Pile hole model risk assessment method

Abstract

The invention innovatively provides a pile hole model risk assessment method, which is characterized by comprising a target detection model establishment step of acquiring pile hole three-dimensional image data in different states, and inputting the pile hole condition into a neural network for training, verification and testing to obtain a target detection model; a pile hole anomaly analysis step: forming a pile hole three-dimensional diagram from the collected pile hole data by using a 3D construction strategy component, inputting the pile hole three-dimensional diagram into a preset target detection model to obtain a pile anomaly point, and mapping the anomaly point into the pile hole three-dimensional diagram as a pile hole anomaly diagram; an abnormal point analysis step: obtaining abnormal data and normal data of the abnormal point through an abnormal point analysis strategy, and obtaining an abnormal probability according to the abnormal data and a target detection model; and the scheme correction step is used for obtaining normal data and abnormal data in the pile hole three-dimensional diagram, and obtaining a pile hole adjustment scheme through a scheme adjustment strategy according to historical data in the target detection model.

Description

Technical Field

[0001] The present invention relates to the technical field of pile hole models, and more particularly to a risk assessment method for pile hole models. Background Art

[0002] Bored piles are one of the main forms of line structure foundations, but the current bored pile hole detection is generally carried out after the concrete pouring is completed. If it is unqualified, the foundation will be scrapped, resulting in a lot of waste. If pre-testing is adopted, it can not only guide the construction process, but also provide important parameters for pile detection. There are many technical methods such as contact type and laser ranging type for pre-testing of bored pile holes in concrete, forming a series of hole measurement equipment. Among them, the ultrasonic hole detector has the ability to penetrate turbid muddy water and can perform non-contact accurate distance measurement, which is widely used in engineering. The equipment can use two pairs of orthogonally installed ultrasonic ranging sensors to measure the four walls of the wellbore, so as to make a general estimate of the parameters such as the wellbore diameter and verticality; after the pile hole model is established, it is currently impossible to evaluate the abnormalities and risks that may exist in the pile hole during the actual construction process. Summary of the invention

[0003] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a pile hole model risk assessment method to overcome the above-mentioned defects in the prior art.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] Pile hole model risk assessment method, including

[0006] The target detection model establishment step is to obtain the three-dimensional image data of the pile hole in different states, and input it into the neural network for training, verification and testing according to the pile hole conditions to obtain the target detection model;

[0007] The pile hole anomaly analysis step comprises: using the collected pile hole data to form a pile hole three-dimensional map using a 3D construction strategy component, inputting the pile hole three-dimensional map into a preset target detection model to obtain pile anomaly points, and mapping the anomaly points into the pile hole three-dimensional map as a pile hole anomaly map;

[0008] An outlier analysis step, wherein the outlier is analyzed by an outlier analysis strategy to obtain abnormal data and normal data of the outlier, and obtains an abnormal probability according to the abnormal data and the target detection model;

[0009] The scheme correction step is used to obtain normal data and abnormal data in the pile hole three-dimensional map, and obtain the pile hole adjustment scheme through the scheme adjustment strategy according to the historical data in the target detection model.

[0010] Preferably, the outlier analysis strategy includes a data acquisition step, a data calculation step and a data evaluation step.

[0011] The data acquisition step is used to obtain a three-dimensional pile hole map, and obtain pile hole data and soil layer data based on the three-dimensional pile hole map;

[0012] The data calculation step uses the pile hole data and soil layer data to calculate and obtain the pile hole bearing capacity and settlement analysis data;

[0013] The data evaluation step obtains normal data and abnormal data through an abnormal analysis algorithm based on the pile hole bearing capacity and settlement analysis data.

[0014] Preferably, the anomaly analysis algorithm includes

[0015]

[0016]

[0017]

[0018] Among them, R(x,y,z) represents the risk assessment value of the abnormal point, G(xu,yv) is used to simulate the spatial distribution characteristics of soil layer data, and σ is the standard deviation, representing the dispersion degree of soil layer data, H(z,u,v) combines the pile hole design parameters and historical maintenance parameters, n is the number of factors considered, and w i is the weight of each parameter, f i (z,u,v) is the function corresponding to the i-th factor, and the threshold of R(x,y,z) is [0,+∞).

[0019] Preferably, the scheme adjustment strategy includes analyzing the cause of the anomaly based on the normal data, abnormal data and historical data of the abnormal point, and determining the location and range of the abnormal point, and generating several groups of adjustment schemes based on the abnormal point information and the model in the database, and evaluating the generated adjustment schemes and sorting them in descending order according to feasibility.

[0020] Preferably, the evaluation algorithm is:

[0021]

[0022] Where f1(P,M) and f2(A) are the actual design requirements of pile hole design parameters, historical data and abnormal point information, a1 and a2 are weight coefficients, and g i (P, M, A) are the pile hole design parameters, historical data and abnormal point information discovery and evaluation, b i is with g iThe corresponding weight coefficient, n is the considered factor, G represents the soil layer information, P represents the pile hole design parameters, M represents the historical data, A represents the data range of the abnormal point location, and R represents the risk, etc.

[0023] Preferably, the scheme adjustment strategy is also provided with a verification sub-strategy, and the verification sub-strategy is provided with a simulation model, and suitable convenient conditions are applied to the simulation model, and the model is tested using a numerical analysis method, and the stress condition of the pile hole is calculated. The feasibility value of the model is obtained by calculating the result, and a suitable adjustment scheme is screened according to the size of the feasibility value.

[0024] Preferably, the 3D component strategy includes:

[0025] Data input, used to input relevant parameters of pile holes, including pile diameter, pile length, pile type and soil layer parameters;

[0026] Data processing is used to process the input pile hole parameters and generate and output a three-dimensional pile hole diagram through a 3D rendering strategy.

[0027] Preferably, the 3D rendering strategy includes:

[0028] Generate point cloud data representing the pile points and divide the space;

[0029] Process the points in the point cloud, remove abnormal points, and project the point cloud data onto a two-dimensional screen;

[0030] Rendering is performed using the graphics rendering pipeline, vertex data is processed using a vertex shader, fragment color is processed using a fragment shader, vertex data is converted into pixels on the screen through rasterization, visibility is processed through depth testing, and the nearest pixel is rendered.

[0031] The beneficial effects of the present invention are as follows: after the three-dimensional pile hole map component is completed, the target detection model is used to detect abnormal points in the three-dimensional pile hole map, and the risk status of the abnormal points is judged through the normal data and abnormal data of the abnormal points, and the adjustment plan of the abnormal points is determined in combination with historical data, and the adjustment plan is evaluated based on the adjustment plan to determine the best adjustment plan. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a control flow chart of the present invention. DETAILED DESCRIPTION

[0033] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] It should be noted that when a component is referred to as being "fixed to" another component, it may be directly on the other component or there may also be a component centered. When a component is considered to be "connected to" another component, it may be directly connected to the other component or there may also be a component centered. When a component is considered to be "set on" another component, it may be directly set on the other component or there may also be a component centered. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are for illustrative purposes only.

[0035] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0036] The embodiments of the present invention are further described in detail below with reference to the accompanying drawings:

[0037] Pile hole model risk assessment method, including

[0038] The target detection model establishment step is to obtain the three-dimensional image data of pile holes in different states, and input them into the neural network for training, verification and testing according to the pile hole conditions to obtain the target detection model. A large amount of three-dimensional image data of pile holes in different states is collected to ensure that the data covers normal and abnormal conditions. The collected data is pre-processed, including denoising, contrast enhancement, etc. The pre-processed data is input into the neural network for training. By adjusting the network parameters, the model can accurately identify the targets in the pile hole, such as the pile body, soil layer, etc. The model is verified and tested using an independent verification set and actual practice to ensure the accuracy of the model.

[0039] The pile hole anomaly analysis step is to use the 3D construction strategy component to form a pile hole three-dimensional map of the collected pile hole data, input the pile hole three-dimensional map into a preset target detection model to obtain pile abnormal points, and map the abnormal points into the pile hole three-dimensional map as a pile hole abnormal map; use the 3D Hou Jian strategy to construct the collected pile hole data into a three-dimensional image, input the constructed three-dimensional image into a preset target detection model, calculate and identify the abnormal points in the pile hole according to the target detection model, and map the identified abnormal points into the three-dimensional image to form pile hole abnormal points, which is convenient for more intuitive observation and analysis;

[0040] In the outlier analysis step, the outlier point obtains the abnormal data and normal data of the outlier point through the outlier analysis strategy, and obtains the anomaly probability based on the abnormal data and the target detection model; collects the data set of the outlier point, preprocesses the data set, and analyzes the abnormal data according to the outlier analysis strategy. According to the specific requirements of the strategy and the characteristics of the data, it may be necessary to adjust the parameters or thresholds to optimize the effect of outlier detection, and verify whether it is a real outlier by comparing the characteristics of the outlier point with the normal data point; and the target detection model can accurately identify the outliers and normal points in the data, and give the corresponding prediction probability, and input the received abnormal data into the target detection model for summary, and output the abnormal probability of each outlier point according to the characteristics of the input data and the calculation principle of the target detection algorithm.

[0041] The outlier analysis strategy includes data acquisition steps, data calculation steps, and data evaluation steps.

[0042] The data acquisition step is used to obtain a three-dimensional pile hole diagram, and obtain pile hole data and soil layer data according to the three-dimensional pile hole diagram, obtain the three-dimensional pile hole diagram in the pile hole anomaly analysis step, and extract the specific location, shape and size of the pile hole and other aggregate data in the three-dimensional pile hole, and extract the soil layer data, the soil layer data includes soil layer type, thickness, physical properties, etc.;

[0043] Data calculation step: using pile hole data and soil layer data, calculate the pile hole bearing capacity and settlement analysis data. Based on the pile hole data and soil layer data, calculate the pile hole bearing capacity, which is a key indicator for evaluating the stability of the pile hole. Combined with the Chengjiang prediction model, the settlement of the pile hole and its surrounding soil layer is collected and distributed;

[0044] Data evaluation step, according to the pile hole bearing capacity and settlement analysis data, obtain normal data and abnormal data through the abnormal analysis algorithm, according to the pile hole bearing capacity and settlement analysis data, according to the creation of the empty bearing capacity and settlement analysis data, and select the abnormal analysis algorithm. The pile hole bearing capacity and settlement analysis data are processed using the abnormal analysis algorithm to identify normal data and abnormal data. Normal data should meet the expected pile hole bearing capacity and settlement range, while abnormal data may exceed these ranges, and the severity and impact of the abnormal points are evaluated.

[0045] Anomaly analysis algorithms include

[0046]

[0047]

[0048]

[0049] Among them, R(x,y,z) represents the risk assessment value of the abnormal point, G(xu,yv) is used to simulate the spatial distribution characteristics of soil layer data, and σ is the standard deviation, representing the dispersion degree of soil layer data, H(z,u,v) combines the pile hole design parameters and historical maintenance parameters, n is the number of factors considered, and w i is the weight of each parameter, f i (z,u,v) is the function corresponding to the i-th factor, and the threshold of R(x,y,z) is [0,+∞). The risk assessment algorithm is used to conduct risk assessment on the outliers to determine the abnormal risk coefficient of the outliers. A risk threshold is preset. When the risk assessment value is greater than the risk threshold, it is determined to be dangerous data.

[0050] The scheme correction step is used to obtain normal data and abnormal data in the pile hole three-dimensional map, and according to the historical data in the target detection model, the pile hole adjustment scheme is obtained through the scheme adjustment strategy, and data is extracted from the pile hole three-dimensional map, including key data such as depth, diameter, and installation, and the three-dimensional data is divided into normal data and abnormal data. The normal and abnormal data are divided according to the anomaly analysis algorithm. Deep learning is used to perform anomaly detection on the pile hole data, and abnormal points are automatically identified. The identified abnormal points can be manually reviewed to ensure the accuracy of anomaly detection. Historical data similar to the current pile hole is obtained from the historical database of the target detection model, and the processing method and effect of abnormal points in the historical data are analyzed, and a reference is provided for the current user to process.

[0051] The scheme adjustment strategy includes analyzing the cause of the anomaly based on the normal data, abnormal data and historical data of the abnormal point, and determining the location and range of the abnormal point. According to the abnormal point information, combined with the model in the database, several groups of adjustment schemes are generated, and the generated adjustment schemes are evaluated and sorted in order from high to low according to feasibility. According to the normal data, abnormal data and historical characteristics of the abnormal point, the cause of the anomaly is analyzed, such as changes in geological conditions, construction errors, etc., and the location and range of the abnormal point are accurately determined through spatial analysis and data processing technology. Combined with the grinding in the database, several groups of feasible adjustment schemes are generated. The schemes include adjustment methods, such as grouting reinforcement, hole expansion, etc., and adjustment parameters, such as grouting volume, hole expansion diameter, etc., and the generated adjustment schemes are evaluated according to the evaluation algorithm, and the schemes are sorted from high to low according to the evaluation results, and the best scheme is selected and implemented according to the best scheme.

[0052] The evaluation algorithm is;

[0053]

[0054] Where f1(P,M) and f2(A) are the actual design requirements of pile hole design parameters, historical data and abnormal point information, a1 and a2 are weight coefficients, and g i (P, M, A) are the pile hole design parameters, historical data and abnormal point information discovery and evaluation, b i is with g i The corresponding weight coefficient, n is the considered factor, G represents the soil layer information, P represents the pile hole design parameters, M represents the historical data, A represents the data range of the abnormal point location, and R represents the risk level. The higher the value of R, the higher the risk.

[0055] The scheme adjustment strategy also has a verification sub-strategy. The verification sub-strategy has a simulation model. Appropriate conditions are applied to the simulation model, and the model is tested using numerical analysis methods to calculate the stress of the pile hole. The feasibility value of the model is obtained by calculating the results, and the appropriate adjustment scheme is selected according to the feasibility value. In the component model in the verification sub-strategy, according to the actual size of the pile hole, geological conditions, construction parameters and other information, a high-precision three-dimensional simulation model of the component is used. Conditions that are consistent with the actual situation, such as boundary constraints and load conditions, are applied in the simulation model. These data are based on technical engineering experience and actual data to ensure the accuracy and reliability of the simulation results, and appropriate vertical analysis methods such as finite element method and finite difference method are selected to monitor the simulation model. The numerical analysis algorithm can accurately calculate the stress of the pile hole, such as stress, strain, displacement and other key data, and calculate and analyze the stress of the pile hole at different adjustment ends. According to the calculation results, the feasibility value of the model is calculated, such as safety factor, stability index, etc. According to the feasibility value, the generated adjustment schemes are screened, and the adjustment schemes with higher feasibility values ​​are preferentially selected for implementation.

[0056] 3D component strategies include:

[0057] Data input, used to input relevant parameters of pile holes, including pile diameter, pile length, pile type, and soil layer parameters such as soil layer type, density, bearing capacity, etc.;

[0058] Data processing is used to process the input pile hole parameters, generate and output a three-dimensional pile hole diagram through a 3D rendering strategy, process the input pile hole adoption data, such as data verification, unit conversion, etc., and generate a three-dimensional pile hole diagram using a 3D rendering strategy.

[0059] 3D rendering strategies include:

[0060] Generate point cloud data representing the pile points and divide the space. According to the parameters of the pile hole, generate point cloud tools representing various parts of the pile hole and divide the three-dimensional space into multiple small areas to facilitate subsequent point cloud processing and rendering;

[0061] Process the points in the point cloud, remove abnormal points, and project the point cloud data onto a two-dimensional screen. Use algorithms to identify and propose abnormal points in the point cloud, such as noise points and outliers, and project the processed point cloud data onto a two-dimensional screen to prepare for subsequent graphics rendering.

[0062] Use the graphics rendering pipeline for rendering, use the vertex shader to process vertex data, use the fragment shader to process the fragment color, convert the vertex data into pixels on the screen through rasterization, and process visibility through depth testing to render the nearest pixel. Use the graphics rendering pipeline to perform vertex processing, fragment processing, rasterization, depth testing and other steps, and use the vertex color finder to process the vertex data, such as position, color, texture group table, etc., use the fragment shader to process the fragment color, such as calculating the final color based on lighting, material, etc., and convert the vertex data into pixels on the screen through rasterization, and perform depth testing to process visibility to ensure that the nearest pixel is rendered.

[0063] The above are only preferred embodiments of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A pile hole model risk assessment method, characterized in that: include The target detection model establishment step is to obtain the three-dimensional image data of the pile hole in different states, and input it into the neural network for training, verification and testing according to the pile hole conditions to obtain the target detection model; The pile hole anomaly analysis step comprises: using the collected pile hole data to form a pile hole three-dimensional map using a 3D construction strategy component, inputting the pile hole three-dimensional map into a preset target detection model to obtain pile anomaly points, and mapping the anomaly points into the pile hole three-dimensional map as a pile hole anomaly map; An outlier analysis step, wherein the outlier is analyzed by an outlier analysis strategy to obtain abnormal data and normal data of the outlier, and obtains an abnormal probability according to the abnormal data and the target detection model; The scheme correction step is used to obtain normal data and abnormal data in the pile hole three-dimensional map, and obtain the pile hole adjustment scheme through the scheme adjustment strategy according to the historical data in the target detection model.

2. The pile hole model risk assessment method according to claim 1, characterized in that: The outlier analysis strategy includes a data acquisition step, a data calculation step, and a data evaluation step. The data acquisition step is used to obtain a three-dimensional pile hole map, and obtain pile hole data and soil layer data based on the three-dimensional pile hole map; The data calculation step uses the pile hole data and soil layer data to calculate and obtain the pile hole bearing capacity and settlement analysis data; The data evaluation step obtains normal data and abnormal data through an abnormal analysis algorithm based on the pile hole bearing capacity and settlement analysis data.

3. The pile hole model risk assessment method according to claim 2, characterized in that: Anomaly analysis algorithms include Among them, R(x,y,z) represents the risk assessment value of the abnormal point, G(xu,yv) is used to simulate the spatial distribution characteristics of soil layer data, and σ is the standard deviation, representing the dispersion degree of soil layer data, H(z,u,v) combines the pile hole design parameters and historical maintenance parameters, n is the number of factors considered, and w i is the weight of each parameter, f i (z,u,v) is the function corresponding to the i-th factor, and the threshold of R(x,y,z) is [0,+∞).

4. The pile hole model risk assessment method according to claim 1, characterized in that: The scheme adjustment strategy includes analyzing the cause of the anomaly based on the normal data, abnormal data and historical data of the abnormal point, and determining the location and range of the abnormal point, and generating several groups of adjustment schemes based on the abnormal point information and the model in the database, and evaluating the generated adjustment schemes through an evaluation algorithm, and sorting them in descending order according to feasibility.

5. The pile hole model risk assessment method according to claim 4, characterized in that: The evaluation algorithm is: Where f1(P,M) and f2(A) are the actual design requirements of pile hole design parameters, historical data and abnormal point information, a1 and a2 are weight coefficients, and g i (P, M, A) are the pile hole design parameters, historical data and abnormal point information discovery and evaluation, b i is with g i The corresponding weight coefficient, n is the considered factor, G represents the soil layer information, P represents the pile hole design parameters, M represents the historical data, A represents the data range of the abnormal point location, and R represents the risk level.

6. The pile hole model risk assessment method according to claim 4, characterized in that: The scheme adjustment strategy is also provided with a verification sub-strategy, and the verification sub-strategy is provided with a simulation model. Appropriate convenient conditions are applied to the simulation model, and the model is tested using a numerical analysis method to calculate the stress condition of the pile hole. The feasibility value of the model is obtained by calculating the results, and the appropriate adjustment scheme is screened according to the size of the feasibility value.

7. The pile hole model risk assessment method according to claim 1, characterized in that: The 3D component strategy includes: Data input, used to input relevant parameters of pile holes, including pile diameter, pile length, pile type and soil layer parameters; Data processing is used to process the input pile hole parameters and generate and output a three-dimensional pile hole diagram through a 3D rendering strategy.

8. The pile hole model risk assessment method according to claim 1, characterized in that: The 3D rendering strategy includes: Generate point cloud data representing the pile points and divide the space; Process the points in the point cloud, remove abnormal points, and project the point cloud data onto a two-dimensional screen; Rendering is performed using the graphics rendering pipeline, vertex data is processed using a vertex shader, fragment color is processed using a fragment shader, vertex data is converted into pixels on the screen through rasterization, visibility is processed through depth testing, and the nearest pixel is rendered.

Images