CN113204830B - Correction method for determining cohesive soil foundation bearing capacity through shear wave velocity based on BP network - Google Patents
Correction method for determining cohesive soil foundation bearing capacity through shear wave velocity based on BP network Download PDFInfo
- Publication number
- CN113204830B CN113204830B CN202110762156.3A CN202110762156A CN113204830B CN 113204830 B CN113204830 B CN 113204830B CN 202110762156 A CN202110762156 A CN 202110762156A CN 113204830 B CN113204830 B CN 113204830B
- Authority
- CN
- China
- Prior art keywords
- wave velocity
- curve
- network
- fitting
- bearing capacity
- 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.)
- Active
Links
Images
Classifications
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/13—Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Architecture (AREA)
- Civil Engineering (AREA)
- Structural Engineering (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Investigation Of Foundation Soil And Reinforcement Of Foundation Soil By Compacting Or Drainage (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The application provides a correction method for determining cohesive soil foundation bearing capacity by shear wave velocity based on a BP network, which comprises the steps of obtaining a shear wave velocity sample value and a standard penetration number sample value of a field; fitting a regression curve to obtain a first fitting curve; acquiring a measured value of the transverse wave velocity of the site, inputting the measured value into a BP network, and obtaining a second fitting curve according to a predicted value output by the BP network; and correcting the first fitted curve by using the second fitted curve to obtain a calibration curve, searching for a corresponding penetration number, and determining the bearing capacity of the foundation by using a standard table according to the obtained penetration number and the obtained actually-measured pore ratio. The invention utilizes the neural network to fully consider the nonlinear factors after linear regression fitting, and corrects the fitting curve. Compared with the conventional exploration method, the method can greatly weaken the influence of external factors on the ground exploration progress, reduce the workload of in-situ test objects, and provide a new technical means for determining the bearing capacity of the cohesive soil foundation better and faster.
Description
Technical Field
The invention relates to the technical field of determining the bearing capacity of a cohesive soil foundation of a high-speed railway, in particular to a correction method for determining the bearing capacity of the cohesive soil foundation by the wave velocity of a shear wave based on a BP network.
Background
The railway construction is changed from a quantity scale type to a quality benefit type, the construction area is gradually expanded to areas such as difficult and complex mountainous areas, forest areas, high and steep terrains, high altitude and the like, more and more newly constructed railways pass through mountainous areas with very complex terrain and geological conditions, and meanwhile, the requirements of increasing environmental protection and construction period are met, and the conventional exploration means are increasingly difficult to meet the exploration requirements of high precision and tight construction period of high-speed rails. Moreover, the geological conditions of the mountainous terrain are generally complex, the types of unfavorable geologic bodies are various, the exploration precision is improved, and the major challenge of engineering investigation is to avoid missing major unfavorable geology. Basic investigation needs to be strengthened to provide basic guarantee for building fine railway engineering.
The bearing capacity of the foundation is one of important indexes of high-speed railway design and is an important geotechnical parameter required to be provided by engineering investigation. The engineering cost difference of the bridge, the roadbed and other projects with different foundation bearing capacities is great, and the inaccurate foundation bearing capacity can cause inaccurate project capital budget, so that the project is difficult to advance. Different foundation bearing capacities also influence the selection of construction technology and protective measures, and the inaccurate foundation bearing capacity brings great quality and safety problems. The bearing capacity of the foundation is the basis for selecting a construction method, and is the basis for scientific management, correct evaluation of economic benefit, determination of loads on the structure, determination of the type and size of the structure, establishment of labor quota, material consumption standard and the like.
Rock and soil are generally divided into cover layers and bedrocks, the bearing capacity of the bedrocks is generally high, and the impact on the safety of engineering construction is small. Cohesive soil is used as common rock soil, the bearing capacity of the cohesive soil is often uneven, the influence of the moisture content, the pore ratio, the cause and the formation age is large, and the influence on the engineering cost and the measure selection is large.
In the investigation and design stage, three methods are mainly used for determining the bearing capacity of the cohesive soil foundation: (1) in-situ test method: determining the bearing capacity through a field direct test (such as flat plate load and a standard penetration test); (2) theoretical formula method: calculating the bearing capacity by using a theoretical formula according to the actually measured shear strength index of the soil; (3) and (3) standard table method: and obtaining the bearing capacity by checking the table listed in the specification according to the indoor test index, the field test index or the field identification index (such as compactness and porosity).
All three methods rely on the success of drilling in situ by the drilling rig. The in-situ test has limited measurement depth, consumes time and labor; the theoretical formula method and the standard table method also depend on samples obtained by drilling and in-situ test results, the workload is large, the test period is long, the limitation by the terrain is obvious, and the problem encountered by the current railway engineering exploration is difficult to solve. Moreover, the methods are usually in a point-to-surface mode, continuous geological results cannot be obtained, the requirements of design precision and adjustment cannot be met, and the foundation bearing capacity of a specified position cannot be determined under the condition of lacking drilling.
For higher and higher high-speed rail geological survey requirements, the existing common technology for determining the bearing capacity of the foundation has the defects of different degrees, and a simple, convenient and quick practical method is urgently needed to solve the problem.
Disclosure of Invention
In view of the above survey problems and needs, embodiments of the present application provide a correction method for determining the bearing capacity of a cohesive soil foundation by using a shear wave velocity based on a BP network, the relationship between the shear wave velocity and a standard penetration number is obtained by performing linear regression on the shear wave velocity, a linear regression curve is corrected by using the BP network, the survey data is linearly fitted, meanwhile, nonlinear factors of the data are fully considered, the standard penetration number can be accurately predicted, the foundation bearing capacity is determined by using a specification table, the influence of factors such as air temperature, forest zones, traffic and topography on the ground survey progress can be greatly reduced, the workload of in-situ test objects is reduced, and the geological survey efficiency is greatly improved.
The invention provides a correction method for determining the bearing capacity of a cohesive soil foundation by the wave velocity of a transverse wave based on a BP network, which comprises the following steps: s1, acquiring a shear wave velocity sample value and a standard penetration number sample value of the site;
s2, performing regression curve fitting on the shear wave velocity sample value and the standard penetration number sample value to obtain a first fitting curve;
s3, acquiring a measured value of the transverse wave velocity of the site, inputting the measured value into a trained BP network, and performing curve fitting according to a predicted value output by the BP network to obtain a second fitting curve;
s4, correcting the first fitting curve by using the second fitting curve to obtain a calibration curve;
and S5, searching the corresponding penetration number by using a calibration curve according to the actual measured value of the transverse wave velocity of the site, and determining the bearing capacity of the foundation by using a standard lookup table according to the obtained penetration number and the obtained actual measured pore ratio.
In one embodiment of the invention, the relation between the shear wave velocity and the standard penetration number is obtained by performing linear regression on the shear wave velocity, a linear regression curve is corrected by using a BP network, the survey data is subjected to linear fitting, meanwhile, the nonlinear factors of the data are fully considered, the standard penetration number can be predicted more accurately, and thus the foundation bearing capacity is determined through a specification table.
Preferably, in S1, when the site shear wave velocity is obtained, the method includes:
collecting a transverse wave data set by adopting small intervals, and processing and inverting the transverse wave data in the transverse wave data set to obtain a high-precision transverse wave velocity value;
calculating the intersection point of the field transverse wave velocity numerical values for the high-precision transverse wave velocity numerical values, and extracting the transverse wave velocity of the specified area;
and carrying out statistical analysis on the obtained field transverse wave velocity, carrying out normal distribution fitting on the field transverse wave velocity, and filtering out transverse wave velocity abnormal points falling in a small probability interval.
Further preferably, when the transverse wave data are collected, the collection track interval is set to be 5-10m, the consistency of the detector is kept to be greater than 95%, and the signal-to-noise ratio of the original data is greater than 2.
More preferably, in step S2, when performing regression curve fitting, the shear wave velocity sample value and the standard penetration number sample value are input to EXCEL to form a scattergram graph, a trend line is added to the scattergram graph, and a curve type is selected from the R2 value to form a first fitted curve.
Further preferably, in S3, the BP network is trained according to the following procedure:
dividing the shear wave velocity sample value and the standard penetration number sample value into a training set and a testing set after data cleaning;
inputting shear wave velocity sample values and standard penetration number sample values in a training set into a constructed BP network for nonlinear fitting training;
the BP network is a three-layer neural network and comprises an input layer, a hidden layer and an output layer; transfer functions from the input layer to the hidden layer and from the hidden layer to the output layer respectively adopt logsig functions and purelin functions;
inputting a shear wave velocity sample value of the test set by using the trained network parameters, performing network prediction, and performing error calculation according to an output standard penetration number predicted value and a standard penetration number sample value corresponding to the test set; and when the error meets a preset threshold value, finishing the training.
In this embodiment, a BP network is used to perform nonlinear fitting, and after fitting the discrete data of the input shear wave velocity sample value and the target penetration number sample value, the nonlinear factor of the data is fully considered, so that the target penetration number can be predicted more accurately.
Further preferably, in S3, when curve fitting is performed according to the predicted value output by the BP network to obtain the second fitting curve, the degree of fitting is calculated by using the following formula
Wherein R is2 2For the fitness of the second fitted curve, SSE is the sum of the squares of the residuals of the input data and the output predicted values, SST is the sum of the squares of the sums of the squares of the residuals of the input data and the output predicted values.
Further preferably, the correcting the first fitted curve by using the second fitted curve includes:
s401, placing the first fitting curve and the second fitting curve into the same coordinate system;
s402, scanning the two curves by adopting a sliding threshold window, searching two original vertical coordinates corresponding to the horizontal coordinates of the point location when any point location appears and the interval between the two curves is larger than the threshold value of the sliding threshold window, calculating a new vertical coordinate according to a preset algorithm, and updating the coordinate values of the point location; forming a new first fitting curve as the coordinate value of the point position in the corrected first fitting curve;
and S403, taking the central track of the sliding threshold window as a new second fitted curve, reducing the threshold of the sliding threshold window, scanning the new second fitted curve and the new first fitted curve in the S402 again, repeating the S402, and finally forming a correction curve.
In this embodiment, a sliding threshold window is adopted to merge two curves, and smooth merging of the two curves is gradually realized through repeated operations, so that smooth transition at a point position where the difference between the two curves is large is realized.
Further preferably, the preset algorithm is calculated by using the following formula:
wherein, yAs a new ordinate, R1 2Degree of fit, R, of the first fitted curve2 2Degree of fit, y, of the second fitted curve1Is the ordinate, y, of the point in the first fitted curve2The ordinate of the point in the second fitted curve.
In the preset algorithm provided by this embodiment, the fitting degree of the first and second fitting curves is used as a weight coefficient for calculating the ordinate, and fitting degree optimization is performed on the basis of maintaining data linear fitting, so that smooth transition is realized at a point location where the difference between the two curves is large.
Further preferably, in step S5, the foundation bearing capacity is determined according to the normalized table by the following method:
searching a specification table according to the standard penetration number to obtain the plastic state and the liquidity index of the cohesive soil corresponding to the standard penetration number;
determining the range of the cohesive soil porosity ratio of the designated area according to the measured value or the area empirical value;
and determining the foundation bearing capacity of the cohesive soil according to the liquidity index and the porosity ratio.
The method for determining the bearing capacity of the cohesive soil foundation by the transverse wave velocity provided by the embodiment of the application; compared with the prior art, the method has stronger applicability and quantificational property, overcomes the terrain constraint and the construction period limitation of complex terrain on exploration, and makes up the defect that drilling cannot be continuously explored; direct connection between the transverse wave velocity and the standard penetration technology is established, restraint is carried out through measured data and a standard, and the foundation bearing capacity of the cohesive soil is determined through table lookup according to the standard. Compared with the conventional exploration method, the method has the advantages that the influence of factors such as air temperature, forest regions, traffic and topography on the land exploration can be greatly weakened, the workload of in-situ test objects is reduced, the geological exploration efficiency is greatly improved, the exploration period is shortened by over 60 percent, a new technical means is provided for determining the bearing capacity of the cohesive soil foundation better and faster, and a reliable basis is provided for engineering design of bridges, roadbeds and the like of high-speed railways better and faster.
Drawings
Fig. 1 is a flowchart illustrating a method for determining a bearing capacity of a cohesive soil foundation according to a shear wave velocity according to an embodiment of the present disclosure;
fig. 2 is a schematic diagram illustrating a regression curve of the transverse wave velocity and the penetration number according to an embodiment of the present application;
fig. 3 is a second fitting curve obtained by the BP network according to an embodiment of the present application;
FIG. 4 is a graph illustrating a sliding threshold window versus a merged calibration graph of a first fitted curve and a second fitted curve according to an embodiment of the present disclosure;
FIG. 5 is a schematic diagram of a calibration curve generated by the present application;
fig. 6 is a surface wave survey result of a certain bridge engineering provided in the embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
The geophysical prospecting transverse wave detection technology is mature, the instrument is light and convenient, the efficiency is high, the requirement on the field is low, continuous results can be obtained, and the method is the preferred method for solving the problems. The transverse wave velocity is the same as the penetration value, so that the hardness and softness of the soil can be reflected, the hardness and softness of the soil are related, and the hardness and softness of the soil are the macroscopic expression of the bearing capacity of the soil body. Therefore, a simple mathematical model can be found theoretically and mathematically, the relation between the field transverse wave velocity and the standard penetration number of the foundation soil layer is established, and the bearing capacity of the cohesive soil is deduced and determined according to the existing standard and the actual measurement data correction and constraint.
The invention provides a correction method for determining the bearing capacity of a cohesive soil foundation by the wave velocity of a transverse wave based on a BP network, which comprises the following steps:
s1, acquiring a shear wave velocity sample value and a standard penetration number sample value of the site; preferably, when the field transverse wave velocity is obtained, the method comprises the following steps: collecting a transverse wave data set by adopting small intervals, and processing and inverting the transverse wave data in the transverse wave data set to obtain a high-precision transverse wave velocity value; and calculating the intersection point of the engineering shear wave velocity numerical values for the high-precision shear wave velocity numerical values, and extracting the shear wave velocity of the specified area. Specifically, commercial wafer software is used for calculating the intersection points of the engineering transverse wave velocity numerical values of bridges, roadbeds and the like, and extracting the transverse wave velocity of the designated area. And when the transverse wave data is collected, the collection track interval is set to be 5-10 m. When the transverse wave data is collected, the data collection is preferably carried out at quiet night, the consistency of the detector is kept to be more than 95%, and the signal-to-noise ratio of the original data is more than 2. The number of the statistical shear wave velocity values is more than 40, and the more the data samples are, the closer the regression formula is to the real situation.
The method also comprises the following steps of, for the obtained shear wave velocity sample value of the field, removing the wave velocity abnormal point by normal distribution: and carrying out statistical analysis on the obtained field transverse wave velocity, carrying out normal distribution fitting on the field transverse wave velocity, and filtering out transverse wave velocity abnormal points falling in a small probability interval.
Because the transverse wave velocities of different rock-soil are greatly influenced by differences of influence of lithology, hydrological conditions, weathering and the like, the transverse wave velocity of a single lithology is subject to normal distribution under the condition of large data volume. According to the theory of normal distribution, when the value of the probability density function is smaller than a certain value, the probability density function can be understood as a small probability event, and the small probability event is almost impossible to occur according to the mathematical theory, so that the value of the wave speed of the transverse wave of the original data which is positioned outside the interval is deleted, and the data outside the normal distribution is considered as an abnormal value.
As shown in fig. 2, S2, performing regression curve fitting on the shear wave velocity sample value and the standard penetration number sample value to obtain a first fitting curve; the regression curve is that when the regression curve is fitted in EXCEL by least square method, the shear wave velocity sample value and the standard penetration number sample value are input into EXCEL to form a scatter diagram chart, a trend line is added in the scatter diagram chart, and R is used for fitting the regression curve2The values select the type of curve, forming a first fitted curve.
In one embodiment of the present invention, a correlation curve (fig. 2) between the velocity of the shear wave (a kind of shear wave) and the penetration number of the cohesive soil is fitted by least squares regression based on the relationship between the measured shear wave (a kind of shear wave) and the measured cohesive soil penetration number of 45 sets (8 sets of abnormal values are removed from 53 sets of measured data) under study. After comparing various relations such as linearity, exponent, polynomial and the like, the logarithmic relation with the best fitting degree is adopted as follows:
Wherein:
is the number of standard penetration, v is the measured shear wave velocity, R2Is the goodness of fit (the closer to 1, the better the fit).
S3, acquiring a measured value of the transverse wave velocity of the site, inputting the measured value into a trained BP network, and performing curve fitting according to a predicted value output by the BP network to obtain a second fitting curve;
as shown in fig. 3, the BP network is trained according to the following process:
dividing the shear wave velocity sample value and the standard penetration number sample value into a training set and a testing set after data cleaning;
inputting shear wave velocity sample values and standard penetration number sample values in a training set into a constructed BP network for nonlinear fitting training; preferably, a MATLAB correlation algorithm is adopted for neural network construction and training.
The BP network is a three-layer neural network and comprises an input layer, a hidden layer and an output layer; the transfer functions from the input layer to the hidden layer and from the hidden layer to the output layer respectively adopt logsig functions and purelin functions;
inputting a shear wave velocity sample value of the test set by using the trained network parameters, performing network prediction, and performing error calculation according to an output standard penetration number predicted value and a standard penetration number sample value corresponding to the test set; and when the error meets a preset threshold value, finishing the training.
In one embodiment of the invention, the network fabric is initialized:
net=newff(v,
, [10,1]{ ' logsig ', ' purelin) }) where newff () is the BP network function, v represents the input data,
represents the output data, [10,1 ]]Representing that the number of neurons of a hidden layer of a BP network is 10, the number of neurons of an output layer is 1, { 'logsig', 'purelin' } represents that the transfer functions of the network from an input layer to the hidden layer and from the hidden layer to the output layer respectively adopt logsig functions and purelin functions;
net = train (net, inputn, outputn) for network training;
an = sim (net, input _ test) to perform network prediction;
setting the hidden layer activation function to relu (v) ═ max (v,0), controls the activation of neurons.
In S3, when curve fitting is performed according to the predicted value output by the BP network to obtain a second fitting curve, the degree of fitting is calculated using the following formula
(formula 2)
Wherein R is2 2For the fitness of the second fitted curve, SSE is the sum of the squares of the residuals of the input data and the output predicted values, SST is the sum of the squares of the sums of the squares of the residuals of the input data and the output predicted values.
As shown in fig. 4, S4, the first fitted curve is corrected by using the second fitted curve to obtain a calibration curve; the method specifically comprises the following steps of when the second fitting curve is used for correcting the first fitting curve:
s401, placing the first fitting curve and the second fitting curve in the same coordinate system;
s402, scanning the two curves by adopting a sliding threshold window, searching two original vertical coordinates corresponding to the horizontal coordinates of the point location when any point location appears and the interval between the two curves is larger than the threshold value of the sliding threshold window, calculating a new vertical coordinate according to a preset algorithm, and updating the coordinate values of the point location; forming a new first fitting curve as the coordinate value of the point position in the corrected first fitting curve;
and S403, taking the central track of the sliding threshold window as a new second fitted curve, reducing the threshold of the sliding threshold window, scanning the new second fitted curve and the new first fitted curve in S402 again, repeating S402, and finally forming a correction curve L as shown in FIG. 5.
The preset algorithm is calculated by adopting the following formula:
Wherein y is the new ordinate, R1 2Degree of fit, R, of the first fitted curve2 2Degree of fit, y, of the second fitted curve1Is the ordinate, y, of the point in the first fitted curve2The ordinate of the point in the second fitted curve.
Further, in the embodiment of the application, a fitting formula according with the variation trend of the calibration curve can be inverted according to the calibration curve, and the penetration number is calculated by adopting the formula.
And S5, determining the bearing capacity of the foundation by using a specification lookup table according to the penetration number and the pore ratio. Further preferably, in step S5, the foundation bearing capacity is determined according to the normalized table by the following method: searching a specification table according to the standard penetration number to obtain the plastic state and the liquidity index of the cohesive soil corresponding to the standard penetration number; determining the range of the cohesive soil porosity ratio of the designated area according to the measured value or the area empirical value; and determining the foundation bearing capacity of the cohesive soil according to the liquidity index and the porosity ratio. In a specific embodiment, the plastic state of the cohesive soil corresponding to the penetration number is obtained by using the penetration number lookup table 1 obtained in S4:
TABLE 1 Clay soil plasticity State partitioning
On the basis of the obtained plastic state and liquidity index of the cohesive soil, the pore ratio range of the cohesive soil in a specified area needs to be determined through actual measurement or regional experience. And finally, determining the foundation bearing capacity of the cohesive soil layer through table look-up (table 2) according to the liquidity index and the porosity ratio.
In practical application, the less favorable condition in the liquidity index determined in the step five is adopted to determine the range of the bearing capacity of the foundation according to the standard table lookup. And fifthly, calculating to obtain the plastic state of the cohesive soil as soft plastic, and during table lookup, selecting the liquidity index according to 0.9-1.0 (unfavorable condition in the soft plastic range). The void ratio is also considered to be in a less favorable range of 0.9 to 1.0. The foundation bearing capacity of the cohesive soil layer is calculated according to the mode to be within the range of 110-140 kPa (an inclined part in the table 2):
TABLE 2Q4Basic bearing capacity sigma of fluvial and flood cohesive soil foundation0 (kPa)
Example 1
And (3) counting the penetration number of the drilled shear wave and the cohesive soil which have already finished exploration in the northeast region, collecting transverse wave data corresponding to the region, wherein the counting is carried out on 53 layers of cohesive soil sample points of 18 drilled holes, and 8 obvious abnormal points need to be removed in the counting process.
The collected data is normally distributed, and data outside the probability density function of 99.7% needs to be removed.
The method comprises the steps of collecting transverse wave data by a transverse wave seismograph, enabling the point distance to be 5m, conducting filtering and deconvolution, finally conducting inversion to obtain a transverse wave velocity profile of the area, conducting gridding by sufer software according to a Kriging difference method, enabling the gridding distance to be 5m x 5m, utilizing MapGen software to extract transverse wave velocities of projects such as bridges and roadbeds, enabling the interval of the transverse wave velocities to be 5m, and conducting statistical result analysis according to the table 1. The engineering mileage of bridges, roadbeds and the like is obtained through actual measurement, and the lithology is obtained according to drilling data.
And (3) obtaining the penetration number of the cohesive soil at different positions through a correction curve according to the actually measured surface wave result (SW-X is a geophysical transverse wave velocity test point, and figure 6). According to the comparison between the in-situ test and the standard penetration number calculated by the novel method, the data shows that only 1 sample with the error of more than 20 percent in 10 comparison samples (the accuracy rate is 90 percent), the calculated value is generally smaller than the measured value and is more conservative, and the calculation precision of the formula can meet the requirement of the preliminary design of the high-speed railway:
TABLE 3 comparison table of penetration number of cohesive soil
In addition to the above-described methods and apparatus, embodiments of the present application may also be a computer program product comprising computer program instructions that, when executed by a processor, cause the processor to perform the steps in the methods according to the various embodiments of the present application described in the "exemplary methods" section of this specification, above.
The block diagrams of devices, apparatuses, systems referred to in this application are only given as illustrative examples and are not intended to require or imply that the connections, arrangements, configurations, etc. must be made in the manner shown in the block diagrams. These devices, apparatuses, devices, systems may be connected, arranged, configured in any manner, as will be appreciated by those skilled in the art. Words such as "including," "comprising," "having," and the like are open-ended words that mean "including, but not limited to," and are used interchangeably therewith. The words "or" and "as used herein mean, and are used interchangeably with, the word" and/or, "unless the context clearly dictates otherwise. The word "such as" is used herein to mean, and is used interchangeably with, the phrase "such as but not limited to".
It should also be noted that in the devices, apparatuses, and methods of the present application, the components or steps may be decomposed and/or recombined. These decompositions and/or recombinations are to be considered as equivalents of the present application.
The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
The foregoing description has been presented for purposes of illustration and description. Furthermore, the description is not intended to limit embodiments of the application to the form disclosed herein. While a number of example aspects and embodiments have been discussed above, those of skill in the art will recognize certain variations, modifications, alterations, additions and sub-combinations thereof.
Claims (9)
1. A correction method for determining the bearing capacity of a cohesive soil foundation by the wave velocity of a transverse wave based on a BP network is characterized by comprising the following steps:
s1, acquiring a shear wave velocity sample value and a standard penetration number sample value of the site;
s2, performing regression curve fitting on the shear wave velocity sample value and the standard penetration number sample value to obtain a first fitting curve;
s3, acquiring a measured value of the transverse wave velocity of the site, inputting the measured value into a trained BP network, and performing curve fitting according to a predicted value output by the BP network to obtain a second fitting curve;
s4, correcting the first fitting curve by using the second fitting curve to obtain a calibration curve;
and S5, searching the corresponding penetration number by using a calibration curve according to the actual measured value of the transverse wave velocity of the site, and determining the bearing capacity of the foundation by using a standard lookup table according to the obtained penetration number and the obtained actual measured pore ratio.
2. The method for correcting the determination of the bearing capacity of the cohesive soil foundation based on the BP network and the shear wave velocity as recited in claim 1, wherein in S1, when the site shear wave velocity is obtained, the method comprises the following steps:
collecting a transverse wave data set by adopting small intervals, and processing and inverting the transverse wave data in the transverse wave data set to obtain a high-precision transverse wave velocity value;
calculating the intersection point of the field transverse wave velocity numerical values for the high-precision transverse wave velocity numerical values, and extracting the transverse wave velocity of the specified area;
and carrying out statistical analysis on the obtained field transverse wave velocity, carrying out normal distribution fitting on the field transverse wave velocity, and filtering out transverse wave velocity abnormal points falling in a small probability interval.
3. The correction method for determining the bearing capacity of the cohesive soil foundation through the transverse wave velocity based on the BP network according to claim 2, wherein when the transverse wave data is collected, the collection track interval is set to be 5-10m, the consistency of the detector is kept to be greater than 95%, and the signal-to-noise ratio of the original data is greater than 2.
4. The method for calibrating the bearing capacity of a cohesive soil foundation determined by the shear wave velocity based on the BP network as claimed in claim 1, wherein in step S2, when performing regression curve fitting, the shear wave velocity sample value and the standard penetration number sample value are input into EXCEL to form a scattergram graph, and a trend line is added to the scattergram graph according to R2The values select the type of curve, forming a first fitted curve.
5. The method for correcting the determination of the bearing capacity of the cohesive soil foundation based on the BP network according to the shear wave velocity of the claim 2, wherein in S3, the BP network is trained according to the following procedures:
dividing the shear wave velocity sample value and the standard penetration number sample value into a training set and a testing set after data cleaning;
inputting shear wave velocity sample values and standard penetration number sample values in a training set into a constructed BP network for nonlinear fitting training;
the BP network is a three-layer neural network and comprises an input layer, a hidden layer and an output layer; transfer functions from the input layer to the hidden layer and from the hidden layer to the output layer respectively adopt logsig functions and purelin functions;
inputting a shear wave velocity sample value of the test set by using the trained network parameters, performing network prediction, and performing error calculation according to an output standard penetration number predicted value and a standard penetration number sample value corresponding to the test set; and when the error meets a preset threshold value, finishing the training.
6. The method for correcting the determination of the bearing capacity of the cohesive soil foundation based on the BP network and the shear wave velocity according to claim 1, wherein in S3, when performing curve fitting according to the predicted value output by the BP network to obtain the second fitting curve, the fitting degree is calculated by using the following formula:
wherein R is2 2For the fitness of the second fitted curve, SSE is the sum of the squares of the residuals of the input data and the output predicted values, SST is the sum of the squares of the sums of the squares of the residuals of the input data and the output predicted values.
7. The method for correcting the determination of the bearing capacity of the cohesive soil foundation based on the BP network and the shear wave velocity as recited in claim 6, wherein the correction of the first fitted curve by using the second fitted curve comprises:
s401, placing the first fitting curve and the second fitting curve in the same coordinate system;
s402, scanning the two curves by adopting a sliding threshold window, searching two original vertical coordinates corresponding to the horizontal coordinates of the point location when any point location appears and the interval between the two curves is larger than the threshold value of the sliding threshold window, calculating a new vertical coordinate according to a preset algorithm, and updating the coordinate values of the point location; forming a new first fitting curve as the coordinate value of the point position in the corrected first fitting curve;
and S403, taking the central track of the sliding threshold window as a new second fitted curve, reducing the threshold of the sliding threshold window, scanning the new second fitted curve and the new first fitted curve in the S402 again, repeating the S402, and finally forming a correction curve.
8. The correction method for determining the bearing capacity of the cohesive soil foundation through the shear wave velocity based on the BP network according to claim 7, wherein the preset algorithm is calculated by adopting the following formula:
wherein y is the new ordinate, R1 2Degree of fit, R, of the first fitted curve2 2Degree of fit, y, of the second fitted curve1Is the ordinate, y, of the point location in the first fitted curve2Is the ordinate of the point in the second fitted curve.
9. The method for correcting the determination of the bearing capacity of the cohesive soil foundation based on the BP network based on the shear wave velocity as claimed in claim 1, wherein in step S5, the bearing capacity of the foundation is determined according to the table look-up of the specification by the following method:
searching a specification table according to the standard penetration number to obtain the plastic state and the liquidity index of the cohesive soil corresponding to the standard penetration number;
determining the range of the cohesive soil porosity ratio of the designated area according to the measured value or the area empirical value;
and determining the foundation bearing capacity of the cohesive soil according to the liquidity index and the porosity ratio.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110762156.3A CN113204830B (en) | 2021-07-06 | 2021-07-06 | Correction method for determining cohesive soil foundation bearing capacity through shear wave velocity based on BP network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110762156.3A CN113204830B (en) | 2021-07-06 | 2021-07-06 | Correction method for determining cohesive soil foundation bearing capacity through shear wave velocity based on BP network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113204830A CN113204830A (en) | 2021-08-03 |
| CN113204830B true CN113204830B (en) | 2021-08-27 |
Family
ID=77022733
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110762156.3A Active CN113204830B (en) | 2021-07-06 | 2021-07-06 | Correction method for determining cohesive soil foundation bearing capacity through shear wave velocity based on BP network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113204830B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113640877B (en) * | 2021-08-11 | 2023-04-07 | 中国铁路设计集团有限公司 | Method and system for calculating proportionality coefficient of soil horizontal reaction coefficient |
| CN114154725A (en) * | 2021-12-06 | 2022-03-08 | 南京工业大学 | Foundation bearing capacity characteristic value prediction method based on machine learning algorithm |
| CN114348895B (en) * | 2021-12-17 | 2023-12-12 | 湖北工业大学 | Intelligent jack device for prestress steel beam test |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107784191A (en) * | 2017-12-12 | 2018-03-09 | 中国地质大学（武汉） | Anisotropic rock joint peak shear strength Forecasting Methodology based on neural network model |
| CN109763474A (en) * | 2018-12-07 | 2019-05-17 | 东南大学 | A kind of shear wave velocity evaluation method based on standard penetration test (SPT) |
| CN112360548A (en) * | 2020-11-24 | 2021-02-12 | 西安科技大学 | Roadside concrete filling body full-service period stability monitoring and early warning system and method |
-
2021
- 2021-07-06 CN CN202110762156.3A patent/CN113204830B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107784191A (en) * | 2017-12-12 | 2018-03-09 | 中国地质大学（武汉） | Anisotropic rock joint peak shear strength Forecasting Methodology based on neural network model |
| CN109763474A (en) * | 2018-12-07 | 2019-05-17 | 东南大学 | A kind of shear wave velocity evaluation method based on standard penetration test (SPT) |
| CN112360548A (en) * | 2020-11-24 | 2021-02-12 | 西安科技大学 | Roadside concrete filling body full-service period stability monitoring and early warning system and method |
Non-Patent Citations (4)
| Title |
|---|
| Uncertainty of in-situ slide zone soil shear test condition and shear strength parameters estimation based on BP network;Xiong Shihu 等;《Rock and Soil Mechanics》;20061027;全文 * |
| 含水量和压实度对黏性土地基承载力的影响试验研究;徐方 等;《铁道科学与工程学报》;20111031;全文 * |
| 基于FOA_GRNN长螺旋钻孔压灌混凝土桩质量预测;王荣华;《中国优秀硕士学位论文全文数据库》;20190115;全文 * |
| 软土地基堆载预压联合强夯的固结变形与承载性状研究;朱常志;《中国博士学位论文全文数据库》;20190115;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113204830A (en) | 2021-08-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113204830B (en) | Correction method for determining cohesive soil foundation bearing capacity through shear wave velocity based on BP network | |
| CN109584510B (en) | Road high slope landslide hazard early warning method based on evaluation function training | |
| CN109145463B (en) | Deformation analysis method used in tunnel excavation process | |
| KR101642951B1 (en) | GIS-based real time earthquake prediction method | |
| US11834947B2 (en) | 3D in-situ characterization method for heterogeneity in generating and reserving performances of shale | |
| CN110795878B (en) | Tunnel water inflow prediction method | |
| Admassu et al. | Evaluating selected factors affecting the depth of undercutting in rocks subject to differential weathering | |
| CN110517794A (en) | A method of that establishes shale gas reservoir buries-thermal evolution history figure | |
| CN111596364A (en) | Seismic sedimentation microphase combination analysis method based on high-precision sequence stratigraphic framework | |
| Johnson et al. | Statistical comparison of methods for estimating sediment thickness from horizontal-to-vertical spectral ratio (HVSR) seismic methods: An example from Tylerville, Connecticut, USA | |
| CN112686522A (en) | Newmark correction model earthquake landslide risk assessment method | |
| CN114036831A (en) | Real-time detection method for geotechnical parameters of side slope of engineering field to be detected | |
| Schoonbeek | Land subsidence as a result of natural gas extraction in the province of Groningen | |
| CN113250172B (en) | Method for determining bearing capacity of cohesive soil foundation based on shear wave velocity | |
| CN107103165A (en) | A kind of method of CALCULATING OILFIELD reservoir steep dip shearing crack line density | |
| CN116862432A (en) | Dynamic adjustment method and system for tunnel shallow buried section construction scheme | |
| CN116050917A (en) | Road surface three-dimensional texture uniformity evaluation method based on grid distribution characteristics | |
| La Pointe et al. | Estimation of effective block conductivities based on discrete network analyses using data from the Äspö site | |
| CN110618455B (en) | Quantitative evaluation method for sedimentary basin structure uplift | |
| CN113240357A (en) | Rapid evaluation method for stability of highway side slope in red layer area | |
| CN114066271A (en) | Tunnel water inrush disaster monitoring and management system | |
| CN112147713A (en) | Shale total organic carbon content segmented prediction method | |
| CN109408920A (en) | Rectifying of leaned building method and deviation-rectifying system | |
| CN115479540B (en) | Karst water bursting disaster magnitude assessment method and system based on laser scanning | |
| Gallichand et al. | Optimal sampling density of hydraulic conductivity for subsurface drainage in the Nile delta |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |