CN104833884A - Fault detection method of voltage class equipment - Google Patents
Fault detection method of voltage class equipment Download PDFInfo
- Publication number
- CN104833884A CN104833884A CN201510252419.0A CN201510252419A CN104833884A CN 104833884 A CN104833884 A CN 104833884A CN 201510252419 A CN201510252419 A CN 201510252419A CN 104833884 A CN104833884 A CN 104833884A
- Authority
- CN
- China
- Prior art keywords
- neural network
- network model
- fourier transform
- signal
- voltage
- 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
Abstract
The invention discloses a fault detection method of voltage class equipment. Through the training process of a third neural network model and the test process of the third neural network model, the output voltage signals of the voltage class equipment are subjected to Fourier transform to obtain voltage harmonic signals. The neural network models are used to classify the voltage harmonic signals obtained through the Fourier transform. The fault types corresponding to various voltage harmonic signals obtained through Fourier transform are determined. The method can be applied to various signals by using the Fourier transform, thus the detection efficiency is raised, the application range is increased, the neural network models are used to classify the voltage harmonic signals obtained through the Fourier transform so as to determine the fault types corresponding to various types of voltage harmonic signals, thus the fault types corresponding to output voltage signals of the voltage class equipment are determined, and the detection efficiency is improved.
Description
Technical field
The present invention relates to electrical equipment technical field, specifically the fault detection method of electric pressure equipment.
Background technology
Electric power be with
electric energyas power
the energy.The electrical energy production that electric system is made up of links such as generating, transmission of electricity, power transformation, distribution and electricity consumptions and consume system.Its function is that natural primary energy is changed into electric energy by generation power device, then supplies power to each user through transmission of electricity, power transformation and distribution.For realizing this function; electric system also has corresponding information and control system at links and different levels; the production run of electric energy is measured, regulates, controls, protects, communicates and dispatched, to ensure that user obtains the electric energy of safety, economy, high-quality.
Electric pressure (voltage class) is the rated voltage rank series of electric system and power equipment.Rated voltage is the normal voltage that electric system and power equipment specify, namely relevant with electric system and some operation characteristic of power equipment nominal voltage.The actual motion voltage of electric system each point allows to depart from its rated voltage to a certain extent, and in this leeway, various power equipment and electric system itself still can normally run.
The electric pressure that current China is conventional: 220V, 380V, 6kV, 10kV, 35kV, 110kV, 220kV, 330kV, 500kV, 1000kV.Electric system is generally be made up of generating plant, transmission line of electricity, electric substation, distribution line and consumer.Usually the voltage circuit of more than 35kV is called power transmission sequence.35kV and following voltage circuit thereof are called distribution line.Be called " high voltage " by specified more than 1kV voltage, rated voltage is called " low-voltage " at below 1kV voltage.
China's regulation safe voltage is 42V, 36V, 24V, 12V, 6V five kinds.
In alternating voltage grade, usually by 1kV and hereinafter referred to as low pressure, more than 1kV, 20kV and hereinafter referred to as middle pressure, more than 20kV, 330kV hereinafter referred to as high pressure, 330kV and above, 1000kV hereinafter referred to as UHV (ultra-high voltage), 1000kV and be called extra-high voltage above.In DC voltage level, ± 800kV hereinafter referred to as high pressure, ± 800kV and be called extra-high voltage above
The classification of meticulous electric pressure makes the number of electric pressure equipment also increase thereupon, make the structure of circuit and control mode more complicated, make the failure rate of electric pressure equipment increase, thus the fault detect of electric pressure equipment is particularly important simultaneously.The fault detect of existing electric pressure equipment detection method mainly knowledge based and experience, but the detection efficiency of the method is lower, and range of application is narrower.
Summary of the invention
The object of the present invention is to provide the fault detection method of the electric pressure equipment improving detection efficiency, to solve the problem proposed in above-mentioned background technology.
For achieving the above object, the invention provides following technical scheme:
The fault detection method of electric pressure equipment, comprising:
The training process of step 1, three neural network models, this training process comprises:
1) output signal under preset failure state of electric pressure equipment is carried out Fourier transform, obtain voltage harmonic signal;
2) by the input layer of the harmonic signal input neural network model in described step 1) after Fourier transform;
3) determine the weights of neural network model according to the input signal of neural network model and the default output signal of neural network model, to determine the classification mechanism of neural network model, the default output signal of neural network model is corresponding with preset failure state;
The test process of step 2, three neural network models, this test process comprises:
1) value of the modulation ratio of electric pressure equipment is adjusted to the value being different from modulation ratio corresponding in training process, and Fourier transform is carried out to the output signal of electric pressure equipment under preset failure;
2) by the harmonic signal input neural network model after Fourier transform;
3) whether the real output signal of comparative neural network model is consistent with default output signal, and if so, then neural network model is trained successfully, carries out next step; If not, then neural network model failure to train;
Step 3, the output voltage signal of electric pressure equipment is carried out Fourier transform, to obtain voltage harmonic signal;
1) analog voltage signal of the output of electric pressure equipment is converted to digital voltage signal;
2) digital voltage signal after conversion is carried out Fourier transform;
Step 4, neural network model is utilized to classify to the voltage harmonic signal after Fourier transform;
1) input signal of neural network model is normalized;
2) value after normalization is carried out dimensionality reduction;
Step 5, determine the fault type corresponding to voltage harmonic signal after various types of Fourier transform.
Compared with prior art, the invention has the beneficial effects as follows: the present invention utilizes Fourier transform can be applied to various different signal, thus detection efficiency improves, and the scope of application increases to some extent; Utilize neural network model to classify to determine the fault type corresponding to all types of voltage harmonic signal to voltage harmonic signal, and then determine the fault type corresponding to output voltage signal of electric pressure equipment, improve detection efficiency.
Embodiment
Below in conjunction with the embodiment of the present invention, be clearly and completely described the technical scheme in the embodiment of the present invention, obviously, described embodiment is only the present invention's part embodiment, instead of whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art, not making the every other embodiment obtained under creative work prerequisite, belong to the scope of protection of the invention.
Embodiment 1
In the embodiment of the present invention, the fault detection method of electric pressure equipment, this fault detection method comprises:
The training process of step 1, three neural network models, this training process comprises: 1) output signal under preset failure state of electric pressure equipment is carried out Fourier transform, obtain voltage harmonic signal; 2) by the input layer of the harmonic signal input neural network model in described step 1) after Fourier transform; 3) determine the weights of neural network model according to the input signal of neural network model and the default output signal of neural network model, to determine the classification mechanism of neural network model, the default output signal of neural network model is corresponding with preset failure state.
The test process of step 2, three neural network models, this test process comprises: 1) value of the modulation ratio of electric pressure equipment is adjusted to the value being different from modulation ratio corresponding in training process, and carries out Fourier transform to the output signal of electric pressure equipment under preset failure; 2) by the harmonic signal input neural network model after Fourier transform; 3) whether the real output signal of comparative neural network model is consistent with default output signal, and if so, then neural network model is trained successfully, carries out next step; If not, then neural network model failure to train.
Step 3, the output voltage signal of electric pressure equipment is carried out Fourier transform, to obtain voltage harmonic signal; 1) analog voltage signal of the output of electric pressure equipment is converted to digital voltage signal; 2) digital voltage signal after conversion is carried out Fourier transform.
Step 4, neural network model is utilized to classify to the voltage harmonic signal after Fourier transform; 1) input signal of neural network model is normalized; 2) value after normalization is carried out dimensionality reduction.
Neural network model is utilized to classify to the voltage harmonic signal after Fourier transform.Particularly, the input layer of neural network model can be provided with 20 ~ 50 input nodes, the each harmonic signal obtained after Fourier transform is inputted each input node successively, such as, fundamental signal is inputted first input node, first harmonic signal inputs second input node, and second harmonic signal inputs the 3rd input node, by that analogy.The classification number of the fault that the quantity of the output node of the output layer of neural network model may be able to occur according to electric pressure equipment and arranging.5 subordinate voltage grade equipment are comprised for electric pressure equipment, when only considering that one of them subordinate voltage grade establishes situation about breaking down, the malfunction of described electric pressure equipment has 5 kinds, be respectively: first subordinate voltage grade is established and broken down, second subordinate voltage grade is established and is broken down, and the 3rd subordinate voltage grade is established and broken down etc.In this case, the quantity of the output node of described neural network model can be set to 3, each node can export 0,1 two kind of signal, and thus described neural network model can export 8 kinds of different signals altogether, and wherein 5 kinds different signals are corresponding with 5 kinds of fault types.Such as, when the output of described neural network model is 0001, can judges that first subordinate voltage grade of electric pressure equipment is established and break down; When the output of described neural network model is 005, can judges that second subordinate voltage grade of electric pressure equipment is established and break down, by that analogy.
Step 5, determine the fault type corresponding to voltage harmonic signal after various types of Fourier transform.Neural network model can have multiple output form, such as multiple node can be set at output layer, multiple sub neural network model also can be set, the output layer of each sub neural network model arranges an output node, as long as can export different signals to distinguish different malfunctions.The benefit applying neural network model in the present invention is, can improve classification effectiveness, and the electric pressure equipment of various faults may occur for complex structure, and application neural network model can improve detection efficiency.
To those skilled in the art, obviously the invention is not restricted to the details of above-mentioned one exemplary embodiment, and when not deviating from spirit of the present invention or essential characteristic, the present invention can be realized in other specific forms.Therefore, no matter from which point, all should embodiment be regarded as exemplary, and be nonrestrictive, scope of the present invention is limited by claims instead of above-mentioned explanation, and all changes be therefore intended in the implication of the equivalency by dropping on claim and scope are included in the present invention.
In addition, be to be understood that, although this instructions is described according to embodiment, but not each embodiment only comprises an independently technical scheme, this narrating mode of instructions is only for clarity sake, those skilled in the art should by instructions integrally, and the technical scheme in each embodiment also through appropriately combined, can form other embodiments that it will be appreciated by those skilled in the art that.
Claims (1)
1. the fault detection method of electric pressure equipment, is characterized in that, comprising:
The training process of step 1, three neural network models, this training process comprises:
1) output signal under preset failure state of electric pressure equipment is carried out Fourier transform, obtain voltage harmonic signal;
2) by the input layer of the harmonic signal input neural network model in described step 1) after Fourier transform;
3) determine the weights of neural network model according to the input signal of neural network model and the default output signal of neural network model, to determine the classification mechanism of neural network model, the default output signal of neural network model is corresponding with preset failure state;
The test process of step 2, three neural network models, this test process comprises:
1) value of the modulation ratio of electric pressure equipment is adjusted to the value being different from modulation ratio corresponding in training process, and Fourier transform is carried out to the output signal of electric pressure equipment under preset failure;
2) by the harmonic signal input neural network model after Fourier transform;
3) whether the real output signal of comparative neural network model is consistent with default output signal, and if so, then neural network model is trained successfully, carries out next step; If not, then neural network model failure to train;
Step 3, the output voltage signal of electric pressure equipment is carried out Fourier transform, to obtain voltage harmonic signal;
1) analog voltage signal of the output of electric pressure equipment is converted to digital voltage signal;
2) digital voltage signal after conversion is carried out Fourier transform;
Step 4, neural network model is utilized to classify to the voltage harmonic signal after Fourier transform;
1) input signal of neural network model is normalized;
2) value after normalization is carried out dimensionality reduction;
Step 5, determine the fault type corresponding to voltage harmonic signal after various types of Fourier transform.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510252419.0A CN104833884A (en) | 2015-05-18 | 2015-05-18 | Fault detection method of voltage class equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510252419.0A CN104833884A (en) | 2015-05-18 | 2015-05-18 | Fault detection method of voltage class equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104833884A true CN104833884A (en) | 2015-08-12 |
Family
ID=53811888
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510252419.0A Pending CN104833884A (en) | 2015-05-18 | 2015-05-18 | Fault detection method of voltage class equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104833884A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101034389A (en) * | 2007-03-19 | 2007-09-12 | 江西省电力科学研究院 | Electrical energy power quality disturbance automatic identification method and system based on information fusion |
| CN101782625A (en) * | 2009-01-16 | 2010-07-21 | 复旦大学 | Power electronic system fault diagnostic method based on Gradation-boosting algorithm |
| CN102466566A (en) * | 2010-11-03 | 2012-05-23 | 财团法人工业技术研究院 | Power equipment abnormality detection device and detection method thereof |
| CN104049159A (en) * | 2014-05-16 | 2014-09-17 | 北京京东方能源科技有限公司 | Fault detection method and device of inverter |
-
2015
- 2015-05-18 CN CN201510252419.0A patent/CN104833884A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101034389A (en) * | 2007-03-19 | 2007-09-12 | 江西省电力科学研究院 | Electrical energy power quality disturbance automatic identification method and system based on information fusion |
| CN101782625A (en) * | 2009-01-16 | 2010-07-21 | 复旦大学 | Power electronic system fault diagnostic method based on Gradation-boosting algorithm |
| CN102466566A (en) * | 2010-11-03 | 2012-05-23 | 财团法人工业技术研究院 | Power equipment abnormality detection device and detection method thereof |
| CN104049159A (en) * | 2014-05-16 | 2014-09-17 | 北京京东方能源科技有限公司 | Fault detection method and device of inverter |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103454521B (en) | A kind of wind power plant power grid operation simulator | |
| CN102801166B (en) | Safely and economically coordinated automatic voltage control method based on cooperative game theory | |
| CN203037769U (en) | Wind power and photovoltaic grid-connected testing platform | |
| CN104330980A (en) | RT-LAB based microgrid simulation testing system | |
| CN103151839B (en) | Intelligent substation test research system | |
| CN112260261B (en) | Method for evaluating and improving vulnerability of electricity-gas comprehensive energy system | |
| CN109103879A (en) | A kind of full dimension closed loop network appraisal procedure | |
| CN103969534A (en) | Islanding testing method for wind generating set | |
| CN105471389B (en) | A kind of the fault traversing detection system and method for photovoltaic plant/photovoltaic DC-to-AC converter | |
| CN105809576A (en) | Power distribution network oriented power supply safety load hierarchical evaluation method | |
| CN106655150A (en) | Optimizing method for the running mode of a power distribution network | |
| Zheng et al. | Fault identification scheme for UPFC compensated transmission line based on characteristic voltage active injection | |
| He et al. | Study on the overvoltage and insulation coordination of flexible DC power distribution network | |
| CN104833884A (en) | Fault detection method of voltage class equipment | |
| CN107769247A (en) | A kind of RLC load simulation systems and its control method for the detection of anti-isolated island | |
| Adail et al. | Safety and reliability of the electrical system based on optimal FACTS allocation for a research reactor | |
| Avendano-Mora et al. | Comparison of control strategies for multiterminal VSC-HVDC systems for offshore wind farm integration | |
| Miao et al. | Power supply reliability indices computation with consideration of generation systems, transmission systems and sub-transmission systems' load transfer capabilities | |
| CN102427228B (en) | Wind power generation system reliability estimating method considering voltage drop of power grid | |
| CN105353258B (en) | A kind of test method of test transformer differential protection reactance transformer excitation surge current | |
| CN103956743B (en) | A kind of 10KV power distribution network optimization method based on phase difference and device | |
| CN204012716U (en) | A kind of 10kV power distribution network optimization device based on phase difference | |
| Ivry et al. | An evaluation of V2G for distribution network harmonic suppression | |
| Nassaj et al. | Confronting with time delays in tap changers for dynamic voltage control by multi agent systems | |
| TWI610508B (en) | Method for adjusting three-phase balance for distribution transformer wiring |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150812 |