CN104569747A - System and method for checking insulativity of power-off cable - Google Patents
System and method for checking insulativity of power-off cable Download PDFInfo
- Publication number
- CN104569747A CN104569747A CN201410548016.6A CN201410548016A CN104569747A CN 104569747 A CN104569747 A CN 104569747A CN 201410548016 A CN201410548016 A CN 201410548016A CN 104569747 A CN104569747 A CN 104569747A
- Authority
- CN
- China
- Prior art keywords
- neural network
- fish
- cable
- behavior
- rvm
- 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
Landscapes
- Testing Relating To Insulation (AREA)
Abstract
The invention relates to a system and method for checking the insulativity of a power-off cable, and belongs to the technical field of cables. The system comprises a voltage signal preprocessing module, a characteristic value extraction module, a neural network trainer and a neural network identifier which are connected in sequence. A BP (back propagation) neural network in the method has the capabilities of nonlinear mapping, adaptive learning and fault tolerance, the wavelet decomposition has the multi-resolution property, the wavelet decomposition and the time-domain feature quantity are combined and input into the neural network, the advantages of the wavelet decomposition and the time-domain feature quantity are brought into full play, the identification of cable discharge is facilitated, and the insulativity is checked.
Description
Technical field
The invention belongs to field of cable technology, be specifically related to a kind of power-off cable insulation checking system and method.
Background technology
Along with society is to the continuous increase of electricity needs, power industry is rapidly developed, and the safe operation of large-scale power transmission network becomes the significant problem of power industry concern.Along with the continuous transformation of China's urban distribution network, cable is own through being widely used in transmission line of electricity and power distribution network as power cable, therefore, detects seem very necessary and important to the state of insulation of cable.
Summary of the invention
In order to overcome the deficiencies in the prior art, the invention provides a kind of power-off cable insulation checking system and method.
Technical scheme of the present invention is: a kind of power-off cable insulation checking system, comprise voltage signal pretreatment module, characteristics extraction module, neural metwork training device and neural network recognizer, voltage signal pretreatment module, characteristics extraction module, neural metwork training device are connected successively with neural network recognizer.Comprise the steps: step one: tested cable is connected direct supply, forms a LCR oscillatory circuit with inductance, switch, direct supply is charged to tested cable; Step 2: complete on the basis of charging, switch conduction, makes LCR oscillation circuit carry out damped oscillation, gathers oscillating voltage waveform; Step 3: build Neural Network Diagnosis model, the oscillating voltage waveform that input gathers, judges the insulating property of cable.In described step 3, the idiographic flow of Neural Network Diagnosis model construction is: shake voltage wave signal to a large amount of normal operation of different cable with electric discharge and carry out real-time sampling, characteristics extraction is carried out to sampled result and as the learning sample of neural network, by in its input neural network, network is trained, using the recognizer that the neural network trained is checked as insulativity, if recognizer detects that electric discharge phenomena just control trip gear tripping circuit, otherwise just start new round detection.Described eigenwert is the eigenwert of high frequency coefficient as neural network of wavelet transformation.Described neural network model adopts the back-propagation neural network algorithm of momentum arithmetic to train network, selects Sigmoid function as the transport function of this node.
The present invention has following good effect: BP neural network has Nonlinear Mapping, adaptive learning and fault-tolerant ability, wavelet decomposition has multi-resolution characteristics, wavelet decomposition and temporal signatures amount being combined is input in neural network, give full play to respective advantage, be conducive to the identification of cable discharge, carry out insulativity inspection.
Accompanying drawing explanation
Fig. 1 is specific embodiment of the invention cable insulation detection system;
Fig. 2 is specific embodiment of the invention Mallat algorithm structure;
Fig. 3 is specific embodiment of the invention electric discharge neural network recognizer model.
Embodiment
Contrast accompanying drawing below, by the description to embodiment, the specific embodiment of the present invention is as the effect of the mutual alignment between the shape of involved each component, structure, each several part and annexation, each several part and principle of work, manufacturing process and operation using method etc., be described in further detail, have more complete, accurate and deep understanding to help those skilled in the art to inventive concept of the present invention, technical scheme.
Main thought of the present invention is: verifying attachment of the present invention comprises direct supply, switch and inductance, direct supply, switch, inductance are connected with power-off cable, form a LCR and shake loop, when the switches are opened, in charge circuit, direct supply can carry out DC charging to tested cable.Inductance, is connected with described direct supply, and is connected with power-off cable.Direct supply, inductance and tested cable three are connected and can form a charge circuit.On the basis completing charging, when the switch is closed, LCR oscillation circuit can carry out damped oscillation.LCR oscillation circuit carries out damped oscillation can produce oscillating voltage ripple, and oscillating voltage ripple can produce oscillating wave voltage on tested cable.By applying the discharge signal that oscillating wave voltage can inspire cable latent defect place on tested cable.The present invention is by gathering oscillating voltage waveform, and utilize neural network to carry out analyzing and processing to waveform, judge whether cable exists leaky, whether insulating property are good.
1. insulativity Cleaning Principle
Cable insulation recognition system based on neural network provided by the invention as shown in Figure 1.System comprises voltage signal pretreatment module, characteristics extraction module, neural metwork training device and neural network recognizer, voltage signal pretreatment module, and characteristics extraction module, neural metwork training device are connected successively with neural network recognizer.System is shaken voltage wave signal to a large amount of normal operation of different cable with electric discharge first respectively and is carried out real-time sampling, characteristics extraction is carried out to sampled result and as the learning sample of neural network, by in its input neural network, network is trained, using the recognizer of the neural network trained as nonisulated property.Training for neural network completes in a computer, and remainder all completes in pick-up unit.During plant running, carry out identification needing the voltage signal identified to be input in the neural network trained.If recognizer detects that electric discharge phenomena just control trip gear tripping circuit, otherwise just start new round detection.
The extraction of 2 eigenwerts
Insulativity based on neural net method is checked, and key determines the input quantity of neural network and the eigenwert of voltage signal.Because wavelet transformation has spatial locality, energy " focusing ", in the partial structurtes of signal, utilizes wavelet transformation can determine the singularity position of signal.Herein using the class input feature vector amount of the high frequency coefficient of wavelet transformation as neural network, using line current at the mechanical periodicity of time domain as a class input feature vector amount of neural network.
3. based on the characteristics extraction of wavelet transformation
There is approximation signal and detail signal in the voltage signal of cable, after generation electric discharge phenomena, the amplitude of two-layer detail signal is all significantly increased after two-layer wavelet transform.Wavelet transformation is a kind of signal analysis method with time domain and frequency domain combined analytical characteristics, modulus maximum point in wavelet transformation detail coefficients can catastrophe point well in corresponding electric discharge original waveform, so wavelet analysis is the effective ways describing electric discharge phenomena, and the detail signal of wavelet transformation also reflects the feature of voltage when producing electric discharge effectively.
Calculate wavelet transformation time, the related conclusions directly quoting the Mallat algorithm of wavelet analysis to calculate the high frequency coefficient of wavelet transformation, if scaling function race
, form metric space V
jorthonormal basis, family of functions
form wavelet space W
jorthonormal basis.Definition digital filter h (n) and g (n) are
Definition QUOTE
(n)=h (-n), QUOTE
n ()=g (-n), wherein uses wave filter h (n) and g (n) to carry out wavelet reconstruction, uses wave filter QUOTE
(n) and QUOTE
n () carries out wavelet decomposition.Mallat wavelet decomposition algorithm can be expressed as
A in formula
j(k)---the approximation signal of jth layer scattering wavelet decomposition
A
j+1the approximation signal of (k)---jth+1 layer scattering wavelet decomposition
D
j+1the detail signal of (n)---jth+1 layer scattering wavelet decomposition
As the fast algorithm of wavelet transform---Mallat algorithm structure is as shown in Figure 2.
When electric discharge phenomena detect, adopt sym2 wavelet function, and get the detail signal energy of per semiperiod first and second layer of wavelet decomposition, as two input quantities of neural network, can be expressed as
Carry out the filter coefficient of the sym2 wavelet function required for wavelet decomposition in the microcontroller, as shown in table 1.
Table one sym2 wavelet function filter coefficient
In test, different loads is normally run and gather with voltage data when discharging, and according to table 1 with under formula (3) ~ (6) calculate different loads after normalization, the energy of per semiperiod wavelet decomposition detail signal.
4. neural network recognizer
Neural network has stronger non-linear mapping capability, learning content can be remembered in the weights of network adaptively, can ensure correctly to classify to required object of classification after training, even or the object of noise pollution can be had correctly to classify to the object do not learnt.The present invention adopts back-propagation (the Back Propagation of momentum arithmetic, BP) neural network algorithm is trained network, the input layer of neural network is made up of the energy of first and second layer of detail signal of each cycle voltage signal wavelet decomposition and the variable quantity of each cycle voltage peak and mean value, there are 4 neurons, and normalized is done to voltage signal.The Output rusults of network is divided into " normally running " and " electric discharge " two class state, and output layer only needs 1 neuron just can meet the demands, and selection Sigmoid function, as the transport function of this node, makes the output quantity of network be limited in [ 0,1 ] scope.For making there is certain optimizing space during network training, the desired output Y of sample is changed in the following manner
When identifying, export with 0.5 for boundary, [ 0,0.5 ] is expressed as normal condition, and [ 0.5,1 ] has been expressed as electric discharge and has produced.According to above elaboration, build neural network recognizer model as shown in Figure 3.
Above by reference to the accompanying drawings to invention has been exemplary description; obvious specific implementation of the present invention is not subject to the restrictions described above; as long as have employed the improvement of the various unsubstantialities that method of the present invention is conceived and technical scheme is carried out; or design of the present invention and technical scheme directly applied to other occasion, all within protection scope of the present invention without to improve.
Claims (3)
1. based on a Fault Diagnosis of Engine of RVM, it is characterized in that: comprise the steps:
Step one: pre-service: engine operational speed and the amplitude of correspondence thereof, the priori data of frequency samples are normalized, and set up the amplitude of travelling speed and correspondence thereof, the corresponding relation between frequency and engine condition;
Step 2: machine is trained: select suitable kernel function and carry out fish-swarm algorithm optimization training to its hyper parameter, setting up suitable RVM model;
Step 3: fault diagnosis: adopt " one to one " RVM sorter to carry out sample to be tested fault diagnosis and Output rusults.
2. the Fault Diagnosis of Engine based on RVM according to claim 1, is characterized in that: in described step 2, kernel function is gaussian radial basis function kernel function.
3. the Fault Diagnosis of Engine based on RVM according to claim 1, is characterized in that: the implementation procedure of described fish-swarm algorithm is:
A. the crossing-over rate in initialization shoal of fish population number, each stage and aberration rate, maximum iteration time;
B. use mutual information calculates the mutual information between each variable;
C. adopt MWST algorithm to generate initial non-directed graph, and specify any one node to be that root node generates initial population;
D. calculate the BIC scoring of all initial population, find out the maximum score value of scoring and individuality;
E. the result of d is put into the bulletin board after initialization;
While(iterations < maximum iteration time)
for k=1:fishnum
If f. the fish individuality of this numbering meets condition of bunching, then perform behavior of bunching; Otherwise execution foraging behavior;
If g. the fish individuality of this numbering meets the condition that knocks into the back, then perform behavior of knocking into the back; Otherwise execution foraging behavior;
H. the fish individuality executing bunch behavior and behavior of knocking into the back is compared, obtain the mode that this fish individuality " predation " is optimum, and record;
I. the optimum individual in the fishinum after " predation " fish individuality is used once to upgrade bulletin board;
J. judge whether to reach maximum iteration time, if reach maximum iteration time, then algorithm terminates; Otherwise proceed;
K. export the top score value after calculating, obtain optimized parameter.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410548016.6A CN104569747A (en) | 2014-10-16 | 2014-10-16 | System and method for checking insulativity of power-off cable |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410548016.6A CN104569747A (en) | 2014-10-16 | 2014-10-16 | System and method for checking insulativity of power-off cable |
Publications (1)
Publication Number | Publication Date |
---|---|
CN104569747A true CN104569747A (en) | 2015-04-29 |
Family
ID=53086278
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410548016.6A Pending CN104569747A (en) | 2014-10-16 | 2014-10-16 | System and method for checking insulativity of power-off cable |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104569747A (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113269952A (en) * | 2020-02-14 | 2021-08-17 | 美光科技公司 | Drive-by-wire sensor monitoring in a vehicle |
US11498388B2 (en) | 2019-08-21 | 2022-11-15 | Micron Technology, Inc. | Intelligent climate control in vehicles |
US11586943B2 (en) | 2019-08-12 | 2023-02-21 | Micron Technology, Inc. | Storage and access of neural network inputs in automotive predictive maintenance |
US11586194B2 (en) | 2019-08-12 | 2023-02-21 | Micron Technology, Inc. | Storage and access of neural network models of automotive predictive maintenance |
US11635893B2 (en) | 2019-08-12 | 2023-04-25 | Micron Technology, Inc. | Communications between processors and storage devices in automotive predictive maintenance implemented via artificial neural networks |
US11650746B2 (en) | 2019-09-05 | 2023-05-16 | Micron Technology, Inc. | Intelligent write-amplification reduction for data storage devices configured on autonomous vehicles |
US11693562B2 (en) | 2019-09-05 | 2023-07-04 | Micron Technology, Inc. | Bandwidth optimization for different types of operations scheduled in a data storage device |
US11702086B2 (en) | 2019-08-21 | 2023-07-18 | Micron Technology, Inc. | Intelligent recording of errant vehicle behaviors |
US11709625B2 (en) | 2020-02-14 | 2023-07-25 | Micron Technology, Inc. | Optimization of power usage of data storage devices |
US11748626B2 (en) | 2019-08-12 | 2023-09-05 | Micron Technology, Inc. | Storage devices with neural network accelerators for automotive predictive maintenance |
US11775816B2 (en) | 2019-08-12 | 2023-10-03 | Micron Technology, Inc. | Storage and access of neural network outputs in automotive predictive maintenance |
US11830296B2 (en) | 2019-12-18 | 2023-11-28 | Lodestar Licensing Group Llc | Predictive maintenance of automotive transmission |
US11853863B2 (en) | 2019-08-12 | 2023-12-26 | Micron Technology, Inc. | Predictive maintenance of automotive tires |
-
2014
- 2014-10-16 CN CN201410548016.6A patent/CN104569747A/en active Pending
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11853863B2 (en) | 2019-08-12 | 2023-12-26 | Micron Technology, Inc. | Predictive maintenance of automotive tires |
US11586943B2 (en) | 2019-08-12 | 2023-02-21 | Micron Technology, Inc. | Storage and access of neural network inputs in automotive predictive maintenance |
US11586194B2 (en) | 2019-08-12 | 2023-02-21 | Micron Technology, Inc. | Storage and access of neural network models of automotive predictive maintenance |
US11635893B2 (en) | 2019-08-12 | 2023-04-25 | Micron Technology, Inc. | Communications between processors and storage devices in automotive predictive maintenance implemented via artificial neural networks |
US11775816B2 (en) | 2019-08-12 | 2023-10-03 | Micron Technology, Inc. | Storage and access of neural network outputs in automotive predictive maintenance |
US11748626B2 (en) | 2019-08-12 | 2023-09-05 | Micron Technology, Inc. | Storage devices with neural network accelerators for automotive predictive maintenance |
US11702086B2 (en) | 2019-08-21 | 2023-07-18 | Micron Technology, Inc. | Intelligent recording of errant vehicle behaviors |
US11498388B2 (en) | 2019-08-21 | 2022-11-15 | Micron Technology, Inc. | Intelligent climate control in vehicles |
US11650746B2 (en) | 2019-09-05 | 2023-05-16 | Micron Technology, Inc. | Intelligent write-amplification reduction for data storage devices configured on autonomous vehicles |
US11693562B2 (en) | 2019-09-05 | 2023-07-04 | Micron Technology, Inc. | Bandwidth optimization for different types of operations scheduled in a data storage device |
US11830296B2 (en) | 2019-12-18 | 2023-11-28 | Lodestar Licensing Group Llc | Predictive maintenance of automotive transmission |
US11709625B2 (en) | 2020-02-14 | 2023-07-25 | Micron Technology, Inc. | Optimization of power usage of data storage devices |
CN113269952A (en) * | 2020-02-14 | 2021-08-17 | 美光科技公司 | Drive-by-wire sensor monitoring in a vehicle |
CN113269952B (en) * | 2020-02-14 | 2023-10-10 | 美光科技公司 | Method for predictive maintenance of a vehicle, data storage device and vehicle |
US11531339B2 (en) | 2020-02-14 | 2022-12-20 | Micron Technology, Inc. | Monitoring of drive by wire sensors in vehicles |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104569747A (en) | System and method for checking insulativity of power-off cable | |
CN110221200B (en) | Universal circuit breaker accessory fault diagnosis method based on deep learning | |
CN111879397B (en) | Fault diagnosis method for energy storage mechanism of high-voltage circuit breaker | |
CN112241605B (en) | Breaker energy storage process state identification method for constructing CNN feature matrix by using acoustic vibration signals | |
CN113408341B (en) | Load identification method and device, computer equipment and storage medium | |
Wang et al. | A novel wind speed forecasting method based on ensemble empirical mode decomposition and GA-BP neural network | |
CN109214356A (en) | A kind of fan transmission system intelligent fault diagnosis method based on DCNN model | |
CN113295702B (en) | Electrical equipment fault diagnosis model training method and electrical equipment fault diagnosis method | |
CN109359665A (en) | A kind of family's electric load recognition methods and device based on support vector machines | |
CN111563827A (en) | Load decomposition method based on electrical appliance physical characteristics and residential electricity consumption behaviors | |
Wang et al. | Improved prediction method of PV output power based on optimised chaotic phase space reconstruction | |
Zhu et al. | Complex disturbances identification: A novel PQDs decomposition and modeling method | |
Anh et al. | A benchmark of deep learning models for multi-leaf diseases for edge devices | |
Rizvi | Time series deep learning for robust steady-state load parameter estimation using 1D-CNN | |
CN106296434A (en) | A kind of Grain Crop Yield Prediction method based on PSO LSSVM algorithm | |
Ye et al. | Single pole‐to‐ground fault location method for mmc‐hvdc system using wavelet decomposition and dbn | |
CN115017828A (en) | Power cable fault identification method and system based on bidirectional long-short-time memory network | |
Wang et al. | A wind speed prediction method based on improved empirical mode decomposition and support vector machine | |
Lun Gaoa et al. | Research and application of data mining feature selection based on relief algorithm | |
Wang et al. | Power quality disturbance recognition method in park distribution network based on one-dimensional vggnet and multi-label classification | |
Zhang et al. | Residual current fault type recognition based on S3VM and KNN cooperative training | |
CN115600754A (en) | Method, system, equipment and storage medium for predicting potential safety hazard of user power utilization | |
CN115293195A (en) | Analysis and prediction method and device applied to coal-to-electricity load characteristics | |
Xing et al. | A method of classified HV circuit breaker fault signal based on EEMD and SOM neural network | |
Zhuo et al. | The evaluation of wind power time series multi-scale modeling methods |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20150429 |
|
WD01 | Invention patent application deemed withdrawn after publication |