CN113239618A - Gas pipeline detection and identification method based on acoustic signal characteristic analysis - Google Patents
Gas pipeline detection and identification method based on acoustic signal characteristic analysis Download PDFInfo
- Publication number
- CN113239618A CN113239618A CN202110490645.8A CN202110490645A CN113239618A CN 113239618 A CN113239618 A CN 113239618A CN 202110490645 A CN202110490645 A CN 202110490645A CN 113239618 A CN113239618 A CN 113239618A
- Authority
- CN
- China
- Prior art keywords
- gas flow
- flow noise
- signal
- sound pressure
- gas
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Computational Mathematics (AREA)
- Biophysics (AREA)
- Pure & Applied Mathematics (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Algebra (AREA)
- Probability & Statistics with Applications (AREA)
- Operations Research (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)
Abstract
The invention discloses a gas pipeline detection and identification method based on acoustic signal characteristic analysis, which comprises the steps of obtaining a gas flow noise sound pressure pulse signal of a target gas pipeline; decomposing the gas flow noise sound pressure pulse signal according to a Hilbert-Huang transform algorithm, analyzing time-frequency and marginal spectrum characteristics of the gas flow noise sound pressure pulse signal, determining characteristic parameters of the gas flow noise sound pressure pulse signal, and normalizing to form a gas flow noise signal characteristic database; determining a gas flow noise signal classification recognition model; and inputting the test data into the trained gas flow noise signal classification and identification model, and further judging whether the acoustic signal is a gas flow acoustic signal. According to the method, the EMD is utilized to decompose two signals to obtain a time domain waveform of an IMF component, a characteristic database of a gas pipeline flow noise signal is obtained, and pattern recognition is carried out through an optimized BP neural network; and the gas pipeline acoustic signal recognition rate of the method reaches 97.5%.
Description
Technical Field
The invention relates to a gas pipeline detection and identification method based on acoustic signal characteristic analysis.
Background
With the rapid development of the gas industry in cities and towns in China, the number and the scale of urban underground pipelines are larger and larger. In the process of continuous reconstruction and extension of town gas pipelines, because the gas pipelines are constructed in different times, the abandoned gas pipelines and the in-service pipelines are buried underground together, the underground pipeline system is disordered, and the pipeline coordinate position is inaccurate, so that liability accidents such as pipeline damage, gas leakage, personal safety and the like frequently occur in the process of reconstruction, new construction or municipal engineering excavation of the gas pipelines. The gas accidents have great influence on life, personal safety and the like of people, so that a method suitable for urban pipeline detection and identification is used in the process of building, rebuilding or excavating a gas pipeline or municipal engineering, and the gas accidents such as gas leakage and the like in the actual construction process are avoided. At present, scholars at home and abroad research various methods in the aspect of buried pipeline detection, such as: the detection methods are complex in operation and are easily limited by the interference of environmental factors in the detection process, particularly under the complex conditions that buried gas pipelines comprise metal pipelines, nonmetal pipelines, in-service pipelines and waste pipelines, the detection efficiency and accuracy of the existing method are greatly limited, and how to judge whether the underground gas pipelines are in service and whether the gas pipelines are in service is urgently needed to be effectively solved from the aspects of theoretical methods and practical operation.
Disclosure of Invention
In order to overcome the problems in the prior art, the invention provides a gas pipeline detection and identification method based on acoustic signal characteristic analysis.
The technical scheme provided by the invention for solving the technical problems is as follows: a gas pipeline detection and identification method based on acoustic signal characteristic analysis comprises the following steps:
acquiring a gas flow noise sound pressure pulse signal of a target gas pipeline;
decomposing the gas flow noise sound pressure pulse signal according to a Hilbert-Huang transform algorithm, analyzing time-frequency and marginal spectrum characteristics of the gas flow noise sound pressure pulse signal, determining characteristic parameters of the gas flow noise sound pressure pulse signal, and normalizing to form a gas flow noise signal characteristic database;
analyzing important influence factors of the BP neural network according to a BP neural network algorithm and a gas flow noise signal characteristic database, determining optimal parameters of the network, finishing network initialization, network construction and network training of the BP neural network, accurately classifying and identifying gas acoustic signals, and determining a classification and identification model of the gas flow noise signals;
and inputting the test data into the trained gas flow noise signal classification and identification model, and further judging whether the acoustic signal is a gas flow acoustic signal.
The gas flow noise sound pressure pulse signal is obtained through numerical simulation or experimental verification.
The further technical scheme is that the gas flow noise sound pressure pulse signal is decomposed according to a Hilbert-Huang transform algorithm, time-frequency and marginal spectrum characteristics of the gas flow noise sound pressure pulse signal are analyzed, characteristic parameters of the gas flow noise sound pressure pulse signal are determined, and a gas flow noise signal characteristic database is formed by normalization:
performing sound pressure level conversion on the fuel flow noise sound pressure pulse signal to obtain time domain data of the sound pressure level changing along with time;
performing EMD on the time domain data to obtain a plurality of IMF components;
calculating the peak value F and the peak value frequency Fmax of each IMF component which is screened out, and carrying out normalization processing;
hilbert transformation is carried out on IMF components of each order, instantaneous frequency and amplitude obtained after Hilbert decomposition is carried out on signals are time variables, and time-frequency distribution of amplitude energy is recorded as a Hilbert spectrum;
performing time domain integration on the Hilbert spectrum to obtain an expression mode of original signal energy on a time-frequency domain, namely the Hilbert marginal spectrum;
calculating the sum parameter of the energy of the whole large region of the Hilbert marginal spectrum;
and calculating the energy ratio Ei of the marginal spectrum of each IMF component as a characteristic parameter, and normalizing.
The further technical scheme is that the BP neural network algorithm comprises the following steps:
(1) inputting the number of nodes;
(2) the number of nodes of an output layer;
(3) the number of nodes of the hidden layer;
(4) selecting an initial weight value;
(5) selecting a learning rate:
(6) and (4) selecting the expected error.
The further technical scheme is that the step (1) comprises the steps of determining the number of network input nodes according to the extracted feature vector P, and setting the number of the input nodes of the neural network for classifying the gas acoustic signals to be 16 according to the dimension of the feature vector and the classification effect during actual classification.
The further technical scheme is that the step (2) comprises the steps of determining the output type by using a binary coding mode; wherein, the code of the gas flow noise signal inside the gas pipeline is (1,0)T(ii) a The coding of the noise signal of other types of gas flow is (0,1)T(ii) a Judging that 2 output nodes are selected, and outputting Y codes as follows:
in the formula: and Y is output code.
The further technical scheme is that the calculation formula of the number of the nodes of the hidden layer is as follows:
in the formula: m is the number of nodes of the input layer, n is the number of nodes of the output layer, and a is a constant of 1-10.
The further technical scheme is that in the step (4), the initial weight is randomly selected between the intervals of [ -1, 1 ].
The invention has the following beneficial effects: the method takes a gas pipeline flow noise signal obtained by numerical simulation and experimental verification as a research object, and simultaneously acquires an underwater sound signal for comparative analysis; decomposing the two signals by using EMD to obtain a time domain waveform of an IMF component, determining two characteristic parameters of a peak value and a peak frequency, calculating an energy characteristic parameter of a Hilbert marginal spectrum, obtaining a characteristic database of a gas pipeline flow noise signal, and performing mode identification through an optimized BP neural network; and the gas pipeline acoustic signal recognition rate of the method reaches 97.5%.
Drawings
FIG. 1 is an EMD decomposition flow diagram;
FIG. 2 is a feature extraction flow diagram;
FIG. 3 is a neural network algorithm flow diagram;
FIG. 4 is a graph of various parameters as a function of hidden layer nodes;
FIG. 5 is a gas flow noise signal BP neural network classification model diagram.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it should be understood that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention discloses a gas pipeline detection and identification method based on acoustic signal characteristic analysis, which comprises the following steps of:
s1, obtaining a gas flow noise sound pressure pulse signal of the target gas pipeline probe according to numerical simulation and experimental verification;
s2, taking the gas flow noise sound pressure pulse signal as an original signal, and then performing sound pressure level transformation on the original signal to obtain time domain data of which the sound pressure level changes along with time;
s3, performing EMD decomposition on the time domain data to obtain a plurality of inherent mode functions, and screening out orders containing the main characteristic information of the original acoustic signal, wherein an EMD decomposition flow chart is shown in figure 1;
s4, calculating the peak value F and the peak value frequency Fmax of each IMF component which is screened out, and carrying out normalization processing;
s5, performing Hilbert transformation on each order of IMF component obtained by EMD decomposition of an original sound signal, wherein the instantaneous frequency and amplitude obtained after Hilbert decomposition of the signal are time variables, the time-frequency characteristic of the signal can be fully displayed, and the time-frequency distribution of amplitude energy is recorded as a Hilbert spectrum; the Hilbert spectrum of the original signal is obtained by summing the Hilbert spectra of all IMF components;
s6, obtaining an expression mode of the original signal energy on a time-frequency domain by performing time domain integration on the Hilbert spectrum, namely the Hilbert marginal spectrum;
s7, calculating the sum parameter of the energy of the whole large region of the Hilbert marginal spectrum;
s8, calculating the energy ratio Ei of each IMF component marginal spectrum as a characteristic parameter, and carrying out normalization processing; all normalization results form a gas flow noise signal characteristic database;
three characteristic parameters F, Fmax and E extracted based on the Hilbert-Huang method form a characteristic vector P:
P＝[F、Fmax、E]
wherein: f ═ F1、F2、F3,······,Fn}
Fmax＝{Fmax1、Fmax2、Fmax3,······,Fmaxn}
E＝{E1、E2、E3,······,En}
S9, analyzing important influence factors of the BP neural network according to a BP neural network algorithm and a gas flow noise signal characteristic database, determining optimal parameters of the network, completing network initialization, network construction and network training of the BP neural network, accurately classifying and identifying gas acoustic signals, and determining a classification and identification model of the gas flow noise signals;
and S10, inputting the test data into the trained gas flow noise signal classification and identification model, and further judging whether the acoustic signal is a gas flow acoustic signal.
The BP neural network algorithm is shown in fig. 3, and comprises the following steps:
(1) number of input nodes
According to the extracted feature vector P (P ═ F, F)max、E]＝[F1、F2、F3,……,Fn，Fmax1、Fmax2、Fmax3,……,Fmaxn、E1、E2、E3,……,En]) And determining the number of network input nodes, and finally setting the number of the input nodes of the neural network for gas acoustic signal classification to be 16 according to the feature vector dimension and the classification effect during actual classification.
(2) Number of output layer nodes
Determining the output type by using a binary coding mode according to two sound types of the noise signal mode of the gas pipeline flow, wherein the coding of the gas pipeline internal gas flow noise signal is (1,0)T(ii) a The coding of the noise signal of other types of gas flow is (0,1)T(ii) a Therefore, it can be determined that 2 output nodes are taken, and the code of output Y is:
(3) number of hidden layer nodes
The number of hidden layer nodes in the determination generally follows the following principles: under the condition of accurately reflecting the input-output relationship, the number of nodes of the hidden layer is less, so that the network structure is simplified. For a single hidden layer BP neural network, the number of hidden layer nodes can be selected by using the following formula.
In the formula: m is the number of nodes of the input layer, n is the number of nodes of the output layer, and a is a constant of 1-10. And calculating the number of nodes of the hidden layer between [6 and 15] according to a formula.
The optimal number of hidden layer nodes is selected by comparing and analyzing the training error, the testing error and the variation condition of the iteration step number of the network under the condition of different hidden layer node numbers by using a method of controlling variables. The variation curve of each parameter with the number of hidden layer nodes is shown in fig. 4.
As can be seen from fig. 4, the test error and the training error of the network both have a trend of decreasing first and then increasing, and when the number of hidden layer nodes is 10, the network training error and the test error are the minimum, so the number of hidden layer nodes is determined to be 10.
(4) Initial weight selection
The BP neural network of the invention randomly generates an initial weight value in an interval of [ -1, 1 ].
(5) Selection of learning rate
The learning rate has a direct influence on the training duration, the network convergence speed and whether the BP network can be converged. Usually, a plurality of groups of values are selected for learning rate to be compared and verified, then the optimal learning rate is selected for neural network learning training, for the learning rate, the learning rate lr is generally selected to be 0.001-1, and the learning rate lr is set to be 0.001 to ensure the stability of the network.
(6) Selection of the expected error
The lower the value is not selected in the expected error selection process, the higher the accuracy of the neural network classification identification. If the expected error selection is too low in the actual simulation training process, the whole network becomes complicated, and the classification and identification efficiency and accuracy of the neural network are affected. In practical cases, therefore, the expected error only needs to be selected to meet the requirement. Therefore, the expected error of the present invention is set to 0.001.
In summary, the BP neural network classification model built herein for the gas flow noise signal: the three-layer topological structure is adopted, the number of input nodes is 16, the number of hidden layer nodes is 10, and the number of output nodes is 2, so that the topological structure of the network nodes is 16 multiplied by 10 multiplied by 2, and a gas flow noise signal BP neural network classification model is shown in figure 5.
Although the present invention has been described with reference to the above embodiments, it should be understood that the present invention is not limited to the above embodiments, and those skilled in the art can make various changes and modifications without departing from the scope of the present invention.
Claims (8)
1. A gas pipeline detection and identification method based on acoustic signal characteristic analysis is characterized by comprising the following steps:
acquiring a gas flow noise sound pressure pulse signal of a target gas pipeline;
decomposing the gas flow noise sound pressure pulse signal according to a Hilbert-Huang transform algorithm, analyzing time-frequency and marginal spectrum characteristics of the gas flow noise sound pressure pulse signal, determining characteristic parameters of the gas flow noise sound pressure pulse signal, and normalizing to form a gas flow noise signal characteristic database;
analyzing important influence factors of the BP neural network according to a BP neural network algorithm and a gas flow noise signal characteristic database, determining optimal parameters of the network, finishing network initialization, network construction and network training of the BP neural network, accurately classifying and identifying gas acoustic signals, and determining a classification and identification model of the gas flow noise signals;
and inputting the test data into the trained gas flow noise signal classification and identification model, and further judging whether the acoustic signal is a gas flow acoustic signal.
2. The gas pipeline detecting and identifying method based on the acoustic signal characteristic analysis as claimed in claim 1, wherein the gas flow noise sound pressure pulsation signal is obtained through numerical simulation or experimental verification.
3. The method for detecting and identifying a gas pipeline based on the acoustic signal characteristic analysis as claimed in claim 2, wherein the method comprises the steps of decomposing a gas flow noise sound pressure pulsation signal according to a Hilbert-Huang transform algorithm, analyzing time-frequency and marginal spectrum characteristics of the gas flow noise sound pressure pulsation signal, determining characteristic parameters of the gas flow noise sound pressure pulsation signal, and normalizing to form a gas flow noise signal characteristic database, wherein the step of:
performing sound pressure level conversion on the fuel flow noise sound pressure pulse signal to obtain time domain data of the sound pressure level changing along with time;
performing EMD on the time domain data to obtain a plurality of IMF components;
calculating the peak value F and the peak value frequency Fmax of each IMF component which is screened out, and carrying out normalization processing;
hilbert transformation is carried out on IMF components of each order, instantaneous frequency and amplitude obtained after Hilbert decomposition is carried out on signals are time variables, and time-frequency distribution of amplitude energy is recorded as a Hilbert spectrum;
performing time domain integration on the Hilbert spectrum to obtain an expression mode of original signal energy on a time-frequency domain, namely the Hilbert marginal spectrum;
calculating the sum parameter of the energy of the whole large region of the Hilbert marginal spectrum;
and calculating the energy ratio Ei of the marginal spectrum of each IMF component as a characteristic parameter, and normalizing.
4. The gas pipeline detection and identification method based on the acoustic signal characteristic analysis as claimed in claim 3, wherein the BP neural network algorithm comprises the following steps:
(1) inputting the number of nodes;
(2) the number of nodes of an output layer;
(3) the number of nodes of the hidden layer;
(4) selecting an initial weight value;
(5) selecting a learning rate:
(6) and (4) selecting the expected error.
5. The gas pipeline detection and identification method based on acoustic signal feature analysis according to claim 4, wherein the step (1) comprises determining the number of network input nodes according to the extracted feature vector P, and setting the number of input nodes of the neural network for gas acoustic signal classification to 16 according to the feature vector dimension and the classification effect during actual classification.
6. The gas pipeline detection and identification method based on the acoustic signal characteristic analysis is characterized in that the step (2) comprises the steps of determining the output type by using a binary coding mode; wherein, the code of the gas flow noise signal inside the gas pipeline is (1,0)T(ii) a The coding of the noise signal of other types of gas flow is (0,1)T(ii) a Judging that 2 output nodes are selected, and outputting Y codes as follows:
in the formula: and Y is output code.
7. The gas pipeline detection and identification method based on acoustic signal characteristic analysis according to claim 4, wherein the calculation formula of the number of hidden layer nodes is as follows:
in the formula: m is the number of nodes of the input layer, n is the number of nodes of the output layer, and a is a constant of 1-10.
8. The gas pipeline detection and identification method based on acoustic signal characteristic analysis according to claim 4, characterized in that in step (4), the initial weight is randomly selected between the intervals [ -1, 1 ].
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110490645.8A CN113239618A (en) | 2021-05-06 | 2021-05-06 | Gas pipeline detection and identification method based on acoustic signal characteristic analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110490645.8A CN113239618A (en) | 2021-05-06 | 2021-05-06 | Gas pipeline detection and identification method based on acoustic signal characteristic analysis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113239618A true CN113239618A (en) | 2021-08-10 |
Family
ID=77132213
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110490645.8A Pending CN113239618A (en) | 2021-05-06 | 2021-05-06 | Gas pipeline detection and identification method based on acoustic signal characteristic analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113239618A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113958816A (en) * | 2021-11-03 | 2022-01-21 | 西南石油大学 | Gas pipeline flow noise detection device |
| CN114818839A (en) * | 2022-07-01 | 2022-07-29 | 之江实验室 | Deep learning-based optical fiber sensing underwater acoustic signal identification method and device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2007231655A1 (en) * | 2003-04-07 | 2007-11-15 | Silverbrook Research Pty Ltd | Laser scanner using acousto-optic deflectors |
| CN111896616A (en) * | 2020-03-29 | 2020-11-06 | 中国石油大学(华东) | Gas-liquid two-phase flow pattern identification method based on acoustic emission-BP neural network |
-
2021
- 2021-05-06 CN CN202110490645.8A patent/CN113239618A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2007231655A1 (en) * | 2003-04-07 | 2007-11-15 | Silverbrook Research Pty Ltd | Laser scanner using acousto-optic deflectors |
| CN111896616A (en) * | 2020-03-29 | 2020-11-06 | 中国石油大学(华东) | Gas-liquid two-phase flow pattern identification method based on acoustic emission-BP neural network |
Non-Patent Citations (4)
| Title |
|---|
| 彭善碧等: "输油管道负压波法泄漏检测中的误报警排除法", 《西南石油学院院报》 * |
| 徐红梅等: "基于EMD和HHT的内燃机噪声信号时频特性研究", 《内燃机工程》 * |
| 焦敬品等: "基于BP神经网络的管道泄漏声信号识别方法研究", 《仪器仪表学报》 * |
| 胡泽等: "基于希尔伯特-黄变换和神经网络的滚动轴承故障诊断", 《电动工具》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113958816A (en) * | 2021-11-03 | 2022-01-21 | 西南石油大学 | Gas pipeline flow noise detection device |
| CN114818839A (en) * | 2022-07-01 | 2022-07-29 | 之江实验室 | Deep learning-based optical fiber sensing underwater acoustic signal identification method and device |
| CN114818839B (en) * | 2022-07-01 | 2022-09-16 | 之江实验室 | Deep learning-based optical fiber sensing underwater acoustic signal identification method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111626153B (en) | Integrated learning-based partial discharge fault state identification method | |
| CN110185939B (en) | Gas pipeline leakage identification method based on convolutional neural network | |
| Wang et al. | Modeling carbon emission trajectory of China, US and India | |
| CN108426713B (en) | Rolling bearing weak fault diagnosis method based on wavelet transformation and deep learning | |
| CN111853555B (en) | Water supply pipe network blind leakage identification method based on dynamic process | |
| CN100567978C (en) | The supersonic phased array for detecting oil gas pipeline girth weld defect type automatic identifying method | |
| CN109084186B (en) | Pipeline leakage signal identification method based on improved ELMD (ensemble empirical mode decomposition) multi-scale entropy | |
| CN113239618A (en) | Gas pipeline detection and identification method based on acoustic signal characteristic analysis | |
| CN112097126B (en) | Water supply network pipe burst pipeline accurate identification method based on deep neural network | |
| CN107480705A (en) | A kind of running water pipe leakage detection method | |
| CN102364501A (en) | Method for reproducing two-dimensional defect of petroleum pipeline PSO-BP (Particle Swarm Optimization-Back-Propagation) neural network | |
| CN103033567A (en) | Pipeline defect signal identification method based on guided wave | |
| CN105546352A (en) | Natural gas pipeline tiny leakage detection method based on sound signals | |
| CN111022941A (en) | Natural gas pipeline leakage detection method based on LSTM recurrent neural network | |
| CN109886433A (en) | The method of intelligent recognition city gas pipeline defect | |
| CN108491931B (en) | Method for improving nondestructive testing precision based on machine learning | |
| CN111735583B (en) | Pipeline working condition identification method based on LCD-EE pipeline sound wave signal characteristic extraction | |
| CN113011660A (en) | Air quality prediction method, system and storage medium | |
| CN111353640B (en) | Method for constructing wind speed prediction model by combination method | |
| CN113687433A (en) | Bi-LSTM-based magnetotelluric signal denoising method and system | |
| CN109632942B (en) | Inversion method of pipeline defect size based on ensemble learning | |
| CN117390381A (en) | Underground diaphragm wall joint seam leakage prediction method and device based on deep learning | |
| CN116975763A (en) | Water supply network abnormality diagnosis method based on bispectrum and convolutional neural network | |
| Han et al. | Identification of Pipeline Leak Sizes Based on Chaos-Grey Wolf-Support Vector Machine | |
| CN111965255A (en) | Pressure shear slip type karst dangerous rock instability early warning multi-precursor sound method and device |
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 |