CN113946691A

CN113946691A - Foundation soil layering system and method

Info

Publication number: CN113946691A
Application number: CN202111267039.6A
Authority: CN
Inventors: 周泽兵; 于娜; 卢奕; 周玉明; 黄恩兴; 刘月辉; 熊鑫; 马乐民; 沈迎志; 邓向辉; 李平虎; 李宝仁; 翟璐
Original assignee: Stargis Tianjin Technology Development Co ltd; Tianjin Survey And Design Institute Group Co Ltd
Current assignee: Stargis Tianjin Technology Development Co ltd; Tianjin Survey And Design Institute Group Co Ltd
Priority date: 2021-10-28
Filing date: 2021-10-28
Publication date: 2022-01-18

Abstract

The invention discloses a system and a method for layering foundation soil, which belong to the technical field of engineering geological exploration and comprise the following steps: establishing a regional professional resource library; the method specifically comprises the following steps: establishing a professional main database; importing project drilling field collected data and test data; establishing a project reference sequence; recommending and confirming the layering in sequence; verifying the rationality of the sequence; and (4) outputting results and warehousing data. The invention can efficiently and accurately layer the foundation soil.

Description

Foundation soil layering system and method

Technical Field

The invention belongs to the technical field of engineering geological investigation, and particularly relates to a method and a system for layering foundation soil.

Background

As is well known, in the geological exploration process, the stratification of the foundation soil is a very important link, and at present, the traditional method for stratifying the foundation soil is as follows: 1. arranging the original data and the test data of each drilling hole of the project to form an initial single-hole layering; 2. comparing the layering of each drill hole item by item, and summarizing the overall layering in the item range; 3. and traversing each drill hole, and converting each single-hole layer in the drill hole into an integral layer according to the corresponding relation. The method mainly relies on the technical staff to carry out statistics, analysis and classification on the original survey data.

The traditional technical scheme has the following defects: firstly, the efficiency is not high, and the workload is directly related to the project scale; second, the efficiency and accuracy of the layering depends on the experience and ability of the technician, namely: the relationship between the layering result and the technical personnel is large, the layering results of different personnel are possibly different, and the direct result is questionable of the objectivity of the layering result; and thirdly, the industrial technical service capability is closely related to a host, and the expert experience is not favorable for inheritance and improvement.

Disclosure of Invention

Technical purpose

The invention provides a method for layering foundation soil and an identification system; the method can efficiently and accurately stratify the foundation soil.

Technical scheme

The first purpose of the invention is to provide a method for layering foundation soil, which comprises the following steps:

building a regional professional resource library: respectively establishing a professional main database and a special database, and establishing an industry application model knowledge base through a knowledge graph technology;

importing project drilling field collected data and test data: the drilling data acquired by field operation is digitally processed by an informatization means and then is associated with the geotechnical test data to generate test data, and the test data is imported and the general profiles of the natural geography, the topographic features, the bedrock geology, the quaternary geology, the hydrogeology and the engineering geology of the region and the adjacent main cities are analyzed;

establishing a project reference sequence: carrying out external expansion processing on the designated project range, judging whether the designated project range belongs to standard stratum data or not, extracting the drilling data in the external expansion range from the standard stratum data through an engineering survey result library, carrying out statistical summary processing on the drilling data to obtain a fine reference sequence in the external expansion range, and giving a processing requirement label when the designated project range belongs to the special geological condition distribution range;

sequentially recommending and confirming the layering: layering the soil sample in the drill hole data subjected to the informatization processing in a designated area according to field records, soil sample characteristics and geotechnical test data, and determining a main sequence and a secondary sequence;

and (3) verifying the rationality of the sequence: carrying out sequence rationalization verification on the drill hole data after the informatization processing;

and (4) outcome output and data storage: and (4) carrying out achievement on the obtained achievements according to requirements, and storing the related achievements into a formulated database.

The method for layering the foundation soil comprises the following specific steps of building a regional professional resource library:

professional master database: extracting required key information from the existing human resource system, project management system and financial management system according to the set requirement information, wherein the key information comprises: management personnel information, operating personnel information, project participant information, project bibliographic information and space information of projects;

constructing a special database: establishing a basic geological map library, an engineering investigation result library and a standard stratum database according to requirements;

establishing an industry application knowledge graph: and forming a knowledge graph by using professional terms, standard specifications, operation specifications, proprietary technologies and experience models in the engineering exploration industry.

The method for layering the foundation soil comprises the following specific flow steps of importing project drilling field collected data and test data:

importing project drilling field data and processing: the drilling data collected by field operation is processed digitally and imported through an informatization means;

project geotechnical test data: correlating the drilling data after the informatization processing with the geotechnical test data, and importing the results of various test data:

analyzing the geological profile: and analyzing the general profiles of the natural geography, the topographic features, the bedrock geology, the quaternary geology, the hydrogeology and the engineering geology of the region and the adjacent main cities around the region.

The method for layering the foundation soil comprises the following specific flow steps of establishing a project reference sequence:

establishing a basic reference sequence of project ranges: carrying out external expansion processing on the designated project range, and extracting a standard stratum in the external expansion range from a standard stratum database to obtain a reference sequence of the project; extracting the drilling data in the external expansion range from the engineering survey result library, and performing statistical summary processing to obtain a fine reference sequence in the external expansion range;

establishing project-wide special stratum references: and (4) judging whether the geological condition belongs to a special geological condition distribution range or not by performing spatial superposition with the basic geological map in the basic geological map library, and giving a processing requirement label when the geological condition belongs to the special geological condition distribution range.

The method for layering the foundation soil comprises the following specific flow steps of sequentially recommending and confirming layering:

the method comprises the following steps of sequentially processing segmented soil layers from the ground downwards according to the buried depth range of the drilling data soil sample after informatization:

the method comprises the following steps of sequentially recording various attributes of parameter ranges, morphological characteristics, soil layer characteristics, distribution ranges and physical characteristics related to soil layers as { At1, At2, At 3.. Atn };

screening candidate soil layers: according to the buried depth of the top and the bottom of the drilling data soil sample after the informatization processing, comparing with the buried depth range of each layer of the project reference sequence, judging that the two groups of data have intersection, listing the two groups of data in a candidate soil layer, and judging that the two groups of data have no intersection, performing exclusion processing to obtain all possible layers of the segmented soil sample, and recording the layers as { TC1, TC2 }; determining a main sequence and a secondary sequence;

determining a main sequence: dividing a main sequence of layers according to the deposition era, the cause type and the deposition environment attribute respectively, wherein the specific judgment rule is as follows:

when the attribute of the standard sequence soil layer is in a numerical range type, matching the attribute value of the current subsection soil sample with the corresponding attribute of the candidate soil layer, recording the matching degree as 1 in the range, and otherwise, recording the matching degree as 0;

when the attribute of the standard sequence soil layer is a text type, splitting the attribute value of the segmented soil sample and the corresponding attribute text of the candidate soil layer by using Chinese word segmentation, and calculating the proportion of the same word existing in the two split words as the matching degree;

when the attribute of the standard sequence soil layer is a vectorization space distribution type, comparing the current segmented soil sample space position P (x, y) with the corresponding attribute space distribution of the candidate soil layer, recording the matching degree as 1 within the range, recording the matching degree as 0 outside the range by 1km, and converting the matching degree according to the distance within the range outside the distribution range by 1 km;

multiplying the matching degrees of the attributes of the various segmented soil samples to obtain the overall layered matching degree of the segmented soil samples, judging the segmented soil samples to be a standard sequence when the overall layered matching degree of the segmented soil samples is judged to be not less than 85%, comparing the segmented soil samples by using different expert models when the overall layered matching degree of the segmented soil samples is judged to be less than 85%, and judging soil layers to recommend optimal soil layers in sequence by combining similar drilling experiences on the periphery; the recommended layers can be confirmed and adjusted by a host;

determining a sequence of the layers: dividing the sequence of the layers according to the properties such as the lithology, the color and the state, and the like, wherein the specific judgment rule is as follows:

multiplying the matching degrees of the attributes of the various segmented soil samples to obtain the overall layered matching degree of the segmented soil samples, judging the segmented soil samples to be a standard sequence when the overall layered matching degree of the segmented soil samples is judged to be not less than 85%, comparing the segmented soil samples by using different expert models when the overall layered matching degree of the segmented soil samples is judged to be less than 85%, and judging soil layers to recommend optimal soil layers in sequence by combining similar drilling experiences on the periphery; the recommended hierarchies may both be confirmed and adjusted by the moderator.

The method for layering the foundation soil comprises the following specific flow steps of verifying the rationality of the sequence:

checking a horizontal section diagram: forming a representative profile map by layering each drilling hole according to site characteristics, and checking the rationality of the division according to the layer sequence division characteristics;

checking the three-dimensional stratum: importing the layered information-processed drilling data into a three-dimensional stratum, checking the distribution condition of the layered data of the project and the surrounding three-dimensional stratum, and verifying the rationality;

checking a special stratum: manually checking and checking the rationality of local special strata according to the drilling data after special informatization;

and (3) overall checking and confirming: and in the process of auditing and approval, the auditing and approving person reviews the layering process, adjusts the layering when the layering is judged to be unreasonable, and records the adjustment information.

The method for layering the foundation soil comprises the following specific process steps of construction of the special database:

base geological map library: digitally creating a library of maps of geological features in the designated area, wherein the library of basic geological maps comprises: the method comprises a silt distribution diagram of a certain soil layer of a fifth land phase layer (Q3cal), a pile foundation supporting layer top and bottom plate burial depth diagram, a local shallow foundation soil bearing characteristic value distribution diagram, a local liquefied soil layer distribution diagram, a filled trench pit and ancient river distribution diagram, an earthquake fracture zone, an earthquake intensity and basic acceleration zoning diagram, an earthquake disaster zoning diagram, a geological disaster easily-distributing distribution diagram, a geological disaster development degree zoning diagram, a ground accumulated settlement amount diagram and a ground settlement rate diagram;

engineering investigation result library: digitally establishing a library for the completed project engineering investigation result in the designated area according to the project; wherein the digital result library: the method comprises the following steps of (1) reporting, a report drawing (horizontal section drawing), attached form data and an engineering investigation drilling library; on the basis of drilling, gathering drilling data of all projects to build an engineering exploration drilling library and carrying out stratum standardization and space benchmark unification treatment on the drilling data; the engineering investigation drilling library comprises: and gathering and establishing a database of the drilling data of each project, wherein the database comprises drilling positions, test information and hierarchical information. Drilling and reserving history;

standard formation database: judging whether the designated area has issued the ground-based soil sequence dividing procedure or not, and directly quoting when the designated area has issued the ground-based soil sequence dividing procedure; when the foundation soil sequence division rule is judged not to be issued, formulating a standard stratum of the region according to the characteristic drilling data in the region; three-dimensional stratum library: extracting characteristic drill holes from the standardized and layered drill holes according to a 1km grid, forming an urban three-dimensional stratum according to the characteristic drill holes, and constructing an area-level subdivision stratum according to 0.3km on the basis;

the concrete flow steps of the engineering investigation result library comprise:

formation standardization: and adopting different stratum division standards for different exploration times, different exploration units, different exploration sites, different exploration personnel and different exploration purposes, and carrying out stratum standardization treatment according to unified standard stratum surface data.

The space standard is unified: unifying the data of different periods to the current reference system.

A foundation soil layering method is characterized in that the specific process steps of establishing an industry application knowledge graph comprise:

extracting the structural information in a human resource system, a project management system and an engineering investigation result library to obtain various entities, relations and attribute information;

meanwhile, integrating comprehensive knowledge rich in open resources;

associating existing project drilling data and standard layering data with an examining and correcting expert, sequentially carrying out normalization matching on layering results confirmed by the expert in a region and soil characteristic index parameters corresponding to regional foundation soil sequence division standards, taking 60% -70% of data as a training set and the rest 30% -40% as a test set, and obtaining an expert layering recommendation model through machine learning; wherein: c is a hierarchical set, and D is a set of training tuples and their associated class labels; using one n-dimensional attribute vector X ═ X for each tuple_1,x₂,...,x_nRepresents; assume that there are m classes C_1,C₂,...C_m。Given a tuple, the classification method makes the prediction attribute vector X belong to the class with the highest posterior probability; the class conditions are considered to be independent according to industry experience;

x_kis the value of k term, C, of the attribute vector X_iThe ith sequence class of X;

according to attribute A_kWhether classified or continuous values, P (X | C) is calculated separately_i) There are two cases:

attribute A_kIs a categorical attribute, then P (x)_k|C_i) Is attribute A in D_kThe value of (c): p (x)_k|C_i) Is equal to x_kC in_iNumber of tuples of class divided by C in D_iNumber of cell groups | C_i,D|；

Attribute A_kThe method is characterized in that the method is a continuous value attribute and is in normal distribution, the mean value and the variance are calculated through samples to obtain a density function of the normal distribution, and the value is substituted to calculate the value of the density function at a certain point;

prediction of X classification, for each sequence class C_iCalculating P (C)_i|X)P(C_i) The classification method predicts the class of the input tuple as C_i，If and only if: p (X | C)_i)P(C_i)>P(X|C_j)P(C_j) When j is not less than 1 and not more than m and j is not equal to i, the predicted class label is P (X | C)_i)P(C_i) Largest class C_i(ii) a Namely layering.

It is a second object of the present invention to provide a system for layering foundation soil, comprising:

a resource library module: establishing a regional professional resource library; the method specifically comprises the following steps:

establishing a professional main database; the professional master database comprises: an enterprise personnel database, a project first party database and a project information database;

establishing a special database; the specialty database includes: a basic geological map library, an engineering investigation report library, an engineering investigation drilling library, a standard stratum database and a three-dimensional stratum library;

establishing an industry application knowledge graph; the method specifically comprises the following steps: associating existing project drilling data, standard layering data and an examining and correcting expert, sequentially carrying out normalized matching on layering results confirmed by the expert in the region and soil layer characteristic index parameters corresponding to regional foundation soil sequence division standards, taking 60% -70% of data as a training set and the rest 30% -40% as a test set, and obtaining an expert layering recommendation model through machine learning;

the data import module: importing project drilling field collected data and test data;

a reference sequence module: establishing a project reference sequence;

a recommendation layering module: recommending and confirming the layering in sequence; the method specifically comprises the following steps: determining a main sequence and a secondary sequence;

a verification module: verifying the rationality of the sequence;

the checking module: checking and confirming by an expert;

a warehousing module: and (4) outputting results and warehousing data.

The foundation soil layering system is characterized in that the expert layering recommendation model is as follows:

defining: c is a hierarchical set, D is a tuple and associated class label set; using one n-dimensional attribute vector X ═ X for each tuple_1,x₂,...,x_nDenotes, assuming there are m classes C_1,C₂,...C_m(ii) a Given a tuple, the prediction attribute vector X belongs to the class with the highest posterior probability through a classification method; the class conditions are considered to be independent according to industry experience;

(a) if attribute A_kIs a categorical attribute, then P (x)_k|C_i) Is attribute A in D_kValue of (a), P (x)_k|C_i) Is equal to x_kC in_iNumber of tuples of class divided by C in D_iNumber of cell groups | C_i,D|；

(b) If attribute A_kIs a continuous value attribute and is normally distributed, and the mean value and the variance are calculated through the sample to obtainSubstituting the value into a normally distributed density function to calculate the value of the density function at a certain point;

The invention has the advantages and positive effects that:

aiming at a specific target area, the engineering geological change is less in a short period, so that the early engineering survey data of the target area has stronger reference for later exploration; namely, an expert database can be formed by constructing data resources of a target area;

the sequence division process can be reproduced, and based on expert hierarchical checking, the process data is saved, so that later-stage optimization of an expert model is facilitated; and the layering accuracy is improved.

The invention utilizes the existing layering data around the project, the regional foundation soil division sequence is standard, and the project test data is combined to carry out automatic layering, thereby realizing the automatic division work of more than 80 percent.

According to the method and the system, different expert models are built, comprehensive comparison is achieved, contents which tend to be consistent, conventional contents are rapidly determined, difficult and special processing contents are highlighted, and the method and the system are beneficial to rapid growth of new technicians.

Drawings

FIG. 1 is a flow chart of a preferred embodiment of the present invention;

FIG. 2 is a system block diagram of a preferred embodiment of the present invention;

FIG. 3 is a block diagram of the structure of a specialized database in a preferred embodiment of the present invention;

Detailed Description

In order to further understand the contents, features and effects of the present invention, the following embodiments are illustrated and described in detail with reference to the accompanying drawings.

Specific example 1:

as shown in fig. 1: a method for layering foundation soil comprises the following steps:

In the specific implementation case, a regional professional resource library is built: respectively establishing a professional main database and a special database, and establishing an industry application model knowledge base through a knowledge graph technology;

1. professional master database: extracting required key information from the existing human resource system, project management system and financial management system according to the set requirement information, wherein the key information comprises: management personnel information, operating personnel information, project participant information, project bibliographic information and space information of projects;

2. constructing a special database: establishing a basic geological map library, an engineering investigation result library and a standard stratum database according to requirements;

2.1 base geological map library: carrying out digital library building on the map of geological features in the designated area, wherein a basic geological map library comprises the following steps: the method comprises a silt distribution diagram of a certain soil layer of a Vth land phase layer (Q3cal), a pile foundation supporting layer top and bottom plate burial depth diagram, a city shallow foundation soil bearing capacity characteristic value distribution diagram, a city liquefied soil layer distribution diagram, a filled trench pit and ancient river distribution diagram, an earthquake fracture zone, an earthquake intensity and basic acceleration zoning diagram, an earthquake disaster zoning diagram, a geological disaster easily-distributing layout diagram, a geological disaster development degree zoning diagram, a ground accumulated settlement amount diagram and a ground settlement rate diagram.

2.2 engineering investigation result library: digitally establishing a library for the completed project engineering investigation result in the designated area according to the project; wherein the digital result library: the method comprises the following steps of (1) reporting, a report drawing (horizontal section drawing), attached form data and an engineering investigation drilling library; on the basis of drilling, gathering drilling data of all projects to build an engineering exploration drilling library and carrying out stratum standardization and space benchmark unification treatment on the drilling data; the engineering investigation drilling library comprises: and gathering and establishing a database of the drilling data of each project, wherein the database comprises drilling positions, test information and hierarchical information. The borehole retains a history.

2.2.1 formation normalization: and adopting different stratum division standards for different exploration times, different exploration units, different exploration sites, different exploration personnel and different exploration purposes, and carrying out stratum standardization treatment according to unified standard stratum surface data.

2.2.2 spatial reference unification: data in different periods are unified to an existing reference system, wherein the existing reference system comprises a plane datum and an elevation datum, the existing reference system is generally a 2000 national geodetic coordinate system, and the elevation datum is unified to a 1985 national elevation datum.

2.3 Standard formation database: judging whether the designated area has issued the ground-based soil sequence dividing procedure or not, and directly quoting when the designated area has issued the ground-based soil sequence dividing procedure; when the foundation soil sequence division rule is judged not to be issued, formulating a standard stratum of the region according to the characteristic drilling data in the region; the standard stratum achievement divides foundation soil in a certain depth range required by a conventional engineering project in a certain area into standard sequence, and the standard sequence is generally divided according to the standard sequence name, characteristics, a buried depth range, soil layer characteristics, a distribution range (specifically graphical description of characters) and a characteristic stratum distribution diagram; physical characteristics, etc.

2.4 three-dimensional stratigraphic library: and extracting characteristic drill holes from the standardized and layered drill holes according to a 1km grid, forming an urban three-dimensional stratum according to the characteristic drill holes, and constructing an area-level subdivision stratum according to 0.3km on the basis.

3. Establishing an industry application knowledge graph: and forming a knowledge graph by using professional terms, standard specifications, operation specifications, proprietary technologies and experience models in the engineering exploration industry.

3.1 extracting the structural information in the human resource system, the project management system and the engineering investigation result library to obtain various entities, relations and attribute information, determining enterprise personnel entities through the human resource system, extracting professional term entities through an industry professional term dictionary and specification, and integrating standard and standardized resources; the project and the entity of the first party unit can be extracted through the project management system, and the relationship between the project and the person is established through a project host; the basic knowledge base can be obtained based on the existing relational database information.

3.2 integrating comprehensive knowledge rich in open resources;

3.3, associating the existing project drilling data and standard layering data with an examining and correcting expert, sequentially carrying out normalized matching on layering results confirmed by the expert in the region and soil layer characteristic index parameters corresponding to regional foundation soil sequence division standards, taking 60% -70% of data as a training set and the rest 30% -40% as a test set, and obtaining an expert layering recommendation model through machine learning; wherein: c is a hierarchical set, and D is a set of training tuples and their associated class labels; each tuple is represented by an n-dimensional attribute vector X ═ { X1, X2. Assume that there are m classes C1, C2.. Cm. Given a tuple, the classification method makes the prediction attribute vector X belong to the class with the highest posterior probability; the class conditions are considered to be independent according to industry experience;

xk is a k item value of the attribute vector X, and Ci is the ith sequence category of X;

p (X | Ci) is calculated separately depending on whether the attribute Ak is classified or continuous, in two cases:

(a) property Ak is a classification property, then P (xk | Ci) is the value of property Ak in D: p (xk | Ci) is equal to the number of tuples of class Ci in xk divided by the number of tuples | Ci, D | of Ci in D;

(b) the attribute Ak is a continuous value attribute and is in normal distribution, the mean value and the variance are calculated through a sample to obtain a density function of the normal distribution, and the value is substituted to calculate the value of the density function at a certain point;

predicting the classification of X, calculating P (Ci | X) P (Ci) for each rank class Ci, the classification predicting the class of input tuples as Ci if and only if: when P (X | Ci) P (Ci) > P (X | Cj) P (Cj), j is more than or equal to 1 and less than or equal to m, and j is not equal to i, the predicted class label is the class Ci which enables P (X | Ci) P (Ci) to be the largest; namely layering.

The specific operation process comprises the following steps:

determining attributes to form a training sample set: determining characteristic attributes such as a buried depth range (a bottom plate buried depth and a top boundary (plate) buried depth), stratum thickness, rock and soil description information (stratum saturation, stratum color, stratum state, stratum bedding and stratum containing objects), standard penetration test information and physical index information of natural state soil according to requirements, and forming a training sample set.

And generating a classifier, wherein the main work is to calculate the occurrence frequency of each class in a training sample and the conditional probability estimation of each characteristic attribute partition on each class, and record the result.

And (4) comparing and classifying: classifying items to be classified by using a classifier, calculating each class P (X | Ci) P (Ci), wherein the input of the class P (X | Ci) P (Ci) is the classifier and the items to be classified, the output is the mapping relation between the items to be classified and the classes, and the maximum item P (X | Ci) P (Ci) is used as the class to which C belongs and the hierarchy.

The database needs to be built before the new project is implemented.

(II) importing project drilling field collected data and test data: the drilling data acquired by field operation is digitally processed by an informatization means and then is associated with the geotechnical test data to generate test data, and the test data is imported and the general profiles of the natural geography, the topographic features, the bedrock geology, the quaternary geology, the hydrogeology and the engineering geology of the region and the adjacent main cities are analyzed;

1. importing project drilling field data and processing: the drilling data collected by field operation is processed digitally and imported by informatization means,

the drilling field data comprises the following information: drilling hole number, variable layer depth, repeated footage, rock-soil field naming, rock-soil description (color, humidity, state, bedding, iron, clam shell, organic matter and the like), standard penetration test (pre-striking hammering number and actual hammering number) and the like.

2. Project geotechnical test data: correlating the drilling data after the informatization processing with the geotechnical test data, and importing the results of various test data:

wherein: the geotechnical test data comprises a drilling number, a soil sampling depth, physical indexes (water content, density, soil particle proportion, pore ratio and saturation) of natural soil, a liquid limit, shrinkage consolidation, a direct shear test and a soil classification name.

3. Analyzing the geological profile: and analyzing the general profiles of the natural geography, the topographic features, the bedrock geology, the quaternary geology, the hydrogeology and the engineering geology of the region and the adjacent main cities around the region.

And (III) establishing a project reference sequence: carrying out external expansion processing on the designated project range, judging whether the designated project range belongs to standard stratum data or not, extracting the drilling data in the external expansion range from the standard stratum data through an engineering survey result library, carrying out statistical summary processing on the drilling data to obtain a fine reference sequence in the external expansion range, and giving a processing requirement label when the designated project range belongs to the special geological condition distribution range;

1. establishing a basic reference sequence of project ranges: carrying out external expansion processing on the designated project range, and extracting a standard stratum in the external expansion range from a standard stratum database to obtain a reference sequence of the project; extracting the drilling data in the external expansion range from the engineering survey result library, and performing statistical summary processing to obtain a fine reference sequence in the external expansion range;

the urban area standard sequence is a statistical division result of a large-range characteristic sequence, can be referred to a project, is relatively coarse with respect to a specific project, and needs to dynamically establish a small-range reference sequence of the project area through the existing project accumulated data resources. Generally, the project scope is expanded by 3km, and the drilling data in the scope is statistically gathered, for example, the burial depth scope of each standard soil layer can be obtained (the scope is generally smaller than the burial depth scope corresponding to the standard sequence of the urban area, and is more targeted).

2. Establishing project-wide special stratum references: performing spatial superposition with a basic geological map in a basic geological map library, judging whether the geological map belongs to a special geological condition distribution range, and giving a processing requirement label if the geological map belongs to the special geological condition distribution range; counting special geological conditions in an expanded area outside the project range, wherein local soil layer distribution can be used for reference by the project;

and (IV) recommending and confirming the layering in sequence: layering the soil sample in the drill hole data subjected to the informatization processing in a designated area according to field records, soil sample characteristics and geotechnical test data, and determining a main sequence and a secondary sequence;

the method specifically comprises the following steps:

screening candidate soil layers: and comparing the burial depths of the top and the bottom of the soil sample of the drilling data after the informatization processing with the burial depth range of each layer of the project reference sequence, listing the two groups of data into a candidate soil layer if the two groups of data have an intersection, and excluding the two groups of data if the two groups of data do not have the intersection, thereby obtaining all possible layers of the segmented soil sample as { TC1, TC2 }.

Determining a main sequence: the main sequence of layers is divided according to attributes such as the deposition era, the cause type, the deposition environment and the like, and the specific judgment rule is as follows:

1. Determining a main sequence: dividing a main sequence according to the deposition era, the cause type and the deposition environment, wherein the main sequence has a clear corresponding relation with a standard stratum main sequence of a peripheral city;

2. determining a sequence of the layers: dividing the hierarchical sequences according to lithology, color and state, and having a clear corresponding relation with the standard hierarchical sequences of the peripheral region;

1) lithology: the soil is divided into sandy soil, silt soil, clay, silty clay, artificial filling soil and the like.

2) Color: mainly distinguishes oxidation environment series (grey yellow, brown yellow, etc.) and reduction environment series (grey, grey black, etc.).

3) The state is as follows: mainly distinguishes soft and hard degrees (flow molding, soft molding, plasticity, hard molding and the like) and dense degrees (slightly dense, medium dense and dense).

(V) verifying the rationality of the sequence: the drilling data after the informatization processing is verified to be reasonable in sequence,

1. checking a horizontal section diagram: forming a representative profile map by layering each drilling hole according to site characteristics, and checking the rationality of the division according to the layer sequence division characteristics;

2. checking the three-dimensional stratum: and importing the layered information-processed drilling data into the three-dimensional stratum, checking the distribution condition of the item layered data and the surrounding three-dimensional stratum, and verifying the rationality.

3. Checking a special stratum: manually checking and checking the rationality of local special strata according to the drilling data after special informatization;

4. and (3) overall checking and confirming: and in the process of auditing and approval, the auditing and approving person reviews the layering process, adjusts the layering when the layering is judged to be unreasonable, and records the adjustment information.

(VI) result output and data storage: the obtained achievements are subjected to achievement attribution according to requirements, and the related achievements are stored into a formulated database; wherein the achievement includes: original data, layered calculation analysis data, audit checking data and layered results.

As shown in fig. 2, a system for layering foundation soil includes:

a reference sequence module: establishing a project reference sequence;

a verification module: verifying the rationality of the sequence;

the checking module: checking and confirming by an expert;

a warehousing module: and (4) outputting results and warehousing data.

In a specific implementation case, the expert hierarchical recommendation model is as follows:

(b) If attribute A_kThe method is characterized in that the method is a continuous value attribute and is in normal distribution, the mean value and the variance are calculated through samples to obtain a density function of the normal distribution, and the value is substituted to calculate the value of the density function at a certain point;

Specific example 2:

referring to fig. 1 of the drawings, a drawing,

a method for layering foundation soil comprises the following steps:

step one, building a regional professional resource library; wherein:

1. professional master database: the method mainly comprises information of enterprise personnel, project first party, project and the like; matching with the existing human resource system, project management system and financial management system of enterprises.

2. Constructing a special database:

2.1 base geological map library: the digital library building method is characterized in that a map capable of reflecting regional geological features is built in a digital mode, and the digital library building method comprises a V-th land phase layer (Q3cal) soil silt distribution diagram of a certain soil layer, a pile foundation supporting layer top and bottom plate burial depth diagram, a city shallow foundation soil bearing capacity characteristic value distribution diagram, a city liquefied soil layer distribution diagram, a filled trench pit and ancient river distribution diagram, an earthquake fracture zone, an earthquake intensity and basic acceleration zone diagram, an earthquake disaster zone diagram, a geological disaster easily-occurring zone diagram, a geological disaster development degree zone diagram, a ground accumulated settlement amount diagram, a ground settlement rate diagram and the like.

2.1 engineering survey report library: the method comprises the steps of establishing a library according to projects by including data such as engineering survey reports, report drawings (horizontal cutting drawings), project attached tables and the like.

2.2 engineering investigation drilling library: and (4) gathering the drilling data of each project based on drilling to build a library, wherein the library comprises drilling positions, test information and layering information. The borehole retains a history. Formation standardization and spatial reference unification of drilling data are required.

Formation standardization: different stratum division standards are used for different exploration times, different exploration units, different exploration sites, different exploration personnel and different exploration purposes, and all data need to be subjected to stratum standardization according to a unified standard stratum table.

The space standard is unified: data at different periods need to be unified to an existing reference system, including a plane reference and an elevation reference, which is currently generally a 2000 national geodetic coordinate system, and the elevation reference is unified to a 1985 national elevation reference.

2.3 Standard formation database: standard strata of a large area of a city can be obtained or statistically formulated, such as a ground soil sequence division rule; the standard sequence achievement divides foundation soil in a certain depth range required by a conventional engineering project in a certain area into standard sequence, and the standard sequence is generally divided according to the standard sequence name, characteristics, a buried depth range, soil layer characteristics, a distribution range (specifically graphical description of characters) and a characteristic stratum distribution diagram of the area; physical characteristics, etc.

2.4 three-dimensional stratigraphic library: and extracting characteristic drill holes from the standardized and layered drill holes according to a 1km grid, forming an urban three-dimensional stratum according to the characteristic drill holes, and constructing an area-level subdivided stratum (integrally layered according to the city and subdivided into layers according to the area) according to 0.3 km.

3.1 extracting the structural information of the existing project management system, the human resource management system, the reconnaissance result library, the drilling library and the like to obtain various entities, relationships, attributes and the like, for example, determining enterprise personnel entities through the human resource system, extracting professional term entities through an industry professional term dictionary and specification, and integrating standard and standardized resources; the project and the entity of the first party unit can be extracted through the project management system, and the relationship between the project and the person is established through a project host; a basic knowledge base can be obtained based on the information of the existing relational database;

3.2 integrating rich comprehensive knowledge of open resources (such as encyclopedia and universal knowledge map) at the same time;

and 3.3, associating the existing project drilling data and standard layering data with an examining and correcting expert, sequentially carrying out normalized matching on the layering results confirmed by the expert in the region and soil characteristic index parameters corresponding to the regional foundation soil sequence division standard, taking 60-70% of the data as a training set and the rest 30-40% as a test set, and recommending the model by machine learning expert layering. Wherein: c is a hierarchical set, and D is a set of training tuples and their associated class labels; using an n-dimensional attribute for each tupleVector X ═ X_1,x₂,...,x_nRepresents it. Assume that there are m classes C_1,C₂,...C_m。Given a tuple, the classification method makes the prediction attribute vector X belong to the class with the highest posterior probability; the class conditions are considered to be independent according to industry experience;

(a) if attribute A_kIs a categorical attribute, then P (x)_k|C_i) Is attribute A in D_kThe value of (c): p (x)_k|C_i) Is equal to x_kC in_iNumber of tuples of class divided by C in D_iNumber of cell groups | C_i,D|；

The specific operation process comprises the following steps:

And (4) comparing and classifying: classifying the items to be classified by using a classifier, and calculating each class P (X | C)_i)P(C_i) The input is a classifier and an item to be classified, and the output is a mapping relation between the item to be classified and the class, and the mapping relation is represented by P (X | C)_i)P(C_i) The maximum item is used as the category and the hierarchy of C.

Importing project drilling field collected data and test data;

1. project drilling field record:

the field records can be sequentially recorded and sorted in a traditional mode; digital field records can also be acquired by an informatization means;

the field records comprise information such as drilling hole numbers, layer depth changing, recurrent footage, rock-soil field naming, rock-soil description (color, humidity, state, bedding, iron, clam shell, organic matter and the like), standard penetration tests (pre-hammering number, actual hammering number) and the like.

2. Project land test data:

associating geotechnical test data with the borehole; and importing each test data structure.

The geotechnical test data comprises information such as a drilling hole number, a soil sampling depth, physical indexes (such as water content, density, soil particle proportion, pore ratio and saturation) of natural soil, a liquid limit, shrinkage consolidation, a direct shear test and a soil classification name.

3. Analyzing the geological profile:

and analyzing the general profiles of the natural geography, the topographic features, the bedrock geology, the quaternary geology, the hydrogeology and the engineering geology of the region and the adjacent main cities around the region.

Step three, establishing a project reference sequence:

1. the urban area standard sequence is a statistical division result of a large-range characteristic sequence, can be referred to a project, is relatively coarse with respect to a specific project, and needs to dynamically establish a small-range reference sequence of the project area through the existing project accumulated data resources. Generally expanding the project range by 3km, and carrying out statistical summary on drilling data in the range, for example, obtaining the burial depth range of each standard soil layer (the range is generally smaller than the burial depth range corresponding to the standard sequence of the urban area, and is more targeted);

2. through superposition with various basic geological maps (digital distribution maps) and statistics of special geological conditions in the area outside the project range, local soil layer distribution can be used for reference of the project;

step four, recommending and confirming layering in sequence:

1. firstly, determining a main sequence: dividing the main sequence according to the deposition era, the cause type and the deposition environment, and having a clear corresponding relation with the standard stratum main sequence of the peripheral area;

1) lithology: the soil is divided into bedrock, gravel soil, sandy soil, silt, clay, silty clay, artificial filling soil and the like. 2) Color: mainly distinguishes oxidation environment series (grey yellow, brown yellow, etc.) and reduction environment series (grey, grey black, etc.).

The specific sequence division is realized by matching with a standard sequence:

specifically layering: firstly, determining a plurality of possible soil layers according to a range interval, screening the possible soil layers by using morphological characteristics, and removing the impossible soil layers; then, the matching degree of the remaining options is calculated according to factors such as physical characteristics and the like;

calculating the matching degree:

sequentially recording parameters such as parameter ranges, morphological characteristics, soil layer characteristics, distribution ranges and physical characteristics related to soil layers as { a1, a2, a3, };

performing matching parameter classification processing, and judging the data range class, such as water content in physical characteristics; descriptively, performing text similarity calculation such as soil layer characteristics; specific characteristics such as colors and morphological characteristic types are limited and basically form a standard, and complete matching is carried out; after vectorization of the distribution range, judging by combining with spatial analysis; and (4) recording the matching degree of each parameter as Pai, and weighting and summing the matching rate of each parameter according to the importance of the index features to obtain the soil layer probability Pa.

Recommending that soil layer results are relatively clear (the single probability matching rate is not lower than 85%) and can be regarded as single; if the multi-soil-layer results are approximate (the matching rates are different by no more than 10%), further distinction is needed: and comparing by using different expert models, and judging the soil layers by combining peripheral similar drilling experience to recommend the optimal soil layers in turn.

The recommended hierarchies may both be confirmed and adjusted by the moderator.

Step five, verifying the rationality of the sequence;

by utilizing the characteristics of the stratum: verifying the division result by means of a three-dimensional and plano-graph;

the single hole is correlated and compared with the peripheral hole, and is connected with the corresponding soil layer, and the single treatment with mutation is carried out;

2. and (4) leading the layered drill hole into a three-dimensional stratum, checking the distribution condition of the three-dimensional stratum around the layered drill hole, and verifying the reasonability.

3. Aiming at a single-hole local special stratum, the rationality needs to be checked;

step six, checking and confirming by experts;

the host needs to be verified and approved by the verification and approval program, the process is generally performed by experienced technical experts, and the approver is more authoritative than the verifier and can find the problem in the layering of the technical staff, so that the layering result is more scientific and reasonable. The auditing and approving person can reproduce the layering process, and can adjust the layering confirmation result of the host, and the adjustment information needs to be recorded;

step seven, outputting results and warehousing data;

the output of the result of the checking can be used as a part of the engineering investigation result, and can be further researched based on the result;

the data including layered original data, layered calculation analysis data, expert checking data, layered results and the like are required to be brought into a regional data resource library, so that data resources are enriched, and an expert recommendation model is optimized.

Referring to fig. 2 and 3, a system for layering foundation soil includes:

a resource library module: building a regional professional resource library; wherein:

2. Constructing a special database:

2.2 engineering survey report library: the method comprises the steps of establishing a library according to projects by including data such as engineering survey reports, report drawings (horizontal cutting drawings), project attached tables and the like.

2.3 engineering investigation drilling library: and (4) gathering the drilling data of each project based on drilling to build a library, wherein the library comprises drilling positions, test information and layering information. The borehole retains a history. Formation standardization and spatial reference unification of drilling data are required.

2.4 Standard formation database: standard strata of a large area of a city can be obtained or statistically formulated, such as a ground soil sequence division rule; the standard sequence achievement divides foundation soil in a certain depth range required by a conventional engineering project in a certain area into standard sequence, and the standard sequence is generally divided according to the standard sequence name, characteristics, a buried depth range, soil layer characteristics, a distribution range (specifically graphical description of characters) and a characteristic stratum distribution diagram of the area; physical characteristics, etc.

2.5 three-dimensional stratigraphic library: and extracting characteristic drill holes from the standardized and layered drill holes according to a 1km grid, forming an urban three-dimensional stratum according to the characteristic drill holes, and constructing an area-level subdivided stratum (integrally layered according to the city and subdivided into layers according to the area) according to 0.3 km.

3.3 associating the existing project drilling data and standard layering data with the checking expert, and sequentially carrying out normalization matching on layering results confirmed by the expert in the region and soil characteristic index parameters corresponding to the regional foundation soil sequence division standardAnd taking 60% -70% of data as a training set and the rest 30% -40% as a test set, and recommending the model by machine learning experts in a layering way. Wherein: c is a hierarchical set, and D is a set of training tuples and their associated class labels; using one n-dimensional attribute vector X ═ X for each tuple_1,x₂,...,x_nRepresents it. Assume that there are m classes C_1,C₂,...C_m。Given a tuple, the classification method makes the prediction attribute vector X belong to the class with the highest posterior probability; the class conditions are considered to be independent according to industry experience;

The specific operation process comprises the following steps:

1. project drilling field record:

2. Project land test data:

3. Analyzing the geological profile:

A reference sequence module: establishing a project reference sequence:

a recommendation layering module: sequentially recommending and confirming the layering:

1) lithology: the soil is divided into bedrock, gravel soil, sandy soil, silt, clay, silty clay, artificial filling soil and the like.

3) The state is as follows: mainly distinguishes soft and hard degrees (flow plastic, soft plastic, hard plastic) and dense degrees (slightly dense, medium dense, dense).

calculating the matching degree:

A verification module: verifying the rationality of the sequence;

the checking module: checking and confirming by an expert;

a warehousing module: outputting results and warehousing data;

An information data processing terminal for realizing the foundation soil layering method.

A computer-readable storage medium comprising instructions which, when executed on a computer, cause the computer to perform the method for stratifying the subsoil as described above.

In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When used in whole or in part, can be implemented in a computer program product that includes one or more computer instructions. When loaded or executed on a computer, cause the flow or functions according to embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, the computer instructions may be transmitted from one website site, computer, server, or data center to another website site, computer, server, or data center via wire (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL), or wireless (e.g., infrared, wireless, microwave, etc.)). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that includes one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.

The above description is only for the preferred embodiment of the present invention, and is not intended to limit the present invention in any way, and all simple modifications, equivalent changes and modifications made to the above embodiment according to the technical spirit of the present invention are within the scope of the technical solution of the present invention.

Claims

1. A method for layering foundation soil is characterized by comprising the following steps:

2. The method for layering foundation soil according to claim 1, wherein the specific process steps of building a regional professional resource library comprise:

3. The method for layering foundation soil according to claim 1, wherein the specific flow steps of the imported project drilling field collected data and the test data comprise:

4. The method for layering foundation soil according to claim 1, wherein the specific process step of establishing the project reference sequence comprises:

5. The method for layering foundation soil according to claim 1, wherein the step of sequentially recommending and confirming the specific process of layering comprises:

6. The method for stratifying foundation soil according to claim 1, wherein the specific process steps of sequence rationality verification comprise:

7. The method for layering foundation soil according to claim 2, wherein the specific process steps for constructing the special database comprise:

8. The method for layering foundation soil according to claim 2, wherein the step of establishing an industry application knowledge graph specific process comprises:

meanwhile, integrating comprehensive knowledge rich in open resources;

associating existing project drilling data and standard layering data with an examination and correction expert, sequentially carrying out normalization matching on layering results confirmed by the expert in the region and soil characteristic index parameters corresponding to regional foundation soil sequence division standards, taking 60% -70% of the data as a training set, and taking the rest 30% -4% of the data as training sets0% is used as a test set, and an expert hierarchical recommendation model is obtained through machine learning; wherein: c is a hierarchical set, and D is a set of training tuples and their associated class labels; using one n-dimensional attribute vector X ═ X for each tuple_1,x₂,...,x_nRepresents; assume that there are m classes C_1,C₂,...C_m。Given a tuple, the classification method makes the prediction attribute vector X belong to the class with the highest posterior probability; the class conditions are considered to be independent according to industry experience;

9. A system for stratifying foundation soil, comprising:

a reference sequence module: establishing a project reference sequence;

a verification module: verifying the rationality of the sequence;

the checking module: checking and confirming by an expert;

a warehousing module: and (4) outputting results and warehousing data.

10. The system of claim 9, wherein the expert stratigraphic recommendation model is: