CN109063687A - A kind of microseism P wave recognition methods and system based on depth convolutional neural networks - Google Patents
A kind of microseism P wave recognition methods and system based on depth convolutional neural networks Download PDFInfo
- Publication number
- CN109063687A CN109063687A CN201810997081.5A CN201810997081A CN109063687A CN 109063687 A CN109063687 A CN 109063687A CN 201810997081 A CN201810997081 A CN 201810997081A CN 109063687 A CN109063687 A CN 109063687A
- Authority
- CN
- China
- Prior art keywords
- neural networks
- convolutional neural
- microseism
- signal
- wave
- 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
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 35
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000012549 training Methods 0.000 claims abstract description 26
- 230000004913 activation Effects 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000006835 compression Effects 0.000 claims description 3
- 238000007906 compression Methods 0.000 claims description 3
- 230000005284 excitation Effects 0.000 claims description 3
- 239000000284 extract Substances 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 238000004519 manufacturing process Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000012935 Averaging Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 244000309464 bull Species 0.000 description 1
- 230000001427 coherent effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06N3/045—Combinations of networks
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of microseism P wave recognition methods and system based on depth convolutional neural networks, the training of depth convolutional neural networks is completed comprising the training set of microseism useful signal and noise by establishing, again by the Network Recognition microseism P wave after training, the method improves the efficiency and precision of useful signal identification.
Description
Technical field
The present invention relates to microseismic field, in particular to a kind of microseism based on depth convolutional neural networks
The recognition methods of P wave and system.
Background technique
Microseismic is to monitor underground state by observing, analyzing the small seismic events generated by pressure break
Geophysical techniques have great importance to oil field development stable and high yields.Microseism data useful signal energy is weaker, letter
It makes an uproar relatively low, or even be submerged among noise completely.Common seismic data processing method is although numerous, if but directly applying to micro-
Seismic data can not often obtain satisfied effect, this will directly affect the quality and precision of micro-seismic monitoring.Therefore, it finds
Weaker useful signal is the key that microseism datum processing in suitable method identification microseism data.
How accurately and rapidly to identify that P wave has microseism seismic source location, FRACTURE PREDICTION, seismic source rupture mechanism analysis
Important meaning.Under normal conditions, the identification of P wave is based on the difference identification of signal and noise, such as amplitude, frequency, polarization.
At present frequently with method be use for reference earthquake in STA/LTA (Short Term Averaging/Long Term
Averaging)(Allen,R.V.,1978.Automatic earthquake recognition and timing from
single traces.Bull.Seismol.Soc.Am.68,1521–1532.)、AIC(Akaike information
Criterion) the single recognition mode such as criterion, coherent detection.But since microseism Signal-to-Noise is low, these single methods
Often recognition effect is poor, or even erroneous judgement, misjudgement occurs, this undoubtedly seriously affects the accuracy of automatic Picking.Manual identified is micro-
Although earthquake first arrival precision is higher, take a long time, is not able to satisfy the needs that microseism is handled in real time.Therefore, how height is carried out
The useful signal identification of precision, which just seems, to be even more important.
Summary of the invention
In view of the deficiencies of the prior art, the present invention provides a kind of microseism P wave identification based on depth convolutional neural networks
Method and system complete the instruction of depth convolutional neural networks by establishing comprising the training set of microseism useful signal and noise
Practice, recycle trained Network Recognition microseism P wave, provides practical for effectively identification microseism useful signal
Methods and techniques.
To achieve the above object, the present invention discloses a kind of microseism P wave recognition methods based on depth convolutional neural networks,
It the described method comprises the following steps:
S1, it is rolled up with different dominant frequency, the microseism forward modeling signal of different signal-to-noise ratio and the production training of practical microseismograms
The data set of product neural network, and data set is divided into two classes in the form of effective P wave and noise；
S2, data set is trained with convolutional neural networks；
S3, the response for increasing signal with Relu activation primitive；
S4, corresponding tag class is encoded come identification signal by the One-hot converted by Softmax function, will believed
Number it is divided into P wave and noise signal.
In the above-mentioned technical solutions, in the step S2 convolutional neural networks include input layer, convolutional layer, pond layer and
Full articulamentum；
The input layer, the input for network data；
The convolutional layer, the data for that will input carry out feature extraction；
The pond layer makes characteristic pattern become smaller, it is complicated to simplify network query function for compressing the characteristic pattern of input
Degree, while Feature Compression is carried out, extract main feature；
The full articulamentum, for connecting all features and output.
In the above-mentioned technical solutions, the calculation formula of the Relu activation primitive are as follows: f (x)=max (0, x).
In the above-mentioned technical solutions, the training method of convolutional neural networks is as follows in the step S2:
The serial number of each pixel, uses x in S21, setting training setI, jIndicate the i-th row jth column element in the training set；
The serial number of weight, uses y in S22, setting convolution kernelM, nIt indicates the n-th column of m row weight in the convolution kernel, uses ybTable
Show the bias term in convolution kernel, uses aI, jIndicate the i-th row jth column element in the convolution kernel；
S23, it is obtained by the calculation formula of Relu activation primitive
Wherein, aI, jFor the output signal of excitation function.
The present invention discloses a kind of microseism P wave identifying system based on depth convolutional neural networks, and the system comprises notes
Record module, training module, computing module, identification module；
Logging modle, for the microseism forward modeling signal and practical microseismograms with different dominant frequency, different signal-to-noise ratio
The data set of training convolutional neural networks is made, and data set is divided into two classes in the form of effective P wave and noise；
Training module, for being trained with convolutional neural networks to data set；
Computing module, for increasing the response of signal with Relu activation primitive；
Identification module is known for tag class corresponding to the One-hot coding by being converted by Softmax function
Signal is divided into P wave and noise signal by level signal.
A kind of microseism P wave recognition methods and system based on depth convolutional neural networks of the present invention, has beneficial below
Effect: this method is that hydraulic fracturing microseismic indicates an important application direction, promotes the application of the technology
And popularization；And provide accurate to the identification of microseism useful signal and quickly pick up scheme, it is provided more for subsequent microseism positioning
Accurate initial information, compared with manual identified, present invention also improves the efficiency and precision of useful signal identification.
Detailed description of the invention
Fig. 1 is a kind of microseism P wave recognition methods flow chart based on depth convolutional neural networks of the present invention；
Fig. 2 is convolutional neural networks schematic diagram of the present invention；
Fig. 3 is Relu function schematic diagram of the present invention；
Fig. 4 is training set of the present invention and convolution kernel schematic diagram；
Fig. 5 is a kind of microseism P wave identifying system module map based on depth convolutional neural networks of the present invention；
Fig. 6 is actual microseism signal in the embodiment of the present invention；
Fig. 7 is the useful signal of the microseism P wave identification in the embodiment of the present invention based on depth convolutional neural networks.
Specific embodiment
Present invention is further described in detail with reference to the accompanying drawing, and the present invention provides a kind of based on depth convolutional Neural net
The microseism P wave recognition methods of network, the described method comprises the following steps, as shown in Figure 1:
S1, it is rolled up with different dominant frequency, the microseism forward modeling signal of different signal-to-noise ratio and the production training of practical microseismograms
The data set of product neural network, and data set is divided into two classes in the form of effective P wave and noise；
S2, data set is trained with convolutional neural networks；
S3, the response for increasing signal with Relu activation primitive；
S4, corresponding tag class is encoded come identification signal by the One-hot converted by Softmax function, will believed
Number it is divided into P wave and noise signal.
Wherein, as shown in Fig. 2, convolutional neural networks include input layer, convolutional layer, pond layer and complete in the step S2
Articulamentum；
The input layer, the input for network data；
The convolutional layer, the data for that will input carry out feature extraction；
The pond layer makes characteristic pattern become smaller, it is complicated to simplify network query function for compressing the characteristic pattern of input
Degree, while Feature Compression is carried out, extract main feature；
The full articulamentum exports data for connecting all features.
Structure are as follows: INPUT- > [[CONV] * N- > POOL] * M- > [FC] * K, the structure I nput indicates network inputs, N
A Convolution convolutional layer adds the pond a Pooling layer, combines M times described in construction, is finally superimposed K Full
The full articulamentum of Connect.
Wherein, as shown in figure 3, the calculation formula of the Relu activation primitive are as follows: f (x)=max (0, x).
Wherein, as shown in figure 4, the training method of convolutional neural networks is as follows in the step S2:
The serial number of each pixel in S21, setting training set, (a) uses x in figureI, jIndicate that the i-th row jth arranges in the training set
Element；
The serial number of weight, uses y in S22, setting convolution kernel (b)M, nIt indicates the n-th column of m row weight in the convolution kernel, uses
ybIt indicates the bias term in convolution kernel, uses aI, jIndicate the i-th row jth column element in the convolution kernel；
S23, it is obtained by the calculation formula of Relu activation primitive
Wherein, aI, jFor the output signal of excitation function.
The microseism P wave identifying system based on depth convolutional neural networks that the present invention also provides a kind of, the system comprises
Logging modle, training module, computing module, identification module, as shown in Figure 5；
Logging modle, for the microseism forward modeling signal and practical microseismograms with different dominant frequency, different signal-to-noise ratio
The data set of training convolutional neural networks is made, and data set is divided into two classes in the form of effective P wave and noise；
Training module, for being trained with convolutional neural networks to data set；
Computing module, for increasing the response of signal with Relu activation primitive；
Identification module is known for tag class corresponding to the One-hot coding by being converted by Softmax function
Signal is divided into P wave and noise signal by level signal.
The embodiment of the present invention is illustrated in figure 6 actual microseism signal, is sampled using 1ms, sampling time 2000ms.
Thus the useful signal of the microseism identification obtained through the invention as shown in Figure 7 may be used by comparison diagram 6 and Fig. 7
See that the present invention can accurately identify microseism useful signal.
The part not illustrated in specification is the prior art or common knowledge.Present embodiment is merely to illustrate the hair
It is bright, rather than limit the scope of the invention, the modifications such as equivalent replacement that those skilled in the art make the present invention are recognized
To be fallen into invention claims institute protection scope.
Claims (5)
1. a kind of microseism P wave recognition methods based on depth convolutional neural networks, which is characterized in that the method includes following
Step:
S1, training convolutional mind is made with different dominant frequency, the microseism forward modeling signal of different signal-to-noise ratio and practical microseismograms
Data set through network, and data set is divided into two classes in the form of effective P wave and noise；
S2, data set is trained with convolutional neural networks；
S3, the response for increasing signal with Relu activation primitive；
S4, corresponding tag class is encoded by the One-hot converted by Softmax function come identification signal, by signal point
For P wave and noise signal.
2. a kind of microseism P wave recognition methods based on depth convolutional neural networks, feature exist according to claim 1
In convolutional neural networks include input layer, convolutional layer, pond layer and full articulamentum in the step S2；
The input layer, the input for network data；
The convolutional layer, the data for that will input carry out feature extraction；
The pond layer makes characteristic pattern become smaller, simplifies network query function complexity, together for compressing the characteristic pattern of input
Shi Jinhang Feature Compression extracts main feature；
The full articulamentum, for connecting all features and output.
3. a kind of microseism P wave recognition methods based on depth convolutional neural networks, feature exist according to claim 1
In the calculation formula of the Relu activation primitive are as follows: f (x)=max (0, x).
4. a kind of microseism P wave recognition methods based on depth convolutional neural networks, feature exist according to claim 3
In the training method of convolutional neural networks is as follows in the step S2:
The serial number of each pixel, uses x in S21, setting training setI, jIndicate the i-th row jth column element in the training set；
The serial number of weight, uses y in S22, setting convolution kernelM, nIt indicates the n-th column of m row weight in the convolution kernel, uses ybIndicate volume
Bias term in product core, uses aI, jIndicate the i-th row jth column element in the convolution kernel；
S23, it is obtained by the calculation formula of Relu activation primitive
Wherein, aI, jFor the output signal of excitation function.
5. a kind of microseism P wave identifying system based on depth convolutional neural networks, which is characterized in that the system comprises records
Module, training module, computing module, identification module；
Logging modle, for different dominant frequency, the microseism forward modeling signal of different signal-to-noise ratio and the production of practical microseismograms
The data set of training convolutional neural networks, and data set is divided into two classes in the form of effective P wave and noise；
Training module, for being trained with convolutional neural networks to data set；
Computing module, for increasing the response of signal with Relu activation primitive；
Identification module identifies letter for tag class corresponding to the One-hot coding by being converted by Softmax function
Number, signal is divided into P wave and noise signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810997081.5A CN109063687A (en) | 2018-08-29 | 2018-08-29 | A kind of microseism P wave recognition methods and system based on depth convolutional neural networks |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810997081.5A CN109063687A (en) | 2018-08-29 | 2018-08-29 | A kind of microseism P wave recognition methods and system based on depth convolutional neural networks |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109063687A true CN109063687A (en) | 2018-12-21 |
Family
ID=64757777
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810997081.5A Pending CN109063687A (en) | 2018-08-29 | 2018-08-29 | A kind of microseism P wave recognition methods and system based on depth convolutional neural networks |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109063687A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110058294A (en) * | 2019-05-10 | 2019-07-26 | 东北大学 | A kind of tunnel micro seismic monitoring rock rupture event automatic identifying method |
| CN110082822A (en) * | 2019-04-09 | 2019-08-02 | 中国科学技术大学 | The method for carrying out earthquake detection using convolutional neural networks |
| CN110632662A (en) * | 2019-09-25 | 2019-12-31 | 成都理工大学 | An Algorithm for Automatic Microseismic Signal Recognition Using DCNN-Inception Network |
| CN111046737A (en) * | 2019-11-14 | 2020-04-21 | 吉林大学 | Efficient intelligent sensing acquisition method for microseism signal detection |
| CN111427079A (en) * | 2020-05-15 | 2020-07-17 | 北京铁科时代科技有限公司 | Method and device applied to prediction of laboratory seismic waves |
| CN111505705A (en) * | 2020-01-19 | 2020-08-07 | 长江大学 | Microseism P wave first arrival pickup method and system based on capsule neural network |
| CN112464721A (en) * | 2020-10-28 | 2021-03-09 | 中国石油天然气集团有限公司 | Automatic microseism event identification method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106845529A (en) * | 2016-12-30 | 2017-06-13 | 北京柏惠维康科技有限公司 | Image feature recognition methods based on many visual field convolutional neural networks |
| CN107908928A (en) * | 2017-12-21 | 2018-04-13 | 天津科技大学 | A kind of hemoglobin Dynamic Spectrum Analysis Forecasting Methodology based on depth learning technology |
| CN108182410A (en) * | 2017-12-28 | 2018-06-19 | 南通大学 | A kind of joint objective zone location and the tumble recognizer of depth characteristic study |
-
2018
- 2018-08-29 CN CN201810997081.5A patent/CN109063687A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106845529A (en) * | 2016-12-30 | 2017-06-13 | 北京柏惠维康科技有限公司 | Image feature recognition methods based on many visual field convolutional neural networks |
| CN107908928A (en) * | 2017-12-21 | 2018-04-13 | 天津科技大学 | A kind of hemoglobin Dynamic Spectrum Analysis Forecasting Methodology based on depth learning technology |
| CN108182410A (en) * | 2017-12-28 | 2018-06-19 | 南通大学 | A kind of joint objective zone location and the tumble recognizer of depth characteristic study |
Non-Patent Citations (1)
| Title |
|---|
| 陈润航等: "地震和爆破事件源波形信号的卷积神经网络分类研究", 《地球物理学进展》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110082822A (en) * | 2019-04-09 | 2019-08-02 | 中国科学技术大学 | The method for carrying out earthquake detection using convolutional neural networks |
| CN110082822B (en) * | 2019-04-09 | 2020-07-28 | 中国科学技术大学 | Method of earthquake detection using convolutional neural network |
| CN110058294A (en) * | 2019-05-10 | 2019-07-26 | 东北大学 | A kind of tunnel micro seismic monitoring rock rupture event automatic identifying method |
| CN110632662A (en) * | 2019-09-25 | 2019-12-31 | 成都理工大学 | An Algorithm for Automatic Microseismic Signal Recognition Using DCNN-Inception Network |
| CN111046737A (en) * | 2019-11-14 | 2020-04-21 | 吉林大学 | Efficient intelligent sensing acquisition method for microseism signal detection |
| CN111046737B (en) * | 2019-11-14 | 2022-07-08 | 吉林大学 | Efficient intelligent sensing acquisition method for microseism signal detection |
| CN111505705A (en) * | 2020-01-19 | 2020-08-07 | 长江大学 | Microseism P wave first arrival pickup method and system based on capsule neural network |
| CN111505705B (en) * | 2020-01-19 | 2022-08-02 | 长江大学 | Microseism P wave first arrival pickup method and system based on capsule neural network |
| CN111427079A (en) * | 2020-05-15 | 2020-07-17 | 北京铁科时代科技有限公司 | Method and device applied to prediction of laboratory seismic waves |
| CN112464721A (en) * | 2020-10-28 | 2021-03-09 | 中国石油天然气集团有限公司 | Automatic microseism event identification method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109063687A (en) | A kind of microseism P wave recognition methods and system based on depth convolutional neural networks | |
| CN106405640B (en) | Microseismic signals based on depth conviction neural network then automatic pick method | |
| CN111027452B (en) | Microseismic signal arrival and earthquake phase identification method and system based on deep neural network | |
| Benítez et al. | Continuous HMM-based seismic-event classification at Deception Island, Antarctica | |
| CN113792685B (en) | Microseism event detection method based on multi-scale convolutional neural network | |
| CN113568043A (en) | Three-stage seismic phase picking method based on deep convolutional neural network | |
| Li et al. | An improved AIC onset-time picking method based on regression convolutional neural network | |
| CN115436993B (en) | A P-wave detection method for seismic signals based on PPNet network | |
| CN109613610B (en) | Automatic Pickup Method of Microseismic Signal Arrival Time Difference | |
| US20230071743A1 (en) | Quantitative hydraulic fracturing surveillance from fiber optic sensing using machine learning | |
| CN114637045B (en) | Microseismic P-wave first arrival picking method based on UNet++ combined with Clique Block | |
| Song et al. | Recognition of microseismic and blasting signals in mines based on convolutional neural network and stockwell transform | |
| CN115097518A (en) | Micro-seismic arrival time picking method based on fuzzy clustering U-shaped neural network | |
| CN113671564A (en) | Automatic Picking Method of Microseismic Effective Events Based on NARX Dynamic Neural Network | |
| CN109031416A (en) | The method of microseism P wave first break pickup | |
| CN115330007A (en) | Reservoir multi-classification prediction method based on deep fusion model | |
| CN118131200A (en) | Underwater acoustic target recognition method based on metric learning | |
| CN117609874A (en) | Rock fault friction microseismic detection method and system based on integrated deep learning | |
| CN115932947A (en) | Microseism P wave S wave separation method based on CNN-LSTM and multi-head attention mechanism | |
| CN112801073B (en) | Classification and recognition method of acoustic emission first arrival signal based on deep neural network | |
| CN109635738A (en) | A kind of image characteristic extracting method and system | |
| CN114139579B (en) | Deep learning microseism event detection method based on variance fractal | |
| Liao et al. | Automatic first arrival picking combining transfer learning and the short-time Fourier transform | |
| He et al. | A study of microseismic first arrival pickup based on image semantic segmentation | |
| CN114429162B (en) | Seismic horizon tracing method based on deep convolutional neural network and random forest |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181221 |