CN102174992A - Pressure curve feature extraction method for pressure pipeline - Google Patents
Pressure curve feature extraction method for pressure pipeline Download PDFInfo
- Publication number
- CN102174992A CN102174992A CN2011100268492A CN201110026849A CN102174992A CN 102174992 A CN102174992 A CN 102174992A CN 2011100268492 A CN2011100268492 A CN 2011100268492A CN 201110026849 A CN201110026849 A CN 201110026849A CN 102174992 A CN102174992 A CN 102174992A
- Authority
- CN
- China
- Prior art keywords
- pressure
- logarithm
- pipeline
- delta
- temperature
- 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
Images
Abstract
The invention discloses a pressure curve feature extraction method for a pressure pipeline. The method is characterized by comprising the following steps of: representing a pressure sequence {Pn l n=1,..., n} by using a function P=alpha.t<beta> + epsilon in a relative short time delta T for the pressure pipeline, wherein the alpha, the beta and the epsilon are fitting parameters; performing logarithm domain transform on the acquired pressure data of the time period delta T, namely log (P) = alog (t) + b, wherein a and b are pressure logarithm curve-based system fitting parameters; dividing the time period delta T of the logarithm pressure curve into smaller time intervals delta t, setting M sampling points, selecting Qi as a representative point of the M sampling points, giving the Qi and the M sampling points in the time intervals delta t by using a median filter, namely Qi is equal to a median after the M sampling points are ordered; taking continuous L time intervals delta t, obtaining L pressure logarithm values {Qi l i=1,..., L} according to the steps, and combining the L values to form a pressure logarithm combination vector n= (Q1Q2,..., and QL)T serving as a feature vector of the logarithm pressure curve; and learning and training an identification engine on site or off line through the extracted pressure curve feature to accurately identify the seepage or leakage condition of the pipeline.
Description
Technical field
The present invention relates to pipeline leakage, leak the detection technology field, relate in particular to a kind of pressure diagram feature extracting method of pressure piping.
Background technique
Pressure piping, be meant and utilize certain pressure, be used for conveying gas or liquid, its scope dictates is the Maximum operating pressure gas that is greater than or equal to 0.1MPa (megapascal (MPa)), liquid gas, steam medium or liquid medium flammable, explosive, poisonous, that be corrosive etc., and nominal diameter is greater than the pipeline of 25mm.Pressure piping has in industrial and agricultural production very widely to be used.Because in the actual motion of pressure piping, the welded joint of pipeline, compressor, pump machine etc. the position be relatively weak position; And the partial pressure pipeline is operated in the pipeline section of operating conditions harshnesses such as high temperature, high pressure most probably; These all cause its possibility of accident occurrence bigger.
From safety in production and point of view of environment protection, the monitoring of the integrity of pressure piping is extremely important.The integrity of monitoring pressure pipeline, its most important mode be exactly the pressure size of real-time monitoring pressure pipeline, by Surveillance center the pressure diagram of the pressure piping that records is done and correspondingly to be analyzed and to handle, extract the pressure diagram feature of pressure piping, whether normal to judge pipeline pressure.In various monitor force pipeline integrity system, the pressure diagram feature of effectively extracting pressure piping is one of Core Feature of system, also directly has influence on monitor system performance.
Summary of the invention
For this reason, the present invention has designed a kind of pressure diagram feature extracting method of pressure piping, basic functional principle of the present invention is: arrive pressure data by sensor acquisition, with the pressure data that collects (perhaps do after the logarithmic transformation pressure to logarithmic data) grouping, select the characteristic point of a representative point as this grouping, some groups of sequential selections, obtain corresponding characteristic point, this certain characteristics point is joined together, as this section pressure data (or pressure is to logarithmic data) associating vector, unite the characteristic vector of vector as this section pressure data (or pressure is to logarithmic data) with this.
Real system operation proof adopts technology provided by the invention, and by monitoring with analyze and continue 20～50 minutes pressure data, the short distance pressure pipeline leakage aperture that system can detect closed at both ends is 0.1 millimeter a minute leakage.
One aspect of the present invention relates to a kind of pressure diagram feature extracting method of pressure piping, it is characterized in that: for pressure piping, and in the relatively short Δ T time, the pressure sequence Pn|n=1 ..., n} available functions P=α .t
β+ ε characterizes, and α, β, ε are fitting parameter, the pressure data of Δ T time period of being obtained is carried out the log-domain conversion, that is:
Log (P)=alog (t)+b, a wherein, b is the system's fitting parameter based on the pressure logarithmic curve;
With the time period Δ T of described logarithm pressure diagram, be divided into littler time lag Δ t, be provided with M sampling point, with Q
iBe the representative point of this M sampling point, Q
iProvide with median filter with this M sampling point in the Δ t time lag:
Q
iAfter=M the sampling point ordering, get intermediate value,
Get continuous L the Δ t time lag, obtain L pressure logarithm value { Q by above-mentioned steps
i| i=1 ..., L} joins together this L numerical value, constitutes a pressure logarithm associating vector Π=(Q
1Q
2... Q
L)
TCharacteristic vector as this logarithm pressure diagram; By the pressure diagram feature of extracting, to discerning the learning training that engine carries out scene or off-line, with the seepage or the leak case of correct identification pipeline.
On the other hand, relate to according to said method and carry out the system that level ground, airport Aviation Fuel pipeline network leak detects, it is characterized in that comprising: temperature transducer, pressure transducer, leak detector, and upper-position unit; Wherein, temperature transducer is used for the temperature of measurement line place environment, being used to the revising influence of ambient temperature to pipeline pressure, and sends this temperature value signal to leak detector by wired or wireless mode; Pressure transducer, be used for detecting in real time ducted pressure changing, pressure signal and temperature signal are converted to electrical signal, send leak detector to by wired or wireless mode, pressure transducer, it is installed on the conveyance conduit, contacts with sound source, be used for the sensing sound wave forms compression and sparse alternate when air is propagated acoustic pressure, obtain the higher pressure signal of precision; Leak detector, operation machine level ground pipeline network leak detection system on it; Upper-position unit, it is the monitor terminal of system.
Description of drawings
Below in conjunction with drawings and embodiments the present invention is described in further detail.
Fig. 1 independently monitors and data capture (SCADA) the system machine level ground pipe network overall system block diagram that leaks hunting;
Fig. 2 airport level ground short distance closed conduct system core handling process of leaking hunting;
Fig. 3 pressure diagram and pressure logarithmic curve (schematic representation);
The associating vector leaching process of Fig. 4 logarithm pressure diagram;
Fig. 5 makes up schematic representation based on machine level ground, airdrome seepage, the leakage search engine of neuron network and pressure diagram feature.
Embodiment
The pressure diagram feature extracting method of pressure piping of the present invention is applicable to that all utilization pipelines carry out the pipe detection system of liquid, gas transmission, in order to set forth the concrete steps of its method better, with based on the machine level ground pipeline network leak detection system of pressure monitoring method in order to describe, but those skilled in the art know, and it does not limit Applicable scope of the present invention.
As shown in Figure 1, it is the general structure block diagram based on the machine level ground pipeline network leak detection system of pressure monitoring method, and this independent supervisory control and data acqui sition system (SCADAS) machine level ground pipe network system of leaking hunting comprises temperature transducer, pressure transducer, leak detector, and upper-position unit.Wherein, temperature transducer is used for the temperature of measurement line place environment, being used to the revising influence of ambient temperature to pipeline pressure, and sends this temperature value signal to leak detector by wired or wireless mode; Pressure transducer, be used for detecting in real time ducted pressure changing, pressure signal and temperature signal are converted to electrical signal, send leak detector to by wired or wireless mode, it is installed on pressure transducer (being also referred to as pressure transmitter sometimes) on the conveyance conduit, contact with sound source, be used for the sensing sound wave forms compression and sparse alternate when air is propagated acoustic pressure, obtain the higher pressure signal of precision.Usually, carried acoustic signals in the pressure signal, but because pressure signal is a bigger force value, and the pressure surge that sound wave causes is a very little force value.For example, its value of a standard atmospheric pressure (air pressure that refers to sea level under standard atmosphere condition) is 101325 Pascals in air, is about 10
5Pascal, and the air pressure variations that the minimum sound that people's ear can be heard causes is about 2*10
-5Pascal); Leak detector, leak detector are the corn module of whole system, operation machine level ground pipeline network leak detection system on it; Upper-position unit (also claiming monitor terminal), it is the monitor terminal of system, operator and caretaker can realize that systems management and parameter are provided with etc. operation and management are when pipe leakage (or leakage) taking place, in time send reports to the police and start emergency processing.
About temperature transducer, pressure transducer, leak detector, and the information between the upper-position unit connects, transmission means, for art technology, intuitivism apprehension with reference to the accompanying drawings just can be learnt, belongs to the common practise of related domain, just repeats no more here.
As shown in Figure 2, it is to run on level ground, airport short distance closed conduct in the pipeline network leak detection system of the operation machine level ground system core handling process of leaking hunting, it comprises following treatment step: (1) is finished the signal conversion that will receive by sensor and front end circuit and is become electrical signal, passes to function block 2 after the amplification; (2) by audio data AD conversion and data pretreatment, finish the AD conversion, and carry out the work of data necessary robustness processing, such as adaptive-filtering, go dry etc.; (3) noise cancellation module eliminate to be disturbed and background noise, the acoustic disturbance that produces as normal switch valve etc.; (4) signal after handling by pressure data calibration and pressure diagram signature search engine, carries out pressure signal feature extraction and identification, thus the seepage of finishing or incident of leakage identification, and determine roughly time of origin of incident, distance etc.Wherein, the pipeline pressure data calibration need to be considered the influence to pipeline pressure of ambient temperature, caliber, pipeline wall thickness, thereby is reduced the influence that extraneous objective factor is analyzed pressure data.
For the pipe leakage detection system, by above-mentioned workflow as can be known, the feature extraction of its pressure diagram, play important effect in its system determination, directly influence the work validity and the work accuracy of system, the present invention is for the pressure diagram feature extraction, designed a kind of pressure diagram feature extracting method of pressure piping, with effective extraction pressure diagram feature, greatly improved the accuracy that characterizes the pressure diagram variation, improved the precision of revealing event recognition.Its extracting method specifies as follows:
Pressure piping pressure data characteristic vector extracting method
Because actual pressure (pressure is calibrated) follows a kind of nonlinear mode to decay, for pressure piping, in the relatively short Δ T time, the pressure sequence Pn|n=1 ..., n} available functions P=α .t
β+ ε, α, β, ε are fitting parameter.Because of the rather inconvenience when the problem analysis of above-mentioned exponential form, thus the pressure data of Δ T time period of being obtained is carried out the log-domain conversion, that is:
Log (P)=alog (t)+b, a wherein, b is the system's fitting parameter based on the pressure logarithmic curve
After handling like this, curvilinear motion is mild relatively, the also more or less freely discontinuity of seeing pressure drop, as shown in Figure 3, and such log-domain pressure conversion curve, its feature is more obvious, more helps follow-up feature extraction and trains, discerns.
In addition, in order to save computing time, to reduce memory space etc.,, also can not need pressure data is done the logarithm computing according to the actual design needs.
Pressure piping pressure logarithm data characteristics vector formation method
Because most pressure piping supervisory systems requires high reliability, too extensive if be criterion with the pressure logarithmic decrement simply just, can't satisfy the practical application needs at all, as easy as rolling off a log causing failed to judge, judged by accident.Given this, data are through after the above-mentioned processing, must be through process of refinement more.For characterizing the characteristics of logarithm pressure diagram, improve system reliability, the present invention proposes first, with the time period Δ T of described logarithm pressure diagram, is divided into the littler time lag (Δ t is provided with M sampling point), with Q
iBe the representative point of this M sampling point, as shown in Figure 4.
For reducing the influence of System noise, Q
iProvide with median filter with this M sampling point in the Δ t time lag:
Q
iAfter=M the sampling point ordering, get intermediate value,
Get continuous L the Δ t time lag, obtain L pressure logarithm value { Q by above-mentioned steps
i| i=1 ..., L} joins together this L numerical value, constitutes a pressure logarithm associating vector Π=(Q
1Q
2... Q
L)
TCharacteristic vector as this logarithm pressure diagram.
Construction method based on the pressure diagram signature search engine of neuron network
Neural network model is widely used in the System Discrimination field with its powerful parallel data Processing capacity and learning ability, can describe complicated signal model effectively, has obtained using widely in pallern identification and coupling field.The present invention is with pressure logarithm associating vector Π=(Q
1Q
2... Q
L)
T, or only working pressure is united vector, is the neuron network input end, and as shown in Figure 5, it is machine level ground, airdrome seepage, leakage search engine structure structural drawing based on neuron network and pressure diagram feature, can obtain from Fig. 5, and network is output as
w
iBe weight factor, φ
i(Q
1Q
2... Q
L) be the neuronal kernel function, native system neural network model storehouse is by parameter set { w
i, φ
i(x) } determine.By the pressure diagram feature of extracting, neuron network is carried out the learning training of scene or off-line, can correctly discern the seepage or the leak case of pipeline.Adopt technology provided by the invention, by monitoring with analyze to continue 20～50 minutes pressure and temperature data, it is 0.1 millimeter minute leakage that system can detect the aperture.
The present invention obtains the advantage of following several respects at least:
Extracted the feature of pressure piping pressure diagram effectively.
Greatly improved and characterized the accuracy that pressure diagram changes, the leakage aperture of 0.1 millimeter of the short distance pressure piping of closed at both ends can be detected.
Claims (2)
1. the pressure diagram feature extracting method of a pressure piping is characterized in that: for pressure piping, and in the relatively short Δ T time, the pressure sequence Pn|n=1 ..., n} available functions P=α .t
β+ ε characterizes, and α, β, ε are fitting parameter, the pressure data of Δ T time period of being obtained is carried out the log-domain conversion, that is:
Log (P)=alog (t)+b, a wherein, b is the system's fitting parameter based on the pressure logarithmic curve;
With the time period Δ T of described logarithm pressure diagram, be divided into littler time lag Δ t, be provided with M sampling point, with Q
iBe the representative point of this M sampling point, Q
iProvide with median filter with this M sampling point in the Δ t time lag:
Q
iAfter=M the sampling point ordering, get intermediate value,
Get continuous L the Δ t time lag, obtain L pressure logarithm value { Q by above-mentioned steps
i| i=1 ..., L} joins together this L numerical value, constitutes a pressure logarithm associating vector Π=(Q
1Q
2... Q
L)
TCharacteristic vector as this logarithm pressure diagram; By the pressure diagram feature of extracting, to discerning the learning training that engine carries out scene or off-line, with the seepage or the leak case of correct identification pipeline.
2. carry out the system that level ground, airport Aviation Fuel pipeline network leak detects according to the method for claim 1, it is characterized in that comprising: temperature transducer, pressure transducer, leak detector, and upper-position unit; Wherein, temperature transducer is used for the temperature of measurement line place environment, being used to the revising influence of ambient temperature to pipeline pressure, and sends this temperature value signal to leak detector by wired or wireless mode; Pressure transducer, be used for detecting in real time ducted pressure changing, pressure signal and temperature signal are converted to electrical signal, send leak detector to by wired or wireless mode, pressure transducer, it is installed on the conveyance conduit, contacts with sound source, be used for the sensing sound wave forms compression and sparse alternate when air is propagated acoustic pressure, obtain the higher pressure signal of precision; Leak detector, operation machine level ground pipeline network leak detection system on it; Upper-position unit, it is the monitor terminal of system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011100268492A CN102174992A (en) | 2011-01-25 | 2011-01-25 | Pressure curve feature extraction method for pressure pipeline |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011100268492A CN102174992A (en) | 2011-01-25 | 2011-01-25 | Pressure curve feature extraction method for pressure pipeline |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102174992A true CN102174992A (en) | 2011-09-07 |
Family
ID=44518207
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011100268492A Pending CN102174992A (en) | 2011-01-25 | 2011-01-25 | Pressure curve feature extraction method for pressure pipeline |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102174992A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105889763A (en) * | 2014-10-10 | 2016-08-24 | 保定市金迪科技开发有限公司 | Detecting device and detecting method for pipeline leakage |
| CN106641737A (en) * | 2016-12-20 | 2017-05-10 | 重庆国际复合材料有限公司 | Heat exchange tube system and pressure maintaining and testing device and method of shell-and-tube heat exchanger |
| CN108199954A (en) * | 2018-01-23 | 2018-06-22 | 南京万荣立体绿化工程有限公司 | Irrigate intelligent monitor system |
| CN110823338A (en) * | 2018-08-10 | 2020-02-21 | 河南工业大学 | Granary detection method and system based on bottom surface single-circle standard deviation logarithm model |
| CN111578149A (en) * | 2020-05-25 | 2020-08-25 | 重庆西美仪器仪表有限公司 | Gas pipeline leakage monitoring method, device and system and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS58100730A (en) * | 1981-12-11 | 1983-06-15 | Sumitomo Metal Ind Ltd | Method and device for detecting leakage of pipeline |
| EP0417763A2 (en) * | 1989-09-12 | 1991-03-20 | Fabriques De Tabac Reunies S.A. | Procedure to assess the existence of a fixation between to end-points of pipes |
| CN1435678A (en) * | 2002-07-04 | 2003-08-13 | 东北大学 | Method and device for intelligent diagnosis and location of leakage fault of fluid delivery pipeline |
| US20070251325A1 (en) * | 2006-04-26 | 2007-11-01 | Joseph King | Impulse Response Pressure Transducer |
-
2011
- 2011-01-25 CN CN2011100268492A patent/CN102174992A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS58100730A (en) * | 1981-12-11 | 1983-06-15 | Sumitomo Metal Ind Ltd | Method and device for detecting leakage of pipeline |
| EP0417763A2 (en) * | 1989-09-12 | 1991-03-20 | Fabriques De Tabac Reunies S.A. | Procedure to assess the existence of a fixation between to end-points of pipes |
| CN1435678A (en) * | 2002-07-04 | 2003-08-13 | 东北大学 | Method and device for intelligent diagnosis and location of leakage fault of fluid delivery pipeline |
| US20070251325A1 (en) * | 2006-04-26 | 2007-11-01 | Joseph King | Impulse Response Pressure Transducer |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105889763A (en) * | 2014-10-10 | 2016-08-24 | 保定市金迪科技开发有限公司 | Detecting device and detecting method for pipeline leakage |
| CN105889763B (en) * | 2014-10-10 | 2017-12-15 | 保定市金迪科技开发有限公司 | The detection device and its detection method of a kind of pipe leakage |
| CN106641737A (en) * | 2016-12-20 | 2017-05-10 | 重庆国际复合材料有限公司 | Heat exchange tube system and pressure maintaining and testing device and method of shell-and-tube heat exchanger |
| CN108199954A (en) * | 2018-01-23 | 2018-06-22 | 南京万荣立体绿化工程有限公司 | Irrigate intelligent monitor system |
| CN110823338A (en) * | 2018-08-10 | 2020-02-21 | 河南工业大学 | Granary detection method and system based on bottom surface single-circle standard deviation logarithm model |
| CN111578149A (en) * | 2020-05-25 | 2020-08-25 | 重庆西美仪器仪表有限公司 | Gas pipeline leakage monitoring method, device and system and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102168809A (en) | Method and system for detecting leakage of parking apron aviation gasoline pipe network based on pressure and temperature analysis | |
| US9797799B2 (en) | Intelligent adaptive system and method for monitoring leakage of oil pipeline networks based on big data | |
| CN101684894B (en) | Method and device for monitoring pipeline leakage | |
| CN101832472B (en) | System implementing pipeline leak detection by utilizing infrasonic wave | |
| CN101561081B (en) | Method for detecting and positioning leakage of oil-gas pipeline by utilizing autonomous navigation robot | |
| CN201273457Y (en) | Pipe leakage monitoring device | |
| CN103836347B (en) | Leakage monitoring device and method for crude oil gathering pipelines | |
| CN100456010C (en) | Method for detecting leakage of oil gas pipe based on pressure signal knee | |
| CN102174992A (en) | Pressure curve feature extraction method for pressure pipeline | |
| CN202082629U (en) | Natural gas pipeline leakage monitoring system | |
| CN101625071A (en) | Method for measuring and locating leakage of gas pipelines | |
| CN101592288B (en) | Method for identifying pipeline leakage | |
| AU2017393649B2 (en) | Leakage positioning method based on speed difference | |
| CN103438359A (en) | Oil pipeline leakage detection and positioning system | |
| CN111271610B (en) | Liquid pipeline leakage detection early warning device and method | |
| CN202469500U (en) | Natural gas transmission pipeline leakage monitoring and positioning device | |
| CN104500984A (en) | Sub-high pressure A gas pipeline leakage monitoring system | |
| CN202209524U (en) | Airport aviation fuel pipe network leak detecting system based on pressure temperature analysis | |
| CN104100842A (en) | Pipeline monitoring device and system based on distributed fiber sensors and acoustic wave | |
| CN103032626A (en) | System and method for diagnosing fault of adjusting valve | |
| CN201723985U (en) | Infrasonic-wave signal monitoring device of built-in voiceprint recognition algorithm | |
| CN206756926U (en) | A kind of Detector for Stray Currents | |
| CN107504374A (en) | Gas pipeline acoustic monitoring system | |
| CN211478167U (en) | Fault diagnosis device and fault diagnosis system | |
| CN104100840B (en) | Monitoring pipeline blocking system and under strong jamming the detection method of jam signal |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20110907 |