CN117848475A - Oil and gas pipeline vibration monitoring and identifying method and system - Google Patents
Oil and gas pipeline vibration monitoring and identifying method and system Download PDFInfo
- Publication number
- CN117848475A CN117848475A CN202311767847.8A CN202311767847A CN117848475A CN 117848475 A CN117848475 A CN 117848475A CN 202311767847 A CN202311767847 A CN 202311767847A CN 117848475 A CN117848475 A CN 117848475A
- Authority
- CN
- China
- Prior art keywords
- signal
- long
- time
- vibration
- short
- 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
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000012544 monitoring process Methods 0.000 title claims abstract description 21
- 230000002159 abnormal effect Effects 0.000 claims abstract description 37
- 238000001514 detection method Methods 0.000 claims abstract description 21
- 239000013307 optical fiber Substances 0.000 claims abstract description 20
- 238000000253 optical time-domain reflectometry Methods 0.000 claims abstract description 12
- 230000006378 damage Effects 0.000 claims abstract description 11
- 230000007774 longterm Effects 0.000 claims description 31
- 238000013528 artificial neural network Methods 0.000 claims description 13
- 238000012549 training Methods 0.000 claims description 12
- 230000004913 activation Effects 0.000 claims description 9
- 210000002569 neuron Anatomy 0.000 claims description 9
- 239000013598 vector Substances 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 6
- 238000005070 sampling Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 5
- 230000007613 environmental effect Effects 0.000 claims description 3
- 239000007789 gas Substances 0.000 description 22
- 230000006872 improvement Effects 0.000 description 8
- 230000006870 function Effects 0.000 description 6
- 230000003287 optical effect Effects 0.000 description 5
- 238000012706 support-vector machine Methods 0.000 description 5
- 230000004044 response Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 239000003345 natural gas Substances 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 239000003208 petroleum Substances 0.000 description 1
- 239000003209 petroleum derivative Substances 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H9/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves by using radiation-sensitive means, e.g. optical means
- G01H9/004—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves by using radiation-sensitive means, e.g. optical means using fibre optic sensors
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention provides a method and a system for monitoring and identifying vibration of an oil and gas pipeline, wherein the method comprises the steps of firstly, acquiring vibration signals near the oil and gas pipeline in real time through a phi-OTDR signal detection system; then short-time resolution feature detection is carried out on the vibration signal, and if abnormal events are detected, long-time resolution features of the vibration signal are extracted; then, the extracted long-time analysis features are input into a plurality of trained network identification classifiers and classification results are output respectively, if the classification result output by each network identification classifier is an abnormal event, the current vibration signal is finally judged to be an artificial destruction event, otherwise, the current vibration signal is judged to be an interference event; the invention can effectively extract and distinguish the disturbance signal characteristics, thereby improving the recognition precision of the distributed optical fiber vibration sensing system and reducing the false alarm rate.
Description
Technical Field
The invention relates to the technical field of distributed optical fiber sensing signal identification, in particular to an oil and gas pipeline vibration monitoring and identifying method and an identifying system.
Background
Oil and gas energy sources such as petroleum and natural gas play an important role in national economy development, and among a plurality of oil and gas transportation modes, pipeline transportation is one of the important transportation modes of oil and gas transportation. Because the pipeline spans a plurality of complicated areas and terrains, the route is long, once the phenomenon of stealing occurs, firstly, the economic loss which is difficult to measure is generated, and meanwhile, if accidents such as oil and gas pipeline breakage and the like are caused during the stealing, the social hazard is not ignored, so that the method has great practical significance on how to effectively monitor the phenomenon of stealing and excavating nearby the oil pipeline. The distributed optical fiber vibration sensing can well solve the problems, has high sensitivity and is passive in the whole process, and dynamic change information such as vibration or sound on an optical cable can be continuously detected.
Although the distributed optical fiber vibration sensing system can effectively monitor the petroleum pipeline and locate abnormal events, the field environment is quite complex, various types of interference events exist, false alarm or missing alarm of the sensing system is easy to occur, and the application of the sensing system is severely restricted.
Therefore, there is a need for an oil and gas pipeline vibration monitoring and identifying method and an identifying system, which can effectively extract and distinguish disturbance signal characteristics, so as to improve the identifying precision of a distributed optical fiber vibration sensing system and reduce the false alarm rate.
Disclosure of Invention
The invention aims to provide an oil and gas pipeline vibration monitoring and identifying method and an identifying system, and aims to solve the technical problems that the traditional distributed optical fiber vibration sensing system is low in identifying precision and is easy to generate false alarm.
In order to achieve the above object, in a first aspect, the present invention provides a method for monitoring and identifying vibration of an oil and gas pipeline, comprising the steps of:
s1, acquiring vibration signals near an oil gas pipeline in real time through a phi-OTDR signal detection system;
s2, short-time resolution feature detection is carried out on the vibration signal, and if an abnormal event is detected, long-time resolution features of the vibration signal are extracted;
s3, inputting the extracted long-term analysis features into a plurality of trained network identification classifiers and respectively outputting classification results, if the classification result output by each network identification classifier is an abnormal event, finally judging the current vibration signal as an artificial destruction event, and otherwise, judging the current vibration signal as an interference event.
As a further improvement of the above, in step S2, short-time main impact strength discrimination characteristics a of the vibration signal are first extracted at the time of short-time discrimination characteristic detection 1 If the main impact strength is short, distinguishing characteristic A 1 If the vibration signal is smaller than the experience threshold thr, judging that the current vibration signal is an environmental noise signal;
if the main impact strength is short, distinguishing characteristic A 1 And if the vibration signal is larger than the experience threshold thr, the current vibration signal is primarily judged to be an abnormal event signal.
As a further improvement of the scheme, the short-time main impact strength distinguishing characteristic A 1 The extraction method comprises the following steps:
will vibrate the signal w i The amplitude interval of (2) is divided into n consecutive sub-intervals, each of which is denoted s l ；
Then, the vibration signal w is counted i At each subinterval s l Number of sampling points m l And obtaining a corresponding histogram, normalizing the histogram and marking the normalized histogram as p, p= { p l (l＝1,2...n)}，Wherein M is the total sampling point number;
in interval s l In the case of p l (l∈[1,n]) Taking the maximum value at l=q, the short-time main impact strength discrimination characteristic a 1 The following formula is shown:
as a further optimization of the above scheme, subinterval s l The amplitude intervals of (a) are expressed as follows:
s l ＝[(l-1)*averlength+min(w i ),l*averlenth+min(w i )](1≤l≤n)
wherein averlength is the width of each subinterval, w i The amplitude interval is [ min (w) i ),max(w i )]，min(w i ) Represents the minimum amplitude, max (w i ) Representing the maximum amplitude.
As a further improvement of the above-described aspect, in step S2, the step of determining that the vibration signal is an abnormal event signal includes:
distinguishing characteristic A according to short-time main impact strength 1 Judging whether the current signal has instant impact or not;
if instantaneous impact exists, the short-time impact strength distinguishing characteristic A of the current signal is extracted 2 ；
Discriminating characteristics A based on short time impact strength 2 And judging whether the signal is an abnormal event signal or not.
As a further improvement of the above solution, in step S2, the long-term analysis feature includes:
long term main impact strength discrimination feature B 1 Long time impact strength distinguishing character mark B 2 The long-term amplitude proportion characteristic is marked as B 3 Long-term signal frequency characteristics B 4 Long-term background intensity resolution feature B 5 Number of severe impacts of long-term signal B 6 ；
The long-time analysis feature corresponding to each long-time signal forms a feature vector T= [ B ] 1 ,B 2 ,B 3 ,B 4 ,B 5 ,B 6 ]。
As a further improvement of the above solution, in step S3, the network identification classifier includes an SVM classifier, a BP neural network classifier, and a bayesian network classifier, and the training process of each classifier is separately and independently completed.
As a further improvement of the scheme, when the BP neural network classifier trains the BP neural network, the depth of the neural network is set to be 4 layers;
the first hidden layer is provided with 5 neuron nodes, and the activation function is set as tan sig;
the second hidden layer is provided with 3 neuron nodes, and the activation function is set as tan sig;
the output layer is set as two neuron nodes, and the activation function is set as purelin;
the category of the artificial destroy event is identified as 1, and the category of the disturbance event is identified as 2.
As a further improvement of the above solution, the extracted long-term analysis features need to be discretized before being input into the bayesian network classifier.
As a further improvement of the above solution, before step S3, training each network identification classifier is further included, where the training step includes:
collecting vibration signals near the oil gas pipeline through a phi-OTDR signal detection system, preprocessing the vibration signals,
constructing a long-time signal sample set for the preprocessed vibration information;
dividing the long-time signal into a plurality of short-time signals;
extracting short-time resolution features of the short-time signals;
extracting and constructing a long-time resolution feature vector;
and respectively inputting the extracted long-time analysis feature vectors into each network identification classifier, and independently carrying out learning training on each network identification classifier to obtain the optimization parameters corresponding to each network identification classifier.
In a second aspect, the invention also provides an oil and gas pipeline vibration monitoring and identifying system, which comprises
The signal acquisition module is used for acquiring an optical fiber vibration signal caused by disturbance along the oil and gas pipeline;
the short-time resolution feature detection module is used for detecting whether the optical fiber vibration signal is an abnormal event signal or not;
the long-time feature extraction module is used for extracting long-time analysis features of the optical fiber vibration signals;
the network identification classifier module comprises a plurality of network identification classifiers and is used for identifying and classifying abnormal event signals according to the long-time resolution features;
and the judging module comprehensively judges the artificial destruction event signals and the interference signals according to the classification output results of the network identification classifiers.
Due to the adoption of the technical scheme, the invention has the beneficial effects that:
the invention provides a method for monitoring and identifying vibration of an oil and gas pipeline, which comprises the steps of firstly, collecting vibration signals near the oil and gas pipeline in real time through a phi-OTDR signal detection system; then short-time resolution feature detection is carried out on the vibration signal, and if abnormal events are detected, long-time resolution features of the vibration signal are extracted; then, the extracted long-time analysis features are input into a plurality of trained network identification classifiers and classification results are output respectively, if the classification result output by each network identification classifier is an abnormal event, the current vibration signal is finally judged to be an artificial destruction event, otherwise, the current vibration signal is judged to be an interference event; according to the invention, the short-time distinguishing features and the long-time distinguishing features of the vibration signals in the time domain and the transformation domain are extracted by combining the short-time signals with the long-time signals, the abnormal event is detected primarily through the short-time distinguishing features, the long-time distinguishing features of the extracted signals are input into a plurality of trained network identification classifiers, and finally the classification and identification results of the network identification classifiers are fused, so that whether the event is a human destruction event or an interference event is judged, and the false alarm rate of the abnormal event signals can be reduced, and the identification precision of the distributed optical fiber vibration sensing system is improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that other drawings may be obtained from the structures shown in these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram of a phi-OTDR signal detection system of the present disclosure;
FIG. 2 is a schematic diagram of a fusion model of a plurality of network identification classifiers disclosed in the present invention;
FIG. 3 is a schematic diagram of a frame of an oil and gas pipeline vibration monitoring and identifying method provided by the invention.
The achievement of the object, functional features and advantages of the present invention will be further described with reference to the drawings in connection with the embodiments.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all embodiments of the invention. All other embodiments, based on the embodiments of the invention, which are apparent to those of ordinary skill in the art without inventive faculty, are intended to be within the scope of the invention.
The technical solutions of the embodiments of the present invention may be combined with each other, but it is necessary to base the implementation of those skilled in the art, and when the combination of the technical solutions contradicts or cannot be implemented, it should be considered that the combination of the technical solutions does not exist and is not within the scope of protection claimed by the present invention.
Example 1:
referring to fig. 1, in order to better explain the technical scheme of the present invention, referring to fig. 1, firstly, a phi-OTDR signal detection system needs to be built near an oil pipeline to collect samples of artificial oil theft events and traffic vehicle interference events occurring near the oil pipeline, that is, a plurality of optical fiber vibration sensors need to be arranged along the oil pipeline, the optical fiber vibration sensors transmit received vibration signals to an optical signal demodulation device through a communication optical cable, and the optical signal demodulation device transmits the demodulated signals to a signal processing host for recognition;
referring to fig. 2 and fig. 3, the signal processing host executes the method for monitoring and identifying the vibration of the oil and gas pipeline, which comprises the following steps:
s1, acquiring vibration signals near an oil gas pipeline in real time through a phi-OTDR signal detection system;
specifically, the process of collecting vibration signals is as follows:
s11, light based on backward Rayleigh scatteringThe signal demodulation equipment takes a periodic trigger pulse as a signal acquisition period, the signal acquired in the time unit is an exponentially decaying OTDR track distributed along a detection optical cable, and the signal acquired in the kth trigger pulse period is set as follows: z is Z k ＝{t ki (i＝1,2,...,N)＝[t k1 ,t k2 ,...,t k3 ]I is a sequence number of a space acquisition point, and N is a data acquisition length in the whole monitoring range;
s12, triggering the collected pulse period, periodically refreshing the OTDR track of the space distribution, continuously accumulating M collected original OTDR tracks along with the increase of k, and constructing a space-time signal response matrix of space N dimension and time M dimension, namely: xy= { t ki (k＝1,2,...,M；i＝1,2,...,N)}；
S13, the time response signal of the ith space sensing node or acquisition node is expressed as:
w i ＝{t ki (k＝1,2,...,M)}；
transverse time sequence X of said time response signal k The spatial sequence signals representing the spatial points at the same moment, wherein the amplitude of each point in the spatial sequence corresponds to the information of the corresponding spatial point, and if a certain point in the spatial sequence is invaded, the amplitude of the corresponding point in the spatial sequence becomes larger;
longitudinal time sequence Y of the time response signal i A corresponding time series signal representing a certain point in space, if a certain point in space is invaded at a certain moment, the amplitude of the corresponding longitudinal time series at the certain moment suddenly becomes larger, and the corresponding longitudinal time series suddenly has an impact at the certain moment;
s2, short-time resolution feature detection is carried out on the vibration signal, and if an abnormal event is detected, long-time resolution features of the vibration signal are extracted;
specifically, in the detection of short-time resolution features, first, short-time main impact strength resolution feature A of vibration signal is extracted 1 If the main impact strength is short, distinguishing characteristic A 1 If the vibration signal is smaller than the experience threshold thr, judging that the current vibration signal is an environmental noise signal;
if the main impact strength is short, distinguishing characteristic A 1 If the vibration signal is larger than the experience threshold thr, the current vibration signal is primarily judged to be an abnormal event signal;
short time main impact strength discrimination feature A 1 The extraction method comprises the following steps:
will vibrate the signal w i The amplitude interval of (2) is divided into n continuous subintervals, each subinterval is denoted as sl, s l ＝[(l-1)*averlength+min(w i ),l*averlenth+min(w i )](1≤l≤n)；
Wherein averlength is the width of each subinterval, w i The amplitude interval is [ min (w) i ),max(w i )]，min(w i ) Represents the minimum amplitude, max (w i ) Representing the maximum amplitude;
then, the vibration signal w is counted i At each subinterval s l Number of sampling points m l And obtaining a corresponding histogram, normalizing the histogram and marking the normalized histogram as p, p= { p l (l＝1,2...n)}，Wherein M is the total sampling point number;
in interval s l In the case of p l (l∈[1,n]) Taking the maximum value at l=q, the short-time main impact strength discrimination characteristic a 1 The following formula is shown:
s3, inputting the extracted long-term analysis features into a plurality of trained network identification classifiers and respectively outputting classification results, if the classification result output by each network identification classifier is an abnormal event, finally judging the current vibration signal as an artificial destruction event, otherwise judging the current vibration signal as an interference event;
specifically, in this embodiment, the network identification classifier includes an SVM classifier, a BP neural network classifier, and a bayesian network classifier, when the output classification results of the three network identification classifiers are simultaneously artificial oil theft mining events, the abnormal event is judged to be an artificial oil theft mining event, otherwise, the abnormal event is judged to be a traffic vehicle interference event;
according to the invention, the short-time distinguishing features and the long-time distinguishing features of the vibration signals in the time domain and the transformation domain are extracted by combining the short-time signals with the long-time signals, the abnormal event is detected primarily through the short-time distinguishing features, then the long-time distinguishing features of the extracted signals are input into a plurality of trained network identification classifiers, finally the classification and identification results of the network identification classifiers are fused, and finally the artificial oil theft event and the traffic vehicle interference event are judged, so that the false alarm rate of the abnormal event signals can be reduced, and the identification precision of the distributed optical fiber vibration sensing system is improved.
As a preferred embodiment, in step S2, the step of determining that the vibration signal is an abnormal event signal includes:
distinguishing characteristic A according to short-time main impact strength 1 Judging whether the current signal has instant impact or not;
if instantaneous impact exists, the short-time impact strength distinguishing characteristic A of the current signal is extracted 2 ；
Discriminating characteristics A based on short time impact strength 2 Whether the abnormal event signal contains an artificial oil theft event and a traffic vehicle interference event is distinguished.
As a preferred embodiment, in step S2, the long-term analysis feature includes:
long term main impact strength discrimination feature B 1 Long time impact strength distinguishing character mark B 2 The long-term amplitude proportion characteristic is marked as B 3 Long-term signal frequency characteristics B 4 Long-term background intensity resolution feature B 5 Number of severe impacts of long-term signal B 6 The method comprises the steps of carrying out a first treatment on the surface of the Then the long-time analysis feature corresponding to each long-time signal forms a feature vector T= [ B ] 1 ,B 2 ,B 3 ,B 4 ,B 5 ,B 6 ]；
Long term main impact strength discrimination feature B 1 The overall impact strength of the long-time signal is effectively measured; from the frequency perspectiveStarting, extracting energy characteristics of long-time wavelet packet to reflect frequency characteristics of long-time signal B 4 The method comprises the steps of carrying out a first treatment on the surface of the Long-term background intensity resolution feature B 5 The background noise intensity of two typical abnormal events is effectively measured.
As a preferred embodiment, in step S3, the training process of each classifier is separately and independently completed, and the same sample may be used to train each classifier;
in particular, when training a Support Vector Machine (SVM) sub-classifier, the structure and parameters of the SVM classifier are trained by long-term analysis of a large number of samples,
in the embodiment, when the BP neural network classifier trains the BP neural network, the classification of the artificial destruction event is marked as 1, the classification of the interference event is marked as 2, and the depth of the neural network is set as 4 layers;
the first hidden layer is provided with 5 neuron nodes, and the activation function is set as tan sig;
the second hidden layer is provided with 3 neuron nodes, and the activation function is set as tan sig;
the output layer is set as two neuron nodes, and the activation function is set as purelin;
the learning rate was set to 0.01, the maximum training was set to 60000, the learning accuracy was set to 0.0001,
and adopts a ten-fold cross verification mode to evaluate the performance of the BP neural network sub-classifier;
in the real-time vibration monitoring and identifying stage, the categories of the abnormal events can be automatically identified and classified only by inputting the extracted long-time analysis features into the BP neural network after training.
As a preferred embodiment, the identification classification is performed by a Bayesian network, since in the Bayesian network the individual network nodes are discrete, whereas all long-term analysis features extracted by the present invention except B 6 The method is continuous, so that before the extracted long-term analysis features are input into a Bayesian network classifier, the extracted long-term analysis features are subjected to discretization treatment and then are trained among partitions;
the structure learning and the parameter learning of the Bayesian network are the cores of the Bayesian network, and the invention analyzes the mutual dependency relationship between each node from the mutual information and the conditional mutual information angle between each node, and when the mutual information of two nodes is larger, the correlation degree of the two nodes is larger;
when the Bayesian network sub-classification is constructed, only abnormal events (artificial oil theft events and traffic vehicle interference events) are considered as well; in the embodiment, the performance of the Bayesian network sub-classifier is also evaluated by adopting a ten-fold empirical crossover mode;
in the real-time vibration monitoring and identifying stage, the extracted long-time analysis features are input into a trained Bayesian network sub-classifier after discretization, so that the categories of abnormal events can be automatically identified and classified.
As a preferred embodiment, before step S3, training each network identification classifier is further included, see fig. 3, where the training step includes:
collecting vibration signals near the oil gas pipeline through a phi-OTDR signal detection system, preprocessing the vibration signals,
constructing a long-time signal sample set for the preprocessed vibration information;
dividing the long-time signal into a plurality of short-time signals;
extracting short-time resolution features of the short-time signals;
extracting and constructing a long-time resolution feature vector;
and respectively inputting the extracted constructed long-term analysis feature vectors into each network identification classifier, and independently carrying out learning training on each network identification classifier to obtain the optimization parameters corresponding to each network identification classifier.
Example 2:
the invention also provides an oil and gas pipeline vibration monitoring and identifying system, which comprises:
the signal acquisition module is used for acquiring an optical fiber vibration signal caused by disturbance along the oil and gas pipeline;
the short-time resolution feature detection module is used for detecting whether the optical fiber vibration signal is an abnormal event signal or not;
the long-time feature extraction module is used for extracting long-time analysis features of the optical fiber vibration signals;
the network identification classifier module comprises a plurality of network identification classifiers and is used for identifying and classifying abnormal event signals according to the long-time resolution features;
and the judging module is used for comprehensively judging the artificial destruction event signals and the interference signals by merging the classification output results of the network identification classifiers.
According to the oil-gas pipeline vibration monitoring and identifying system, short-time resolution features and long-time resolution features of vibration signals in time domains and transformation domains are extracted, then the long-time resolution features of the extracted signals are input into a trained network identification classifier module, finally the classification and identification results of all the network identification classifiers are fused, and finally whether an artificial destruction event or an interference event is judged, so that the false alarm rate of abnormal event signals can be reduced, and the identification precision of the distributed optical fiber vibration sensing system is improved.
The foregoing description of the preferred embodiments of the present invention should not be construed as limiting the scope of the invention, but rather as utilizing equivalent structural changes made in the description of the present invention and the accompanying drawings or directly/indirectly applied to other related technical fields under the inventive concept of the present invention.
Claims (10)
1. The method for monitoring and identifying the vibration of the oil and gas pipeline is characterized by comprising the following steps:
s1, acquiring vibration signals near an oil gas pipeline in real time through a phi-OTDR signal detection system;
s2, short-time resolution feature detection is carried out on the vibration signal, and if an abnormal event is detected, long-time resolution features of the vibration signal are extracted;
s3, inputting the extracted long-term analysis features into a plurality of trained network identification classifiers and respectively outputting classification results, if the classification result output by each network identification classifier is an abnormal event, finally judging the current vibration signal as an artificial destruction event, and otherwise, judging the current vibration signal as an interference event.
2. The method according to claim 1, wherein in step S2, short-time main impact strength resolution characteristic a of the vibration signal is first extracted when short-time resolution characteristic detection is performed 1 If the main impact strength is short, distinguishing characteristic A 1 If the vibration signal is smaller than the experience threshold thr, judging that the current vibration signal is an environmental noise signal;
if the main impact strength is short, distinguishing characteristic A 1 And if the vibration signal is larger than the experience threshold thr, the current vibration signal is primarily judged to be an abnormal event signal.
3. The method for monitoring and identifying vibration of oil and gas pipeline according to claim 2, wherein the short-time main impact strength distinguishing characteristic A 1 The extraction method comprises the following steps:
will vibrate the signal w i The amplitude interval of (2) is divided into n consecutive sub-intervals, each of which is denoted s l ；
Then, the vibration signal w is counted i At each subinterval s l Number of sampling points m l And obtaining a corresponding histogram, normalizing the histogram and marking the normalized histogram as p, p= { p l (l＝1,2...n)}，Wherein M is the total sampling point number;
in interval s l In the case of p l (l∈[1,n]) Taking the maximum value at l=q, the short-time main impact strength discrimination characteristic a 1 The following formula is shown:
4. a method of monitoring and identifying vibration of an oil and gas pipeline according to claim 3, wherein the method comprises the steps ofInterval s l The amplitude intervals of (a) are expressed as follows:
s l ＝[(l-1)*averlength+min(w i ),l*averlenth+min(w i )](1≤l≤n)
wherein averlength is the width of each subinterval, w i The amplitude interval is [ min (w) i ),max(w i )]，min(w i ) Represents the minimum amplitude, max (w i ) Representing the maximum amplitude.
5. The method according to claim 2, wherein in step S2, the step of determining that the vibration signal is an abnormal event signal includes:
distinguishing characteristic A according to short-time main impact strength 1 Judging whether the current signal has instant impact or not;
if instantaneous impact exists, the short-time impact strength distinguishing characteristic A of the current signal is extracted 2 ；
Discriminating characteristics A based on short time impact strength 2 And judging whether the signal is an abnormal event signal or not.
6. The method according to any one of claims 1-5, wherein in step S2, the long-term analysis feature comprises:
long term main impact strength discrimination feature B 1 Long time impact strength distinguishing character mark B 2 The long-term amplitude proportion characteristic is marked as B 3 Long-term signal frequency characteristics B 4 Long-term background intensity resolution feature B 5 Number of severe impacts of long-term signal B 6 ；
The long-time analysis feature corresponding to each long-time signal forms a feature vector T= [ B ] 1 ,B 2 ,B 3 ,B 4 ,B 5 ,B 6 ]。
7. The method according to any one of claims 1 to 5, wherein in step S3, the network identification classifier includes an SVM classifier, a BP neural network classifier, and a bayesian network classifier, and the training process of each classifier is separately and independently completed.
8. The method for monitoring and identifying vibration of an oil and gas pipeline according to claim 7, wherein the depth of the BP neural network classifier is set to be 4 layers when the BP neural network is trained;
the first hidden layer is provided with 5 neuron nodes, and the activation function is set as tan sig;
the second hidden layer is provided with 3 neuron nodes, and the activation function is set as tan sig;
the output layer is set as two neuron nodes, and the activation function is set as purelin;
the category of the artificial destroy event is identified as 1, and the category of the disturbance event is identified as 2.
9. The method of claim 7, wherein the extracted long-term analysis features are discretized before being input into a bayesian network classifier.
10. The utility model provides an oil gas pipeline vibration monitoring identification system which characterized in that includes
The signal acquisition module is used for acquiring an optical fiber vibration signal caused by disturbance along the oil and gas pipeline;
the short-time resolution feature detection module is used for detecting whether the optical fiber vibration signal is an abnormal event signal or not;
the long-time feature extraction module is used for extracting long-time analysis features of the optical fiber vibration signals;
a network identification classifier module including a plurality of network identification classifiers for classifying abnormal event signals based on the long-term analysis feature identification,
and the judging module comprehensively judges the artificial destruction event signals and the interference signals according to the classification output results of the network identification classifiers.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311767847.8A CN117848475A (en) | 2023-12-21 | 2023-12-21 | Oil and gas pipeline vibration monitoring and identifying method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311767847.8A CN117848475A (en) | 2023-12-21 | 2023-12-21 | Oil and gas pipeline vibration monitoring and identifying method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117848475A true CN117848475A (en) | 2024-04-09 |
Family
ID=90541120
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311767847.8A Pending CN117848475A (en) | 2023-12-21 | 2023-12-21 | Oil and gas pipeline vibration monitoring and identifying method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117848475A (en) |
-
2023
- 2023-12-21 CN CN202311767847.8A patent/CN117848475A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107425906B (en) | Distributing optical fiber sensing signal processing method towards underground pipe network safety monitoring | |
| CN107256635B (en) | Vehicle identification method based on distributed optical fiber sensing in intelligent traffic | |
| CN103456175B (en) | Accompanying vehicle real-time detection method based on vehicle registration plate recognition and meshing monitoring | |
| Liu et al. | A sound monitoring system for prevention of underground pipeline damage caused by construction | |
| CN111537056A (en) | Pipeline along-line third-party construction dynamic early warning method based on SVM and time-frequency domain characteristics | |
| JP7539582B2 (en) | Distributed Intelligent SNAP Informatics | |
| CN103226028B (en) | Method for identifying and detecting disturbance signals of phase-sensitive optical time domain reflectometer | |
| CN112985574B (en) | High-precision classification identification method for optical fiber distributed acoustic sensing signals based on model fusion | |
| CN103196465A (en) | Method for separating noise of sensing signal and extracting signal of phase-sensitive optical time domain reflectometer | |
| CN103994817A (en) | Vibration source identification method based on long-distance optical fiber frequent occurring events | |
| CN112032575B (en) | Pipeline safety monitoring method and system based on weak grating and storage medium | |
| CN103473932A (en) | Sound signal vehicle type identification system combined with oscillation mark line | |
| US20220329068A1 (en) | Utility Pole Hazardous Event Localization | |
| CN105678263A (en) | Method and system for object recognition | |
| CN105139869A (en) | Baby crying detection method based on interval difference features | |
| CN109612568A (en) | A kind of mobile method for interference source identification of vibration source | |
| US20240003717A1 (en) | Novel ai-driven approach to detect adn localize vehicle emergency stops based on vehicle trajectory using distributed fiber optic sensing (dfos) data | |
| US20220381607A1 (en) | Contrastive learning of utility pole representations from distributed acoustic sensing signals | |
| CN109614926B (en) | Distributed optical fiber sensing signal mode identification method and system based on prediction model | |
| CN117848475A (en) | Oil and gas pipeline vibration monitoring and identifying method and system | |
| CN116204784B (en) | DAS-based subway tunnel external hazard operation intrusion recognition method | |
| CN103994816A (en) | Identification method based on optical fiber multiple events | |
| US20230024104A1 (en) | Identification of false transformer humming using machine learning | |
| CN107870359A (en) | Micro-seismic event recognition methods and device | |
| US12051326B2 (en) | Road traffic extraction for unknown anomaly detection using distributed fiber optic sensing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |