CN204925409U - Super high -density resistivity method inverting device based on principal component neural network - Google Patents
Super high -density resistivity method inverting device based on principal component neural network Download PDFInfo
- Publication number
- CN204925409U CN204925409U CN201520673270.9U CN201520673270U CN204925409U CN 204925409 U CN204925409 U CN 204925409U CN 201520673270 U CN201520673270 U CN 201520673270U CN 204925409 U CN204925409 U CN 204925409U
- Authority
- CN
- China
- Prior art keywords
- module
- neural network
- inverting
- super high
- super
- 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.)
- Expired - Fee Related
Links
Abstract
The utility model discloses a super high -density resistivity method inverting device based on principal component neural network, the device include microprocessor ARM, WIFI module, just drill module, current potential data memory, electrode array data generator, PCA feature extraction module, sample memory, IO drive module, SPI interface, keyboard, display screen, BP neural network inverting module, parameter storage, power module and clock module. The device low cost, simple structure, convenient to carry can adopt the PCA technique to carry out feature extraction and reduction of dimensionality to the higher -dimension observed data that super high -density resistivity method gathered to utilize the non - linear inversion model of BP neural network fast settling, have higher inverting speed and formation of image precision.
Description
Technical field
The utility model relates to geophysics super-high density Electrical Data inversion interpretation field, particularly relates to a kind of super-high density electrical method inverting device based on major component neural network.
Background technology
Super-high density electrical method is the improvement of high-density electric on acquisition mode, its exploration principle is identical with conventional electrical method, all based on the electrical property difference of rock and ore, solve the hydrology, environment and engineering geological problems by observing and studying the regularity of distribution manually setting up consistent electric field.But in traditional multi-electrode resistivity imaging survey, the device of different arrangement type has different resolution and depth of exploration, in identical geologic structure, there is larger difference in the pseudo-cross section of apparent resistivity of different device, the data volume that electrod-array can gather cannot be made full use of, be unfavorable for later stage high precision inverting.And super-high density electrical method is not subject to the restriction of conventional electrode device and data acquisition modes, in same time, the electrical potential information likely combined between this section electrode can be gathered, ensure that the precision of late time data process and inversion result.Although the case of super-high density electrical method in engineer applied is more, the apparatus design carrying out forward simulation and inversion interpretation for super-high density electrical method is actually rare.Therefore the non-linear inversion device studying super-high density electrical method has positive theoretical and practical significance to the further genralrlization of super-high density electrical method and application.
Summary of the invention
The purpose of this utility model is to provide a kind of super-high density electrical method non-linear inversion device based on major component neural network, this device adopts the higher-dimension observation data of principal component analysis (PCA) PCA to the collection of super-high density electrical method to carry out feature extraction and dimensionality reduction, utilize BP neural network to carry out the Inverse modeling of super-high density electrical method, thus realize the quick of super-high density electrical method and high precision inverting.
For achieving the above object, the technical scheme that the utility model adopts is:
Based on the super-high density electrical method inverting device of major component neural network, it is characterized in that: this device comprises microprocessor ARM, WIFI module, just drills module, potential data storer, electrode spread number generator, PCA characteristic extracting module, sample storage, I/O driver module, SPI interface, keyboard, display screen, BP Neural Network Inversion module, parameter storage, power module and clock module; Wherein microprocessor ARM with just drill module and be connected with WIFI module, be connected with display screen with keyboard by I/O driver module, be connected with BP Neural Network Inversion module by SPI interface; Potential data storer respectively with just drill module, electrode spread number generator is connected, and to be connected with sample storage by PCA characteristic extracting module; BP Neural Network Inversion module is connected with parameter storage with sample storage respectively; Clock module provides synchronous clock for device, and power module provides power supply for device.
The beneficial effects of the utility model are: should be cheap based on the super-high density electrical method inverting installation cost of major component neural network, structure is simple, easy to carry, the higher-dimension observation data of PCA technology to the collection of super-high density electrical method can be adopted to carry out feature extraction and dimensionality reduction, and utilize BP neural network to set up non-linear inversion model fast, there is higher inversion speed and imaging precision.
Accompanying drawing explanation
Fig. 1 is the structured flowchart of the utility model based on the super-high density electrical method inverting device of major component neural network.
Embodiment
As shown in Figure 1, the super-high density electrical method inverting device based on major component neural network comprises microprocessor ARM, WIFI module, just drills module, potential data storer, electrode spread number generator, PCA characteristic extracting module, sample storage, I/O driver module, SPI interface, keyboard, display screen, BP Neural Network Inversion module, parameter storage, power module and clock module.Wherein microprocessor ARM with just drill module and be connected with WIFI module, be connected with display screen with keyboard by I/O driver module, be connected with BP Neural Network Inversion module by SPI interface; Potential data storer respectively with just drill module, electrode spread number generator is connected, and to be connected with sample storage by PCA characteristic extracting module; BP Neural Network Inversion module is connected with parameter storage with sample storage respectively; Clock module provides synchronous clock for device, and power module provides power supply for device.
This device is divided into training and inverting two kinds of mode of operations.In training mode, microprocessor ARM controls current potential matrix when just drilling the exploration of CMOS macro cell electrode, and being stored in potential data storer successively, electrode spread number generator requires in potential data storer, to take out corresponding potential data successively as primary data during neural metwork training according to the electrode spread of super-high density electrical method.After these data carry out feature extraction and dimensionality reduction by PCA characteristic extracting module, form the training sample of neural network and be stored in sample storage.Meanwhile, microprocessor ARM enters the training stage by SPI Interface Controller BP Neural Network Inversion module.BP Neural Network Inversion module reads sample successively, adopts BP algorithm to train, and keeps the neural network parameter after training to parameter storage.Under inverting mode, microprocessor ARM reads in outside survey data by WIFI module, arranges inverted parameters, then enters inversion stage by SPI Interface Controller BP Neural Network Inversion module.BP Neural Network Inversion module reads the neural network parameter in parameter storage, carries out direct inversion, and the result of inverting is back to microprocessor ARM to the survey data of input.Inversion result is carried out imaging display by I/O driver module by microprocessor ARM on a display screen, and carries out the operation such as optimum configurations and data processing by keyboard.
Claims (1)
1. based on a super-high density electrical method inverting device for major component neural network, it is characterized in that: this device comprises microprocessor ARM, WIFI module, just drills module, potential data storer, electrode spread number generator, PCA characteristic extracting module, sample storage, I/O driver module, SPI interface, keyboard, display screen, BP Neural Network Inversion module, parameter storage, power module and clock module; Wherein microprocessor ARM with just drill module and be connected with WIFI module, be connected with display screen with keyboard by I/O driver module, be connected with BP Neural Network Inversion module by SPI interface; Potential data storer respectively with just drill module, electrode spread number generator is connected, and to be connected with sample storage by PCA characteristic extracting module; BP Neural Network Inversion module is connected with parameter storage with sample storage respectively; Clock module provides synchronous clock for device, and power module provides power supply for device.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201520673270.9U CN204925409U (en) | 2015-09-02 | 2015-09-02 | Super high -density resistivity method inverting device based on principal component neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201520673270.9U CN204925409U (en) | 2015-09-02 | 2015-09-02 | Super high -density resistivity method inverting device based on principal component neural network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN204925409U true CN204925409U (en) | 2015-12-30 |
Family
ID=54974522
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201520673270.9U Expired - Fee Related CN204925409U (en) | 2015-09-02 | 2015-09-02 | Super high -density resistivity method inverting device based on principal component neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN204925409U (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106443795A (en) * | 2016-07-21 | 2017-02-22 | 安徽惠洲地质安全研究院股份有限公司 | Method for deducting and synthesizing AM data into ABM data by means of dual mode network parallel electrical method |
CN106570227A (en) * | 2016-10-20 | 2017-04-19 | 湖南师范大学 | Electrode arrangement optimization method and device for ultra-high density resistivity method |
CN112987125A (en) * | 2021-02-22 | 2021-06-18 | 中国地质大学(北京) | Shale brittleness index prediction method based on logging data |
-
2015
- 2015-09-02 CN CN201520673270.9U patent/CN204925409U/en not_active Expired - Fee Related
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106443795A (en) * | 2016-07-21 | 2017-02-22 | 安徽惠洲地质安全研究院股份有限公司 | Method for deducting and synthesizing AM data into ABM data by means of dual mode network parallel electrical method |
CN106443795B (en) * | 2016-07-21 | 2018-05-15 | 安徽惠洲地质安全研究院股份有限公司 | A kind of parallel electrical method AM data of dual-mode network deduce synthesis ABM data methods |
CN106570227A (en) * | 2016-10-20 | 2017-04-19 | 湖南师范大学 | Electrode arrangement optimization method and device for ultra-high density resistivity method |
CN106570227B (en) * | 2016-10-20 | 2019-09-24 | 湖南师范大学 | A kind of electrode arrangement optimization method and device of ultra high density electrical method |
CN112987125A (en) * | 2021-02-22 | 2021-06-18 | 中国地质大学(北京) | Shale brittleness index prediction method based on logging data |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN204925409U (en) | Super high -density resistivity method inverting device based on principal component neural network | |
CN204086568U (en) | A kind of super-high density electrical method data collector | |
CN103615238B (en) | A kind of cavernous formation dual laterolog physical simulating device of scaled down and experimental technique | |
CN105334541B (en) | Suitable for TBM portable resistor rate method advanced forecasting system and its method | |
CN105842751B (en) | Shale reservoir fracture evaluation method | |
CN110824563A (en) | Reservoir lithology prediction method based on Xgboost algorithm | |
CN103628816B (en) | The method of seismic properties Guidance Levels well guiding and device thereof | |
CN104360402A (en) | Well-to-ground joint electrical-method testing method and system | |
CN113720880A (en) | Resistivity monitoring system and method for underground three-dimensional migration process of pollutants | |
CN110119586B (en) | Axial conductivity anisotropy transient electromagnetic three-component three-dimensional FDTD forward modeling method | |
Yu et al. | Exploration of digital twin design mechanism of the deep in situ rock insulation coring device | |
CN105301647A (en) | Method for distinguishing calcareous mudstone from sandstone | |
CN104501942A (en) | Explosion wave measuring device | |
CN115879647B (en) | Stratum lithology prediction method based on combination of domain knowledge and deep learning | |
Gong-Xiang et al. | Methods and practices of reservoir geologic modelling constrained by seismic impedance. | |
Mudge et al. | Evolving inversion methods in geophysics with cloud computing-a case study of an escience collaboration | |
CN105604538A (en) | Compensated neutron logging instrument with controllable neutron source and control circuit of same | |
CN201110896Y (en) | High resolution mine geology detector | |
CN105005078A (en) | Seismic data quick time depth conversion method based on speed model | |
CN201306166Y (en) | Signal simulator for mud logging unit | |
CN204256187U (en) | Parallel electrical method test macro is combined on a kind of well ground | |
CN104502986B (en) | Physical prospecting induced electrical sounding data chromatography processing method | |
CN209212166U (en) | A kind of geological exploration in coal mine vehicle | |
Damsleth et al. | Geostatistical approaches in reservoir evaluation | |
CN112649881A (en) | Numerical simulation method and system for anisotropic medium transient electromagnetic field |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20151230 Termination date: 20160902 |