Disclosure of Invention
The invention aims to overcome the problems existing in the prior art and greatly improve the technical effect on the basis of the prior art; to this end, the present invention provides a method of analyzing soil horizon profile from drilling data, the method comprising:
drilling holes for the first time at proper positions of engineering sites, and carrying out sectional treatment on a drilled soil sample according to drilling results to divide the drilled soil sample into n sections; the principle of the segmentation treatment is that drilling results are segmented as much as possible according to the length average, and the same segmentation only contains one soil sample;
calculating the average value of each parameter in the soil sample of each segment, taking the calculated average value of each parameter as the final parameter extraction value of the soil sample of each segment, wherein the extracted parameters comprise: soil density, elastic modulus, shear modulus, specific volume, porosity and soil sample classification;
generating a parameter table influencing the classification of the total soil sample through the extracted parameters, and classifying the parameter table through a naive Bayes classifier; the classification is carried out by a naive Bayes classifierThe method comprises the following steps: a) Firstly, estimating the probability P (c) of each soil sample classification; b) Calculating the conditional probability P (x i C), each attribute refers to each parameter of the classified soil sample; c) Substituting the obtained conditional probability into a Bayesian classification criterion, and determining the corresponding relation between the parameter range and the soil sample classification; the Bayesian classification criterion is as follows:wherein x is a parameter vector formed by soil parameters of each segment, and x is i For each element in x, c represents the classification of each segmented soil sample, P (c) is the probability of each soil sample classification, d is the number of parameters in the segment other than the soil sample classification, argmax represents the value of c when the formula reaches the maximum, i.e., h nb (x) Representing classification results, namely representing the correspondence between the classification of the soil samples and the parameter vector x;
according to the result of the Bayes classification rule, acquiring the value range of each parameter in each classified soil sample, and generating a value range table of each classified soil sample parameter;
four holes are drilled near the periphery of the first hole, and an included angle formed by each three holes is 90 degrees; and extracting parameters of each classified soil sample according to the steps, comparing the extracted parameters of each classified soil sample with the obtained value range table of each classified soil sample parameter, determining whether drilling data with the same height are in the value range of the same classified soil sample parameter with the drilling data for the first time, and further determining whether the drilling performed again is consistent with the soil layer distribution result obtained by the drilling for the first time.
Further, the step of carrying out the sectional treatment on the drilled soil sample comprises the following steps: segmenting from the distribution condition of the drilled soil sample from top to bottom in the original engineering site; according to the observed distribution of the soil samples, as many segments as possible are averaged according to length, and each segment only contains one soil sample, i.e. each segment corresponds to one soil sample classification.
Further, the generating the parameter table affecting the total soil sample classification includes: the generated parameter table affecting the total soil sample classification comprises parameters of each segment and soil sample classifications, and the parameters of each segment correspond to one soil sample classification.
Further, the estimating the probability P (c) of each soil sample classification includes: the probability P (c) of each soil sample classification is calculated from the number of segments processed per segment of the soil sample classification, where c represents the classification of the soil sample.
Further, the method calculates the conditional probability P (x i The c) includes: x is x i Representative parameters are soil density, elastic modulus, shear modulus, specific volume and porosity, respectively; the passing bayesian classification criterion further comprises: h is a nb (x) If the probability of which classification is the largest, it is determined that the soil sample belongs to the classification with the largest probability.
Further, the obtaining the value range of each parameter in each classified soil sample includes: and determining the value range of each parameter in each classified soil sample according to the result of the Bayesian classification criterion and the value of each parameter.
Further, the comparing the extracted each classified soil sample parameter with the obtained value range table of each classified soil sample parameter comprises: according to the drilling result, determining whether each parameter corresponding to drilling data with the same height as that of the first drilling is in the value range of the same classified soil sample parameter; if all parameters corresponding to drilling data with the same height of the secondary drilling are in the value range of the parameters of the classified soil samples, determining that soil layer distribution results obtained by the secondary drilling and the primary drilling are consistent; if all parameters corresponding to drilling data with the same height of the secondary drilling are not in the value range of the parameters of the classified soil samples, determining that soil layer distribution results obtained by the secondary drilling and the primary drilling are inconsistent.
The beneficial effects of the invention are as follows:
the invention provides a method for analyzing soil layer distribution by drilling data, which comprises the steps of firstly, carrying out drilling detection for the first time on an engineering area, segmenting a soil sample detected by drilling, and then carrying out classification treatment on the segmented soil sample by a Bayesian classifier to obtain a value range table of each parameter of each soil sample classification; drilling the area adjacent to the first drilling, substituting the drilling data into a value range table of each parameter of each soil sample classification, and determining the classification of the drilling soil samples; comparing the data of the equal height of the drilling soil sample with the data of the first drilling, and determining whether the drilling performed again is consistent with the soil layer distribution result obtained by the first drilling; according to the method, the distribution situation of soil layers of the whole engineering site can be approximately known only through drilling detection, the operation is simple and convenient, and fitting analysis of the distribution of soil layers in the later period is facilitated; the adopted naive Bayes classification rule is used for processing the parameters of soil layer distribution, and has more scientific basis than direct observation by naked eyes.
Detailed Description
Specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the particular embodiments presented herein are illustrative and explanatory only and are not restrictive of the invention.
It should be noted that numerous specific details are set forth in the following description in order to provide a thorough understanding of the present invention, however, that other embodiments of the invention and variations thereof are possible and, therefore, the scope of the invention is not limited by the specific examples disclosed below.
As shown in fig. 1, a method of analyzing soil horizon distribution according to the drilling data according to the embodiment of the invention includes: step S100, drilling holes for the first time at proper positions of engineering sites, and carrying out sectional treatment on a drilled soil sample according to drilling results to divide the drilled soil sample into n sections; the principle of the segmentation treatment is that drilling results are segmented as much as possible according to the length average, and the same segmentation only contains one soil sample; step S101, the average value of each parameter in each segmented soil sample is obtained, the obtained average value of each parameter is used as the final soil sample parameter extraction value of each segment, and the extracted parameters comprise: soil density, elastic modulus, shear modulus, specific volume, porosity and soil sample classification; step S102, generating a parameter table influencing the classification of the total soil sample through the extracted parameters, and classifying the parameter table through a naive Bayes classifier; step S103, obtaining the value range of each parameter in each classified soil sample according to the result of the Bayes classification rule, and generating a value range table of each classified soil sample parameter; step S104, four holes are drilled on the periphery adjacent to the first hole, and an included angle formed by every three holes is 90 degrees; and extracting parameters of each classified soil sample according to the steps, comparing the extracted parameters of each classified soil sample with the obtained value range table of each classified soil sample parameter, determining whether drilling data with the same height are in the value range of the same classified soil sample parameter with the drilling data for the first time, and further determining whether the drilling performed again is consistent with the soil layer distribution result obtained by the drilling for the first time.
In the embodiment, specifically, the method comprises the steps of firstly selecting a proper area of an engineering area to perform primary drilling detection, segmenting a soil sample detected by primary drilling, and then classifying the segmented soil sample through a Bayesian classifier to obtain a value range table of each parameter of each soil sample classification; then, drilling the area adjacent to the first drilling, substituting the drilling data into a value range table of each parameter of each soil sample classification, and determining the classification of the drilling soil samples; and comparing the data of the equal height of the drilled soil sample with the data of the first drilled hole, and determining whether the repeated drilled hole is consistent with the soil layer distribution result obtained by the first drilled hole.
Step S100, drilling holes for the first time at proper positions of engineering sites, and carrying out sectional treatment on a drilled soil sample according to drilling results to divide the drilled soil sample into n sections; the principle of the segmentation treatment is to segment the drilling result as much as possible in average length, and the same segment only contains one soil sample.
Specifically, the step of carrying out sectional treatment on the drilled soil sample comprises the following steps: segmenting from the distribution condition of the drilled soil sample from top to bottom in the original engineering site; according to the observed distribution of the soil samples, as many segments as possible are averaged according to length, and each segment only contains one soil sample, i.e. each segment corresponds to one soil sample classification.
Step S101, the average value of each parameter in each segmented soil sample is obtained, the obtained average value of each parameter is used as the final soil sample parameter extraction value of each segment, and the extracted parameters comprise: soil density, elastic modulus, shear modulus, specific volume, porosity, and soil sample classification.
Specifically, according to the overall distribution condition of the soil sample of each segment, the average value of each parameter in the soil sample of each segment is calculated, and the final parameter for determining the soil sample classification of each segment is determined; the parameters include: soil density, elastic modulus, shear modulus, specific volume, porosity, and soil sample classification.
Step S102, generating a parameter table influencing the classification of the total soil sample through the extracted parameters, and classifying the parameter table through a naive Bayes classifier.
Specifically, the generating the parameter table affecting the total soil sample classification includes: the generated parameter table affecting the total soil sample classification comprises parameters of each segment and soil sample classifications, and the parameters of each segment correspond to one soil sample classification.
In the foregoing embodiment, specifically, the classifying by the naive bayes classifier includes: a) Firstly, estimating the probability P (c) of each soil sample classification; b) Calculating the conditional probability P (x i C), each attribute refers to each parameter of the classified soil sample; c) Substituting the obtained conditional probability into a Bayesian classification criterion, and determining the corresponding relation between the parameter range and the soil sample classification; the Bayesian classification criterion is as follows:wherein x is a parameter vector formed by soil parameters of each segment, and x is i For each element in x, c represents the classification of each segmented soil sample, P (c) is the probability of each soil sample classification, d is the number of parameters in the segment other than the soil sample classification, argmax represents the value of c when the formula reaches the maximum, i.e., h nb (x) The classification result is represented by the correspondence between the classification of the soil sample and the parameter vector x.
In the above-described embodiment, specifically, the probability P (c) of each soil sample classification is calculated from the number of segments of the segmentation process of each soil sample classification, where c represents the classification of the soil sample.
In the above embodiment, in particular, the calculation of the conditional probability P (x i The c) includes: x is x i Representative parameters are soil density, elastic modulus, shear modulus, specific volume and porosity, respectively; the bayesian classification criterion further comprises: h is a nb (x) If the probability of the classification is the largest, the soil sample is judged to belong to the classification with the largest probability, namely, the corresponding relation between the classification of the soil sample and the parameter vector x is shown.
Step S103, according to the result of the Bayes classification rule, obtaining the value range of each parameter in each classified soil sample, and generating a value range table of each classified soil sample parameter.
Specifically, the obtaining the value range of each parameter in each classified soil sample includes: and determining the value range of each parameter in each classified soil sample according to the corresponding relation between the result of the Bayesian classification criterion and the value of each parameter.
In the above embodiment, preferably, the present invention may further obtain more parameter values corresponding to each classified soil sample by looking up the data, expand the value range of the parameters, classify all the parameter values by using a naive bayes classifier, and determine the corresponding relationship between each soil sample classification and the expanded parameters.
Step S104, four holes are drilled on the periphery adjacent to the first hole, and an included angle formed by every three holes is 90 degrees; and extracting parameters of each classified soil sample according to the steps, comparing the extracted parameters of each classified soil sample with the obtained value range table of each classified soil sample parameter, determining whether drilling data with the same height are in the value range of the same classified soil sample parameter with the drilling data for the first time, and further determining whether the drilling performed again is consistent with the soil layer distribution result obtained by the drilling for the first time.
Specifically, the comparing the extracted parameters of each classified soil sample with the obtained value range table of each classified soil sample comprises: according to the drilling result, determining whether each parameter corresponding to drilling data with the same height as that of the first drilling is in the value range of the same classified soil sample, and if each parameter corresponding to drilling data with the same height of the second drilling is in the value range of the same classified soil sample parameter as that of the first drilling, determining that the soil layer distribution results obtained by the second drilling and the first drilling are consistent; if all parameters corresponding to drilling data with the same height of the secondary drilling are not in the value range of the parameters of the classified soil samples, determining that soil layer distribution results obtained by the secondary drilling and the primary drilling are inconsistent; if the soil layers are inconsistent, the soil layer distribution of the engineering site is complex, the early detection is required to be enlarged, and the safety of later construction is ensured.
In the above embodiment, it is preferable that a plurality of holes, including but not limited to 4 holes, be selectively drilled in the surrounding vicinity of the first hole; selecting positions of a plurality of drilling holes to form a regular polygon shape as much as possible; analyzing the results of the plurality of drilling holes, and determining whether each parameter corresponding to the drilling data of the same height of the results of the plurality of drilling holes compared with the results of the first drilling holes is in the value range of the same classified soil sample; if the drilling results are in the same value range, the same heights of the drilling results and the first drilling results are the same soil sample, and the soil layer distribution results obtained by the drilling results are consistent; the soil layer distribution results obtained by drilling the hole again are consistent with the soil layer distribution results obtained by drilling the hole for the first time, but the parameter values are different, and the overall soil layer distribution condition of the engineering site can be determined by combining the data fitting analysis through the different parameter values.
In the foregoing embodiment, it should be noted that, preferably, the first drilling is performed at a suitable location of the engineering site, and the first drilling is performed by selecting a plurality of suitable areas of the engineering site at the same time, that is, the first drilling is not performed, and only one drilling is selected; after the first drilling, a plurality of drilling holes are formed in the surrounding adjacent area of each first drilling hole, and the results of the plurality of drilling holes are compared with the corresponding results of the first drilling holes, so that the final results are obtained.
It is to be understood that the above-described embodiments are one or more embodiments of the invention, and that many other embodiments and variations thereof are possible in accordance with the invention; variations and modifications of the invention, which are intended to be within the scope of the invention, will occur to those skilled in the art without any development of the invention.