CN110414675A - A kind of underground shallow layer seismic source location method based on deep learning - Google Patents
A kind of underground shallow layer seismic source location method based on deep learning Download PDFInfo
- Publication number
- CN110414675A CN110414675A CN201910823790.6A CN201910823790A CN110414675A CN 110414675 A CN110414675 A CN 110414675A CN 201910823790 A CN201910823790 A CN 201910823790A CN 110414675 A CN110414675 A CN 110414675A
- Authority
- CN
- China
- Prior art keywords
- deep learning
- seismic source
- source location
- sample
- sensor
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V1/00—Seismology; Seismic or acoustic prospecting or detecting
- G01V1/28—Processing seismic data, e.g. analysis, for interpretation, for correction
- G01V1/282—Application of seismic models, synthetic seismograms
-
- 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
Abstract
The underground shallow layer seismic source location method based on deep learning that the present invention relates to a kind of, comprising the following steps: the distributed shock sensor array of laying, generation learning sample, the corresponding hypocentrum cartridge position of setting three-dimensional energy field picture sample, which are used as, trains label, building deep learning network frame, trains network, position to practical hypocenter of the explosion.The present invention reduces the intermediate steps such as extracting of positional parameters, location model modeling and location model resolving during traditional shallow-layer seismic source location, greatly improve seismic source location efficiency, eliminate blind location area, the dependence of seismic source location precision monitored area channel reconstructing precision is reduced, provides a kind of new seismic source location method for underground shallow layer seismic source location.
Description
Technical field
The invention belongs to blasting vibration test technical fields, and in particular to a kind of underground shallow layer focus based on deep learning
Localization method hits positioning suitable for the fixed point group under the conditions of unknown geologic structure.
Background technique
Underground shallow layer focus refers to that the event occurred in space of the subsurface source depth no more than 100m, Passive Positioning are
It solves high value ammunition underground in military field and fries the main paths such as point location and penetration trajectory measurement；It is to realize civil field
The monitoring of middle geology, engineering explosion, historical relic anti-theft monitoring, coal mine prospecting, the exploration of surface infrastructure component analysis, geologic structure,
The important means such as the exploration of underground rare mineral matter, petroleum detection excavation.
With big region, big equivalent, big depth, the long-time seismic source location such as deep seismic, coal mine deep production, oil exploration
It compares, such seismic source location has the following characteristics that (1) subsurface source depth is shallower, is usually no more than 100m, shallow-layer geology knot
Structure is complicated and unknown, can not use for reference deep layer Model of the crustal structure and establish shallow-layer velocity field model；(2) shock wave wave group aliasing is multiple
Miscellaneous, explosion near field soil constitutive bchavior is in elastoplasticity, and elastic wave is larger by ground return, refractive effect, and Seismic Facies Characteristics are unobvious；
(3) underground medium is complicated, and transmission channel is more complicated and uneven, and it is more to frequently include surface dust, rock, sandstone, closely knit soil etc.
Seed type；(4) positioning accuracy request is higher, and stereoscopic localized error is less than 1m in the dispersion zone of 100m, is different from big depth
Hundred meters of position error requirements, belong to zonule high accuracy positioning problem in earthquakes location.
Currently, being positioned mainly for deep seismic in underground space positioning, not being directly used in the shallow-layer zonule
Localization method, and deep seismic localization method not can be used directly in shallow-layer positioning, be primarily present following problem:
(1) localization method when walking based on Geiger, this method emphasis are solved in geology uniformly and under speed known conditions
Fire point orientation problem, but for the complex geological conditions such as stratiform, hollow, speed parameter is only capable of being set as average value, positioning
Error is larger；
(2) it is based on polarization angle localization method, this method can be realized fire point using a small amount of observer nodes and quickly position,
But under shallow-layer complex geological condition, when especially there is strong reflection interface, surface layer incident angle is unable to characterize focus and sensor
True ray path between receiving point, therefore there are locating artifacts；
(3) more focus and velocity structure inversion position (SSH) model, and this method is suitable for unknown geologic structure
Under the conditions of fixed point group hit orientation problem, wherein detonation spot placement accuracy is by velocity field modeling accuracy, focus excitation number
It influences.Since single detonates point location, big gun penetrates that data for projection is limited, and rate matrices are sparse, although it is about in a helpless situation that regularization etc. is added
Section, but positioning accuracy is still difficult to ensure.
Summary of the invention
The present invention provides a kind of underground shallow layer seismic source location method based on deep learning, and technical problems to be solved are:
Shallow-layer zonule target can not high accuracy positioning the problem of.
In order to solve the above technical problems, the present invention provides a kind of underground shallow layer seismic source location side based on deep learning
Method, which is characterized in that specifically includes the following steps:
S1. lay distributed shock sensor array: in the monitoring random preset coordinate origin in region, by n sensor with
Certain angle is rotation steps, is to increase radius with certain length, shock sensor is laid in earth's surface, forms spiral battle array
Column, and obtain each sensor coordinates information (xj,yj,zj) (j=1,2,3, K, n)；
S2. it generates learning sample: grid dividing is carried out to monitoring region, preset hypocentrum cartridge in monitoring region, obtain one
After hypocentrum cartridge explosion, three-dimensional energy field picture sequence corresponding to all sampled points, as learning sample；
S3. using the corresponding hypocentrum cartridge position of three-dimensional energy field picture sample as training label；
S4. deep learning network frame is constructed；
S5. the sample that S2 is generated is paired into data set to the corresponding hypocentral location determined S3, training network is trained
Good deep learning network；
S6. seismic source location: the vibration signal that practical explosion generates is obtained using sensor array, it is corresponding to obtain practical focus
Three-dimensional energy field picture sequence, randomly select sample therein and be sent into trained deep learning network, obtained output knot
Fruit is focus coordinate.
The utility model has the advantages that
(1) sensor is obtained signal reconstruction using monitoring region as black box by the method that the present invention uses deep learning
Three-dimensional energy field as input, using hypocentral location as export, foundation in energy field hypocentral location identification learning method,
Compared with prior art, reduce extracting of positional parameters during traditional shallow-layer seismic source location, location model modeling and positioning mould
The intermediate steps such as type resolving, greatly improve seismic source location efficiency, while reducing seismic source location precision monitored area letter
The dependence of road reconstruction precision provides a kind of new seismic source location method for underground shallow layer seismic source location.
(2) by sensor array with spiral cloth station, compared with prior art, the range of receiving of vibration signal is expanded,
Seismic phase information redundance of the different quick-fried hearts away under is increased, blind location area is eliminated, improves energy field degree of focus, expanded axis
To the data sample amount of, shearing, data are provided for seismic source location and are guaranteed；
(3) to monitoring region grid division, each grid is set as virtual focus, the method being superimposed with reverse-time migration amplitude
Grid is filled, building contains " seismic source information " three-dimensional energy field picture, and the time serial message of signal is obtained using sensor, is obtained
To the three-dimensional energy field picture sequence containing seismic source information, compared with prior art, three-dimensional energy field sample is greatly expanded
Quantity and diversity increase the generalization ability of depth of focus study, are single goal and few target seismic source location problem, provide
A kind of generation method of redundancy, diversified location information.
Detailed description of the invention
Fig. 1 is positioning flow figure of the invention；
Fig. 2 is sensor cloth station figure；
Fig. 3 is monitoring region grid dividing schematic diagram；
Fig. 4 is deep learning network structure.
Specific embodiment
To keep the purpose of the present invention, content and advantage clearer, a specific embodiment of the invention is made into one below
Step detailed description.
A kind of underground shallow layer seismic source location method based on deep learning proposed by the present invention, which is characterized in that specific packet
Include following steps:
S1. distributed shock sensor array is laid
It is to increase radius with 1m by n sensor with 10 ° for rotation steps in the monitoring random preset coordinate origin in region,
According to clockwise, shock sensor is laid in earth's surface, forms spiral array, obtains each sensing using high-precision Beidou
Device coordinate information (xj,yj,zj) (j=1,2,3, K, n)；N=44；
S2. learning sample is generated
S2.1: monitoring area grid is divided:
As shown in figure 3, carrying out grid dividing to monitored region, (by positioning accuracy request) divides the space into N number of big
Small identical cube grid, obtains each mesh coordinate (hi,li,ki) (i=1,2,3, K, N)；
S2.2: default source signal is obtained:
In monitoring 4, region quadrant, 1 hair hypocentrum cartridge, the hypocentrum cartridge for the first quartile that detonates, using described are preset respectively
Shock sensor array obtains the vibration signal that explosion generates；
S2.3 extracts preliminary wave then:
Then believed using the preliminary wave preliminary wave that then extracting method (such as long short time-window method) obtains all the sensors signal
Breath；
To the hypocentrum cartridge of fourth quadrant, repeat the above steps S2.3, obtains different quadrants for S2.4, second quadrant that successively detonates
Hypocentrum cartridge detonation after corresponding preliminary wave then；
S2.5: monitoring region velocity field model is established:
The preliminary waves of 4 explosions is obtained then after information, by preliminary wave then information and its corresponding sensor position
The location information of information and four hypocentrum cartridges is as input parameter, and chromatography establishes underground shallow layer speed when being walked using preliminary wave
, revised velocity field information is obtained by Quadratic interpolation Shortest path ray tracing method (PTISPR)；
S2.6: information when walking of each grid is calculated
After obtaining velocity field information, with each grid (hi,li,ki) it is used as virtual focus, it calculates it and propagates to each sensing
Device (xj,yj,zj) (j=1,2,3, K, n) information t when walkingij；
S2.7: the Zhang San generated under a hypocentrum cartridge explosive event ties up energy field picture
By taking i-th of grid as an example, it is assumed that i-th of grid is virtual focus, and it is M that the sampling of each sensor, which is always counted, will
J-th of collected vibration signal of sensor institute, according to above-mentioned acquired time tijInverse offset is carried out, i.e., is sensed j-th
The collected vibration signal of device institute, pushes back t to 0 moment direction from time shaftij, i.e. removal tijSample point before moment, In
The sample tail portion supplements the 0 of respective numbers, is always counted with keeping sampling as M, and so on, the signal that n sensor is acquired
It carries out reverse-time migration the sensor array signal after offset is overlapped with the corresponding amplitude of sampled point m, obtains i-th of net
Lattice are in the corresponding energy summation of sampled point m
In formula: aj(m)The amplitude size for being j-th of sensor at sampled point m；N is number of sensors；T is acquisition signal
Time window length；
All grids in region are looped through, energy filling is successively carried out, obtains three-dimensional corresponding to m-th of sampled point
Energy diagram；
S2.8: the corresponding three-dimensional energy field picture sequence of the default focus of generation one
With sample rate 1/fsIt is successively progressive for step-length, S2.7 is repeated, successively carries out amplitude superposition again in different sampled points
Filling, after obtaining a hypocentrum cartridge explosion, three-dimensional energy field picture sequence corresponding to all sampled points；
S2.9: enlarged sample capacity
S2.7-2.8 is repeated, the corresponding three-dimensional energy field picture sample sequence of No. 4 hypocentrum cartridges explosions is obtained；
S3. using hypocentrum cartridge position corresponding to three-dimensional energy field picture sample as training label；
S4. construction depth learning network frame
As shown in table 2, depth of focus learning network is by the first convolutional layer, the first close gang mould block, the second convolutional layer, the first pond
Change layer, the second close gang mould block, third convolutional layer, the second pond layer, the close gang mould block of third, Volume Four lamination, third pond layer according to
It is secondary to be formed by connecting.
Wherein close gang mould block is made of 5 convolutional layers, by connecting 10 kinds of connection types of composition two-by-two.In addition to the first convolution
Layer, remaining convolutional layer have all used hidden Regularization method such as by probability 20%: random inactivation (dropout algorithm).
2 depth of focus learning network structure of table
S5. network is trained
S5.1, the learning sample that S2 is generated is paired into data set (S2, S3) to the corresponding hypocentral location determined S3, at random
80% is used as training sample in extraction data set, and residue 20% is as verifying sample；
Training sample is input in deep learning network by S5.2, obtains each layer parameter information of deep learning network frame；
Above-mentioned parameter is substituted into deep learning network frame by S5.3, and verifying sample input deep learning network is predicted
Source position information, and hypocentral location corresponding with verifying sample compares to obtain the precision of prediction of model；
If S5.4 precision of prediction is undesirable, continue the size and quantity that adjust each layer convolution kernel, until reaching expected mesh
Mark, obtains trained deep learning network.
S6. seismic source location
After terminating above-mentioned repetitive exercise, start practical seismic source location.
S6.1, the vibration signal that practical explosion generates is obtained using sensor array；
On the basis of S6.2, the monitoring region gridding information obtained using above-mentioned steps 2 and velocity field model, step is repeated
S2.3, S2.6-S2.8 obtain the corresponding three-dimensional energy field picture sequence of practical focus.
S6.3, the trained deep learning network of 50 samples feeding therein is randomly selected；By accordingly export 50 groups
Positioning result obtains accurate focus coordinate in a manner of average weighted.
The above is only a preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, without departing from the technical principles of the invention, several improvement and deformations can also be made, these improvement and deformations
Also it should be regarded as protection scope of the present invention.
Claims (10)
1. a kind of underground shallow layer seismic source location method based on deep learning, specifically includes the following steps:
S1. distributed shock sensor array is laid: in the monitoring random preset coordinate origin in region, by n sensor with certain
Angle is rotation steps, is to increase radius with certain length, shock sensor is laid in earth's surface, forms spiral array, and
Obtain each sensor coordinates information (xj,yj,zj) (j=1,2,3, K, n)；
S2. it generates learning sample: grid dividing is carried out to monitoring region, preset hypocentrum cartridge in monitoring region, obtain a focus
Three-dimensional energy field picture sequence corresponding to all sampled points after bullet explosion, and as learning sample；
S3. using hypocentrum cartridge position corresponding to learning sample as training label；
S4. deep learning network frame is constructed；
S5. the training label determined the S2 learning sample generated and S3 is input in the network of S4 building and instructs as data set
Practice network, obtains trained deep learning network；
S6. seismic source location: the vibration signal that practical explosion generates is obtained using sensor array, obtains practical focus corresponding three
Energy field image sequence is tieed up, sample therein is randomly selected and is sent into trained deep learning network, obtained output result is i.e.
For focus coordinate.
2. a kind of underground shallow layer seismic source location method based on deep learning according to claim 1, which is characterized in that S1
In, it is to increase radius with 1m by n sensor with 10 ° for rotation steps in the monitoring random preset coordinate origin in region, according to
Clockwise, shock sensor is laid in earth's surface, forms spiral array.
3. a kind of underground shallow layer seismic source location method based on deep learning according to claim 1, which is characterized in that S2
Specifically includes the following steps:
S2.1: monitoring area grid is divided:
Grid dividing is carried out to monitoring region, divides the space into the identical cube grid of N number of size, each grid is obtained and sits
Mark (hi,li,ki) (i=1,2,3, K, N)；
S2.2: default source signal is obtained:
In monitoring 4, region quadrant, 1 hair hypocentrum cartridge is preset respectively, and the hypocentrum cartridge for the first quartile that detonates utilizes the vibration
Sensor array obtains the vibration signal that explosion generates；
S2.3: the preliminary wave then information of all the sensors signal is extracted；
S2.4: to the hypocentrum cartridge of fourth quadrant, repeat the above steps second quadrant that successively detonates S2.3, obtains the shake of different quadrants
Bounce quick-fried rear corresponding preliminary wave then in source；
S2.5: monitoring region velocity field model is established；
S2.6: information when walking of each grid is calculated:
With each grid (hi,li,ki) it is used as virtual focus, it calculates it and propagates to each sensor (xj,yj,zj) (j=1,2,3, K,
N) information t when walkingij；
S2.7: the Zhang San generated under a hypocentrum cartridge explosive event ties up energy field picture；
S2.8: using sample rate as step-length, repeating S2.7, successively progressive, successively carries out amplitude superposition in different sampled points and fills out again
It fills, after obtaining a hypocentrum cartridge explosion, three-dimensional energy field picture sequence corresponding to all sampled points；
S2.9: repeating S2.7-2.8, obtains the corresponding three-dimensional energy field picture sample sequence of No. 4 hypocentrum cartridges explosions.
4. a kind of underground shallow layer seismic source location method based on deep learning according to claim 3, which is characterized in that
In S2.7, it is assumed that i-th of grid is virtual focus, and it is M that the sampling of each sensor, which is always counted, and j-th of sensor is acquired
The vibration signal arrived, according to time t when acquiring awayijInverse offset is carried out, t is removedijSample point before moment, and at this
Sample tail portion supplements the 0 of respective numbers, to keep sampling total points as M, and so on, the signal that n sensor is acquired into
Row reverse-time migration is overlapped with the corresponding amplitude of sampled point m to the sensor array signal after offset, obtains i-th of grid
In the corresponding energy summation of sampled point m
In formula: aj(m)The amplitude size for being j-th of sensor at sampled point m；N is number of sensors；
All grids in region are looped through, energy filling is successively carried out, obtains three-dimensional energy corresponding to m-th of sampled point
Figure.
5. a kind of underground shallow layer seismic source location method based on deep learning according to claim 4, which is characterized in that S4
In, depth of focus learning network is by the first convolutional layer, the first close gang mould block, the second convolutional layer, the first pond layer, the second close gang mould
Block, third convolutional layer, the second pond layer, the close gang mould block of third, Volume Four lamination, third pond layer are connected in sequence.
6. a kind of underground shallow layer seismic source location method based on deep learning according to claim 5, which is characterized in that institute
It states close gang mould block to be made of 5 convolutional layers, by connecting 10 kinds of connection types of composition two-by-two.
7. a kind of underground shallow layer seismic source location method based on deep learning according to claim 1-6, special
Sign is, S5 specifically includes the following steps:
S5.1, the learning sample that S2 is generated is paired into data set to the corresponding hypocentral location determined S3, randomly selects data set
In a part as training sample, residue is as verifying sample；
Training sample is input in deep learning network by S5.2, obtains each layer parameter information of deep learning network frame；
Above-mentioned parameter is substituted into deep learning network frame by S5.3, and verifying sample input deep learning network is obtained prediction focus
Location information, and true hypocentral location corresponding with verifying sample compares to obtain the precision of prediction of model.
8. a kind of underground shallow layer seismic source location method based on deep learning according to claim 7, which is characterized in that such as
Fruit S5.3 precision of prediction is undesirable, and the size and quantity for continuing to adjust each layer convolution kernel are trained until reaching target
Good deep learning network.
9. a kind of underground shallow layer seismic source location method based on deep learning according to claim 8, which is characterized in that S6
In, it randomly selects 50 samples therein and is sent into trained deep learning network, by accordingly export 50 groups of positioning results, with
Average weighted mode obtains focus coordinate.
10. special according to a kind of described in any item underground shallow layer seismic source location methods based on deep learning of claim 3-6
Sign is, in S2.5, passes through the preliminary wave of acquisition then information and its corresponding sensor position information and hypocentrum cartridge
For location information as input parameter, chromatography establishes underground shallow layer velocity field when being walked using preliminary wave, most short by Quadratic interpolation
Path-ray back tracking method obtains revised velocity field information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910823790.6A CN110414675B (en) | 2019-09-02 | 2019-09-02 | Underground shallow seismic source positioning method based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910823790.6A CN110414675B (en) | 2019-09-02 | 2019-09-02 | Underground shallow seismic source positioning method based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110414675A true CN110414675A (en) | 2019-11-05 |
| CN110414675B CN110414675B (en) | 2022-05-27 |
Family
ID=68370002
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910823790.6A Active CN110414675B (en) | 2019-09-02 | 2019-09-02 | Underground shallow seismic source positioning method based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110414675B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111505708A (en) * | 2020-04-28 | 2020-08-07 | 西安交通大学 | Deep learning-based strong reflection layer stripping method |
| CN112051611A (en) * | 2020-09-07 | 2020-12-08 | 中北大学 | Underground shallow layer detonation point positioning method based on deep reinforcement learning |
| CN112114361A (en) * | 2020-09-07 | 2020-12-22 | 中北大学 | Vibration field space-time tomography method in shallow underground complex space |
| CN112231974A (en) * | 2020-09-30 | 2021-01-15 | 山东大学 | TBM rock breaking seismic source seismic wave field characteristic recovery method and system based on deep learning |
| CN112686259A (en) * | 2020-12-16 | 2021-04-20 | 中国石油大学(北京) | Rock image intelligent identification method and device based on deep learning and storage medium |
| CN112748463A (en) * | 2020-12-21 | 2021-05-04 | 中国科学技术大学 | Local offset imaging method based on deep learning illumination analysis |
| CN113762481A (en) * | 2021-09-13 | 2021-12-07 | 中国地震局地球物理研究所 | Tomography method and system based on deep learning |
| CN115951404A (en) * | 2023-02-06 | 2023-04-11 | 中南大学 | Seismic source positioning method and system based on historical data |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101770038A (en) * | 2010-01-22 | 2010-07-07 | 中国科学院武汉岩土力学研究所 | Intelligent positioning method of mine microquake sources |
| CN102841371A (en) * | 2012-09-20 | 2012-12-26 | 中北大学 | Compound intelligent vibration sensor and vibration source test and location method |
| CN102841373A (en) * | 2012-08-23 | 2012-12-26 | 中国石油集团川庆钻探工程有限公司地球物理勘探公司 | Microseism positioning method based on azimuth angle constraint |
| US20130054147A1 (en) * | 2011-07-07 | 2013-02-28 | Nanoseis Llc | System and Method for Narrow Beam Scanning Microseismic Monitoring |
| CN103605151A (en) * | 2013-11-20 | 2014-02-26 | 中北大学 | Distributed group wave shallow-layer slight shock positioning method based on phase measuring |
| CN104166159A (en) * | 2014-07-15 | 2014-11-26 | 刘改成 | Fracture conformation processing method of four-dimensional microseismic monitoring and system thereof |
| CN104678427A (en) * | 2015-03-11 | 2015-06-03 | 山东大学 | Tunnel boring machine rock breaking seismic source three-dimensional seismic advanced detection device and method |
| CN104820240A (en) * | 2015-05-15 | 2015-08-05 | 江南大学 | Seismic tomography method based on WSNs |
| CN105093297A (en) * | 2015-07-10 | 2015-11-25 | 中联煤层气有限责任公司 | Quality control method for minor tremor positioning precision |
| CN105954795A (en) * | 2016-04-25 | 2016-09-21 | 吉林大学 | Grid successive dissection method used for microseismic positioning |
| CN106461803A (en) * | 2014-03-20 | 2017-02-22 | 施蓝姆伯格技术公司 | Reconstructing impulsive source seismic data from time distributed firing airgun array data |
| CN107132578A (en) * | 2017-04-06 | 2017-09-05 | 吉林大学 | A kind of microseism ground monitoring velocity model corrections algorithm |
| CN108646295A (en) * | 2018-06-29 | 2018-10-12 | 深圳市汇沣世纪数据工程有限公司 | Determination method, apparatus, equipment and the storage medium of investigation depth |
| CN109212597A (en) * | 2018-10-16 | 2019-01-15 | 中国矿业大学(北京) | Microquake sources automatic positioning method based on deepness belief network and scanning superposition |
| CN109974833A (en) * | 2017-12-27 | 2019-07-05 | 中国人民解放军63653部队 | Quasi-distributed underground explosion earthquake motion measuring system based on fiber grating sensing technology |
| CN110018516A (en) * | 2019-05-07 | 2019-07-16 | 西安石油大学 | A kind of decoupling wave field microseism inverse time interference localization method |
-
2019
- 2019-09-02 CN CN201910823790.6A patent/CN110414675B/en active Active
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101770038A (en) * | 2010-01-22 | 2010-07-07 | 中国科学院武汉岩土力学研究所 | Intelligent positioning method of mine microquake sources |
| US20130054147A1 (en) * | 2011-07-07 | 2013-02-28 | Nanoseis Llc | System and Method for Narrow Beam Scanning Microseismic Monitoring |
| CN102841373A (en) * | 2012-08-23 | 2012-12-26 | 中国石油集团川庆钻探工程有限公司地球物理勘探公司 | Microseism positioning method based on azimuth angle constraint |
| CN102841371A (en) * | 2012-09-20 | 2012-12-26 | 中北大学 | Compound intelligent vibration sensor and vibration source test and location method |
| CN103605151A (en) * | 2013-11-20 | 2014-02-26 | 中北大学 | Distributed group wave shallow-layer slight shock positioning method based on phase measuring |
| CN106461803A (en) * | 2014-03-20 | 2017-02-22 | 施蓝姆伯格技术公司 | Reconstructing impulsive source seismic data from time distributed firing airgun array data |
| CN104166159A (en) * | 2014-07-15 | 2014-11-26 | 刘改成 | Fracture conformation processing method of four-dimensional microseismic monitoring and system thereof |
| CN104678427A (en) * | 2015-03-11 | 2015-06-03 | 山东大学 | Tunnel boring machine rock breaking seismic source three-dimensional seismic advanced detection device and method |
| CN104820240A (en) * | 2015-05-15 | 2015-08-05 | 江南大学 | Seismic tomography method based on WSNs |
| CN105093297A (en) * | 2015-07-10 | 2015-11-25 | 中联煤层气有限责任公司 | Quality control method for minor tremor positioning precision |
| CN105954795A (en) * | 2016-04-25 | 2016-09-21 | 吉林大学 | Grid successive dissection method used for microseismic positioning |
| CN107132578A (en) * | 2017-04-06 | 2017-09-05 | 吉林大学 | A kind of microseism ground monitoring velocity model corrections algorithm |
| CN109974833A (en) * | 2017-12-27 | 2019-07-05 | 中国人民解放军63653部队 | Quasi-distributed underground explosion earthquake motion measuring system based on fiber grating sensing technology |
| CN108646295A (en) * | 2018-06-29 | 2018-10-12 | 深圳市汇沣世纪数据工程有限公司 | Determination method, apparatus, equipment and the storage medium of investigation depth |
| CN109212597A (en) * | 2018-10-16 | 2019-01-15 | 中国矿业大学(北京) | Microquake sources automatic positioning method based on deepness belief network and scanning superposition |
| CN110018516A (en) * | 2019-05-07 | 2019-07-16 | 西安石油大学 | A kind of decoupling wave field microseism inverse time interference localization method |
Non-Patent Citations (6)
| Title |
|---|
| HONN KAO ET AL: "The Source-Scanning Algorithm: mapping the distribution of seismic sources in time and space", 《GEOPHYSICAL JOURNAL INTERNATIONAL》 * |
| LI JIAN ET AL: "A PSO microseismic localization method based on group waves time difference information", 《JOURNAL OF MEASUREMENT SCIENCE AND INSTRUMENTATION》 * |
| YANGXU WU ET AL: "Research on Microseismic Source Locations Based on Deep Reinforcement Learning", 《IEEE ACCESS》 * |
| 李萌等: "微地震逆时聚焦定位算法的模拟实验研究", 《中国石油大学学报(自然科学版)》 * |
| 王纪程: "基于网格剖分的微地震地面监测定位技术研究", 《中国优秀硕士学位论文全文数据库 基础科学辑》 * |
| 雷朝阳等: "震源逆时定位的成像条件分析", 《地球物理学进展》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111505708B (en) * | 2020-04-28 | 2021-04-20 | 西安交通大学 | Deep learning-based strong reflection layer stripping method |
| CN111505708A (en) * | 2020-04-28 | 2020-08-07 | 西安交通大学 | Deep learning-based strong reflection layer stripping method |
| US11243320B2 (en) | 2020-04-28 | 2022-02-08 | Xi'an Jiaotong University | Method of stripping strong reflection layer based on deep learning |
| CN112051611A (en) * | 2020-09-07 | 2020-12-08 | 中北大学 | Underground shallow layer detonation point positioning method based on deep reinforcement learning |
| CN112114361A (en) * | 2020-09-07 | 2020-12-22 | 中北大学 | Vibration field space-time tomography method in shallow underground complex space |
| CN112231974B (en) * | 2020-09-30 | 2022-11-04 | 山东大学 | Deep learning-based method and system for recovering seismic wave field characteristics of rock breaking seismic source of TBM (Tunnel boring machine) |
| CN112231974A (en) * | 2020-09-30 | 2021-01-15 | 山东大学 | TBM rock breaking seismic source seismic wave field characteristic recovery method and system based on deep learning |
| CN112686259A (en) * | 2020-12-16 | 2021-04-20 | 中国石油大学(北京) | Rock image intelligent identification method and device based on deep learning and storage medium |
| CN112686259B (en) * | 2020-12-16 | 2023-09-26 | 中国石油大学(北京) | Rock image intelligent recognition method and device based on deep learning and storage medium |
| CN112748463A (en) * | 2020-12-21 | 2021-05-04 | 中国科学技术大学 | Local offset imaging method based on deep learning illumination analysis |
| CN113762481A (en) * | 2021-09-13 | 2021-12-07 | 中国地震局地球物理研究所 | Tomography method and system based on deep learning |
| CN113762481B (en) * | 2021-09-13 | 2024-03-12 | 中国地震局地球物理研究所 | Tomographic imaging method and system based on deep learning |
| CN115951404A (en) * | 2023-02-06 | 2023-04-11 | 中南大学 | Seismic source positioning method and system based on historical data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110414675B (en) | 2022-05-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110414675A (en) | A kind of underground shallow layer seismic source location method based on deep learning | |
| CN110516650A (en) | A kind of blind space seismic source location system of shallow-layer based on shock sensor array | |
| Parker et al. | Active‐source seismic tomography at the Brady geothermal field, Nevada, with dense nodal and fiber‐optic seismic arrays | |
| Steeples et al. | Seismic reflection methods applied to engineering, environmental, and ground-water problems | |
| CN110261900B (en) | Underground shallow layer microseism positioning system based on speed information | |
| CN105510993B (en) | The buried extrusion pressing type complexity gypseous-salt rock identification of foreland basin and distribution forecasting method | |
| CN105386756B (en) | A method of brittle formation porosity is calculated using dependent variable | |
| Klein et al. | Microearthquakes on the mid‐Atlantic plate boundary on the Reykjanes Peninsula in Iceland | |
| CN110261902B (en) | Underground shallow seismic source positioning method based on multi-spectrum energy synthesis | |
| CN110261903B (en) | Underground seismic source passive positioning method based on reverse-time energy focusing | |
| Langet et al. | Joint focal mechanism inversion using downhole and surface monitoring at the Decatur, Illinois, CO2 injection site | |
| Obermann et al. | Combined large‐N seismic arrays and DAS fiber optic cables across the Hengill Geothermal Field, Iceland | |
| Toledo et al. | Ambient seismic noise monitoring and imaging at the Theistareykir geothermal field (Iceland) | |
| Toney et al. | Joint body‐and surface‐wave tomography of Yucca Flat, Nevada, using a novel seismic source | |
| McCalpin et al. | New developments in onshore paleoseismic methods, and their impact on Quaternary tectonic studies | |
| Molnar et al. | A microearthquakes survey in the Ethiopian Rift | |
| Krauss et al. | A Long‐Term Earthquake Catalog for the Endeavour Segment: Constraints on the Extensional Cycle and Evidence for Hydrothermal Venting Supported by Propagating Rifts | |
| Awad et al. | Focal mechanism of earthquakes from the June 1987 swarm in Aswan, Egypt, calculated by the moment tensor inversion | |
| VanDeMark et al. | Evaluation of seismic-acoustic analysis methods for a real-time UXO monitoring system | |
| Kenedi et al. | Microseismicity and 3-d mapping of an active geothermal field, kilauea lower east rift zone, puna, hawaii | |
| Goertz-Allmann et al. | Integrated active and passive seismic monitoring at the Decatur CCS project | |
| Begovic | Structure and physical properties of the subduction plate boundary | |
| Banghar | New Tables of Angels of Incidence of P Waves as a Function of Epicentral Distance | |
| Dando et al. | Enhancing CO2 monitoring at the Decatur CCS site through improved microseismic location constraints | |
| Gaucher et al. | Propagation of Velocity Model Errors in Earthquake Absolute Locations: Application to the Rittershoffen Geothermal Field |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |