CN110632662A - Algorithm for automatically identifying microseism signals by using DCNN-inclusion network - Google Patents
Algorithm for automatically identifying microseism signals by using DCNN-inclusion network Download PDFInfo
- Publication number
- CN110632662A CN110632662A CN201910910886.6A CN201910910886A CN110632662A CN 110632662 A CN110632662 A CN 110632662A CN 201910910886 A CN201910910886 A CN 201910910886A CN 110632662 A CN110632662 A CN 110632662A
- Authority
- CN
- China
- Prior art keywords
- network
- signals
- dcnn
- inclusion
- algorithm
- 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 claims abstract description 24
- 238000012545 processing Methods 0.000 claims abstract description 9
- 239000011159 matrix material Substances 0.000 claims description 11
- 238000011176 pooling Methods 0.000 claims description 11
- 238000000605 extraction Methods 0.000 claims description 5
- 238000004880 explosion Methods 0.000 claims 1
- 238000000034 method Methods 0.000 abstract description 21
- 238000010586 diagram Methods 0.000 description 13
- 238000012544 monitoring process Methods 0.000 description 12
- 238000005422 blasting Methods 0.000 description 8
- 238000012549 training Methods 0.000 description 8
- 238000004458 analytical method Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000002068 genetic effect Effects 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 239000011435 rock Substances 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 238000012800 visualization Methods 0.000 description 2
- 208000025174 PANDAS Diseases 0.000 description 1
- 208000021155 Paediatric autoimmune neuropsychiatric disorders associated with streptococcal infection Diseases 0.000 description 1
- 240000000220 Panda oleosa Species 0.000 description 1
- 235000016496 Panda oleosa Nutrition 0.000 description 1
- 238000009412 basement excavation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000010348 incorporation Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
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. for interpretation or for event detection
- G01V1/288—Event detection in seismic signals, e.g. microseismics
Landscapes
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Environmental & Geological Engineering (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- Emergency Management (AREA)
- Business, Economics & Management (AREA)
- Acoustics & Sound (AREA)
- Geology (AREA)
- General Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Geophysics (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention discloses an algorithm for automatically identifying microseism signals by utilizing a DCNN-inclusion network, which comprises the following steps: (S1) collecting and processing the micro-seismic signal dataset; (S2) building a convolutional neural network on the processed data set; (S3) establishing a DCNN-inclusion network to judge the processed data set. Through the scheme, the method achieves the purposes of high precision, fast time, high automation degree, high signal-to-noise ratio and almost no influence of signal complexity on identifying the micro-seismic signals, and has very high practical value and popularization value.
Description
Technical Field
The invention belongs to the technical field of microseism monitoring, and particularly relates to an algorithm for automatically identifying microseism signals by using a DCNN-inclusion network.
Background
The microseism monitoring technology is widely applied to the fields of safety monitoring of dam mines, hydraulic fracturing monitoring in shale gas exploitation, stability monitoring of deep-buried tunnels and the like, and a plurality of remarkable research results are obtained, the safety problem is one of important problems faced in a plurality of large projects such as dams and bridges, so how to effectively apply the microseism detection technology in the large projects such as underground caverns and a plurality of large structural bodies plays an important role in extracting, identifying and classifying microseism signals with important influence in long-term projects at high precision and high speed. Therefore, accurate, fast and automatic identification of the micro-seismic signals is a necessary task.
For the identification method of micro-seismic signals, there have been many related researches by scholars at home and abroad, such as SVM method mentioned in the research published by the former soviet union scholars vladimir n. In 1980, an akaikelnformationcriterion (AIC akachi information criterion) algorithm, which was established and developed by japan statistician hongchi research, was frequently used in the identification classification of micro-seismic signals. In 1982, the french geophysicist j. morlet proposed wavelet transform linear time-frequency representation, which was remodeled by several other french scholars to create a signal analysis tool that lays a solid foundation for signal analysis processing. In 1986, scientists including Rumelhart and McClelland proposed BP genetic networks. In 1992, Coifman, R.R and m.v. wickerhauser proposed the concept of wavelet packets, and the wavelet analysis method can be a more detailed method of analyzing signals than the wavelet analysis. The method is characterized in that the algorithm of the range and the scale-free fractal box size is established by studying the fractal characteristics of microseismic signals by the national Zhu-Quanjie et al to identify the mechanical vibration waveform, the blasting waveform and the rockburst waveform of an SVM network, and the AIC algorithm is used in the initial pickup of microseismic human signals by the King flood, such as the application of the AIC in a seismic quick reporting system. The Zhanyanping professor decomposes the monitored seismic signals and artificial blasting signals by small-envelope wave transformation, analyzes the regularity of the kurtosis coefficient, the signal energy, the signal zero crossing rate and the time-frequency characteristic, and obtains an analysis result, and the Bianju professor applies the genetic neural network algorithm to natural earthquakes and artificial blasting events. With the development of science and technology, the recognition effect of the convolutional neural network is superior to that of the artificial neural network.
Microseismic monitoring the stability of the rock mass is monitored by analysing the microseismic waveforms during production using the identification method described above. Due to the complexity of the geographic environment, the monitored microseismic signals are often disturbed by engineering noise. The existing commonly used microseism identification method is a method for referencing seismic waves in natural seismic data processing, adopts a time-frequency analysis method, and eliminates interference noise by adopting methods such as fast Fourier transform, wavelet denoising, (Wiener) Wiener filtering, polarization denoising, Hilbert transform and the like due to the non-stationarity of the waveform of a microseism signal, thereby achieving the purpose of identification. Although the method is successfully applied in some projects, compared with the traditional seismic data, the microseismic data often contains various interference signals such as mechanical vibration, environmental noise, electromagnetic noise, artificial blasting and the like, the signal-to-noise ratio is generally low, and the ideal effect of identifying the microseismic signals by adopting a single algorithm is difficult to achieve.
Disclosure of Invention
In order to overcome the defects in the prior art, the invention provides an algorithm for automatically identifying microseismic signals by using a DCNN-incorporation network, the method is less influenced by a signal-to-noise ratio, and the method can quickly and accurately screen out real microseismic signals from complex mixed signals.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
an algorithm for automatically identifying microseismic signals by utilizing a DCNN-inclusion network comprises the following steps:
(S1) collecting and processing the micro-seismic signal dataset;
(S2) building a convolutional neural network on the processed data set;
(S3) establishing a DCNN-inclusion network to judge the processed data set.
Further, the micro-seismic signal dataset in the step (S1) includes an artificial blasting signal, a micro-seismic signal, a mechanical vibration signal and a random noise signal, wherein the data obtained after the micro-seismic signal dataset is subjected to the preliminary processing and the extraction is a basis for performing a micro-seismic signal algorithm network.
Further, in the step (S2), the convolutional neural network includes an input layer, four convolutional layers, four pooling layers, two full-link layers, and an output layer, where data in the input layer is a processed data set, the convolutional layers mainly perform a convolution operation on the processed data set, and a convolution formula is as follows:
wherein f (x, y) is a convolved matrix, x, y are matrix element coordinate values, g (m, n) is a convolution kernel, m, n are element coordinate values, and z (x, y) is a convolution result.
The pooling layers are periodically inserted between the convolution layers of the convolution network and used for extracting data characteristics and combining the extracted characteristics;
the fully-connected layer is a multilayer sensor network and is used for connecting all the characteristics, fully connecting the characteristics of the convolutional layer and the pooling layer and transmitting the result to the output layer;
the output layer adopts Softmax, and the mathematical expression of the output layer is as follows:
wherein C is the number of categories, C represents the C-th category, wcA weight vector representing class c, x being a given sample, and the Softmax function having a value ranging from 0 to 1, i.e. the output corresponding to each input can be considered as the probability that the input is of class c.
Specifically, in the step (S3), the DCNN-inclusion network is to add an inclusion model structure based on the convolutional neural network, and the inclusion structure is to add convolution kernels with different sizes after the CNN network is subjected to convolution pooling and before the CNN network enters the full connection layer. The sizes of convolution kernels mainly adopt 1x1, 3x3 and 5x5, dimensionality reduction is carried out through the convolution kernels of 1x1, then output features of different scales are fused, and finally the output features enter a full-connection layer. The inclusion structure is shown in fig. 2, and when the time complexity and the space complexity of the network structure are increased, the parameters are reduced, and the added filtering structure can more finely extract features, so that the accuracy is improved.
Compared with the prior art, the invention has the following beneficial effects:
the method can screen out real microseism signals from complex mixed signals acquired in the microseism monitoring process, is effective to three-component and single-component microseism mixed signal data sets acquired through a sensor in the identification process, improves the identification precision by 2.4 percent compared with a CNN structure by adopting a DCNN-inclusion network structure, reaches 93 percent, and has higher identification speed than a CNN algorithm.
Drawings
FIG. 1 is a flow chart of the system of the present invention.
Fig. 2 is a diagram of the inclusion model structure of the present invention.
Fig. 3 is a network visualization structure diagram of the present invention.
FIG. 4 is a schematic diagram of the deployment of a microseismic monitoring geophone and field devices in accordance with an embodiment of the present invention.
FIG. 5 is a diagram of a four-class data set used in the practice of the present invention.
Fig. 6 is a diagram of the network structure of the present invention.
Fig. 7 is a graph comparing the effect of using the present invention and CNN network identification in the examples.
Detailed Description
The present invention is further illustrated by the following figures and examples, which include, but are not limited to, the following examples.
Examples
As shown in fig. 1, an algorithm for automatically identifying microseismic signals by using a DCNN-inclusion network includes the following steps:
(1) and collecting and processing data sets required by the algorithm, wherein the data sets are divided into four types including artificial blasting signals, micro seismic signals, mechanical vibration signals and random noise signals. All data are in an ASC format, the data comprise continuous three-channel waveforms in a period of time, the time interval is 0.000167s, three channels are set to be X, Y and Z channels in the algorithm, and the data in the original ASC format are converted into waveform data and a waveform diagram through a pandas library in python language so as to observe the distribution characteristics and the waveform characteristics of the data. And then carrying out primary processing and extraction on data, wherein the data is the basis of the algorithm network.
(2) Establishing a convolutional neural network (CNN network) for the processed data set, wherein the CNN network comprises an input layer, four convolutional layers, four pooling layers, two full-link layers and an output layer, the data in the input layer is the processed data set, the convolutional layers mainly perform convolution operation on the processed data set, and the convolution formula is as follows:
wherein f (x, y) is a convolved matrix, x, y are matrix element coordinate values, g (m, n) is a convolution kernel, m, n are element coordinate values, and z (x, y) is a convolution result, and in the formula, the ordering of the three matrices starts from 0.
The pooling layers are periodically inserted between the convolution layers of the convolution network and used for extracting data characteristics and combining the extracted characteristics;
the fully-connected layer is a multilayer sensor network and is used for connecting all the characteristics, fully connecting the characteristics of the convolutional layer and the pooling layer and transmitting the result to the output layer;
the output layer adopts Softmax, and the mathematical expression of the output layer is as follows:
wherein C is the number of categories, C represents the C-th category, wcA weight vector representing class c, x being a given sample, and the Softmax function having a value ranging from 0 to 1, i.e. the output corresponding to each input can be considered as the probability that the input is of class c.
Meanwhile, the convolutional neural network can extract more features more quickly through local perception of multiple convolutional kernels and global sharing of parameters, and the training speed is higher under the condition of good feature classification effect.
(3) The DCNN-inclusion network is established to judge the processed data set, the network adds an inclusion model structure on the basis of the CNN network, achieves the effect of sparse parameter reduction and also utilizes the performance of dense matrix optimization in hardware, thereby increasing the time complexity and the space complexity of the network structure, reducing the parameters, increasing the multi-core convolution filtering structure of the inclusion, being capable of extracting the characteristics more finely and improving the accuracy.
The algorithm can be used for accurately identifying the precision and the speed of the microseism, the algorithm adopts four layers of convolution networks and pooling layers, data reshape is required to be formed into a two-dimensional matrix before entering a full-link layer, a dropout method is required to be added for randomly removing a certain proportion of neurons, so that the neurons do not participate in forward propagation, the calculation and iteration speed is higher, the possibility of overfitting is effectively reduced, an Adam optimizer is used as a gradient descent algorithm for updating parameters of the neural network in the process of backward propagation, and the method for updating the step length is calculated, so that the training and identifying speed is obtained. And performing forward propagation calculation on the network by using the data of the verification set every 20 iterations in the network training process, wherein dropout and backward propagation are not needed at the moment, only the feature extraction capability of the current model is verified, whether the classification can be correctly performed or not is verified, and meanwhile, the accuracy and loss of the current iteration times on the training set and the verification set are recorded so as to draw an accuracy curve at a later stage. The structure diagram of the visualization of the DCNN-inclusion network is shown in FIG. 3.
For a better understanding of the invention, the invention is further illustrated below using experimental examples:
the reliability of the invention is verified by adopting microseismic monitoring data of rock burst monitoring of tail water tunnel excavation of a certain hydropower station in southwest of China, the instrument adopts 20-channel microseismic monitoring equipment of IMS, 10 microseismic detectors are arranged on site, wherein, 5 detectors are respectively arranged in a single-component detector and a three-component detector, the sampling frequency is 6kHz, and the schematic diagram of the arrangement of the detectors and the field equipment is shown in figure 4. The acquired microseism mixed signal triggers 4 detectors which comprise 2 three-component detectors and 2 single-component detectors, the contained microseism data time window comprises 3000 sampling points, and the time length is 0.5 s.
Firstly, monitoring data are divided into four types, namely, artificial blasting signals, micro seismic signals, mechanical vibration signals and random noise signals. As shown in fig. 5, (a) to (c) are microseismic waveform diagrams, (d) to (f) are artificial blasting waveform diagrams, (g) to (i) are mechanical vibration waveform diagrams, and (j) to (m) are random noise waveform diagrams. 6000 data per class were selected as datasets and preprocessed with python. Then, 80% of the data was randomly drawn as a training set, 10% of the data was drawn as a validation set, and 10% of the data was drawn as a test set. The network structure diagram of DCNN-acceptance is shown in FIG. 6. The input sample data is 2304 x 3x3 three-dimensional matrix, and after the input sample data is processed by the first four-layer convolutional network and pooled, the dimension of the output matrix is 36x 36x 48. And changing the dimension of the Inceptation structure matrix into 36x 36x 192. And finally, classifying the data and updating the parameters of back propagation through two full connection layers. In the back propagation process, an Adam optimizer is used as a gradient descent algorithm of neural network parameter updating, and finally, softmax is used as an output layer, so that the class to which the sample belongs is predicted. Fig. 7(a) and (b) are accuracy graphs of a test set drawn by a CNN network and a network of the present invention according to training parameters, respectively, and it can be seen from comparison that the network structure of the present invention achieves higher accuracy than the CNN network; FIGS. 7(c) and (d) are graphs of loss rates of the CNN network and the network of the present invention according to the test set drawn by the training parameters, and it can be seen by comparison that the loss rate achieved by the network of the present invention is lower; table 1 shows the comparison of the CNN network algorithm calculation time cases; table 2 the present invention compares the recognition results of CNN network algorithms on actual data.
TABLE 1 actual data calculation time (unit: min) by different methods
TABLE 2 recognition results of actual data by different methods (event: data Length 3000)
From the analysis of the above data, it can be seen that a microseismic event usually triggers multiple detectors at the same time, a data waveform with low signal-to-noise ratio exists in each event, and many interference signals exist, so it is necessary to identify the microseismic signal from the mixed signal. Although the DCNN-inclusion structure is longer than a CNN network in training time, the characteristic fitting capacity of the DCNN-inclusion structure on time sequence data is strong, the characteristic extraction capacity is stronger, and the DCNN-inclusion structure has higher automation degree.
The above embodiments are only preferred embodiments of the present invention, and are not intended to limit the scope of the present invention, but all changes that can be made by applying the principles of the present invention and performing non-inventive work on the basis of the principles shall fall within the scope of the present invention.
Claims (4)
1. An algorithm for automatically identifying microseismic signals by utilizing a DCNN-inclusion network is characterized by comprising the following steps of:
(S1) collecting and processing the micro-seismic signal dataset;
(S2) building a convolutional neural network on the processed data set;
(S3) establishing a DCNN-inclusion network to judge the processed data set.
2. The algorithm for automatically identifying micro-seismic signals using the DCNN-inclusion network as claimed in claim 1, wherein the micro-seismic signal data set in said step (S1) comprises artificial explosion signals, micro-seismic signals, mechanical vibration signals and random noise signals, wherein the data obtained after the preliminary processing and extraction of the micro-seismic signal data set is the basis for the micro-seismic signal algorithm network.
3. The algorithm for automatically identifying microseismic signals using DCNN-inclusion network as claimed in claim 1, wherein the convolutional neural network in the step (S2) comprises an input layer, four convolutional layers, four pooling layers, two fully-connected layers and an output layer, wherein the data in the input layer is a processed data set, the convolutional layers mainly perform convolution operation on the processed data set, and the convolution formula is as follows:
wherein f (x, y) is a convolved matrix, x, y are matrix element coordinate values, g (m, n) is a convolution kernel, m, n are element coordinate values, and z (x, y) is a convolution result.
The pooling layers are periodically inserted between the convolution layers of the convolution network and used for extracting data characteristics and combining the extracted characteristics;
the fully-connected layer is a multilayer sensor network and is used for connecting all the characteristics, fully connecting the characteristics of the convolutional layer and the pooling layer and transmitting the result to the output layer;
the output layer adopts Softmax, and the mathematical expression of the output layer is as follows:
wherein C is the number of categories, C represents the C-th category, wcWeight representing class cThe weight vector, x, is a given sample, and the value of the Softmax function ranges between 0 and 1, i.e., the output corresponding to each input can be regarded as the probability that the input is of the c-th class.
4. The algorithm for automatically identifying microseismic signals by using the DCNN-inclusion network as claimed in claim 1, wherein the DCNN-inclusion network in the step (S3) is obtained by adding an inclusion model structure on the basis of a convolutional neural network, so that the time complexity and the space complexity of the network structure are increased, the parameters are reduced, the filter structure of the multi-core convolution of the inclusion is increased, the features can be extracted more finely, and the accuracy is improved.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910910886.6A CN110632662A (en) | 2019-09-25 | 2019-09-25 | Algorithm for automatically identifying microseism signals by using DCNN-inclusion network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910910886.6A CN110632662A (en) | 2019-09-25 | 2019-09-25 | Algorithm for automatically identifying microseism signals by using DCNN-inclusion network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110632662A true CN110632662A (en) | 2019-12-31 |
Family
ID=68974192
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910910886.6A Pending CN110632662A (en) | 2019-09-25 | 2019-09-25 | Algorithm for automatically identifying microseism signals by using DCNN-inclusion network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110632662A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111190225A (en) * | 2020-01-13 | 2020-05-22 | 吉林大学 | Rock physical mechanism and earthquake prediction method for salt-space shale oil reservoir |
| CN112380198A (en) * | 2020-10-29 | 2021-02-19 | 吉林大学 | Seismic receiving function automatic selection method based on deep learning |
| CN112487952A (en) * | 2020-11-27 | 2021-03-12 | 东北大学 | Mine microseismic signal automatic identification method based on deep learning |
| CN112712007A (en) * | 2020-12-25 | 2021-04-27 | 天津城建大学 | Algorithm for identifying automobile exhaust remote sensing data based on DCNN-inclusion network |
| CN114545504A (en) * | 2022-02-18 | 2022-05-27 | 吉林大学 | Desert seismic exploration random noise modeling method based on deep learning |
| CN114777947A (en) * | 2022-04-06 | 2022-07-22 | 成都理工大学 | Distributed optical fiber temperature abnormal event positioning method |
| CN116299684A (en) * | 2023-05-17 | 2023-06-23 | 成都理工大学 | Novel microseismic classification method based on bimodal neurons in artificial neural network |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150316674A1 (en) * | 2013-01-22 | 2015-11-05 | Cgg Services Sa | Systems and methods for multi-volume directional de-noising |
| CN107784276A (en) * | 2017-10-13 | 2018-03-09 | 中南大学 | Microseismic event recognition methods and device |
| CN109063687A (en) * | 2018-08-29 | 2018-12-21 | 长江大学 | A kind of microseism P wave recognition methods and system based on depth convolutional neural networks |
| 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 |
-
2019
- 2019-09-25 CN CN201910910886.6A patent/CN110632662A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150316674A1 (en) * | 2013-01-22 | 2015-11-05 | Cgg Services Sa | Systems and methods for multi-volume directional de-noising |
| CN107784276A (en) * | 2017-10-13 | 2018-03-09 | 中南大学 | Microseismic event recognition methods and device |
| CN109063687A (en) * | 2018-08-29 | 2018-12-21 | 长江大学 | A kind of microseism P wave recognition methods and system based on depth convolutional neural networks |
| 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 |
Non-Patent Citations (3)
| Title |
|---|
| 刘雨桐 等: "改进卷积神经网络在遥感图像分类中的应用", 《计算机应用》 * |
| 彭桂力 等: "微地震监测混合信号的CNN算法识别研究", 《第二届地球物理信息前沿技术研讨会论文摘要集》 * |
| 童浩然 等: "基于卷积神经网络的图像分类技术", 《科技视界》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111190225A (en) * | 2020-01-13 | 2020-05-22 | 吉林大学 | Rock physical mechanism and earthquake prediction method for salt-space shale oil reservoir |
| CN112380198A (en) * | 2020-10-29 | 2021-02-19 | 吉林大学 | Seismic receiving function automatic selection method based on deep learning |
| CN112487952A (en) * | 2020-11-27 | 2021-03-12 | 东北大学 | Mine microseismic signal automatic identification method based on deep learning |
| CN112712007A (en) * | 2020-12-25 | 2021-04-27 | 天津城建大学 | Algorithm for identifying automobile exhaust remote sensing data based on DCNN-inclusion network |
| CN114545504A (en) * | 2022-02-18 | 2022-05-27 | 吉林大学 | Desert seismic exploration random noise modeling method based on deep learning |
| CN114777947A (en) * | 2022-04-06 | 2022-07-22 | 成都理工大学 | Distributed optical fiber temperature abnormal event positioning method |
| CN116299684A (en) * | 2023-05-17 | 2023-06-23 | 成都理工大学 | Novel microseismic classification method based on bimodal neurons in artificial neural network |
| CN116299684B (en) * | 2023-05-17 | 2023-07-21 | 成都理工大学 | Novel microseismic classification method based on bimodal neurons in artificial neural network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110632662A (en) | Algorithm for automatically identifying microseism signals by using DCNN-inclusion network | |
| Huang et al. | Micro-seismic event detection and location in underground mines by using Convolutional Neural Networks (CNN) and deep learning | |
| CN108805269B (en) | Method for picking seismic facies arrival time based on LSTM recurrent neural network | |
| Dong et al. | New suppression technology for low-frequency noise in desert region: The improved robust principal component analysis based on prediction of neural network | |
| CN111580151B (en) | SSNet model-based earthquake event time-of-arrival identification method | |
| CN103487832A (en) | Method for classifying supervised waveforms in three-dimensional seismic signal | |
| CN112884134B (en) | Time domain-based convolutional neural network model for seismic facies identification and application | |
| CN109613610B (en) | Automatic picking method for microseism signal arrival time difference | |
| CN113687433B (en) | Bi-LSTM-based magnetotelluric signal denoising method and system | |
| CN112231974B (en) | Deep learning-based method and system for recovering seismic wave field characteristics of rock breaking seismic source of TBM (Tunnel boring machine) | |
| CN110895348A (en) | Method, system and storage medium for extracting low-frequency information of seismic elastic impedance | |
| CN113792685A (en) | Microseism event detection method based on multi-scale convolution neural network | |
| CN112464721A (en) | Automatic microseism event identification method and device | |
| Zhao et al. | A hybrid recognition model of microseismic signals for underground mining based on CNN and LSTM networks | |
| CN110837111B (en) | Seismic data interpolation method and system | |
| Peng et al. | Recognition of rock micro-fracture signal based on deep convolution neural network inception algorithm | |
| Li et al. | Magnetotelluric data denoising method combining two deep-learning-based models | |
| Zhang et al. | Microseismic source location based on improved artificial bee colony algorithm: Performance analysis and case study | |
| CN113419278B (en) | Well-seismic joint multi-target simultaneous inversion method based on state space model and support vector regression | |
| CN115932959A (en) | Surface wave frequency dispersion quality rapid evaluation method based on UNet | |
| CN114706128A (en) | Microseism data noise suppression method and system, storage medium and seismic information processing equipment | |
| Dai et al. | Stratigraphic automatic correlation using SegNet semantic segmentation model | |
| Zhu et al. | Deep learning for seismic event detection of earthquake aftershocks | |
| CN112578439B (en) | Seismic inversion method based on space constraint | |
| Pulli | Extracting and processing signal parameters for regional seismic event identification |
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 |
Application publication date: 20191231 |
|
| RJ01 | Rejection of invention patent application after publication |