CN116452891A - Automatic identification method for type of seafloor soil based on naive Bayes algorithm - Google Patents
Automatic identification method for type of seafloor soil based on naive Bayes algorithm Download PDFInfo
- Publication number
- CN116452891A CN116452891A CN202310452265.4A CN202310452265A CN116452891A CN 116452891 A CN116452891 A CN 116452891A CN 202310452265 A CN202310452265 A CN 202310452265A CN 116452891 A CN116452891 A CN 116452891A
- Authority
- CN
- China
- Prior art keywords
- soil
- type
- calculating
- end resistance
- cone end
- 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
- 239000002689 soil Substances 0.000 title claims abstract description 60
- 238000000034 method Methods 0.000 title claims abstract description 16
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 7
- 238000001914 filtration Methods 0.000 claims abstract description 20
- 230000035515 penetration Effects 0.000 claims description 22
- 238000012549 training Methods 0.000 claims description 15
- 239000011148 porous material Substances 0.000 claims description 11
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 7
- 230000003068 static effect Effects 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 5
- 239000000523 sample Substances 0.000 claims description 4
- 239000013535 sea water Substances 0.000 claims description 3
- 239000002002 slurry Substances 0.000 claims description 3
- 239000004927 clay Substances 0.000 description 4
- 239000011159 matrix material Substances 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 239000004576 sand Substances 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/36—Applying a local operator, i.e. means to operate on image points situated in the vicinity of a given point; Non-linear local filtering operations, e.g. median filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Multimedia (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Databases & Information Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Nonlinear Science (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Investigation Of Foundation Soil And Reinforcement Of Foundation Soil By Compacting Or Drainage (AREA)
Abstract
The invention discloses a naive Bayes algorithm-based automatic identification method for the type of the seafloor soil. The method comprises the following basic steps: 1) CPT data acquisition; 2) Calculating attribute characteristics; 3) Calculating fluctuation characteristics of the normalized cone end resistance; 4) Filtering characteristic parameters; 5) The type of the seafloor soil is identified using a bayesian classifier. The method has the advantages of simplicity, small calculated amount, good real-time performance, manpower saving, easy realization and the like. The method is suitable for automatic identification of the seabed soil type.
Description
Technical Field
The invention relates to the technical fields of ocean engineering, ocean mapping, ocean oil exploration and development, ocean investigation and the like, in particular to a naive Bayesian algorithm-based automatic identification method for the type of the seabed soil.
Background
The type and the characteristics of the seabed soil are one of important contents of ocean engineering, ocean census and ocean oil exploration and development, are important reference indexes for ocean oil exploitation and offshore wind power construction, and are the precondition of fully knowing the type and the characteristics of the seabed soil to exploit ocean resources.
And through a static cone penetration test (CPT for short), a signal fed back after the static cone penetration probe contacts the seabed soil quality can be obtained, and the data of three attributes including cone end resistance, side wall friction and pore water pressure are included, so that the information about the soil quality attribute of the measuring point is reflected. Drawing a soil property classification plate by using two of the three properties as classification basis for classifying soil property; further, the three attributes are modified through formula derivation, normalized attribute features are proposed and applied to the earthen classification. Compared with the original three attributes, the modified and normalized attribute data are adopted to conduct soil property classification research, the influence of the depth of the detection equipment and the soil layer on the soil property classification is eliminated, and the data structure and distribution are more reasonable.
The traditional soil property classification work is generally based on soil property attribute data, and is manually judged by combining with a division criterion provided by a classification plate. In recent years, studies have been made on a method for automatically classifying soil properties by machine learning. And forming a training data set based on the soil property attribute data, and training by utilizing an algorithm to obtain a machine learning model, thereby realizing automatic identification of soil lithology. The method has the advantages of simple operation, high working efficiency and the like, and can realize online, real-time and automatic prediction of the type of the seabed soil.
Although there are some studies on soil classification methods at present, most of such studies are focused on the field of land soil type recognition, and few studies on the identification of the type of seafloor soil are involved. Meanwhile, in the existing method, the algorithm simplicity, classification accuracy and other performances still need to be improved.
Disclosure of Invention
The invention aims to automatically identify the type of the seafloor soil by using CPT data. The method has the advantages of simplicity, small calculated amount, good real-time performance, manpower saving, easy realization and the like. The method is suitable for automatic identification of the seabed soil type.
The invention comprises the following steps:
(1) CPT data acquisition:
the cone end resistance q is obtained through a static cone penetration test c Friction force f of side wall s Pore water pressure u 2 (unit: MPa), probe penetration depth d (unit: m);
(2) Calculating attribute characteristics:
according toCalculating to obtain pore pressure parameter ratio B q According to->Calculating to obtain normalized cone end resistance Q t According to->Calculating to obtain normalized friction-resistance ratio F r Wherein Δu=u 2 -γ w X d is the excess pore water pressure, gamma w Is the volume weight of the seawater or the slurry; q t ＝q c +(1+α)u 2 Alpha is the specification of the cone head for correcting the cone end resistance; sigma'. vo Let γ '×d be the effective overburden pressure, γ' be the average effective volume weight of overburden, γ w Manually setting gamma' according to the regional characteristics, wherein alpha is CPT equipment parameters and is determined according to CPT equipment;
(3) Calculate Q t Wave characteristics:
for Q t Average filtering is carried out to obtain normalized cone end resistance Q after filtering t ' according to DeltaQ t ＝|Q t -Q t ' I is calculated to obtain normalized cone end resistance deviation delta Q t If DeltaQ t_j < 0.5, deltaQ t_j ＝ΔQ t_j-1 Wherein DeltaQ t_j-1 And DeltaQ t_j The penetration depths are respectively j-1 and jA cone end resistance deviation, j=2, 3, … n, n is the maximum penetration depth;
(4) Characteristic parameter filtering:
pair B q 、Q t 、F r 、ΔQ t Sliding median filtering, wherein the width of a filtering window is w 1 ，w 1 Is odd;
(5) Identifying the type of the seafloor soil by using a Bayes classifier:
according toCalculating to obtain characteristic parameter x k Estimated conditional probability G corresponding to the ith soil type i (x k ) According to->Calculating to obtain probability that the soil at the point to be identified belongs to the ith soil type, and max { GP } i The corresponding i is the soil type of the point to be identified, wherein p i The prior probability of the ith soil type is determined offline according to training set data; mu (mu) i Sum sigma i Respectively is the characteristic parameter x corresponding to the ith soil type in the training set k Is determined offline according to the mean value and standard deviation of the training set data; k=1, 2,..4, x 1 、x 2 、x 3 、x 4 Respectively B q 、Q t 、F r 、ΔQ t The value of the point to be identified; i represents a soil type number, i=1, 2,..8.
Drawings
FIG. 1 is CPT-02-01 well q c 、f s 、u 2 A depth variation graph is penetrated along with penetration;
FIG. 2 is a CPT-02-01 well B q 、Q t 、F r A depth variation graph is penetrated along with penetration;
FIG. 3 is a graph of normalized cone end resistance deviation of CPT-02-01 well as a function of penetration depth;
FIG. 4 is a graph of CPT-02-01 well characteristic parameters prior to median filtering;
FIG. 5 is a graph of CPT-02-01 well characteristic parameters after median filtering;
FIG. 6 shows prior probabilities of soil types, means and standard deviations of feature parameters in a training dataset;
FIG. 7 is a graph comparing the result of automatic identification of CPT-02-01 well soil quality with the result of manual classification;
Detailed Description
According to basic data obtained by a static cone penetration test, the pore pressure parameter ratio, the normalized cone end resistance, the normalized friction resistance ratio and the normalized cone end resistance deviation are calculated, and the naive Bayes method is adopted to realize automatic identification of the type of the subsoil.
The method comprises the following specific steps:
(1) CPT data acquisition:
the cone end resistance q is obtained through a static cone penetration test c Friction force f of side wall s Pore water pressure u 2 (unit: MPa), probe penetration depth d (unit: m);
in this embodiment, taking CPT-02-01 well in a sea area of China as an example, q is collected c 、f s 、u 2 The variation with penetration depth d is shown in FIG. 1.
(2) Calculating attribute characteristics:
according toCalculating to obtain pore pressure parameter ratio B q According to->Calculating to obtain normalized cone end resistance Q t According to->Calculating to obtain normalized friction-resistance ratio F r Wherein Δu=u 2 -γ w X d is the excess pore water pressure, gamma w Is the volume weight of the seawater or the slurry; q t ＝q c +(1+α)u 2 Alpha is the specification of the cone head for correcting the cone end resistance; sigma'. vo Let γ '×d be effective overburden pressure, γ' be overburden soil layerAverage effective volume weight of (gamma) w Manually setting gamma' according to the regional characteristics, wherein alpha is CPT equipment parameters and is determined according to CPT equipment;
in this embodiment, α=0.75, γ w ＝10.05KN/m 3 ，γ'＝0.01MN/m 3 Calculation of B for CPT-02-01 well q 、Q t 、F r Attribute features, B q 、Q t 、F r The variation with penetration depth is shown in figure 2.
(3) Calculate Q t Wave characteristics:
for Q t Average filtering is carried out to obtain normalized cone end resistance Q after filtering t ' according to DeltaQ t ＝|Q t -Q t ' I is calculated to obtain normalized cone end resistance deviation delta Q t If DeltaQ t_j < 0.5, deltaQ t_j ＝ΔQ t_j-1 Wherein DeltaQ t_j-1 And DeltaQ t_j Normalized cone end resistance deviation at penetration depths j-1 and j, j=2, 3, … n, n being the maximum penetration depth;
in this embodiment, the window width of the mean filter is 30, and the normalized cone end resistance deviation DeltaQ of CPT-02-01 well is calculated t ，ΔQ t The variation with penetration depth is shown in figure 3.
(4) Characteristic parameter filtering:
pair B q 、Q t 、F r 、ΔQ t Sliding median filtering, wherein the width of a filtering window is w 1 ，w 1 Is odd;
in the present embodiment, w 1 =31, characteristic parameters B for CPT-02-01 wells respectively q 、Q t 、F r 、ΔQ t And sliding median filtering is carried out, and the variation curves of characteristic parameters before and after filtering along with penetration depth are shown in fig. 4 and 5.
(5) Identifying the type of the seafloor soil by using a Bayes classifier:
according toCalculating to obtain characteristic parameter x k Corresponding to the ith kindEstimated conditional probability G of soil type i (x k ) According to->Calculating to obtain probability that the soil at the point to be identified belongs to the ith soil type, and max { GP } i The corresponding i is the soil type of the point to be identified, wherein p i The prior probability of the ith soil type is determined offline according to training set data; mu (mu) i Sum sigma i Respectively is the characteristic parameter x corresponding to the ith soil type in the training set k Is determined offline according to the mean value and standard deviation of the training set data; k=1, 2,..4, x 1 、x 2 、x 3 、x 4 Respectively B q 、Q t 、F r 、ΔQ t The value of the point to be identified; i represents a soil type number, i=1, 2,..8.
In this embodiment, the total number of collected data n= 34090 in a sea area of our country. The training data set is a matrix with the scale of 4102×5, and the first 4 columns of the matrix are characteristic parameters B q 、Q t 、F r 、ΔQ t Randomly extracting all data acquired from CPT equipment, wherein the 5 th column of the matrix is the soil type corresponding to the first 4 columns of data, and analyzing in a laboratory. The data amount of the ith soil type in the training data set isN i I=1, 2,..8, wherein 1 is clay, 2 is silt, 3 is fine sand, 4 is silty clay, 5 is silt, 6 is sandy silt, 7 is sandy clay, 8 is clay sand, and i.e. the data amount of the i-th soil type in the total data volume collected.
Determining mu of the ith soil type from the training dataset i Sum sigma i According toThe prior probability of the ith soil type is calculated, and specific numerical values are shown in fig. 6.
The result of identifying the type of the seabed soil of the CPT-02-01 well is shown in figure 7, the time is 0.104 seconds, and the identification accuracy is 90.49%.
Claims (1)
1. The automatic identification method of the type of the seabed soil based on the naive Bayes algorithm is characterized by comprising the following steps of:
(1) CPT data acquisition: the cone end resistance q is obtained through a static cone penetration test c Friction force f of side wall s Pore water pressure u 2 (unit: MPa), probe penetration depth d (unit: m);
(2) Calculating attribute characteristics: according toCalculating to obtain pore pressure parameter ratio B q According to->Calculating to obtain normalized cone end resistance Q t According to->Calculating to obtain normalized friction-resistance ratio F r Wherein Δu=u2- γ w X d is the excess pore water pressure, gamma w Is the volume weight of the seawater or the slurry; q t ＝q c +(1+α)u 2 Alpha is the specification of the cone head for correcting the cone end resistance; sigma'. vo Let γ '×d be the effective overburden pressure, γ' be the average effective volume weight of overburden, γ w Manually setting gamma' according to the regional characteristics, wherein alpha is CPT equipment parameters and is determined according to CPT equipment;
(3) Calculate Q t Wave characteristics: for Q t Average filtering is carried out to obtain normalized cone end resistance Q after filtering t ' according to DeltaQ t ＝|Q t -Q t ' I is calculated to obtain normalized cone end resistance deviation delta Q t If DeltaQ t_j < 0.5, deltaQ t_j ＝ΔQ t_j-1 Wherein DeltaQ t_j-1 And DeltaQ t_j Normalized cone end resistance deviation at penetration depths j-1 and j, j=2, 3, … n, n being the maximum penetration depth;
(4) Characteristic parameter filtering: pair B q 、Q t 、F r 、ΔQ t Sliding median filtering, wherein the width of a filtering window is w 1 ，w 1 Is odd;
(5) Identifying the type of the seafloor soil by using a Bayes classifier: according toCalculating to obtain characteristic parameter x k Estimated conditional probability G corresponding to the ith soil type i (x k ) According to->Calculating to obtain probability that the soil at the point to be identified belongs to the ith soil type, and max { GP } i The corresponding i is the soil type of the point to be identified, wherein p i The prior probability of the ith soil type is determined offline according to training set data; mu (mu) i Sum sigma i Respectively is the characteristic parameter x corresponding to the ith soil type in the training set k Is determined offline according to the mean value and standard deviation of the training set data; k=1, 2,..4, x 1 、x 2 、x 3 、x 4 Respectively B q 、Q t 、F r 、ΔQ t The value of the point to be identified; i represents a soil type number, i=1, 2,..8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310452265.4A CN116452891A (en) | 2023-04-25 | 2023-04-25 | Automatic identification method for type of seafloor soil based on naive Bayes algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310452265.4A CN116452891A (en) | 2023-04-25 | 2023-04-25 | Automatic identification method for type of seafloor soil based on naive Bayes algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116452891A true CN116452891A (en) | 2023-07-18 |
Family
ID=87130036
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310452265.4A Pending CN116452891A (en) | 2023-04-25 | 2023-04-25 | Automatic identification method for type of seafloor soil based on naive Bayes algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116452891A (en) |
-
2023
- 2023-04-25 CN CN202310452265.4A patent/CN116452891A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109214084B (en) | Pore pressure static sounding seabed soil layer dividing method and system | |
| CN102011583B (en) | Method for identifying reservoir by combining electric imaging with reef geologic model | |
| CN111705774B (en) | Saturated sandy soil liquefaction detection method based on standard penetration test | |
| CN110378921B (en) | Intelligent identification method for substrate layer boundary of channel based on floating mud rheological property and gray level co-occurrence matrix | |
| CN113011094B (en) | Soil pressure balance shield tunneling machine muck improvement method based on mixed GBDT and random forest algorithm | |
| Bargawa et al. | Iron ore resource modeling and estimation using geostatistics | |
| CN116305850B (en) | Stratum thermal conductivity prediction method based on random forest model | |
| CN111622201B (en) | Soil layer equivalent shear wave velocity direct calculation method based on surface wave frequency dispersion characteristics | |
| CN115049173A (en) | Deep learning and Eaton method coupling driving stratum pore pressure prediction method | |
| Abubakar et al. | Deep learning applications for wind farms site characterization and monitoring | |
| CN112836393B (en) | Method for analyzing reservoir heterogeneity based on multi-scale entropy | |
| Rajabi et al. | Sensitivity analysis of effective factors for estimating formation pore pressure using a new method: the LSSVM-PSO algorithm | |
| CN110118994B (en) | Continental facies hydrocarbon source rock quantitative prediction method based on seismic inversion and machine learning | |
| CN116452891A (en) | Automatic identification method for type of seafloor soil based on naive Bayes algorithm | |
| CN117093922A (en) | Improved SVM-based complex fluid identification method for unbalanced sample oil reservoir | |
| CN110552693A (en) | layer interface identification method of induction logging curve based on deep neural network | |
| CN109669211B (en) | Method and device for predicting seabed lithologic thin sand body reservoir, storage medium and terminal | |
| CN116482757A (en) | Stratum structure identification analysis method based on shallow well-logging earthquake combination | |
| CN104834934A (en) | Nuclear body capturing method used for identifying reservoir fluid | |
| CN112906465B (en) | Coal measure stratum acoustic curve reconstruction method and system based on stratum factors | |
| CN115434696A (en) | Lithology identification database construction method and lithology identification method | |
| Cho et al. | Development of locally specified soil stratification method with CPT data based on machine learning techniques | |
| CN116881780A (en) | Automatic identification method for type of seafloor soil based on BP neural network algorithm | |
| Ching et al. | Identification of soil degradation during earthquake excitations by Bayesian inference | |
| CN116451141A (en) | Seafloor soil type identification method based on support vector machine |
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 |