Soil load analysis method and device

Abstract

The invention discloses a soil load analysis method and device. The form of the soil structure at the position of the soil load to be studied is converted into a matrix. Ellipsoids are established at stress positions according to soil categories, and the influence of different stress angles on soil loads can be considered. And obtaining the soil load through big data and deep learning according to the matrix indicating the ellipsoid position and the soil overall structure. The technical effect that different stress positions, stress angles and stress pressures are accurately judged, and different soil loads are obtained for different types and different forms of soil is achieved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular, to a method and device for soil load analysis. Background Art

[0002] In geotechnical engineering, bearing capacity refers to the ability of soil to support the load applied to the ground. The bearing capacity of soil is the maximum average contact pressure between the foundation and the soil, and this pressure should not cause shear failure of the soil, so that soil deformation does not occur. Soil load represents the load that the soil can bear. Currently, the analysis of soil load is usually carried out by analyzing the structure of the soil itself, such as the cohesion influence coefficient of the soil, the shape coefficient of the soil, and the unit weight influence coefficient of the soil, without considering the overall structure and the direction of force of the soil, resulting in inaccurate analysis of soil load. Summary of the Invention

[0003] The purpose of the present invention is to provide a method and device for soil load analysis to solve the above problems existing in the prior art.

[0004] In a first aspect, an embodiment of the present invention provides a method for soil load analysis, including:

[0005] Obtain the soil type, soil surface shape matrix, force application position, force application angle, and force application pressure; the soil type represents the type of soil; the soil surface shape matrix is a matrix including the three-dimensional structure of the soil area; the force application pressure represents the pressure exerted by an object on the soil; the force application angle represents the angle of the force application pressure; the force application position represents the position of the force application pressure on the soil;

[0006] Construct a three-dimensional coordinate axis according to the soil surface shape matrix to obtain a soil three-dimensional coordinate system; the soil three-dimensional coordinate system includes an abscissa, an ordinate, and a vertical coordinate;

[0007] According to the soil three-dimensional coordinate system, use the cosine value of the force application angle as the slope, and take the straight line passing through the force application position as the force application straight line;

[0008] Through the soil force detection network, based on the soil type and force application pressure, judge the pressure change in the soil three-dimensional coordinate system to obtain the soil straight line position; the soil straight line position represents the position where the force application threshold is reached;

[0009] Based on the soil type, force application position, and soil straight line position, construct an ellipsoid on the soil surface shape matrix in the soil three-dimensional coordinate system to obtain a soil force matrix;

[0010] Through the shape discrimination network, based on the soil surface shape matrix, judge the force application state of the soil to obtain the soil load.

[0011] Optionally, the shape discrimination network determines the stress state of the soil based on the soil surface shape matrix to obtain the soil load, including:

[0012] The shape discrimination network includes a plurality of three-dimensional convolutional kernels;

[0013] Find the positions overlapping with the stress line in the soil surface shape matrix and mark them to obtain the stress positions;

[0014] Match the center points of the three-dimensional convolutional kernels with the stress positions and perform convolution sequentially with a step size of 1 to obtain feature vectors; the feature vectors contain the features of the soil state around the stress line;

[0015] Input the feature vectors into a classification network to judge the bearing capacity of the soil and obtain the soil load.

[0016] Optionally, constructing an ellipsoid on the soil surface shape matrix in the soil three-dimensional coordinate system based on the soil category, stress position, and soil line position to obtain the soil stress matrix, including:

[0017] Input the soil category into a soil category network to judge the ratio of the major axis to the minor axis of the ellipsoids corresponding to multiple stress threshold positions to obtain the major-to-minor axis ratio; the stress threshold positions represent multiple positions where the stress threshold is reached at different stress angles;

[0018] Based on the stress position, soil line position, and the major-to-minor axis ratio, make a judgment in the soil three-dimensional coordinate system to obtain the major axis length, middle axis length, and minor axis length;

[0019] Construct an ellipsoid in the soil three-dimensional coordinate system according to the major axis length, middle axis length, minor axis length, and stress position.

[0020] Optionally, the judgment in the soil three-dimensional coordinate system based on the stress position, soil line position, and the major-to-minor axis ratio to obtain the major axis length, middle axis length, and minor axis length includes:

[0021] Map the soil line position onto the abscissa, ordinate, and vertical coordinate of the soil three-dimensional coordinate system respectively to obtain the soil horizontal position, soil vertical position, and soil vertical position;

[0022] The major axis length, middle axis length, and minor axis length are obtained by the following ellipsoid formula:

[0023]

[0024] a÷b=d;

[0025] a=c;

[0026] Among them, a represents the length of the major axis; b represents the length of the minor axis; c represents the length of the central axis; d represents the ratio of the major axis to the minor axis; x represents the horizontal position of the soil; y represents the horizontal position of the soil; z represents the horizontal position of the soil.

[0027] Optionally, the length of the major axis of the ellipsoid is equal to the length of the central axis; the length of the minor axis is less than the length of the major axis;

[0028] The direction of the minor axis of the ellipsoid is the direction of the vertical coordinate of the soil three-dimensional coordinate system; the direction of the major axis of the ellipsoid is the direction of the horizontal coordinate of the soil three-dimensional coordinate system; the direction of the central axis of the ellipsoid is the direction of the vertical coordinate of the soil three-dimensional coordinate system.

[0029] Optionally, the method for obtaining the labeled data of the soil force detection network includes:

[0030] Obtain the force threshold and the force cancellation rate; the force threshold represents the pressure that changes the shape of the soil to a set state; the force cancellation rate is the plasticity index of the soil corresponding to the soil type;

[0031] According to the force cancellation rate, judge the change of the pressure, find the position where the pressure received on the force straight line is the force threshold and mark it to obtain the labeled soil straight line position.

[0032] Optionally, the method for obtaining the soil straight line position through the soil force detection network based on the soil type and the force pressure includes:

[0033] Normalize the soil type and the force pressure to obtain the normalized soil type value and the normalized force pressure value;

[0034] Construct a detection vector from the normalized soil type value and the normalized force pressure value;

[0035] Input the detection vector into the soil force detection network, extract the force relationship, and obtain the soil straight line position.

[0036] Optionally, the method for constructing an ellipsoid based on the length of the major axis, the length of the central axis, the length of the minor axis, and the force position in the soil three-dimensional coordinate system includes:

[0037] Use the force position as the midpoint of the ellipsoid and construct an ellipsoid in the soil three-dimensional coordinate system to obtain an ellipsoid region;

[0038] Retain the part of the ellipsoid region that coincides with the soil region in the soil surface shape matrix and delete the part that does not coincide.

[0039] Optionally, the method for constructing a three-dimensional coordinate axis based on the soil surface shape matrix to obtain a soil three-dimensional coordinate system includes:

[0040] Set the direction perpendicular to the ground as the direction of the vertical coordinate;

[0041] Set the direction parallel to the ground and parallel to the rows of the soil surface shape matrix as the direction of the horizontal coordinate;

[0042] Set the direction parallel to the ground and parallel to the columns of the soil surface shape matrix as the direction of the vertical coordinate;

[0043] Set the lower left corner of the soil surface shape matrix as the origin to obtain a three-dimensional soil coordinate system.

[0044] In a second aspect, an embodiment of the present invention provides a soil load analysis device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the method according to any one of claims 1-9 are implemented.

[0045] Compared with the prior art, the embodiments of the present invention achieve the following beneficial effects:

[0046] The embodiments of the present invention also provide a soil load analysis method and device. The method includes: obtaining soil type, soil surface shape matrix, force application position, force application angle, and force application pressure; the soil type represents the type of soil; the soil surface shape matrix is a matrix including a three-dimensional structure of a soil area; the force application pressure represents the pressure applied by an object to the soil; the force application angle represents the angle of the force application pressure; the force application position represents the position of the force application pressure on the soil; according to the soil surface shape matrix, construct three-dimensional coordinate axes to obtain a three-dimensional soil coordinate system; the three-dimensional soil coordinate system includes a horizontal coordinate, a vertical coordinate, and a vertical coordinate; according to the three-dimensional soil coordinate system, use the cosine value of the force application angle as the slope, and use the straight line passing through the force application position as the force application straight line; through a soil force detection network, based on the soil type and the force application pressure, judge the pressure change in the three-dimensional soil coordinate system to obtain the soil straight line position; the soil straight line position represents the position reaching the force application threshold; based on the soil type, the force application position, and the soil straight line position, construct an ellipsoid on the soil surface shape matrix in the three-dimensional soil coordinate system to obtain a soil force matrix; through a shape discrimination network, based on the soil surface shape matrix, judge the force application state of the soil to obtain the soil load.

[0047] In the present invention, the morphology of the soil structure at the position of the soil load to be studied is converted into a matrix. An ellipsoid is established according to the force application position and the soil type, which can consider the influence of different force application angles on the soil load. The matrix indicating the position of the ellipsoid and the overall soil structure is used to obtain the soil load through big data and deep learning. The technical effect of accurately judging different soil loads for different types and different morphologies of soil under different force application positions, force application angles, and force application pressures is achieved. Description of the Drawings

[0048] Figure 1 It is a flowchart of a soil load analysis method provided by an embodiment of the present invention. Specific embodiments

[0049] The present invention will be described in detail below with reference to the accompanying drawings.

[0050] Embodiment 1

[0051] As Figure 1 shown, an embodiment of the present invention provides a soil load analysis method, and the method includes:

[0052] S101: Obtain the soil type, soil surface shape matrix, force application position, force application angle, and force application pressure; the soil type represents the type of soil; the soil surface shape matrix is a matrix including the three-dimensional structure of the soil area; the force application pressure represents the pressure applied by an object to the soil; the force application angle represents the angle of the force application pressure; the force application position represents the position of the force application pressure on the soil;

[0053] Among them, the soil type includes clay, fine sand, loose gravel, soft rock, etc.

[0054] Among them, the soil surface shape matrix uses 1 to represent the position where the soil exists and 0 to represent the position where there is no soil. The soil surface shape matrix is equivalent to a perspective view; in this embodiment, the soil surface shape matrix is a matrix in the shape of an inverted triangular hole dug in the soil.

[0055] S102: Construct a three-dimensional coordinate axis according to the soil surface shape matrix to obtain a soil three-dimensional coordinate system; the soil three-dimensional coordinate system includes an abscissa, an ordinate, and a vertical coordinate.

[0056] S103: According to the soil three-dimensional coordinate system, use the cosine value of the force application angle as the slope, and use the straight line passing through the force application position as the force application straight line.

[0057] S104: Through the soil force detection network, based on the soil type and the force application pressure, judge the pressure change in the soil three-dimensional coordinate system to obtain the soil straight line position; the soil straight line position represents the position where the force application threshold is reached.

[0058] S105: Based on the soil type, the force application position, and the soil straight line position, construct an ellipsoid on the soil surface shape matrix in the soil three-dimensional coordinate system to obtain a soil force matrix;

[0059] S106: Through the shape discrimination network, based on the soil surface shape matrix, judge the force application state of the soil to obtain the soil load.

[0060] Optionally, the shape discrimination network determines the stress state of the soil based on the soil surface shape matrix to obtain the soil load, including:

[0061] The shape discrimination network includes a plurality of three-dimensional convolutional kernels;

[0062] Find the positions overlapping with the stress line in the soil surface shape matrix and mark them to obtain the stress positions.

[0063] Among them, the stress position is the position in the three-dimensional coordinate system.

[0064] Match the center point of the three-dimensional convolutional kernel with the stress position and perform convolution sequentially with a step size of 1 to obtain a feature vector; the feature vector contains the features of the soil state around the stress line.

[0065] Among them, the three-dimensional convolutional kernel is convolved with the matching positions in the soil surface shape matrix in ascending order of abscissa, ascending order of ordinate, and ascending order of vertical coordinate to obtain a feature vector.

[0066] Input the feature vector into the classification network to judge the bearing capacity of the soil and obtain the soil load.

[0067] Among them, in this embodiment, the classification network is a fully connected neural network (FCN).

[0068] Optionally, constructing an ellipsoid on the soil surface shape matrix in the soil three-dimensional coordinate system based on the soil category, stress position, and soil line position to obtain the soil stress matrix, including:

[0069] Input the soil category into the soil category network to judge the ratio of the major axis to the minor axis of the ellipsoid corresponding to multiple stress threshold positions to obtain the ratio of the major axis to the minor axis; the stress threshold positions represent multiple positions where the stress threshold is reached at different stress angles.

[0070] Among them, because the distances from the positions where the stress threshold is reached to the stress position are different due to different stress angles of the soil, and due to gravity, the distance reached when the stress angle is perpendicular to the ground is the shortest, and the distance reached when it is parallel to the ground is the longest, so an ellipsoidal area is constructed to represent different bearing capacities of the soil.

[0071] Based on the stress position, soil line position, and the ratio of the major axis to the minor axis, make a judgment in the soil three-dimensional coordinate system to obtain the major axis length, middle axis length, and minor axis length;

[0072] Construct an ellipsoid in the soil three-dimensional coordinate system according to the major axis length, middle axis length, minor axis length, and stress position.

[0073] Optionally, based on the force application position, the straight-line position of the soil, and the long-axis-to-short-axis ratio, a judgment is made in the three-dimensional soil coordinate system to obtain the long-axis length, the middle-axis length, and the short-axis length, including:

[0074] Map the straight-line position of the soil onto the abscissa, ordinate, and vertical coordinate of the three-dimensional soil coordinate system respectively to obtain the horizontal position of the soil, the vertical position of the soil, and the vertical position of the soil.

[0075] The long-axis length, the middle-axis length, and the short-axis length are obtained by the following ellipsoid formula:

[0076]

[0077] a÷b=d;

[0078] a=c;

[0079] Wherein, a represents the long-axis length; b represents the short-axis length; c represents the middle-axis length; d represents the long-axis-to-short-axis ratio; z represents the horizontal position of the soil; y represents the horizontal position of the soil; z represents the horizontal position of the soil.

[0080] Optionally, the long-axis length of the ellipsoid is equal to the middle-axis length; the short-axis length is less than the long-axis length;

[0081] The direction of the short axis of the ellipsoid is the direction of the vertical coordinate of the three-dimensional soil coordinate system; the direction of the long axis of the ellipsoid is the direction of the abscissa of the three-dimensional soil coordinate system; the direction of the middle axis of the ellipsoid is the direction of the ordinate of the three-dimensional soil coordinate system.

[0082] Optionally, a method for obtaining the annotation data of the soil force detection network includes:

[0083] Obtain a force threshold and a force cancellation rate; the force threshold represents the pressure that changes the shape of the soil to a set state; the force cancellation rate is the plasticity index of the soil corresponding to the soil type.

[0084] Among them, the force threshold is related to the type of the soil. In this embodiment, the force threshold of the clay is 5,000 kg / m 2 , and the force cancellation rate is 20.

[0085] According to the force cancellation rate, judge the change of the pressure, and mark the position where the pressure received on the force straight line is the force threshold to obtain the annotated soil straight-line position.

[0086] Among them, because the pressures received at positions different from the force - receiving position are different, so according to the force - cancellation rate, pressure judgment is performed on each position on the force - receiving straight line. When a position where the detected pressure is equal to the force - receiving threshold is detected, the position is set as the marked soil straight - line position.

[0087] Optionally, obtaining the soil straight - line position based on the soil category and the force - receiving pressure through the soil force - receiving detection network includes:

[0088] Normalize the soil category and the force - receiving pressure to obtain a normalized soil - category value and a normalized force - receiving pressure value.

[0089] Among them, in this embodiment, the normalization uses the softmax function.

[0090] Construct a detection vector from the normalized soil - category value and the normalized force - receiving pressure value.

[0091] Among them, in this embodiment, the normalized soil - category value and the normalized force - receiving pressure value are constructed into a one - dimensional vector with 2 elements, that is, the detection vector.

[0092] Input the detection vector into the soil force - receiving detection network, extract the force - receiving relationship, and obtain the soil straight - line position.

[0093] Among them, in this embodiment, the soil force - receiving detection network is a fully - connected neural network (FullyConnectedNetural Network, FCN).

[0094] Optionally, constructing an ellipsoid based on the major - axis length, the middle - axis length, the minor - axis length, and the force - receiving position in the soil three - dimensional coordinate system includes:

[0095] Use the force - receiving position as the mid - point of the ellipsoid, and construct an ellipsoid in the soil three - dimensional coordinate system to obtain an ellipsoid region.

[0096] Retain the part of the ellipsoid region that coincides with the soil region in the soil surface shape matrix, and delete the non - coincident part.

[0097] Optionally, constructing a three - dimensional coordinate axis based on the soil surface shape matrix to obtain a soil three - dimensional coordinate system includes:

[0098] Set the direction perpendicular to the ground as the direction of the vertical coordinate;

[0099] Set the direction parallel to the ground and parallel to the rows of the soil surface shape matrix as the direction of the horizontal coordinate;

[0100] Set the direction parallel to the ground and parallel to the columns of the soil surface shape matrix as the direction of the vertical coordinate;

[0101] Set the lower left corner of the soil surface shape matrix as the origin to obtain a three-dimensional soil coordinate system.

[0102] Embodiment 2

[0103] Based on the above soil load analysis method, an embodiment of the present invention further provides a soil load analysis device, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the above method are implemented.

[0104] In the specification provided herein, a large number of specific details are set forth. However, it is understood that embodiments of the present invention may be practiced without these specific details. In some instances, well-known methods, structures, and technologies have not been shown in detail so as not to obscure the understanding of this specification. Each component embodiment of the present invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art should understand that a microprocessor or a digital signal processor (DSP) may be used in practice to implement some or all of the functions of some or all of the components in the device according to the embodiments of the present invention. The present invention may also be implemented as a device or device program for executing some or all of the methods described herein (e.g., a computer program and a computer program product). Such a program for implementing the present invention may be stored on a computer-readable medium, or may be in the form of one or more signals. Such signals may be downloaded from an Internet website, or provided on a carrier signal, or in any other form.

Claims

1. A soil load analysis method, characterized in that: include: Obtaining soil category, soil surface shape matrix, stress position, stress angle and stress pressure; the soil category indicates the category of soil; The soil surface shape matrix is ​​a matrix containing the three-dimensional structure of the soil area; the force pressure represents the pressure exerted on the soil by an object; the force angle represents the angle of the force pressure; and the force position represents the position of the force pressure on the soil; According to the soil surface shape matrix, a three-dimensional coordinate axis is constructed to obtain a soil three-dimensional coordinate system; the soil three-dimensional coordinate system includes a horizontal coordinate, a vertical coordinate and a vertical coordinate; According to the soil three-dimensional coordinate system, the cosine value of the force angle is taken as the slope, and a straight line passing through the force position is taken as the force straight line; Through the soil stress detection network, based on the soil type and stress pressure, the pressure change is judged in the soil three-dimensional coordinate system to obtain the soil linear position; the soil linear position indicates the position where the stress threshold is reached; Based on the soil type, force position and soil linear position, an ellipsoid is constructed on the soil surface shape matrix in the soil three-dimensional coordinate system to obtain a soil force matrix; The stress state of the soil is judged based on the soil surface shape matrix through a shape discrimination network to obtain the soil load.

2. The soil load analysis method according to claim 1, characterized in that: The method of determining the stress state of the soil based on the soil surface shape matrix through a shape discrimination network to obtain the soil load includes: The shape discrimination network includes a plurality of three-dimensional convolution kernels; Find the position overlapping with the force line in the soil surface shape matrix and mark it to obtain the force position; Match the center point of the three-dimensional convolution kernel with the force position, and perform convolution in sequence with a step size of 1 to obtain a feature vector; the feature vector contains the characteristics of the soil state around the force line; The characteristic vector is input into a classification network to determine the bearing capacity of the soil and obtain the soil load.

3. The soil load analysis method according to claim 1, characterized in that: The method of constructing an ellipsoid on the soil surface shape matrix in the soil three-dimensional coordinate system based on the soil type, stress position and soil linear position to obtain a soil stress matrix includes: The soil category is input into the soil category network, and the ratio of the major and minor axes of the ellipsoid corresponding to multiple stress threshold positions is determined to obtain the ratio of the major and minor axes; the stress threshold position represents multiple positions that reach the stress threshold under different stress angles; Based on the force position, the soil linear position and the major-minor axis ratio, a judgment is made in the soil three-dimensional coordinate system to obtain the major axis length, the median axis length and the minor axis length; An ellipsoid is constructed in the soil three-dimensional coordinate system according to the major axis length, median axis length, minor axis length and force position.

4. The soil load analysis method according to claim 3, characterized in that: The method of determining the length of the major axis, the length of the middle axis and the length of the minor axis based on the force position, the linear position of the soil and the ratio of the major axis to the minor axis in the three-dimensional soil coordinate system includes: Mapping the soil linear position onto the horizontal coordinate, vertical coordinate and vertical coordinate of the soil three-dimensional coordinate system respectively to obtain the soil horizontal position, soil vertical position and soil vertical position; The major axis length, the median axis length and the minor axis length are obtained by the following ellipsoid formula: a÷b=d; a=c; Among them, a represents the length of the major axis; b represents the length of the minor axis; c represents the length of the median axis; d represents the ratio of the major and minor axes; x represents the horizontal position of the soil; y represents the horizontal position of the soil; and z represents the horizontal position of the soil.

5. The soil load analysis method according to claim 4, characterized in that: The length of the major axis of the ellipsoid is equal to the length of the median axis; the length of the minor axis is less than the length of the major axis; The direction of the short axis of the ellipsoid is the direction of the vertical coordinate of the soil three-dimensional coordinate system; the direction of the long axis of the ellipsoid is the direction of the horizontal coordinate of the soil three-dimensional coordinate system; the direction of the central axis of the ellipsoid is the direction of the vertical coordinate of the soil three-dimensional coordinate system.

6. The soil load analysis method according to claim 1, characterized in that: The method for obtaining the labeled data of the soil stress detection network includes: Obtaining a stress threshold and a stress offset rate; the stress threshold represents the pressure that causes the shape of the soil to change to a set state; the stress offset rate is the plasticity index of the soil corresponding to the soil category; According to the force offset rate, the change of pressure is judged, and the position where the pressure is the force threshold is found on the force straight line and marked to obtain the marked soil straight line position.

7. The soil load analysis method according to claim 1, characterized in that: The soil linear position is obtained based on the soil type and stress pressure through the soil stress detection network, including: Normalizing the soil category and stress pressure to obtain a normalized soil category value and a normalized stress pressure value; Constructing a detection vector using the normalized soil category value and the normalized stress pressure value; The detection vector is input into the soil force detection network, the force relationship is extracted, and the linear position of the soil is obtained.

8. The soil load analysis method according to claim 3, characterized in that: The method of constructing an ellipsoid in a soil three-dimensional coordinate system according to the major axis length, the median axis length, the minor axis length and the stress position includes: Taking the force position as the midpoint of the ellipsoid, constructing the ellipsoid in the soil three-dimensional coordinate system to obtain the ellipsoid area; The parts of the ellipsoid region and the soil surface shape matrix that overlap with the soil region are retained, and the parts that do not overlap are deleted.

9. The soil load analysis method according to claim 1, characterized in that: The step of constructing a three-dimensional coordinate axis according to the soil surface shape matrix to obtain a three-dimensional soil coordinate system includes: Set the direction perpendicular to the ground as the direction of the vertical coordinate; The direction of the horizontal coordinate is set as the direction of the horizontal coordinate which is parallel to the ground and parallel to the row of the soil surface shape matrix; The direction parallel to the ground and parallel to the column of the soil surface shape matrix is ​​set as the direction of the ordinate; The lower left corner of the soil surface shape matrix is ​​set as the origin to obtain the soil three-dimensional coordinate system.

10. A soil load analysis device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method according to any one of claims 1 to 9 when executing the program.

Images