A method and system for quickly determining soil strength parameters based on CPT data

Abstract

The present application relates to the field of geotechnical engineering parameters, in particular to a method and system for quickly determining soil strength parameters based on CPT data. The method comprises: obtaining CPT data and drilling data, and obtaining standardized data; constructing a local neighbor sample set according to the spatial relationship between the target CPT query point and the drilling point, and obtaining the KNN weighted representation value; fitting the edge distribution of the KNN weighted representation value of cohesion and internal friction angle respectively, selecting the optimal edge distribution and constructing the prior distribution; constructing the normalized CPT response according to the standardized CPT data, and constructing the main CPT likelihood function; constructing the shear strength weak constraint likelihood function by using the KNN weighted representation value of shear strength; constructing the joint likelihood function based on the two kinds of likelihood functions, and updating the prior distribution to the posterior distribution combined with the Bayesian update, and outputting the posterior parameter and uncertainty evaluation result. The present application solves the problems of insufficient expression of CPT empirical estimation uncertainty and difficulty of updating the target CPT observation with drilling information in the prior art.

Description

Technical Field

[0001] This invention relates to the field of geotechnical engineering parameters, and in particular to a method and system for rapidly determining soil strength parameters based on CPT data, which can be used for the rapid determination and uncertainty evaluation of effective cohesion and effective internal friction angle in soft soil, cohesive soil or similar strata. Background Technology

[0002] Because in-situ and laboratory experiments are complex and time-consuming, measured data on soil and rock parameters at specific sites are usually scarce. Furthermore, soil and rock materials are natural materials with significant uncertainties, making accurate characterization of their parameters particularly difficult.

[0003] Engineering projects often rely on the empirical relationship between CPT response values ​​and laboratory tests to estimate soil strength parameters. However, the CPT response is influenced by factors such as stress state, burial depth, pore pressure, and formation heterogeneity. A single empirical formula can only provide a deterministic estimate and cannot fully express the uncertainty of the parameters. Existing K-nearest neighbor algorithms rely on weighted values ​​from nearby borehole samples, which can reflect local engineering experience in the strata near the target CPT point, but they are biased towards prior estimates and cannot easily update the parameter distribution by incorporating continuous response information from the CPT point. Existing Bayesian methods typically focus on constructing prior parameter distributions or complex correlation structures. If only complex joint distribution forms are used to express parameter correlations without reweighting candidate parameters through the physical relationship between the CPT main response and shear strength, the model structure may be complex but its contribution to engineering interpretation may be limited.

[0004] Therefore, it is necessary to propose a method that combines the physical relationship between CPT response and shear strength to quickly determine soil strength parameters and uncertainty evaluation results, so that CPT observation information can directly screen reasonable parameter combinations, while the shear strength constraint can prevent the subsequent update from deviating from the basic mechanical relationship. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method and system for rapidly determining soil strength parameters based on CPT data, solving the problems of insufficient expression of uncertainty in CPT empirical estimation and difficulty in effectively updating borehole information through target CPT observations in existing technologies.

[0006] To achieve the above objectives, in a first aspect, the present invention provides a method for rapidly determining soil strength parameters based on CPT data, comprising: acquiring CPT data and borehole data, processing them to obtain standardized data; constructing a local nearest neighbor sample set based on the spatial relationship between the target CPT query point and the borehole point, and obtaining KNN weighted representation values; fitting the KNN weighted representation values ​​of cohesion and internal friction angle to the marginal distributions respectively, selecting the optimal marginal distribution and constructing a prior distribution; constructing a normalized CPT response based on the standardized CPT data in the standardized data, and constructing a principal CPT likelihood function; constructing a weakly constrained likelihood function of shear strength using the KNN weighted representation value of shear strength; constructing a joint likelihood function based on the two likelihood functions, and updating the prior distribution to a posterior distribution using Bayesian update, and outputting posterior parameters and uncertainty evaluation results.

[0007] This invention combines the edge prior formed by local borehole samples, the main likelihood function formed by the CPT normalized response, and the weak physical constraint formed by the KNN weighted characterization value of shear strength to quickly determine soil strength parameters, posterior confidence intervals, posterior joint distribution, and parameter uncertainty evaluation results. This allows CPT observation information to directly screen reasonable parameter combinations, while the shear strength constraint prevents posterior updates from deviating from basic mechanical relationships.

[0008] Optionally, the acquisition of CPT data and borehole data, and the processing to obtain standardized data, includes: acquiring static cone penetration test (CPT) data, borehole data, and corresponding spatial coordinates and depth information of the target site; performing coordinate unification, depth benchmark correction, soil layer screening, missing value processing, and outlier removal on the CPT data and borehole data to form standardized data; the standardized data includes the standardized CPT data and standardized borehole data, wherein the standardized CPT data includes the borehole coordinates, test depth, CPT response value, total overburden stress, and effective overburden stress of the CPT boreholes in the target site, and the standardized borehole data includes the borehole coordinates, sampling depth, soil layer type, cohesion, internal friction angle, and shear strength of the borehole samples.

[0009] This invention acquires CPT data and borehole data, performs spatial matching on them to form a standardized dataset, maps sparse borehole test data to a massive number of dense CPT measurement points, realizes spatialized and gridded assignment of soil strength parameters in the test area, and provides a data foundation for the subsequent construction of local nearest neighbor sample sets.

[0010] Optionally, the step of constructing a local nearest neighbor sample set based on the spatial relationship between the target CPT query point and the borehole point, and obtaining KNN weighted representation values ​​includes: calculating the planar distance or spatial distance between the target CPT query point and each borehole point, selecting the K boreholes with the smallest distance as the local nearest neighbor sample set; determining the weight of each nearest sample using the inverse distance weight, exponential decay weight, or other monotonically decreasing weight function, and obtaining KNN weighted representation values ​​of cohesion, internal friction angle, and shear strength based on the weights.

[0011] This invention relies on the spatial proximity relationship between the target CPT query point and the borehole point to screen effective borehole samples and construct a local nearest neighbor sample set. This makes up for the deficiency that static cone penetration CPT cannot directly measure shear parameters, and improves the spatial assignment accuracy of soil cohesion and internal friction angle while saving exploration costs.

[0012] Optionally, the step of fitting marginal distributions to the KNN-weighted characterization values ​​of cohesion and internal friction angle, selecting the optimal marginal distribution, and constructing a prior distribution includes: applying one or more candidate distributions among normal, log-normal, generalized extreme value, Weibull, Johnson-SU, Beta, or Student-t distributions to the KNN-weighted characterization values ​​of cohesion and internal friction angle; selecting the optimal marginal distribution using the AIC criterion; and generating a candidate parameter sample set from the optimal marginal distribution. ,in, For optimal edge distribution cohesion, The internal friction angle is the optimal edge distribution; the prior distribution is characterized as:

[0013] in, The cohesive force Edge prior density, The internal friction angle Edge prior density, Candidate parameters The joint prior density; Candidate parameters are obtained by sampling from the prior distribution. and the candidate parameters joint prior density ,in, For a priori cohesion, This is the a priori internal friction angle.

[0014] This invention performs edge distribution fitting on the KNN weighted characterization values ​​of cohesion and internal friction angle and selects the optimal edge distribution. The optimal edge distribution is selected using the AIC criterion, and a prior distribution reflecting the local engineering experience range is constructed. This construction method avoids forcibly introducing complex joint correlation structures, so that the prior mainly plays the role of limiting the local engineering experience range, and retains sufficient degrees of freedom for subsequent CPT observation updates.

[0015] Optionally, constructing a normalized CPT response based on the normalized CPT dataset in the normalized data includes: the normalized CPT response satisfies:

[0016] in, For the first Normalized CPT response value for each target CPT query point For the first CPT response value for each target CPT query point For the first The total overlying stress of each target CPT query point For the first Effective overlying stress of each target CPT query point Standard atmospheric pressure, The stress normalization index is denoted as .

[0017] This invention constructs a normalized CPT response by utilizing the normalized CPT response value, total overburden stress, and effective overburden stress, thereby unifying data dimensions and evaluation criteria and enhancing the numerical stability of the likelihood function.

[0018] Optionally, constructing the main CPT likelihood function includes: calculating the observed response of the normalized CPT response and the candidate parameters. The error between the corresponding predicted responses, wherein the error is characterized as:

[0019] in, For the first The error between the observed response and the predicted response for each target CPT query point For the first Normalized response of each target CPT query point The observed response can be Taking the logarithm, we get, that is , For the first Candidate parameters for each target CPT query point The corresponding predicted response, This can be expressed as a priori cohesion. a priori internal friction angle and depth The linear or nonlinear function can also be given by an empirical regression model, a local surrogate model, or a pre-calibrated response surface model; The principal CPT likelihood function is constructed by combining the aforementioned errors and is characterized as follows:

[0020] in, The set of observation data for the target CPT query point The Middle Observational data for each target CPT query point For the first The principal CPT likelihood function for each target CPT query point The error scale parameter of the main CPT likelihood function, For the first The error between the observed response and the predicted response of each target CPT query point.

[0021] This invention constructs the main CPT likelihood function by using the observed response of the normalized CPT response and the predicted response corresponding to the candidate parameters. It can reweight the candidate parameter combination using the continuous in-situ test information of the target CPT point, serving as an essential data constraint carrier for Bayesian updates.

[0022] Optionally, constructing the weakly constrained likelihood function of shear strength using the KNN-weighted characterization value of shear strength includes: calculating the candidate parameters according to the Mohr-Coulomb strength criterion. In the Calculated candidate shear strength values ​​for each target CPT query point Characterized as: ,in, For a priori cohesion, For the first Effective overlying stress of each target CPT query point The prior internal friction angle is used; the shear strength residual is calculated between the KNN-weighted characterization value of the shear strength and the calculated shear strength value corresponding to the candidate parameter, and the shear strength residual is characterized as follows:

[0023] in, For the first Shear strength residuals at each target CPT query point For the first The KNN weighted representation of the shear strength of each target CPT query point. For the first Calculated candidate shear strength values ​​for each target CPT query point; A weakly constrained likelihood function for shear strength is constructed by combining the calculated candidate shear strength values. This weakly constrained likelihood function for shear strength is characterized as follows:

[0024] in, The set of observation data for the target CPT query point The Middle Observational data for each target CPT query point For the first Weakly constrained likelihood function of shear strength for each target CPT query point The error scale parameter is the likelihood function of weak shear strength constraint. For the first Shear strength residuals of each target CPT query point.

[0025] This invention constructs a weakly constrained likelihood function for shear strength by using the KNN weighted representation value of shear strength and the calculated shear strength value corresponding to the candidate parameters. This weak constraint does not replace the main CPT likelihood function, but serves as a physical rationality check term to flexibly screen the combination of candidate parameters, so that the posterior update is controlled by the continuous response of CPT and does not deviate from the mechanical range defined by the strength information of neighboring boreholes.

[0026] Optionally, constructing the joint likelihood function based on the two likelihood functions includes: weighting and combining the main CPT likelihood function with the shear strength weak constraint likelihood function to obtain the joint likelihood function, characterized as follows:

[0027] in, The set of observation data for the target CPT query point. For the joint likelihood function, The set of observation data for the target CPT query point The Middle Observational data for each target CPT query point For the first The principal CPT likelihood function for each target CPT query point For the first Weakly constrained likelihood function of shear strength for each target CPT query point The weights of the main CPT likelihood function, The weights are for the likelihood function with weak shear strength constraints. and Values ​​can be determined based on experience, or verified by using a posterior parameter optimization scoring system, posterior confidence interval coverage, posterior distribution shrinkage rate, or the results of a borehole independence test.

[0028] This invention obtains a joint likelihood function by weighting the main CPT likelihood function and the shear strength weak constraint likelihood function. Parameter combinations in the prior samples that can simultaneously explain the CPT normalized response and satisfy the shear strength weak constraint are given higher weights, while parameter combinations that cannot explain the CPT response or that clearly conflict with the KNN weighted representation of shear strength are given lower weights, thereby causing the posterior distribution to shrink compared to the prior distribution.

[0029] Optionally, the step of updating the prior distribution to a posterior distribution using Bayesian update and outputting the posterior parameters and uncertainty evaluation results includes: updating the prior distribution to a posterior distribution using Bayesian update, wherein the posterior distribution is characterized as follows:

[0030] in, The set of observation data for the target CPT query point. To incorporate the posterior distribution updated by Bayes, For the joint likelihood function, Candidate parameters The joint prior density; The posterior parameters include the posterior mean of cohesion and internal friction angle. The posterior distribution output includes the posterior median, the highest posterior density interval, the 95% confidence interval, and the joint posterior distribution. The uncertainty evaluation result is one or more of the following evaluation results obtained by using the 95% confidence interval as the parameter value range: posterior distribution shrinkage rate, KNN weighted parameter confidence interval coverage, or the result of the borehole independence test.

[0031] This invention updates the prior distribution to the posterior distribution through Bayesian update and outputs the joint posterior distribution, marginal posterior distribution, confidence interval coverage, posterior distribution shrinkage rate, and the results of the independent test of the left-one borehole. The posterior distribution can effectively shrink compared to the prior distribution, thereby providing strength parameter estimation results with confidence intervals for engineering design.

[0032] Secondly, the present invention also provides a system for rapidly determining soil strength parameters based on CPT data, comprising: a processor, an input device, an output device, and a memory, wherein the processor, the input device, the output device, and the memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute the method for rapidly determining soil strength parameters based on CPT data as described in any of the preceding aspects of the present invention.

[0033] The present invention provides a rapid soil strength parameter determination system based on CPT data, which is compact, stable, highly integrated, and simple in construction. It can stably execute the rapid soil strength parameter determination method based on CPT data provided in the preceding part of the present invention, further improving the overall applicability and practical application capability of the present invention. Attached Figure Description

[0034] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 This is a flowchart illustrating a method for rapidly determining soil strength parameters based on CPT data according to an embodiment of the present invention. Figure 2 This is a schematic diagram of a system for rapidly determining soil strength parameters based on CPT data, according to an embodiment of the present invention. Detailed Implementation

[0036] Specific embodiments of the present invention will now be described in detail. It should be noted that the embodiments described herein are for illustrative purposes only and are not intended to limit the invention. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the invention. However, it will be apparent to those skilled in the art that these specific details are not necessary to practice the invention. In other instances, well-known circuits, software, or methods have not been specifically described to avoid obscuring the invention.

[0037] Throughout this specification, references to "an embodiment," "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in connection with that embodiment or example is included in at least one embodiment of the invention. Therefore, the phrases "in an embodiment," "in an embodiment," "an example," or "an example" appearing in various places throughout the specification do not necessarily refer to the same embodiment or example. Furthermore, specific features, structures, or characteristics can be combined in one or more embodiments or examples in any suitable combination and / or sub-combination. Moreover, those skilled in the art will understand that the illustrations provided herein are for illustrative purposes and are not necessarily drawn to scale.

[0038] It should be noted in advance that, in one alternative embodiment, except for independent descriptions, the same symbols or letters appearing in all formulas have the same meaning.

[0039] In one optional embodiment, please refer to Figure 1 This invention provides a method for rapidly determining soil strength parameters based on CPT data, the method comprising the following steps: S1. Obtain CPT data and borehole data, and process them to obtain standardized data.

[0040] In this embodiment, static cone penetration test (CPT) data, borehole data, and corresponding spatial coordinates and depth information of the target site are acquired.

[0041] Furthermore, coordinate unification, depth benchmark correction, soil layer screening, missing value processing, and outlier removal were performed on the CPT data and borehole data.

[0042] Specifically, in this embodiment, the target soil layer is first screened based on the soil layer name or soil layer number, such as silty clay layer. Then, the CPT test depth and borehole sampling depth are projected onto a unified depth coordinate system, so that the target CPT query point has consistent spatial and stratigraphic semantics with the adjacent borehole samples.

[0043] Furthermore, standardized data is generated, including standardized CPT data and standardized borehole data. The standardized CPT data includes the coordinates of the CPT borehole locations and the test depth within the target site. CPT response value Total overburden stress and effective overburden stress Standardized borehole data includes borehole coordinates, sampling depth, soil type, cohesion, internal friction angle, and shear strength of the borehole sample.

[0044] S2. Construct a local nearest neighbor sample set based on the spatial relationship between the target CPT query point and the borehole point, and obtain the KNN weighted representation value.

[0045] Specifically, in this embodiment, the planar distance or spatial distance between the target CPT query point and each borehole point is calculated, and the K boreholes with the smallest distance are selected as the local neighbor sample set.

[0046] Furthermore, the weights of each neighboring sample are determined using inverse distance weights, exponentially decaying weights, or other monotonically decreasing weight functions. For the th... For each target CPT query point, the KNN weighted representation value of cohesion can be obtained according to the weight. KNN weighted representation of the internal friction angle KNN weighted characterization of shear strength .

[0047] S3. Fit the edge distribution of the KNN weighted characterization values ​​of cohesion and internal friction angle respectively, select the optimal edge distribution and construct the prior distribution.

[0048] Specifically, in this embodiment, the KNN weighted representation values ​​of cohesion and internal friction angle are subjected to one or more candidate distributions, namely normal distribution, log-normal distribution, generalized extreme value distribution, Weibull distribution, Johnson-SU distribution, Beta distribution, or Student-t distribution.

[0049] Furthermore, the AIC criterion is used to select the optimal marginal distribution, and a candidate parameter sample set is generated from the optimal marginal distribution. ,in, For optimal edge distribution cohesion, Let be the internal friction angle of the optimal edge distribution. Therefore, the prior distribution is characterized as:

[0050] in, For cohesion Edge prior density, Angle of internal friction Edge prior density, Candidate parameters The joint prior density; Furthermore, candidate parameters are obtained by sampling from the prior distribution. and candidate parameters joint prior density ,in, For a priori cohesion, This is the a priori internal friction angle.

[0051] S4. Construct a normalized CPT response based on the standardized CPT data in the standardized data, and construct the main CPT likelihood function.

[0052] Specifically, in this embodiment, the normalized CPT response satisfy:

[0053] in, For the first Normalized CPT response value for each target CPT query point For the first CPT response value for each target CPT query point For the first The total overlying stress of each target CPT query point For the first Effective overlying stress of each target CPT query point Standard atmospheric pressure, The stress normalization index is denoted as .

[0054] Furthermore, the observed response and candidate parameters of the normalized CPT response are calculated. The error between the corresponding predicted responses is characterized as follows:

[0055] in, For the first The error between the observed response and the predicted response for each target CPT query point For the first Normalized response of each target CPT query point The observed response can be Taking the logarithm, we get, that is , For the first Candidate parameters for each target CPT query point The corresponding predicted response, This can be expressed as a priori cohesion. a priori internal friction angle and depth The linear or nonlinear function can also be given by an empirical regression model, a local surrogate model, or a pre-calibrated response surface model.

[0056] Furthermore, the principal CPT likelihood function is constructed by combining the error, and it is characterized as follows:

[0057] in, The set of observation data for the target CPT query point The Middle Observational data for each target CPT query point For the first The principal CPT likelihood function for each target CPT query point The error scale parameter of the main CPT likelihood function, For the first The error between the observed response and the predicted response of each target CPT query point.

[0058] S5. Construct a weakly constrained likelihood function for shear strength using the KNN weighted representation of shear strength.

[0059] Specifically, in this embodiment, candidate parameters are calculated based on the Mohr-Coulomb strength criterion. In the Calculated candidate shear strength values ​​for each target CPT query point It is characterized as follows: ,in, For a priori cohesion, For the first Effective overlying stress of each target CPT query point This is the a priori internal friction angle.

[0060] Furthermore, the shear strength residual is calculated between the KNN-weighted representation of shear strength and the calculated shear strength value corresponding to the candidate parameters. The shear strength residual is characterized as follows:

[0061] in, For the first Shear strength residuals at each target CPT query point For the first The KNN weighted representation of the shear strength of each target CPT query point. For the first Calculated candidate shear strength values ​​for each target CPT query point.

[0062] Furthermore, a weakly constrained likelihood function for shear strength is constructed by combining the calculated values ​​of candidate shear strength. The weakly constrained likelihood function for shear strength is characterized as follows:

[0063] in, The set of observation data for the target CPT query point The Middle Observational data for each target CPT query point For the first Weakly constrained likelihood function of shear strength for each target CPT query point The error scale parameter is the likelihood function with weak shear strength constraints. For the first Shear strength residuals of each target CPT query point.

[0064] S6. Construct a joint likelihood function based on the two likelihood functions, and combine it with Bayesian update to update the prior distribution to the posterior distribution, and output the posterior parameters and uncertainty evaluation results.

[0065] Specifically, in this embodiment, the principal CPT likelihood function and the shear strength weak constraint likelihood function are weighted and combined to obtain the joint likelihood function, which is characterized as follows:

[0066] in, The set of observation data for the target CPT query point. For the joint likelihood function, The set of observation data for the target CPT query point The Middle Observational data for each target CPT query point For the first The principal CPT likelihood function for each target CPT query point For the first Weakly constrained likelihood function of shear strength for each target CPT query point The weights of the main CPT likelihood function, The weights are for the likelihood function with weak shear strength constraints. and Values ​​can be determined based on experience, or verified through PPS parameter optimization scoring, posterior confidence interval coverage, posterior distribution shrinkage rate, or the results of the borehole independence test.

[0067] Furthermore, by combining Bayesian update, the prior distribution is updated to the posterior distribution, which is represented as:

[0068] in, The set of observation data for the target CPT query point. To incorporate the posterior distribution updated by Bayes, For the joint likelihood function, Candidate parameters The joint prior density.

[0069] Furthermore, the output posterior parameters include the posterior mean of cohesion and internal friction angle. The posterior distribution output includes the posterior median, the highest posterior density interval, the 95% confidence interval, and the joint posterior distribution. The uncertainty evaluation result is one or more of the following: the posterior distribution shrinkage rate, the KNN weighted parameter confidence interval coverage, or the result of the borehole independence test, with the 95% confidence interval as the parameter value range.

[0070] In an optional embodiment, a leave-one-hole validation method is used to test the independence of the parameter determination results. Each time, one hole is removed as the validation object, and the remaining holes are used to reconstruct the nearest neighbor samples, edge priors, principal CPT likelihood, and so on. Weak constraints and posterior distributions are used, and the obtained posterior parameter results are compared with the test parameters of the removed boreholes to determine the stability and engineering applicability of the method for parameter determination at unknown borehole locations.

[0071] In another alternative embodiment, a posterior parameter optimization scoring system is used for... and The selection process is as follows. The posterior parameter optimization scoring system can comprehensively consider the degree of posterior distribution contraction and the coverage of the KNN weighted parameters by the posterior confidence interval, so that the final weights can both reflect the effective update of the prior distribution by CPT observations and avoid excessive contraction that leads to the unreasonable exclusion of local borehole information.

[0072] In other embodiments, the distance metric for KNN can be Euclidean distance, three-dimensional spatial distance, weighted distance considering formation interface constraints, or layered distance considering engineering zones. The K value can be fixed to a preset value, or it can be adaptively determined through cross-validation, leave-one-hole validation, or information criteria.

[0073] In other embodiments, the predicted response in the main CPT likelihood function The parameters can be determined by linear regression, nonlinear regression, random forest, support vector regression, neural network surrogate models, or semi-empirical formulas based on theoretical support mechanisms. This is as long as the model can characterize the candidate parameters. CPT normalized response The relationships between them should all fall within the protection scope of this invention.

[0074] In other embodiments, Weak constraints can also be used in conjunction with shear strength proxy relations constructed from CPT responses, for example by... ,depth or effective stress predict Response, then with candidate parameters corresponding A comparison is made. This change does not alter the essence of this invention, which utilizes weak shear strength constraints to screen for physically reasonable combinations of parameters.

[0075] In one optional embodiment, please refer to Figure 2 The present invention also provides a system for rapidly determining soil strength parameters based on CPT data. This system includes a processor, an input device, an output device, and a memory. The processor, input device, output device, and memory are interconnected. The memory stores a computer program, which includes program instructions. The processor is configured to call the program instructions and execute relevant steps in relevant embodiments of the method for rapidly determining soil strength parameters based on CPT data.

[0076] This invention provides a rapid soil strength parameter determination system based on CPT data. The functional components can be integrated into a single processing unit, exist as separate physical entities, or be integrated into a single unit. The integrated components can be implemented in hardware or as software functions.

[0077] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A method for rapidly determining soil strength parameters based on CPT data, characterized in that, include: Acquire CPT data and borehole data, and process them to obtain standardized data; A local nearest neighbor sample set is constructed based on the spatial relationship between the target CPT query point and the borehole point, and the KNN weighted representation value is obtained. Marginal distribution fitting is performed on the KNN weighted characterization values ​​of cohesion and internal friction angle respectively. The optimal marginal distribution is selected and a prior distribution is constructed. The KNN weighted characterization values ​​of cohesion and internal friction angle are respectively subjected to one or more candidate distributions among normal distribution, log-normal distribution, generalized extreme value distribution, Weibull distribution, Johnson-SU distribution, Beta distribution or Student-t distribution. The optimal marginal distribution is selected using the AIC criterion, and a candidate parameter sample set is generated from the optimal marginal distribution. ,in, For optimal edge distribution cohesion, The internal friction angle is the one with the optimal edge distribution. The prior distribution is characterized as follows: in, The cohesive force Edge prior density, The internal friction angle Edge prior density, Candidate parameters The joint prior density; Candidate parameters are obtained by sampling from the prior distribution. and the candidate parameters joint prior density ,in, For a priori cohesion, This is the a priori internal friction angle; Based on the standardized CPT data in the standardized data, a normalized CPT response is constructed, and a master CPT likelihood function is built. The normalized CPT response satisfies: in, For the first Normalized CPT response value for each target CPT query point For the first CPT response value for each target CPT query point For the first The total overlying stress of each target CPT query point For the first Effective overlying stress of each target CPT query point Standard atmospheric pressure, The stress normalization index; Calculate the observed response of the normalized CPT response and the candidate parameters. The error between the corresponding predicted responses, wherein the error is characterized as: in, For the first The error between the observed response and the predicted response for each target CPT query point For the first Normalized response of each target CPT query point The observed response can be Taking the logarithm, we get, that is , For the first Candidate parameters for each target CPT query point The corresponding predicted response; The principal CPT likelihood function is constructed by combining the aforementioned errors and is characterized as follows: in, The set of observation data for the target CPT query point The Middle Observational data for each target CPT query point For the first The principal CPT likelihood function for each target CPT query point The error scale parameter of the main CPT likelihood function, For the first The error between the observed response and the predicted response of each target CPT query point; A weakly constrained likelihood function for shear strength is constructed using the KNN-weighted representation of shear strength, wherein the candidate parameters are calculated according to the Mohr-Coulomb strength criterion. In the Calculated candidate shear strength values ​​for each target CPT query point Characterized as: ,in, For a priori cohesion, For the first Effective overlying stress of each target CPT query point This is the a priori internal friction angle; Calculate the shear strength residual between the KNN-weighted representation of the shear strength and the calculated shear strength value corresponding to the candidate parameter, wherein the shear strength residual is characterized as follows: in, For the first Shear strength residuals at each target CPT query point For the first The KNN weighted representation of the shear strength of each target CPT query point. For the first Calculated candidate shear strength values ​​for each target CPT query point; A weakly constrained likelihood function for shear strength is constructed by combining the calculated candidate shear strength values. This weakly constrained likelihood function for shear strength is characterized as follows: in, The set of observation data for the target CPT query point The Middle Observational data for each target CPT query point For the first Weakly constrained likelihood function of shear strength for each target CPT query point The error scale parameter is the likelihood function with weak shear strength constraints. For the first Shear strength residuals at each target CPT query point; A joint likelihood function is constructed based on the two likelihood functions, and the prior distribution is updated to the posterior distribution by combining Bayesian update, outputting the posterior parameters and uncertainty evaluation results.

2. The method for rapidly determining soil strength parameters based on CPT data according to claim 1, characterized in that, The process of acquiring CPT data and borehole data, and then processing them to obtain standardized data, includes: Acquire static cone penetration test (CPT) data, borehole data, and corresponding spatial coordinates and depth information of the target site; The CPT data and the borehole data are subjected to coordinate unification, depth benchmark correction, soil layer screening, missing value processing and outlier removal to form standardized data; The standardized data includes the standardized CPT data and the standardized borehole data. The standardized CPT data includes the borehole coordinates, test depth, CPT response value, total overburden stress, and effective overburden stress of the CPT boreholes in the target site. The standardized borehole data includes the borehole coordinates, sampling depth, soil type, cohesion, internal friction angle, and shear strength of the borehole samples.

3. The method for rapidly determining soil strength parameters based on CPT data according to claim 1, characterized in that, The process of constructing a local nearest neighbor sample set based on the spatial relationship between the target CPT query point and the borehole point, and obtaining the KNN weighted representation value, includes: Calculate the planar or spatial distance between the target CPT query point and each borehole point, and select the K boreholes with the smallest distance as the local neighbor sample set; The weights of each neighboring sample are determined using the inverse distance weight, exponential decay weight, or other monotonically decreasing weight functions, and the KNN weighted characterization values ​​of cohesion, internal friction angle, and shear strength are obtained according to the weights.

4. The method for rapidly determining soil strength parameters based on CPT data according to claim 1, characterized in that, The construction of the joint likelihood function based on two likelihood functions includes: The combined likelihood function obtained by weighting the principal CPT likelihood function and the shear strength weak constraint likelihood function is as follows: in, The set of observation data for the target CPT query point. For the joint likelihood function, The set of observation data for the target CPT query point The Middle Observational data for each target CPT query point For the first The principal CPT likelihood function for each target CPT query point For the first Weakly constrained likelihood function of shear strength for each target CPT query point The weights of the main CPT likelihood function, The weights are the values ​​for the likelihood function with weak shear strength constraints.

5. The method for rapidly determining soil strength parameters based on CPT data according to claim 1, characterized in that, The step of updating the prior distribution to a posterior distribution using Bayesian update and outputting posterior parameters and uncertainty evaluation results includes: The prior distribution is updated to a posterior distribution using Bayesian update, and the posterior distribution is characterized as follows: in, The set of observation data for the target CPT query point. To incorporate the posterior distribution updated by Bayes, For the joint likelihood function, Candidate parameters The joint prior density; The posterior parameters include the posterior mean of cohesion and internal friction angle. The posterior distribution output includes the posterior median, the highest posterior density interval, the 95% confidence interval, and the joint posterior distribution. The uncertainty evaluation result is one or more of the following evaluation results obtained by using the 95% confidence interval as the parameter value range: posterior distribution shrinkage rate, KNN weighted parameter confidence interval coverage, or the result of the borehole independence test.

6. A system for rapidly determining soil strength parameters based on CPT data, characterized in that, include: The system includes a processor, an input device, an output device, and a memory, all interconnected. The memory stores a computer program, which includes program instructions. The processor is configured to invoke the program instructions to execute a method for rapidly determining soil strength parameters based on CPT data, as described in any one of claims 1 to 5.

Images