the IMFs are converted into a frequency domain by fast Fourier transformation, and k frequency domain windowing intervals are set according to the frequency response characteristics of each mode;
Then sequentially executing kurtosis and amplitude criteria on the modal components of each frequency windowing interval, removing the modal components which do not meet the kurtosis criteria and the amplitude criteria, and simultaneously comparing and selecting main modal components of each frequency domain windowing interval;
finally, the combined feature matrix measures the amplitude, attenuation characteristic, pulse characteristic and marginal spectrum distribution characteristic of the signal respectively, performs marginal spectrum conversion and singular value, variation coefficient, amplitude, kurtosis and spectrum gravity center feature calculation on the k obtained independent modal components, and further constructs the combined feature matrix, wherein the calculation method of each feature quantity is as follows:
Singular values
Coefficient of variation
Kurtosis of
Center of gravity of spectrum
Amplitude a m＝max(xi), i=1..n.
Optionally, the S3 specifically is:
Firstly, a mechanical state identification unit performs model training based on a fingerprint feature database of different loads of GIS mechanical state, poor contact defect of a disconnecting switch, bolt loosening defect of a long conductor base, bolt loosening defect of a molecular sieve adsorbent tray and fatigue loosening defect of a disconnecting switch spring and a multi-core improved multi-classification related vector machine mRVMs algorithm of a mechanical state sample, and generates a series of identification models of related load types aiming at vibration samples of different voltage and current levels;
Then, extracting a combined detail feature matrix by adopting the same flow from an abnormal sound vibration sample of GIS equipment to be evaluated, and matching the most relevant mechanical state identification model according to the recorded voltage and current information of the combined detail feature matrix, so that the effective identification of the mechanical state of the sample GIS equipment is realized;
the specific detail principle of the multi-core improved multi-classification related vector machine algorithm is as follows:
(1) Constructing a multi-core feature mapping kernel function
Assuming a training sample set with a source domain space size of N, and a feature dimension of D; selecting a linear kernel function K line and a Gaussian radial kernel function K rbf to establish a kernel function matrix, and carrying out weighted summation on each term K m according to the Mercer theorem to construct a multi-kernel function K (,) specifically expressed as:
Wherein 1 is greater than or equal to beta m is greater than or equal to 0 and is the weight of an mth basis function K m (,) and
(2) Multi-core mRVMs algorithm
Given source domain sample feature vector and class data setWherein x ε R N×D, t ε {1,2,3 …, C }; the training set adopts a multi-core function K, and each row/>, of the core K m∈RN×D Representing a relationship between the observed value n based on the selected mth kernel function and other observed values of the training set, and/>
The learning process includes the inference of model parameters W ε R N×C, while the magnitude of the W T K value represents the correlation between the sample and the data, acting like a voting system, so that the model has the appropriate discriminative properties;
multi-layer discrimination is realized by introducing an auxiliary variable Y epsilon R N×C, Y is a regression target of W T K, and the distribution accords with a standard noise model The auxiliary variables are endowed with independent standardized Gaussian probability distribution so as to ensure statistical identifiability and realize closed iteration reasoning; the regression factor W represents the weight of the data point as a specific class "vote", the auxiliary variable Y represents a class membership ordering system, and given a sample n, it is classified as class c according to its highest Y value Y cn; the continuity of Y not only allows for multi-class discrimination by polynomial probability likelihood functions, i.e./>At the same time, a probability output is provided for the members of each layer;
wherein: u to N (0, 1); phi is a Gaussian cumulative distribution function;
setting the weight W to obey the prior distribution of standard normalization To ensure sparsity of the model, wherein alpha nc belongs to a scale matrix A epsilon R N×C and obeys gamma distribution with super parameters gamma and v; setting gamma and v at smaller values so as to limit most of weights W to be near 0 values, and further obtaining few non-0 related vectors (RELEVANCE VECTORS, RV) to form thin fluffs of the model; the posterior probability of the weight W is as follows:
Wherein A c is the diagonal matrix derived from column c of scale matrix A; the maximum a posteriori probability (MAP) estimate of the regression factor W is: giving a class c, and updating the class parameters according to the MAP value and the formula (6);
e-step form of the auxiliary variable Y is obtained according to formula (6) for The method comprises the following steps:
For the i-th layer, i.e. c=i,
Using given super parameters gamma and v and gamma distribution function, to average valueAnd (3) estimating:
The training process updates the model parameters in equations (6) - (9) until convergence.
Optionally, the S4 specifically is:
the identification result output and display unit outputs and displays the identification analysis result of the sample in a probabilistic way; the unit provides the probability that the sample belongs to each state, and sets warning according to the defect type with the largest membership probability, and provides related overhaul suggestions to facilitate development of further operation and maintenance work;
The identification result output unit provides a sample waveform and a combined characteristic query function, outputs and displays time domain and frequency domain waveforms of original signals of the detection samples and vibration mode signals decomposed by RPSEMD, and provides amplitude and gravity center frequency characteristic information of the original signals.
Optionally, the step S5 specifically includes:
The wireless communication and cloud service unit supports various communication protocols and communication protocol standard access modes, including NBIOT, WIFI,2G-5G and Bluetooth, integrally conforms to IEC standards, and the data exchange function provides data access and shared uploading interfaces for multiple data sources and multiple data formats, so that further data utilization and information mining can be realized;
The wireless communication and cloud service unit relies on the mass real-time data transmission and screening integration capability of the PaaS-level cloud computing platform to establish a real-time centralized management and control platform of remote data for GIS equipment operation and maintenance related users; the user realizes real-time maintenance of the system through network docking with the cloud server, and updates the GIS mechanical state database and the diagnosis algorithm scheme in real time, so that the state type and the data quantity of the sample database are continuously expanded, and the accuracy of the diagnosis algorithm model is improved.
Optionally, the step S6 specifically includes:
The vibration simulation platform of the laboratory GIS equipment is adopted to simulate 5 mechanical vibration working conditions of the GIS equipment under different load currents in the range of 0-3000A, namely, poor contact defect of the isolating switch, loosening defect of bolts of the long conductor base, loosening defect of bolts of the molecular sieve adsorbent tray and fatigue loosening defect of the isolating switch spring, and meanwhile, the high-sensitivity GIS equipment abnormal sound vibration detection device and the defect identification system perform training and learning work of the data acquisition, combination feature extraction and mechanical state identification system.
The invention has the beneficial effects that:
(1) The invention adopts the regeneration phase movement auxiliary empirical mode decomposition to obtain the intrinsic mode function of the vibration signal, then obtains the independent mode sequence of the vibration signal based on the frequency domain windowing method and the amplitude and kurtosis criterion, further solves the detail combination characteristics of singular value, kurtosis, variation coefficient, amplitude, gravity center frequency and the like of each mode, and can increase the differentiation degree of the vibration signals in different states.
(2) According to the invention, the GIS equipment mechanical defect identification model is established by combining the GIS equipment vibration samples with different voltage levels and multiple mechanical states under load current and the multi-core improved multi-classification related vector machine, a user can carry out interactive input aiming at the load condition and the voltage level of the field test equipment, then the identification model corresponding to the current and voltage conditions is matched, and the adaptability of the model to the samples is improved. Thirdly, the invention provides the integrated operation of displaying and outputting the signal detection and identification result, and is configured with a plurality of communication modes of Bluetooth, wiFi and 4G/5G, so that the sample database and the identification model of the identification system can be uploaded, updated and learned based on the cloud server according to actual requirements, and the diagnosis capability of the model on complex defects is improved.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objects and other advantages of the invention may be realized and obtained by means of the instrumentalities and combinations particularly pointed out in the specification.
Drawings
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in the following preferred detail with reference to the accompanying drawings, in which:
FIG. 1 is a system frame diagram of the present invention;
FIG. 2 is a flow chart of the present invention;
FIG. 3 is a flow chart of a regenerated phase-shifted sine wave assisted empirical mode decomposition;
FIG. 4 is a flow chart of independent mode combination feature extraction based on frequency domain windowing;
FIG. 5 is a flow chart for identifying the mechanical state of GIS equipment;
FIG. 6 is a diagram of a multi-core mRVMs model;
Fig. 7 is a schematic diagram of a GIS device abnormal sound vibration cloud server function;
FIG. 8 is a flowchart of the establishment of a GIS equipment abnormal sound mechanical defect identification system;
Fig. 9 is a vibration signal spectrum of poor contact (defect 1) and loosening of the long conductor contact (defect 2) of the isolating switch;
FIG. 10 is a diagram showing the output of the GIS equipment abnormal sound mechanical defect identification system; (a) loosening the long conductor contact base; (b) is a disconnector contact failure.
Detailed Description
Other advantages and effects of the present invention will become apparent to those skilled in the art from the following disclosure, which describes the embodiments of the present invention with reference to specific examples. The invention may be practiced or carried out in other embodiments that depart from the specific details, and the details of the present description may be modified or varied from the spirit and scope of the present invention. It should be noted that the illustrations provided in the following embodiments merely illustrate the basic idea of the present invention by way of illustration, and the following embodiments and features in the embodiments may be combined with each other without conflict.
Wherein the drawings are for illustrative purposes only and are shown in schematic, non-physical, and not intended to limit the invention; for the purpose of better illustrating embodiments of the invention, certain elements of the drawings may be omitted, enlarged or reduced and do not represent the size of the actual product; it will be appreciated by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.
The same or similar reference numbers in the drawings of embodiments of the invention correspond to the same or similar components; in the description of the present invention, it should be understood that, if there are terms such as "upper", "lower", "left", "right", "front", "rear", etc., that indicate an azimuth or a positional relationship based on the azimuth or the positional relationship shown in the drawings, it is only for convenience of describing the present invention and simplifying the description, but not for indicating or suggesting that the referred device or element must have a specific azimuth, be constructed and operated in a specific azimuth, so that the terms describing the positional relationship in the drawings are merely for exemplary illustration and should not be construed as limiting the present invention, and that the specific meaning of the above terms may be understood by those of ordinary skill in the art according to the specific circumstances.
The combined detail characteristic quantity extraction and intelligent analysis system for identifying abnormal sound vibration defects of GIS equipment is shown in figure 1. The system mainly comprises a data acquisition and waveform display unit of vibration signals, a combination detail characteristic extraction unit, a mechanical state identification unit, an evaluation result output unit and a wireless communication and cloud service unit.
The specific steps of the implementation are shown in fig. 2.
3.1 Construction of data acquisition and display control Unit
The main functions of the data acquisition and display control unit are as follows: firstly, displaying a time domain and a frequency domain of vibration signals collected on site and collected historical data; and secondly, recording GIS equipment operation load information of the acquired signals, wherein the GIS equipment operation load information comprises voltage and current information, and if no current information exists, recording is not needed. And thirdly, inputting and controlling the acquisition signals, wherein the instruction input and control comprises whether sampling data store an instruction, whether the instruction is continuously acquired, an instruction for starting and stopping the acquisition process, setting the sampling time and the sampling rate and the like.
3.2 Construction of the Combined detail feature extraction Unit
The combined detail feature extraction unit is composed of two aspect functions. Firstly, in the model training process, extracting features of a GIS equipment mechanical state database built in a system, and establishing a fingerprint feature database; and secondly, extracting a combined detail feature matrix from the GIS equipment vibration sample to be evaluated.
The combined detail feature extraction unit acquires an intrinsic mode function (IMFs, INTRINSIC MODE FUNCTION COMPONENTS) of the acquired vibration signal sample by using a regenerative phase shift auxiliary empirical mode decomposition algorithm (REGENERATED PHASE-shifted sinusoid-ASSISTED EMPIRICAL mode decomposition, RPSEMD), and further acquires a training sample set source domain feature matrix by adopting a frequency domain windowing independent mode combined feature construction method, and the specific details are as follows:
(1) Regenerated phase-shifted sine wave assisted empirical mode decomposition
The regenerated phase-shifted sine wave assisted empirical mode decomposition (REGENERATED PHASE-shifted sinusoid-ASSISTED EMPIRICAL mode decomposition, RPSEMD) is mainly composed of three steps. First, an automatically generated sine wave is employed as an analysis signal, as shown in formula (1). Where k corresponds to the decomposition extracted kth IMF modality component c k (t). And c k,fk and θ k are amplitude, frequency and phase, respectively.
sk(t|ak,fk,θk)＝akcos(2πfkt+θk) (1)
Then, s k(t)[21] is designed based on IM cluster analysis and modal aliasing criteria. This proactive approach can ensure RPSEMD certainty and more effectively solve the MMP problem.
Finally, changing the position of the extreme points by shifting s k (t) by θ k not only helps to preserve more detail of the individual IMs, but also ensures that the auxiliary signal s k (t) is completely cancelled out in the final result. The overall flow of RPSEMD algorithm is shown in figure 3.
(2) Independent mode combination feature extraction based on frequency domain windowing
According to the independent mode combination characteristic extraction method of frequency domain windowing, RPSEMD decomposition is adopted to perform mode decomposition on a vibration signal x (t) to generate a mode component IMFs, the IMFs are converted into a frequency domain by fast Fourier transformation, and k frequency domain windowing intervals are set according to frequency response characteristics of each mode.
And then sequentially executing kurtosis and amplitude criteria on the modal components of each frequency windowing interval, removing the modal components which do not meet the kurtosis criteria and the amplitude criteria, and simultaneously comparing and selecting the main modal components of each frequency domain windowing interval.
Finally, the combined feature matrix is measured from the characteristics of amplitude, attenuation characteristic, pulse characteristic, marginal spectrum distribution and the like of the signal, the acquired k independent modal components are subjected to marginal spectrum conversion and calculation of characteristics such as singular values, variation coefficients, amplitude values, kurtosis, spectrum center of gravity and the like, and then the combined feature matrix is constructed, and the calculation method of each characteristic is shown in table 1.
Table 1 method for obtaining combined feature quantity
3.3 Construction of GIS device mechanical State identification Unit
The mechanical state identification unit is mainly used for carrying out state classification and identification on vibration signal samples of unknown state GIS equipment, and mainly comprises two steps:
Firstly, a mechanical state identification unit performs model training based on fingerprint feature databases of samples of different loads and mechanical states such as GIS normal operation of GIS equipment mechanical states, poor contact defects of a disconnecting switch, loosening defects of bolts of a long conductor base, loosening defects of bolts of a molecular sieve adsorbent tray, fatigue loosening defects of a disconnecting switch spring and the like and a multi-core improved multi-classification related vector machine (mRVMs) algorithm, and generates a series of identification models of related load types aiming at vibration samples of different voltage and current levels.
And then, extracting a combined detail feature matrix by adopting the same flow for an abnormal sound vibration sample of the GIS equipment to be evaluated, and matching the most relevant mechanical state identification model according to the recorded voltage and current information, so that the effective identification of the mechanical state of the sample GIS equipment is realized.
FIG. 4 is a flow chart of independent mode combination feature extraction based on frequency domain windowing;
FIG. 5 is a flow chart for identifying the mechanical state of GIS equipment;
the specific detail principle of the multi-core improved multi-classification related vector machine algorithm is as follows:
(1) Constructing a multi-core feature mapping kernel function
Let the training sample set with a source domain space size N and the feature dimension D. Selecting a linear kernel function K line and a Gaussian radial kernel function K rbf to establish a kernel function matrix, and carrying out weighted summation on each term K m according to the Mercer theorem to construct a multi-kernel function K (,) specifically expressed as:
Wherein 1 is greater than or equal to beta m is greater than or equal to 0 and is the weight of an mth basis function K m (,) and
(2) Multi-core mRVMs algorithm
The model of mRVMs modified with the multi-core function K (,) is shown in FIG. 6. Given source domain sample feature vector and class data setWhere x ε R N×D, t ε {1,2,3 …, C }. The training set adopts a multi-core function K, and each row of the core K m∈RN×D Representing a relationship between the observed value n based on the selected mth kernel function and other observed values of the training set, an
The learning process includes an inference of the model parameter W ε R N×C, while the magnitude of the W T K value represents the correlation between the sample and the data, acting like a voting system, so that the model has the appropriate discriminative properties.
Multi-layer discrimination is realized by introducing an auxiliary variable Y epsilon R N×C, Y is a regression target of W T K, and the distribution accords with a standard noise modelThe auxiliary variables are given independent standardized Gaussian probability distribution to ensure statistical identifiability and realize closed iteration reasoning. The regression factor W represents the weight of the data point for a particular class "vote", the auxiliary variable Y represents a class membership ranking system, and given a sample n, it is classified as class c based on its highest Y value Y cn. The continuity of Y not only allows for multi-class discrimination by polynomial probability likelihood functions, i.e./>While providing a probability output for the members of each layer.
Wherein: u to N (0, 1); Φ is a gaussian cumulative distribution function.
Setting the weight W to obey the prior distribution of standard normalizationTo ensure sparsity of the model, wherein alpha nc belongs to the scale matrix A epsilon R N×C and obeys gamma distribution with super parameters gamma and v. By setting γ and ν to small values, most of the weights W are limited to be around 0, so that few non-0 correlation vectors (RELEVANCE VECTORS, RV) are obtained to form a lean solution of the model. The posterior probability of the weight W is as follows:
Where A c is the diagonal matrix derived from column c of scale matrix A. The maximum a posteriori probability (MAP) estimate of the regression factor W is: thus, given a class c, the class parameters are updated according to equation (6) based on the MAP value.
E-step form of the auxiliary variable Y can be derived from the equation (), forThe method can obtain the following steps:
For the i-th layer, i.e. c=i,
At the same time, the given super parameters gamma and v and gamma distribution function can be used for the average valueAnd (3) estimating:
The training process updates the model parameters in equations (6) - (9) until convergence.
3.4 Construction of the identification result output Unit
The identification result output and display unit outputs and displays the identification analysis result of the sample in a probabilistic way. The unit provides the probability that the sample belongs to each state, and sets warning according to the defect type with the largest membership probability, and provides related overhaul suggestions to facilitate development of further operation and maintenance work.
The identification result output unit provides a sample waveform and a combined characteristic query function, can output and display time domain and frequency domain waveforms of original signals of the detection sample and vibration mode signals decomposed by RPSEMD, and provides characteristic information such as amplitude, gravity center frequency and the like.
3.5 Wireless communication and cloud service Unit construction
The wireless communication and cloud service unit supports various communication protocols and communication protocol standard access modes, including NBIOT, WIFI,2G-5G, bluetooth and the like, integrally conforms to IEC standards, and has a data exchange function for providing data access and sharing uploading interfaces for multiple data sources and multiple data formats, thereby being beneficial to realizing further data utilization and information mining.
The wireless communication and cloud service unit relies on the mass real-time data transmission and screening integration capability of the PaaS-level cloud computing platform to establish a real-time centralized management and control platform for remote data for users related to GIS equipment operation and maintenance. The user can realize the real-time maintenance of the system by the network docking with the cloud server, and can update the GIS mechanical state database and the diagnosis algorithm scheme in real time, so that the state type and the data volume of the sample database are continuously expanded, and the accuracy of the diagnosis algorithm model is improved.
3.6 Constructing standard sample of database of GIS equipment abnormal sound vibration mechanical defect intelligent identification system
Fig. 7 is a schematic diagram of a GIS device abnormal sound vibration cloud server function; the vibration simulation platform of the laboratory GIS equipment is adopted to simulate 5 mechanical vibration working conditions of the GIS equipment under different load currents in the range of 0-3000A, namely, poor contact defect of the isolating switch, loosening defect of bolts of the long conductor base, loosening defect of bolts of the molecular sieve adsorbent tray and fatigue loosening defect of the isolating switch spring, and meanwhile, the high-sensitivity GIS equipment abnormal sound vibration detection device and the defect identification system perform training and learning work of the data acquisition, combination feature extraction and mechanical state identification system. The specific flow is shown in fig. 8.
3.7 Case analysis and verification
The GIS equipment abnormal sound vibration defect identification combined detail characteristic quantity extraction and intelligent analysis system can analyze and identify the mechanical states of the GIS equipment under different loads. The same sampling rate and time setting are adopted to collect 2400A and 73kV load vibration samples of GIS equipment with two mechanical defects of poor contact (defect 1) and loose contact (defect 2) of a long conductor of a disconnecting switch respectively, and the obtained signal time domain and frequency domain diagrams are shown in figure 9.
Carrying out RPSEMD decomposition and frequency domain windowing on the measured GIS equipment mechanical vibration sample to extract the combination detail characteristics, and obtaining a corresponding combination detail characteristic matrix as shown in table 2:
TABLE 2 combination of detail characteristics of contact failure and loosening defects of long conductor contact
The system automatically brings the generated combined detail characteristic matrix into a GIS equipment abnormal sound mechanical defect identification system, and can automatically identify the mechanical state type. And the identification result and the membership probability of each state are sent to a display and output unit, and are displayed in a combined interface with the modal result and the overhaul advice obtained by RPSEMD decomposition, and the result is shown in fig. 10. (a) loosening the long conductor contact base; GIS mechanical state identification result: the connection piece 70.18% of the long conductor contacting the base has a looseness defect, the operation and maintenance department is recommended to pay important attention, and the fault removal work is rapidly adopted if the abnormality is serious; (b) is a disconnector contact failure. GIS mechanical state identification result: the contact defect exists in 96.78% probability of the contact finger of the isolating switch, the operation and maintenance department is recommended to pay important attention to the fact that whether the contact finger is caused by loosening of an operating mechanism or not is checked, and hidden dangers are eliminated timely.
The identification result output from the detection sample shows that the probability of loosening of the long conductor contact seat is 70.18% in the first defect, the probability of poor contact defect of the isolating switch is 96.78% in the second defect, and the first defect is consistent with the actual defect type. Therefore, the accuracy and the good practicability of the combined detail characteristic quantity extraction and intelligent analysis system for identifying abnormal sound vibration defects of the GIS equipment are verified.
Finally, it is noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications and equivalents may be made thereto without departing from the spirit and scope of the present invention, which is intended to be covered by the claims of the present invention.
Claims (5)
- A detail feature quantity extraction and intelligent analysis method for GIS vibration defect identification is characterized in that: the method comprises the following steps:s1: constructing a data acquisition and display control unit;S2: constructing a combined detail feature extraction unit;s3: constructing a GIS equipment mechanical state identification unit;s4: constructing an identification result output unit;s5: constructing a wireless communication and cloud service unit;s6: constructing a database standard sample of the intelligent identification system of abnormal sound vibration mechanical defects of GIS equipment;The step S2 is specifically as follows:In the model training process, extracting features of a GIS equipment mechanical state database built in the system, and establishing a fingerprint feature database; extracting a combined detail feature matrix from a GIS equipment vibration sample to be evaluated;The combined detail feature extraction unit acquires an intrinsic mode function IMFs of an acquired vibration signal sample by using a regenerative phase shift auxiliary empirical mode decomposition algorithm RPSEMD, and further acquires a training sample set source domain feature matrix by using an independent mode combined feature construction method of frequency domain windowing, and the method comprises the following steps of:(1) Regenerated phase-shifted sine wave assisted empirical mode decompositionFirstly, adopting an automatically generated sine wave as an analysis signal, as shown in a formula (1); wherein k corresponds to the decomposition extracted kth IMF modality component c k (t); and c k,fk and θ k are amplitude, frequency and phase, respectively;sk(tak,fk,θk)＝akcos(2πfkt+θk)(1)Then, designing s k (t) based on IM cluster analysis and modal aliasing criteria;Finally, changing the position of the extreme points by shifting s k (t) by θ k not only helps to preserve more detail of the independent IMs, but also ensures that the auxiliary signal s k (t) is completely cancelled out in the final result;(2) Independent mode combination feature extraction based on frequency domain windowingAccording to the independent mode combination characteristic extraction method of frequency domain windowing, RPSEMD decomposition is adopted to perform mode decomposition on a vibration signal x (t) to generate a mode component IMFs, the IMFs are converted into a frequency domain by fast Fourier transformation, and k frequency domain windowing intervals are set according to the frequency response characteristics of each mode;Then sequentially executing kurtosis and amplitude criteria on the modal components of each frequency windowing interval, removing the modal components which do not meet the kurtosis criteria and the amplitude criteria, and simultaneously comparing and selecting main modal components of each frequency domain windowing interval;finally, the combined feature matrix measures the amplitude, attenuation characteristic, pulse characteristic and marginal spectrum distribution characteristic of the signal respectively, performs marginal spectrum conversion and singular value, variation coefficient, amplitude, kurtosis and spectrum gravity center feature calculation on the k obtained independent modal components, and further constructs the combined feature matrix, wherein the calculation method of each feature quantity is as follows:The step S3 is specifically as follows:Firstly, a mechanical state identification unit performs model training based on a fingerprint feature database of different loads of GIS mechanical state, poor contact defect of a disconnecting switch, bolt loosening defect of a long conductor base, bolt loosening defect of a molecular sieve adsorbent tray and fatigue loosening defect of a disconnecting switch spring and a multi-core improved multi-classification related vector machine mRVMs algorithm of a mechanical state sample, and generates a series of identification models of related load types aiming at vibration samples of different voltage and current levels;Then, extracting a combined detail feature matrix by adopting the same flow from an abnormal sound vibration sample of GIS equipment to be evaluated, and matching the most relevant mechanical state identification model according to the recorded voltage and current information of the combined detail feature matrix, so that the effective identification of the mechanical state of the sample GIS equipment is realized;the specific detail principle of the multi-core improved multi-classification related vector machine algorithm is as follows:(1) Constructing a multi-core feature mapping kernel functionAssuming a training sample set with a source domain space size of N, and a feature dimension of D; selecting a linear kernel function K line and a Gaussian radial kernel function K rbf to establish a kernel function matrix, and carrying out weighted summation on each term K m according to the Mercer theorem to construct a multi-kernel function K (x i,xj), wherein the multi-kernel function K is specifically expressed as:Wherein 1 is greater than or equal to beta m is greater than or equal to 0 and is the weight of an mth basis function K m (,) and(2) Multi-core mRVMs algorithmGiven source domain sample feature vector and class data setWherein x ε R N×D, t ε {1,2,3 …, C }; the training set adopts a multi-core function K, and each row/>, of the core K m∈RN×D Representing a relationship between the observed value n based on the selected mth kernel function and other observed values of the training set, and/>The learning process includes the inference of model parameters W ε R N×C, while the magnitude of the W T K value represents the correlation between the sample and the data, acting like a voting system, so that the model has the appropriate discriminative properties;multi-layer discrimination is realized by introducing an auxiliary variable Y epsilon R N×C, Y is a regression target of W T K, and the distribution accords with a standard noise model The auxiliary variables are endowed with independent standardized Gaussian probability distribution so as to ensure statistical identifiability and realize closed iteration reasoning; the regression factor W represents the weight of the data point as a specific class "vote", the auxiliary variable Y represents a class membership ordering system, and given a sample n, it is classified as class c according to its highest Y value Y cn; the continuity of Y not only allows for multi-class discrimination by polynomial probability likelihood functions, i.e./>At the same time, a probability output is provided for the members of each layer;wherein: u to N (0, 1); phi is a Gaussian cumulative distribution function;setting the weight W to obey the prior distribution of standard normalization To ensure sparsity of the model, wherein alpha nc belongs to a scale matrix A epsilon R N×C and obeys gamma distribution with super parameters gamma and v; setting gamma and v at smaller values so as to limit most of weights W to be near 0 values, and further obtaining few non-0 related vectors (RelevanceVectors, RV) to form thin fluffs of the model; the posterior probability of the weight W is as follows:Wherein A c is the diagonal matrix derived from column c of scale matrix A; the maximum a posteriori probability (MAP) estimate of the regression factor W is: giving a class c, and updating the class parameters according to the MAP value and the formula (6);e-step form of the auxiliary variable Y is obtained according to formula (6) for The method comprises the following steps:For the i-th layer, i.e. c=i,Using given super parameters gamma and v and gamma distribution function, to average valueAnd (3) estimating:The training process updates the model parameters in equations (6) - (9) until convergence.
- 2. The detailed feature quantity extraction and intelligent analysis method for identifying GIS vibration defects according to claim 1, wherein the detailed feature quantity extraction and intelligent analysis method is characterized in that: the S1 specifically comprises the following steps:displaying the time domain and the frequency domain of the vibration signals collected on site and the collected historical data;Recording GIS equipment operation load information of the acquired signals, wherein the GIS equipment operation load information comprises voltage and current information;And carrying out instruction input and control on the acquisition signals, wherein the instruction input and control comprises whether sampling data stores an instruction, whether the instruction is continuously acquired, an instruction for starting and stopping an acquisition process and setting of sampling time and sampling rate.
- 3. The detailed feature quantity extraction and intelligent analysis method for identifying GIS vibration defects according to claim 1, wherein the detailed feature quantity extraction and intelligent analysis method is characterized in that: the step S4 specifically comprises the following steps:The identification result output unit outputs and displays the identification analysis result of the sample in a probabilistic way; the unit provides the probability that the sample belongs to each state, and sets warning according to the defect type with the largest membership probability, and provides related overhaul suggestions to facilitate development of further operation and maintenance work;The identification result output unit provides a sample waveform and a combined characteristic query function, outputs and displays time domain and frequency domain waveforms of original signals of the detection samples and vibration mode signals decomposed by RPSEMD, and provides amplitude and gravity center frequency characteristic information of the original signals.
- 4. The detailed feature quantity extraction and intelligent analysis method for identifying GIS vibration defects according to claim 1, wherein the detailed feature quantity extraction and intelligent analysis method is characterized in that: the step S5 specifically comprises the following steps:The wireless communication and cloud service unit supports various communication protocols and communication protocol standard access modes, including NBIOT, WIFI,2G-5G and Bluetooth, integrally conforms to IEC standards, and the data exchange function provides data access and shared uploading interfaces for multiple data sources and multiple data formats, so that further data utilization and information mining can be realized;The wireless communication and cloud service unit relies on the mass real-time data transmission and screening integration capability of the PaaS-level cloud computing platform to establish a real-time centralized management and control platform of remote data for GIS equipment operation and maintenance related users; the user realizes real-time maintenance of the system through network docking with the cloud server, and updates the GIS mechanical state database and the diagnosis algorithm scheme in real time, so that the state type and the data quantity of the sample database are continuously expanded, and the accuracy of the diagnosis algorithm model is improved.
- 5. The detailed feature quantity extraction and intelligent analysis method for identifying GIS vibration defects according to claim 1, wherein the detailed feature quantity extraction and intelligent analysis method is characterized in that: the step S6 specifically comprises the following steps:The vibration simulation platform of the GIS equipment in a laboratory is adopted to simulate the vibration working conditions of 5 mechanical states of GIS equipment normal operation, poor contact defect of a disconnecting switch, loosening defect of a bolt of a long conductor base, loosening defect of a bolt of a molecular sieve adsorbent tray and fatigue loosening defect of a disconnecting switch spring under different load currents in the range of 0-3000A, and meanwhile, the abnormal sound vibration detection device of the GIS equipment and a defect identification system with high sensitivity are used for carrying out data acquisition, combined feature extraction and training and learning of the mechanical state identification system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011181967.6A CN112307950B (en) | 2020-10-29 | 2020-10-29 | Detail feature quantity extraction and intelligent analysis method for GIS vibration defect identification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011181967.6A CN112307950B (en) | 2020-10-29 | 2020-10-29 | Detail feature quantity extraction and intelligent analysis method for GIS vibration defect identification |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112307950A CN112307950A (en) | 2021-02-02 |
| CN112307950B true CN112307950B (en) | 2024-04-23 |
Family
ID=74331600
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011181967.6A Active CN112307950B (en) | 2020-10-29 | 2020-10-29 | Detail feature quantity extraction and intelligent analysis method for GIS vibration defect identification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112307950B (en) |
Families Citing this family (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112964306A (en) * | 2021-03-17 | 2021-06-15 | 中国电力科学研究院有限公司 | Device and method for wirelessly monitoring vibration and load change of GIS (gas insulated switchgear) |
| CN113110386A (en) * | 2021-04-19 | 2021-07-13 | 重庆大学 | GIS/GIL equipment mechanical vibration state on-line monitoring and mechanical fault cloud diagnosis system |
| CN113654647A (en) * | 2021-06-29 | 2021-11-16 | 国网江苏省电力有限公司电力科学研究院 | A non-contact GIL vibration detection method and device |
| CN113567130A (en) * | 2021-07-28 | 2021-10-29 | 江南大学 | Bearing fault diagnosis method based on multiple working conditions of equipment |
| CN114047410B (en) * | 2021-08-25 | 2025-01-28 | 国网电力科学研究院武汉南瑞有限责任公司 | A GIS/GIL equipment mechanical vibration simulation system and method |
| CN114114001B (en) * | 2021-11-24 | 2023-10-31 | 国网安徽省电力有限公司电力科学研究院 | A method and system for monitoring the mechanical status of GIS equipment isolating switches |
| CN114295343A (en) * | 2021-12-03 | 2022-04-08 | 云南电网有限责任公司红河供电局 | 145kVGIS fault diagnosis method based on acoustic feature method |
| CN114417929B (en) * | 2022-01-21 | 2024-12-13 | 重庆大学 | A method and system for analyzing nonlinear vibration behavior of GIS equipment |
| CN114818978A (en) * | 2022-05-25 | 2022-07-29 | 国网江苏省电力有限公司泰州供电分公司 | A Defect Judgment Method Based on Dynamic Characteristics of High Voltage Switchgear |
| CN114882683A (en) * | 2022-07-11 | 2022-08-09 | 深圳市鼎合丰科技有限公司 | Processing equipment alarm system and method for semiconductor device |
| CN115656437B (en) * | 2022-10-26 | 2024-06-11 | 东北大学 | Soft measurement method for concentration of easily-poisoned gas based on signal multi-feature data |
| CN116383602B (en) * | 2022-12-14 | 2024-05-03 | 国网安徽省电力有限公司电力科学研究院 | GIS isolating switch mechanical defect identification method considering noise and sample size |
| CN116011345A (en) * | 2023-02-07 | 2023-04-25 | 广东电网有限责任公司 | Insulator information prediction model generation method, device, equipment and medium |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20050074205A (en) * | 2004-01-13 | 2005-07-18 | 한국전력공사 | Apparatus for detecting defacts in gas insulated switchgear |
| CN107702908A (en) * | 2017-10-12 | 2018-02-16 | 国网山东省电力公司莱芜供电公司 | GIS mechanical oscillation signal Time-Frequency Analysis Methods based on VMD self adapting morphologies |
| CN107742922A (en) * | 2017-09-15 | 2018-02-27 | 江苏京电电气股份有限公司 | A kind of generalized information system and method with long-range self-diagnostic function |
| CN108981792A (en) * | 2018-05-22 | 2018-12-11 | 南京中高知识产权股份有限公司 | A kind of intelligent GIS equipment air leakage detection system and its working method |
| CN109839263A (en) * | 2019-02-26 | 2019-06-04 | 北京航空航天大学 | A kind of GIS feature extraction and mechanical defect diagnostic method based on vibration information |
| CN109871746A (en) * | 2018-12-30 | 2019-06-11 | 国网辽宁省电力有限公司电力科学研究院 | A Pattern Recognition Method for Diagnosing Insulation Defects Based on GIS |
| CN109974986A (en) * | 2019-03-05 | 2019-07-05 | 国网江苏省电力有限公司电力科学研究院 | A system and method for diagnosing mechanical defects of GIS equipment based on vibration detection |
| 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 |
| CN110703078A (en) * | 2019-09-26 | 2020-01-17 | 河海大学 | GIS fault diagnosis method based on spectrum energy analysis and self-organizing competition algorithm |
| JP2020046202A (en) * | 2018-09-14 | 2020-03-26 | 株式会社東芝 | Partial discharge detection device, partial discharge detection method, partial discharge detection system, and computer program product |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9652723B2 (en) * | 2015-06-05 | 2017-05-16 | Sas Institute Inc. | Electrical transformer failure prediction |
-
2020
- 2020-10-29 CN CN202011181967.6A patent/CN112307950B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20050074205A (en) * | 2004-01-13 | 2005-07-18 | 한국전력공사 | Apparatus for detecting defacts in gas insulated switchgear |
| CN107742922A (en) * | 2017-09-15 | 2018-02-27 | 江苏京电电气股份有限公司 | A kind of generalized information system and method with long-range self-diagnostic function |
| CN107702908A (en) * | 2017-10-12 | 2018-02-16 | 国网山东省电力公司莱芜供电公司 | GIS mechanical oscillation signal Time-Frequency Analysis Methods based on VMD self adapting morphologies |
| CN108981792A (en) * | 2018-05-22 | 2018-12-11 | 南京中高知识产权股份有限公司 | A kind of intelligent GIS equipment air leakage detection system and its working method |
| JP2020046202A (en) * | 2018-09-14 | 2020-03-26 | 株式会社東芝 | Partial discharge detection device, partial discharge detection method, partial discharge detection system, and computer program product |
| CN109871746A (en) * | 2018-12-30 | 2019-06-11 | 国网辽宁省电力有限公司电力科学研究院 | A Pattern Recognition Method for Diagnosing Insulation Defects Based on GIS |
| CN109839263A (en) * | 2019-02-26 | 2019-06-04 | 北京航空航天大学 | A kind of GIS feature extraction and mechanical defect diagnostic method based on vibration information |
| CN109974986A (en) * | 2019-03-05 | 2019-07-05 | 国网江苏省电力有限公司电力科学研究院 | A system and method for diagnosing mechanical defects of GIS equipment based on vibration detection |
| 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 |
| CN110703078A (en) * | 2019-09-26 | 2020-01-17 | 河海大学 | GIS fault diagnosis method based on spectrum energy analysis and self-organizing competition algorithm |
Non-Patent Citations (5)
| Title |
|---|
| Analysis of GIS mechanical vibration signal based on EMD algorithm;Zhigang Wu，等;《IEEE Xplore》;20161203;第853-857页 * |
| Classification Method for Mechanical Defects in GIS Equipment Based on Mode Function Analysis and Improved Relevance Vector Machines;Yao Zhong，等;《CSEE Journal of Power and Energy Systems》;20220506;第9卷(第2期);第790-801页 * |
| 具有远程专家自诊断功能的智能化GIS;白世军，等;《智能电网》;20170131;第5卷(第1期);第75-83页 * |
| 基于多层融合振动数据分析的GIS设备机械缺陷诊断方法研究;冯英，等;《中国电机工程学报》;20230727;第1-13页 * |
| 基于异常振动分析的GIS机械故障诊断技术研究;侯焰;《中国优秀硕士学位论文全文数据库 工程科技II辑》;20171015(第10期);C042-35 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112307950A (en) | 2021-02-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112307950B (en) | Detail feature quantity extraction and intelligent analysis method for GIS vibration defect identification | |
| CN114219147B (en) | Power distribution station fault prediction method based on federal learning | |
| Sun et al. | Fault diagnosis of conventional circuit breaker contact system based on time–frequency analysis and improved AlexNet | |
| CN113110386A (en) | GIS/GIL equipment mechanical vibration state on-line monitoring and mechanical fault cloud diagnosis system | |
| CN110376522B (en) | A data fusion deep learning network based motor fault diagnosis method | |
| CN109187025A (en) | A kind of rolling bearing remaining life prediction technique of integrated KELM | |
| CN106952028A (en) | Dynamoelectric equipment failure is examined and health control method and system in advance | |
| CN111680875B (en) | Unmanned aerial vehicle state risk fuzzy comprehensive evaluation method based on probability baseline model | |
| CN116610998A (en) | Switch cabinet fault diagnosis method and system based on multi-mode data fusion | |
| CN111553495A (en) | Small circuit breaker fault analysis method based on probabilistic neural network | |
| CN112180204A (en) | A method for fault diagnosis of power grid lines based on electrical quantity information | |
| CN112507479B (en) | Oil drilling machine health state assessment method based on manifold learning and softmax | |
| CN113850320A (en) | Transformer fault detection method based on improved support vector machine regression algorithm | |
| Lu et al. | A novel fault diagnosis method for circuit breakers based on optimized affinity propagation clustering | |
| Zu et al. | A simple gated recurrent network for detection of power quality disturbances | |
| CN112906764A (en) | Communication safety equipment intelligent diagnosis method and system based on improved BP neural network | |
| CN115877205A (en) | Intelligent fault diagnosis system and method for servo motor | |
| CN117371207A (en) | Extra-high voltage converter valve state evaluation method, medium and system | |
| CN115098561B (en) | A method for extracting alarm thresholds from cultural relic ontology monitoring data | |
| CN118885944B (en) | Multi-mode fault diagnosis and early warning method | |
| CN113536662B (en) | Error state prediction method of electronic transformer based on firefly optimization LightGBM algorithm | |
| CN115146715B (en) | Method, device, equipment and storage medium for diagnosing potential safety hazard of electricity | |
| CN115169405B (en) | Hotel guest room equipment fault diagnosis method and system based on support vector machine | |
| CN114137915B (en) | A Fault Diagnosis Method for Industrial Equipment | |
| CN114184883B (en) | A distribution network fault detection accuracy calculation method based on distribution network fault simulation |
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 |