CN102494247B - Gas leakage detection method based on voice frequency characteristic recognition - Google Patents
Gas leakage detection method based on voice frequency characteristic recognition Download PDFInfo
- Publication number
- CN102494247B CN102494247B CN 201210002433 CN201210002433A CN102494247B CN 102494247 B CN102494247 B CN 102494247B CN 201210002433 CN201210002433 CN 201210002433 CN 201210002433 A CN201210002433 A CN 201210002433A CN 102494247 B CN102494247 B CN 102494247B
- Authority
- CN
- China
- Prior art keywords
- database
- noise
- leakage
- analysis
- detection method
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Examining Or Testing Airtightness (AREA)
Abstract
The invention discloses a gas leakage detection method based on voice frequency characteristic recognition. The gas leakage detection method comprises the following steps of: constructing a database, extracting noise characteristics, analyzing and comparing noise signals, and optimizing a noise database. According to the gas leakage detection method, the aim of recognizing and detecting leakage points is fulfilled by recognizing the essential characteristics of various noises, establishing the technology of removing background noise which is irrelevant with leakage and establishing a characteristic filter for distinguishing and recognizing a leakage voice frequency.
Description
Technical field
The invention belongs to Leakage Gas method for detecting and equipment technical field, particularly the leakage point for flammable explosive gas detects, and utilizes method and the long-range detecting instrument of fault of audio frequency characteristics identifying method row fast accurate location.
Background technique
It is global industrial knotty problem that the Leakage Gas of high pressure overhead pipeline detects always, still there is no so far good Leakage Gas detection means and instrument.The share less than 20% that solar energy, nuclear power, water power etc. account at present, in world wide, the production of electric power is mainly to derive from fire coal.The U.S. and Chinese power station more than 70% are coal-burning power plants, if the vacuum-pump lines of coal-burning power plant etc. leak (welding damage etc.), the same with the principle of suction engine, along with a large amount of outside air intake lines, vacuum pump draft loss, the miopragia of guiding exhaust steam, cause the exhaust steam of steamer afterbody to be detained accumulation, causes that the afterbody pressure of steamer raises, will cause high pressure steam acting effort, generator speed slows down, and waste coal, cause generation deficiency.Because steamer condenser system vacuum leak, the direct coal-fired waste caused to power plant is surprising, may be up to more than 1,000,000,000.China is when being badly in need of electricity consumption, the power plant's vacuum tightness generated electricity by way of merging two or more grid systems have up to 1000pa, cause great loss, also air has been caused to pollution in various degree when causing huge coal-fired waste, therefore, ensure that the energy-conservation index of each power plant reaches or surpass the standard of developed country by high-tech means, most important; And the industries such as petrochemical industry are in recent years also because vacuum leak, the leakage equifrequency of the poisonous flammable explosive gas living security incident that takes place frequently, brought threat to workmen's personal safety.
At present, general gas flow Leak testtion way mainly contains: bandaging strapping, soap bubble method, qualitative and quantitative detection instrument, semiconductor transducer method, and by implanting helium and locating the technological means such as leakage point by the responsive detection of the leakage of helium.These technological means are large for overhead pipe detection difficulty, must manually waste time and energy by scaffolding, and the leak detection dead angle is many, when the inspector detects, needs staff's Body contact or nearby device.
Conventional method has following shortcoming: dangerous, precision is low, unreliable, efficiency is low, cost is large.So it is high and just can locate leakage point beyond 20 meters to develop a kind of precision, to the vacuum tightness of various sealing pipelines, the leakage point of poisonous flammable explosive gas carries out precisely effective detecting instrument and seems most important.
Summary of the invention
The objective of the invention is in order to solve the defect of above-mentioned existing leak hunting method and leak detector a kind of method of utilizing audio frequency characteristics identifying method row fast accurate to determine leakage point of initiative proposition.
Realize that above-mentioned purpose technological scheme of the present invention is, the gas leak detection method based on audio frequency characteristics identification, the method comprises the following steps:
(1) set up respectively the database that comes from leakage noise and non-leakage noise, by comparing both difference, build the wave filter of distinguishing leakage noise;
(2) carry out feature extraction for the noise signal in the noise data storehouse, be specially the common feature of summarizing in the individualized feature that leaks audio frequency, common feature is the spectrum signature irrelevant with volume, for the sound wave gathered, gather the minimum rule of Euclidean distance of the frequency spectrum of ripple and database element by calculating, determine and gather ripple with respect to the database degree of membership, common feature comprises the frequency spectrum of noise signal, cepstrum, wavelet analysis, the frequency spectrum Fourier transform, resonance peak, pitch analysis or reflective analysis obtain the regularity of statistical analysis, for its regularity of leaks sound, be presented as periodically, harmonious, ,
(3) for the feature extraction of carrying out of the noise signal detected, with noise, the contrast of non-noise data storehouse, whether the noise signal detected is leaks sound respectively;
(4) if the error of the leakage noise signal detected and database is larger, automatic optimization of data storehouse, the main method in optimization data storehouse is: while confirming that current noise is leaks sound, the feature of current noise will directly be added in database, to enrich data standard, and the Euclidean distance between the calculating common feature, reset the numerical value of definition database degree of membership.
In said method, when setting up the noise characteristic database, adopt fuzzy mathematics method, for the new feature detected, carry out Characteristic Contrast, calculate proximity, retain and foundation " next-door neighbour " database, carry out mathematical optimization and approximate analysis with this, FCM Algorithms commonly used in main employing fuzzy mathematics, adjacency adopts the maximum value of the Euclidean distance of database All Eigenvalues group, data for new collection, if through the actual leakage point noise of confirming as, calculate the Euclidean distance with each element of database, if distance is less than adjacency, directly insert database, if be greater than adjacency, after directly inserting database, also to upgrade adjacency, therefore by each leak detection activity, make database constantly upgrade expansion, its ability is enriched and is optimized.。
The step of said method (1) acquisition noise signal is to improve by filter filtering the precision that noise obtains.
The invention also discloses a kind of long-range detecting instrument of Leakage Gas fault based on audio frequency characteristics identification, comprise noise transducer, wave filter, audio frequency storage and analysis module, audio output port and power supply, described noise transducer, wave filter, the audio frequency storage is connected with audio output port successively with analysis module, be connected in series gain switch and show the amount of batteries switch between wave filter and audio frequency storage and analysis module again, power supply is above-mentioned each parts power supply.
Be connected with sensitivity regulator and circuit overcurrent protection module on described audio frequency storage and analysis module.
Lug machine and/or audio frequency storage facilities on described audio output port.
Utilize technological scheme of the present invention by the intrinsic propesties of all kinds of noises of identification, set up for the removal technology with leaking irrelevant background noise, set up the characteristic filtering device of difference and identified leakage audio frequency, reach the purpose of identification and detecting leakage point.
The long-range detecting instrument of Leakage Gas fault based on audio frequency characteristics identification of the present invention utilizes high-precision audio detector, noise identification filtering, the whole ripple of noise, the storage of leaks sound database and analysis, VI volume indicator, gain switch, noise ratio regulators, noise isolation earphone, power supply etc.; Intergrated circuit printing and packaging technology that employing waterproof and antistatic, dustproof, shockproof etc. requires; Employing prevents the circuit design that disturb highfield, to avoid the interference of self noise.
The accompanying drawing explanation
Fig. 1 is the long-range detecting instrument structure of the Leakage Gas fault connection diagram based on audio frequency characteristics identification of the present invention.
Embodiment
Below in conjunction with accompanying drawing, the present invention is specifically described, the gas leak detection method based on audio frequency characteristics identification of the present invention, its step is specially:
(1) set up respectively the database that comes from leakage noise and non-leakage noise, by comparing both difference, build the wave filter of distinguishing leakage noise;
(2) carry out feature extraction for the noise signal in the noise data storehouse, be specially the common feature of summarizing in the individualized feature that leaks audio frequency, common feature is the spectrum signature irrelevant with volume, for the sound wave gathered, gather the minimum rule of Euclidean distance of the frequency spectrum of ripple and database element by calculating, determine and gather ripple with respect to the database degree of membership, common feature comprises the frequency spectrum of noise signal, cepstrum, wavelet analysis, the frequency spectrum Fourier transform, resonance peak, pitch analysis or reflective analysis obtain the regularity of statistical analysis, for its regularity of leaks sound, be presented as periodically, harmonious, ,
(3) for the feature extraction of carrying out of the noise signal detected, with noise, the contrast of non-noise data storehouse, whether the noise signal detected is leaks sound respectively;
(4) if the error of the leakage noise signal detected and database is larger, automatic optimization of data storehouse, the main method in optimization data storehouse is: while confirming that current noise is leaks sound, the feature of current noise will directly be added in database, to enrich data standard, and the Euclidean distance between the calculating common feature, reset the numerical value of definition database degree of membership.
When setting up the noise characteristic database, adopt fuzzy mathematics method, for the new feature detected, carry out Characteristic Contrast, calculate proximity, retain and set up " next-door neighbour " database, carry out mathematical optimization and approximate analysis with this.
Said method step (1) acquisition noise signal is to improve by filter filtering the precision that noise obtains.
Fig. 1 is the long-range detecting instrument structure of the Leakage Gas fault connection diagram based on audio frequency characteristics identification of the present invention, this instrument is by noise transducer, wave filter, the audio frequency storage is with after analysis module, audio output port are connected, serial connection gain switch and demonstration amount of batteries switch between wave filter and audio frequency storage and analysis module, power supply is above-mentioned each parts power supply.
For improving the circuit protection function of instrument, on audio frequency storage and analysis module, be connected with sensitivity regulator and circuit overcurrent protection module.
Lug machine and/or audio frequency storage facilities on audio output port, can be used for directly noise analogy respectively, or it is stored as to data-signal carries out subsequent analysis.
Technique scheme has only embodied the optimal technical scheme of technical solution of the present invention, and those skilled in the art have all embodied principle of the present invention to some changes that wherein some part may be made, within belonging to protection scope of the present invention.
Claims (3)
1. the gas leak detection method based on audio frequency characteristics identification, is characterized in that, the method comprises the following steps:
(1) set up respectively the database of the data that come from leakage noise and non-leakage noise, by comparing both difference, build the wave filter of distinguishing leakage noise;
(2) carry out feature extraction for the noise signal in the noise data storehouse, be specially the common feature of summarizing in the individualized feature that leaks audio frequency, common feature is the spectrum signature irrelevant with volume, for the sound wave gathered, the minimum rule of the Euclidean distance of the sound wave gathered by calculating and the frequency spectrum of database element, determine that the sound wave gathered is with respect to the database degree of membership, common feature comprises the frequency spectrum of noise signal, cepstrum, wavelet analysis, the frequency spectrum Fourier transform, resonance peak, pitch analysis or reflective analysis obtain the regularity of statistical analysis, for its regularity of leaks sound, be presented as periodically, harmonious,
(3) carry out feature extraction for the noise signal detected, with the database of leakage noise and non-leakage noise, contrast respectively, here the parameter of contrast is the common feature parameter, be the regularity that frequency spectrum, cepstrum, wavelet analysis, frequency spectrum Fourier transform, resonance peak, pitch analysis and the reflective analysis of noise signal obtains statistical analysis, determine whether the noise signal detected is leaks sound;
(4) if the common feature parameter error of the leakage noise signal detected and database is larger, automatic optimization of data storehouse, the main method in optimization data storehouse is: while confirming that current noise is leaks sound, the feature of current noise will directly be added in database, to enrich data standard, and the Euclidean distance between the calculating common feature, reset the numerical value of definition database degree of membership.
2. the gas leak detection method based on audio frequency characteristics identification according to claim 1, it is characterized in that, while setting up the noise characteristic database, adopt fuzzy mathematics method, for the new feature detected, carry out Characteristic Contrast, calculate proximity, retain and foundation " next-door neighbour " database, carry out mathematical optimization and approximate analysis with this, FCM Algorithms commonly used in main employing fuzzy mathematics, adjacency adopts the maximum value of the Euclidean distance of database All Eigenvalues group, data for new collection, if through the actual leakage point noise of confirming as, calculate the Euclidean distance of itself and each element of database, if distance is less than adjacency, directly insert database, if be greater than adjacency, after directly inserting database, also to upgrade adjacency, therefore by each leak detection activity, make database constantly upgrade expansion, its ability is enriched and is optimized.
3. the gas leak detection method based on audio frequency characteristics identification according to claim 1, is characterized in that, the noise signal that step (3) detects is to improve by filter filtering the precision that noise obtains.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201210002433 CN102494247B (en) | 2011-10-28 | 2012-01-06 | Gas leakage detection method based on voice frequency characteristic recognition |
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110332963 | 2011-10-28 | ||
| CN201110332963.8 | 2011-10-28 | ||
| CN 201210002433 CN102494247B (en) | 2011-10-28 | 2012-01-06 | Gas leakage detection method based on voice frequency characteristic recognition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102494247A CN102494247A (en) | 2012-06-13 |
| CN102494247B true CN102494247B (en) | 2013-12-18 |
Family
ID=46186093
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201210002433 Expired - Fee Related CN102494247B (en) | 2011-10-28 | 2012-01-06 | Gas leakage detection method based on voice frequency characteristic recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102494247B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104634376A (en) * | 2013-11-15 | 2015-05-20 | 中国石油天然气股份有限公司 | Non-insertion sound wave type pigging passing indicator monitoring device and monitoring method |
| EP3153834B1 (en) | 2015-10-08 | 2020-04-01 | Carrier Corporation | Acoustic profile recognition for discriminating between hazardous emissions and non-hazardous emissions |
| CN111006137B (en) * | 2019-12-18 | 2021-12-17 | 北京无线电计量测试研究所 | Water supply pipeline leakage monitoring and leakage positioning method and system |
| CN111322524B (en) * | 2020-03-02 | 2021-07-09 | 四川盐业地质钻井大队 | Safety detection method, device and detection equipment for drilling platform |
| CN117437740A (en) * | 2023-12-11 | 2024-01-23 | 江苏南工科技集团有限公司 | Industrial fire risk early warning and analyzing method based on big data technology |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2779211Y (en) * | 2005-04-06 | 2006-05-10 | 中国石油天然气股份有限公司 | Pipeline leakage detection device |
| CN101159995A (en) * | 2006-10-06 | 2008-04-09 | 索尼株式会社 | Audio reproducing apparatus and program |
| CN101603630A (en) * | 2009-05-08 | 2009-12-16 | 上海能信石油科技服务有限公司 | The monitoring pipeline vibration system and method |
| CN101960303A (en) * | 2008-11-11 | 2011-01-26 | 韩国电力***开发设计有限公司 | Apparatus for measuring fluid leakage from a valve using ultrasonic wave, sound, and temperature variations, and method for measuring fluid leakage using same |
| US20110232770A1 (en) * | 2010-03-24 | 2011-09-29 | Baggett D David | Irrigation System Leak Controller and Method |
| RU2432558C1 (en) * | 2010-04-09 | 2011-10-27 | Виктор Сергеевич Аносов | Device for searching for leakage points in main pipelines |
-
2012
- 2012-01-06 CN CN 201210002433 patent/CN102494247B/en not_active Expired - Fee Related
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2779211Y (en) * | 2005-04-06 | 2006-05-10 | 中国石油天然气股份有限公司 | Pipeline leakage detection device |
| CN101159995A (en) * | 2006-10-06 | 2008-04-09 | 索尼株式会社 | Audio reproducing apparatus and program |
| CN101960303A (en) * | 2008-11-11 | 2011-01-26 | 韩国电力***开发设计有限公司 | Apparatus for measuring fluid leakage from a valve using ultrasonic wave, sound, and temperature variations, and method for measuring fluid leakage using same |
| CN101603630A (en) * | 2009-05-08 | 2009-12-16 | 上海能信石油科技服务有限公司 | The monitoring pipeline vibration system and method |
| US20110232770A1 (en) * | 2010-03-24 | 2011-09-29 | Baggett D David | Irrigation System Leak Controller and Method |
| RU2432558C1 (en) * | 2010-04-09 | 2011-10-27 | Виктор Сергеевич Аносов | Device for searching for leakage points in main pipelines |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102494247A (en) | 2012-06-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102494247B (en) | Gas leakage detection method based on voice frequency characteristic recognition | |
| CN105156905A (en) | Leakage monitoring system, method and device for pipeline and server | |
| CN101625071B (en) | Method for measuring and locating leakage of gas pipelines | |
| CN102563362B (en) | Compressed air system and intelligent pipe network leakage detecting method for same | |
| CN205090197U (en) | Leak detection system and monitoring facilities of pipeline | |
| CN201992349U (en) | Pipeline leakage monitoring system | |
| CN109854953B (en) | Crude oil conveying pipeline leakage detection system and method under special working condition | |
| CN103631681A (en) | Method for online restoring abnormal data of wind power plant | |
| CN207316488U (en) | A kind of long distance wireless routine for pipe-line transportation system leakage or gas leakage | |
| CN103438359A (en) | Oil pipeline leakage detection and positioning system | |
| CN203929630U (en) | TDLAS concentration of methane gas detection system | |
| CN102004211A (en) | Method for detecting insulation defect of high-voltage cable accessory | |
| CN107906375B (en) | Pipeline leakage detection method and system based on weighted permutation entropy | |
| CN103032626B (en) | System and method for diagnosing fault of adjusting valve | |
| CN102242872B (en) | Oil transportation pipeline network leakage detection method based on generalized fuzzy hyperbolic model | |
| CN103529337B (en) | The recognition methods of nonlinear correlation relation between equipment failure and electric quantity information | |
| CN105675216A (en) | Detection and location method for leaked sound signals | |
| Cai et al. | A data-driven early micro-leakage detection and localization approach of hydraulic systems | |
| CN209524328U (en) | A kind of oil transmission line leak detection system under special operation condition | |
| CN202442118U (en) | Intelligent pipe network leakage detection system for compressed air system | |
| CN112484796A (en) | Experiment platform and method for calibrating flow of regulating valve by sound wave signal | |
| CN111579170A (en) | Comprehensive leak detection testing device and method for steam turbine vacuum system | |
| CN202392444U (en) | Gas leakage detector based on audio feature identification method | |
| CN201477169U (en) | Insulation defect detection device for high voltage cable accessory | |
| CN113670536B (en) | Power plant electricity water monitoring and informationized management method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20131218 Termination date: 20200106 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |