Geophysical prospecting data analysis method for building foundation pile

Abstract

The invention relates to the technical field of foundation pile detection, in particular to a geophysical prospecting data analysis method for a building foundation pile, which comprises the following steps of: extracting path time difference and signal intensity based on pile foundation multi-path signals, screening direct path signals, recording propagation characteristics of the path signals, and obtaining direct path signal results. According to the method, a reflected signal path and a direct signal are effectively separated through a blind signal separation algorithm, accurate extraction and analysis of abnormal signals are ensured, the accuracy of signal processing is improved, interference is eliminated in a complex signal environment, key abnormal path signals are extracted, and a foundation is laid for subsequent defect analysis; the graph neural network performs correlation analysis on stress fluctuation and a defect reflection signal path by extracting stress data in a pile foundation structure, so that accurate positioning of pile foundation defects is realized, defect positioning is performed through the stress fluctuation, and an accurate defect spatial distribution model is generated.

Description

Technical Field

[0001] The invention relates to the technical field of foundation pile detection, and in particular to a geophysical data analysis method for building foundation piles. Background Art

[0002] The field of pile foundation inspection technology aims to ensure the structural integrity and bearing capacity of the pile foundation to meet the safety and stability requirements of the building. Pile foundation inspection technology is often used to identify defects, cracks, settlement problems or other potential structural risks in the pile foundation to ensure the safety of the building under various geological conditions.

[0003] The purpose of the geophysical data analysis method for building foundation piles is to evaluate the health status and potential risks of foundation piles by analyzing the acquired geophysical data, ensure the long-term stability and safety of the building, achieve accurate detection of foundation piles, discover potential problems in a timely manner, and provide a scientific basis for subsequent maintenance and reinforcement.

[0004] When dealing with complex signal interference, traditional methods have difficulty separating direct signals from reflected signals, which makes it difficult to identify abnormal signal paths and affects the accuracy of defect analysis. Traditional methods rely on simple signal comparison and are easily affected by external interference and signal attenuation. They are unable to effectively extract key abnormal signals in pile foundations, have low recognition rates, lack deep correlation analysis with stress data, and are unable to accurately determine the specific location of defects and their impact on the pile foundation structure, resulting in delayed and inaccurate maintenance and reinforcement measures, increasing potential structural risks and affecting the safety and stability of the entire building. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings in the prior art and to propose a geophysical data analysis method for building foundation piles.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a method for analyzing geophysical data for building foundation piles, comprising the following steps:

[0007] Step 1: Based on the pile foundation multipath signal, the path time difference and signal strength are extracted, the direct path signal is screened and the propagation characteristics of the path signal are recorded to obtain the direct path signal result;

[0008] Step 2: Based on the direct path signal result, extract the phase and amplitude of each path signal, perform amplitude difference comparison and phase adjustment operations, correct its phase amplitude according to the signal propagation law, adjust the propagation parameters by matching the phase and amplitude, and generate a signal correction result;

[0009] Step 3: Based on the signal correction result, a blind signal separation algorithm is used to separate the reflected signal path and the direct signal, and the propagation time of the abnormal reflection path is extracted. The amplitude of the abnormal signal of the path is compared by time difference, the propagation characteristics of the abnormal signal path are identified, and the abnormal signal analysis result is generated;

[0010] Step 4: Based on the abnormal signal analysis results, a graph neural network is used to extract stress data in the pile foundation structure model, and the concentrated area is determined according to the stress value fluctuation. The corresponding relationship between the reflection signal path and the stress fluctuation is attributed, and the defect is located and the defect coordinates and type are recorded to generate the defect location result;

[0011] Step 5: Based on the defect location results, the defect position coordinates and stress fluctuation amplitude are extracted, and the damage types are compared and classified. The damage data is assigned according to the coordinates of the defect area to generate a damage analysis result;

[0012] Step 6: Based on the damage analysis results, the defect position and coordinate information are extracted, and a defect spatial distribution model is established according to the three-dimensional coordinates. The position of each defect point and the coordinates of the stress concentration area are assigned to generate a spatial distribution model.

[0013] As a further scheme of the present invention, the direct path signal results include time difference, intensity and path length, the signal correction results include corrected signal phase, signal amplitude and frequency response, the abnormal signal analysis results include time difference of the reflection path, abnormal signal amplitude and path propagation angle, the defect location results include stress concentration position, defect coordinates and defect type, the damage analysis results include damage position, damage type and structural integrity level, and the spatial distribution model includes defect point coordinates, stress concentration area position and three-dimensional distribution information.

[0014] As a further solution of the present invention, the specific steps of generating the direct path signal result are:

[0015] Based on the multi-path signal of the pile foundation, the time difference and signal strength of each path are extracted, and the path time difference is calculated. By comparing the time difference of each signal, the direct path signal is distinguished from other signals, and the signals are recorded to generate the path time difference result.

[0016] Based on the path time difference result, extract the strength value of each path signal, classify the signal strength, filter out the signal strength data of the direct path by comparing the signal strength of each path, record the strength interval, and generate the signal strength classification result;

[0017] Based on the signal strength classification result, the propagation speed of the direct path signal is recorded, the signal propagation path is verified using signal strength and time difference, the propagation speed data of the direct path is obtained and archived, and the direct path signal result is generated.

[0018] As a further solution of the present invention, it is characterized in that the specific steps of generating the signal correction result are:

[0019] Based on the direct path signal result, extract the phase information of the direct path signal, classify the time difference and propagation speed of the signal, record the phase characteristics of each signal, classify the phase according to the path characteristics, and obtain the signal phase result;

[0020] Based on the signal phase result, extract the signal amplitude data of each path, compare the amplitude difference, correct the amplitude characteristics of the signal by comparing the upper and lower limits of the amplitude, record the adjusted amplitude value, and obtain the signal amplitude result;

[0021] Based on the signal amplitude result, a phase and amplitude matching operation of the signal is performed, and the phase and amplitude are corrected in combination with the propagation speed. The corrected phase and amplitude data are synchronously recorded to generate a signal correction result.

[0022] As a further solution of the present invention, the specific steps of generating the abnormal signal analysis result are:

[0023] Based on the signal correction result, a blind signal separation algorithm is used to extract the propagation time of each signal, calculate the time difference between the direct signal and the reflected signal, distinguish the reflected path and the direct path by the time difference, record the time difference and signal type of each path, and generate a signal path separation result;

[0024] Based on the signal path separation result, the signal amplitude of each reflection path is extracted, and the amplitude change analysis is performed. By comparing the amplitude differences of different paths, the abnormal path is identified and the amplitude information is recorded to generate the abnormal signal amplitude result;

[0025] Based on the abnormal signal amplitude result, the propagation speed and angle data of the abnormal path are extracted to perform propagation characteristic analysis. By comparing the speed and angle changes, the characteristics of the abnormal signal path are attributed and recorded to generate the abnormal signal analysis result.

[0026] As a further solution of the present invention, the blind signal separation algorithm is according to the formula:

[0027]

[0028] Where: σ is stress, E is the elastic modulus of the material, ΔL is the change in the length of the reflected signal path, L is the initial signal path length, α is the temperature variation coefficient, T is the temperature variation, β is the adjustment coefficient of the internal structure characteristics of the material, and γ is the signal attenuation rate.

[0029] As a further solution of the present invention, the specific steps of generating the defect location result are:

[0030] Based on the abnormal signal analysis results, the stress data in the pile foundation structure model is extracted to extract stress fluctuations. By analyzing the stress changes in each area, the stress concentration area is identified and recorded to generate stress fluctuation results;

[0031] Based on the stress fluctuation results, a graph neural network is used to extract stress data corresponding to the reflection signal path, compare the signal path with the stress change, and through the corresponding relationship analysis, the association between the attribute signal path and the stress area is generated to generate the path and stress association result;

[0032] Based on the path and stress association results, the coordinate information of the defect position is extracted to locate the defect area. By comparing the stress data and the coordinate information, the type and position of the defect are classified, the relevant information is recorded, and the defect location result is generated.

[0033] As a further solution of the present invention, the graph neural network is according to the formula:

[0034] σ=(E·ò+C·v+k1·α·T+k2·δ)

[0035] Where: σ is stress, E is Young's modulus, is strain, C is the signal propagation velocity, v is the acceleration of the signal path node, k1 is the thermal expansion weight coefficient, α is the thermal expansion coefficient of the material, T is the ambient temperature, k2 is the signal attenuation weight coefficient, and δ is the signal attenuation coefficient.

[0036] As a further solution of the present invention, the specific steps of generating the damage analysis result are:

[0037] Based on the defect location results, the coordinates and stress fluctuation data of each defect are extracted, the coordinates are compared with the stress fluctuations, and the stress changes of the defect points are classified by calculating the differences in stress fluctuation amplitudes at different positions to generate stress fluctuation classification results;

[0038] Based on the stress fluctuation classification result, the stress amplitude data of the defect point is extracted to classify the damage type. By comparing the relationship between the stress fluctuation amplitude and the position, the damage type is determined and attributed to the corresponding position, and the damage classification result is generated;

[0039] Based on the damage classification result, the coordinate data of each damage point is extracted, and the damage data is attributed to each coordinate through the correlation analysis between the coordinates and the damage type, and the damage coordinate relationship is generated and recorded to generate the damage analysis result.

[0040] As a further solution of the present invention, the specific steps of generating the spatial distribution model are:

[0041] Based on the damage analysis results, the three-dimensional coordinates of each defect point are extracted, and the coordinates are assigned to each other. By comparing the spatial positions of the defect points, the three-dimensional coordinates of each defect point are assigned to generate defect coordinate results;

[0042] Based on the defect coordinate results, a three-dimensional space model is constructed to extract the coordinate information of the defect point, and three-dimensional distribution data is generated through the correspondence between the coordinate data of the defect point and the stress fluctuation;

[0043] Based on the three-dimensional distribution data, the corresponding relationship between the three-dimensional coordinates of the defect point and the stress area, by comparing the coordinate distribution of the defect point and the stress concentration area, the defect and stress distribution in the three-dimensional space is generated to generate a spatial distribution model.

[0044] Compared with the prior art, the advantages and positive effects of the present invention are:

[0045] In the present invention, a blind signal separation algorithm is used to effectively separate the reflected signal path and the direct signal, ensuring the accurate extraction and analysis of abnormal signals, improving the accuracy of signal processing, eliminating interference in a complex signal environment, and extracting key abnormal path signals, laying the foundation for subsequent defect analysis. The graph neural network extracts stress data from the pile foundation structure, correlates stress fluctuations with defect reflection signal paths, and achieves accurate positioning of pile foundation defects. Defects are located through stress fluctuations, and an accurate defect spatial distribution model is generated, which improves the accuracy and efficiency of pile foundation detection and facilitates subsequent maintenance and reinforcement. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a schematic diagram of the main steps of the present invention. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0048] Embodiment 1

[0049] See also Figure 1 The present invention provides a technical solution: a method for analyzing geophysical data for building foundation piles, comprising the following steps:

[0050] Step 1: Based on the pile foundation multipath signal, the path time difference and signal strength are extracted, the direct path signal is screened and the propagation characteristics of the path signal are recorded to obtain the direct path signal result;

[0051] Step 2: Based on the direct path signal results, extract the phase and amplitude of each path signal, perform amplitude difference comparison and phase adjustment operations, correct its phase amplitude according to the signal propagation law, adjust the propagation parameters by matching the phase and amplitude, and generate the signal correction result;

[0052] Step 3: Based on the signal correction result, a blind signal separation algorithm is used to separate the reflected signal path and the direct signal, extract the propagation time of the abnormal reflection path, compare the amplitude of the abnormal signal of the path through the time difference, identify the propagation characteristics of the abnormal signal path, and generate the abnormal signal analysis result;

[0053] Step 4: Based on the abnormal signal analysis results, the graph neural network is used to extract the stress data in the pile foundation structure model, and the concentrated area is determined according to the stress value fluctuation. The corresponding relationship between the reflection signal path and the stress fluctuation is attributed, the defect is located, and the defect coordinates and type are recorded to generate the defect location result;

[0054] Step 5: Based on the defect location results, the defect position coordinates and stress fluctuation amplitude are extracted to compare and classify the damage types, and the damage data is assigned according to the coordinates of the defect area to generate damage analysis results;

[0055] Step 6: Based on the damage analysis results, extract the defect location and coordinate information, establish a defect spatial distribution model according to the three-dimensional coordinates, attribute the location of each defect point and the coordinates of the stress concentration area, and generate a spatial distribution model.

[0056] The signal results of the reflection path include time difference, intensity and path length; the signal correction results include corrected signal phase, signal amplitude and frequency response; the abnormal signal analysis results include time difference of the reflection path, abnormal signal amplitude and path propagation angle; the defect location results include stress concentration position, defect coordinates and defect type; the damage analysis results include damage position, damage type and structural integrity level; the spatial distribution model includes defect point coordinates, stress concentration area position and three-dimensional distribution information.

[0057] The specific steps to generate the direct path signal result are:

[0058] Based on the multi-path signal of the pile foundation, the time difference and signal strength of each path are extracted, and the path time difference is calculated. By comparing the time difference of each signal, the direct path signal is distinguished from other signals, and the signals are recorded to generate the path time difference result.

[0059] Based on the path time difference results, extract the strength value of each path signal, classify the signal strength, and filter out the signal strength data of the direct path by comparing the signal strength of each path, record the strength interval, and generate the signal strength classification result;

[0060] Based on the signal strength classification results, the propagation speed of the direct path signal is recorded, the signal propagation path is verified using signal strength and time difference, the propagation speed data of the direct path is obtained and archived, and the direct path signal result is generated;

[0061] Based on the multi-path signal of the pile foundation, the path time difference calculation method is adopted to extract the signal time difference of each path. The specific operation is to record the arrival time of the signal of each path, execute the time difference calculation command, calculate the arrival time difference of each two signals, and the command parameters include the signal receiving time and the preset time difference range. The receiving time is in milliseconds, and the time difference range is 1-10 milliseconds. The time difference between the signals is sorted, and the sorting parameter in the command is set to ascending order. The path signal is classified, the time difference classification result is archived, and the path time difference result is generated;

[0062] Based on the path time difference result, the signal strength classification method is used to extract the signal strength of each path. Specifically, during the signal acquisition process, the maximum strength value of the signal of each path is recorded, and the signal strength classification command is used to compare the strengths. The parameters in the classification command include the signal strength threshold range is set to 0-100, and the strength value unit is decibel. In the comparison operation, the strength classification command divides the signal into three categories according to the strength interval, and performs signal strength classification records, archives the numerical interval of the strength classification, and generates the signal strength classification result;

[0063] Based on the signal strength classification results, the signal propagation path verification method is used to verify the propagation speed of the direct path signal. Specifically, the signal strength data is compared with the time difference data, and the propagation speed parameter in the verification command is used to calculate the direct path signal in the propagation path. The time difference in the verification parameters is set to 1-5 milliseconds, the propagation speed unit is meters per second, and the speed range is 300-500 meters per second. In the verification operation, the signal propagation speed is compared with the preset direct path propagation speed, and the direct path propagation speed is recorded, the verified speed data is archived, and the direct path signal result is generated.

[0064] It is characterized in that the specific steps of generating the signal correction result are:

[0065] Based on the direct path signal results, the phase information of the direct path signal is extracted, the time difference and propagation speed of the signal are classified, the phase characteristics of each signal are recorded, the phase is classified according to the path characteristics, and the signal phase results are obtained;

[0066] Based on the signal phase result, the signal amplitude data of each path is extracted, and the amplitude difference is compared. The amplitude characteristics of the signal are corrected by comparing the upper and lower limits of the amplitude, and the adjusted amplitude value is recorded to obtain the signal amplitude result;

[0067] Based on the signal amplitude result, the phase and amplitude matching operation of the signal is performed, and the phase and amplitude are corrected in combination with the propagation speed. The corrected phase and amplitude data are synchronously recorded to generate the signal correction result;

[0068] Based on the direct path signal result, a phase extraction algorithm is used to extract the phase information of the direct path signal. Specifically, the time difference and propagation speed are read from the signal data, and the time difference and propagation speed parameters in the phase calculation command are used. The time difference is in microseconds, and the propagation speed parameter is set to 300 meters per second. The time difference and propagation speed are classified through the classification command. The phase interval is set to 0 to 2π in the classification operation, and the phase is classified according to the path characteristics. The phase characteristics of each path signal are recorded to generate the signal phase result;

[0069] Based on the signal phase result, the amplitude difference comparison algorithm is used to extract the amplitude data of each path signal. Specifically, the signal amplitude value of each path is read, and the amplitude difference comparison is performed using the amplitude comparison command. The upper and lower limit parameters of the amplitude are set to 0 to 100 decibels. The amplitude characteristics are corrected according to the upper and lower limits in the difference comparison command, and the corrected amplitude value is recorded according to the path. After the amplitude adjustment is performed, the signal amplitude result is generated;

[0070] Based on the signal amplitude result, a phase and amplitude matching correction algorithm is used to match the phase and amplitude of the signal. Specifically, the propagation velocity data is combined to read the phase and amplitude values ​​of the signal, and the phase and amplitude parameters in the matching correction command are used for matching. The matching correction command sets the phase interval to 0 to 2π and the amplitude range to 0 to 100 decibels. The correction operation performs synchronous correction by comparing the matching characteristics of the phase and amplitude, and synchronously records the corrected phase and amplitude data to generate a signal correction result.

[0071] The specific steps to generate abnormal signal analysis results are:

[0072] Based on the signal correction results, a blind signal separation algorithm is used to extract the propagation time of each signal, calculate the time difference between the direct signal and the reflected signal, distinguish the reflected path from the direct path by the time difference, record the time difference and signal type of each path, and generate the signal path separation result;

[0073] Based on the signal path separation results, the signal amplitude of each reflection path is extracted, and the amplitude change analysis is performed. By comparing the amplitude differences of different paths, the abnormal paths are identified and the amplitude information is recorded to generate the abnormal signal amplitude results;

[0074] Based on the abnormal signal amplitude results, the propagation speed and angle data of the abnormal path are extracted to analyze the propagation characteristics. By comparing the speed and angle changes, the characteristics of the abnormal signal path are attributed and recorded to generate the abnormal signal analysis results;

[0075] Based on the signal correction result, a blind signal separation algorithm is used to extract the propagation time of each signal. Specifically, the time difference and signal type in the signal data are read. A signal separation command is used, and the command parameters are set to the signal receiving time range of 0-100 milliseconds. The signal type parameter includes direct signals and reflected signals. In the time difference calculation, the time difference threshold between the direct signal and the reflected signal is set to 10 milliseconds. The reflected path and the direct path are distinguished by the time difference, and the propagation time and signal type of each path signal are recorded to generate the signal path separation result.

[0076] Based on the signal path separation results, the amplitude change analysis algorithm is used to extract the signal amplitude of each reflection path. Specifically, the signal amplitude value in the reflection path is read, and the signal amplitude is compared using the amplitude analysis command. The amplitude value unit in the command parameter is decibel, and the amplitude upper and lower limit parameters set in the amplitude difference comparison command are 0-100 decibels. In the amplitude analysis, the abnormal path is identified through the amplitude difference of different paths, and the identified abnormal path and its signal amplitude value are recorded to generate the abnormal signal amplitude result;

[0077] Based on the abnormal signal amplitude result, the propagation characteristic analysis algorithm is used to extract the propagation speed and angle data of the abnormal path. Specifically, the signal propagation speed and propagation angle data are read. The propagation characteristic analysis command is used. The speed unit in the command parameters is set to meters per second, and the angle unit is degree. The speed range set in the propagation speed and angle change comparison command is 300-500 meters per second, and the angle change range is 0-90 degrees. The propagation characteristics of the abnormal path are attributed to the comparison of speed and angle changes and the propagation characteristic data is recorded to generate the abnormal signal analysis results.

[0078] Blind signal separation algorithm, according to the formula:

[0079]

[0080] Where: σ is stress, E is the elastic modulus of the material, ΔL is the change in the length of the reflected signal path, L is the initial signal path length, α is the temperature variation coefficient, T is the temperature variation, β is the adjustment coefficient of the internal structure characteristics of the material, and γ is the signal attenuation rate;

[0081] Implementation process: First, measure the initial signal path length L of the foundation pile structure and the path length change ΔL after stress, which reflects the changes of the building foundation piles when subjected to different loads. Determine the elastic modulus E of the material through experiments. During the construction of the building foundation column, introduce the temperature change coefficient α to correct the influence of the external temperature change T on the material. The influence of the internal structural characteristics of the material on the signal transmission path needs to be corrected by adjusting the coefficient β. Add the signal attenuation rate γ to correct the energy loss in the signal transmission process, generate the correlation result of the signal path and stress change, and provide accurate data support for the structural health monitoring and analysis of the building foundation piles.

[0082] The specific steps to generate defect location results are:

[0083] Based on the abnormal signal analysis results, the stress data in the pile foundation structure model is extracted to extract stress fluctuations. By analyzing the stress changes in each area, the stress concentration area is identified and recorded to generate stress fluctuation results.

[0084] Based on the stress fluctuation results, the graph neural network is used to extract the stress data corresponding to the reflection signal path, and the signal path and stress changes are compared. Through the corresponding relationship analysis, the association between the attribute signal path and the stress area is obtained, and the path and stress association results are generated;

[0085] Based on the path and stress correlation results, the coordinate information of the defect position is extracted to locate the defect area. By comparing the stress data and coordinate information, the type and location of the defect are classified, the relevant information is recorded, and the defect location results are generated;

[0086] Based on the abnormal signal analysis results, the stress data in the pile foundation structure model is extracted using the stress fluctuation extraction algorithm. Specifically, the stress values ​​of each area of ​​the pile foundation are read. The stress extraction command is used. The stress value unit in the command parameters is set to Pascal, and the stress fluctuation range is set to 1-10 Pascal. The stress changes in each area are analyzed through the stress fluctuation calculation command. The areas are classified according to the amplitude of stress fluctuations and the coordinate information of the stress concentration area is recorded to generate stress fluctuation results.

[0087] Based on the stress fluctuation results, the graph neural network algorithm is used to extract the stress data corresponding to the reflection signal path. Specifically, the stress value of each reflection path is read from the stress fluctuation data, and the graph neural network command is used to compare the signal path with the stress change. The number of neural network layers in the command is set to 3 layers, and the activation function uses the ReLU function. The stress value range is 1-10 Pascals. The path stress correlation analysis command analyzes the correspondence between stress data and path data, records the correlation information between the signal path and the stress area, and generates the path and stress correlation results.

[0088] Based on the path and stress association results, the defect location algorithm is used to extract the coordinate information of the defect position. Specifically, the coordinate information and stress value in the path stress association data are read, and the defect area is located using the defect location command. The coordinate information in the command parameters is in three-dimensional coordinate format, and the stress value range is 1-10 Pascal. Through comparative analysis of stress data and coordinate information, the type and location of the defect are classified, the coordinate and type information of the defect are recorded, and the defect location result is generated.

[0089] Graph neural network, according to the formula:

[0090] σ=(E·ò+C·v+k1·α·T+k2·δ)

[0091] Where: σ is stress, E is Young's modulus, ∈ is strain, C is signal propagation velocity, v is the acceleration of the signal path node, k1 is the thermal expansion weight coefficient, α is the thermal expansion coefficient of the material, T is the ambient temperature, k2 is the signal attenuation weight coefficient, and δ is the signal attenuation coefficient;

[0092] Execution process: First, according to the geophysical data in the signal path, the ratio of the reflected signal to the initial strain in the pile body is extracted and calculated to obtain the strain. Then, the strain value is multiplied by the Young's modulus E of the material to preliminarily estimate the stress σ in the pile body. The speed C of the signal propagating in the pile body and the acceleration v of each path node are introduced. The stress fluctuation caused by the change in the speed of the signal in different nodes is corrected through their ratio to improve the accuracy of the stress estimation of the path node. The thermal expansion effect of the pile body under different temperature environments is calculated by combining the thermal expansion coefficient α of the material and the ambient temperature T. The thermal expansion weight coefficient k1 is adjusted to ensure that the influence of temperature changes on the stress distribution is properly reflected. Finally, according to the signal attenuation coefficient δ, the influence of the energy loss generated during the transmission of the signal on the stress distribution of the pile body is corrected. The weight coefficient k2 is used to balance the attenuation effect to ensure the accurate stress estimation of different nodes in the pile body, and the correlation analysis results between the signal path and the stress area of ​​the foundation pile are obtained.

[0093] The specific steps to generate damage analysis results are:

[0094] Based on the defect location results, the coordinates and stress fluctuation data of each defect are extracted, and the coordinates and stress fluctuations are compared. By calculating the difference in stress fluctuation amplitudes at different locations, the stress changes at the defect points are classified to generate stress fluctuation classification results;

[0095] Based on the stress fluctuation classification results, the stress amplitude data of the defect point is extracted to classify the damage type. By comparing the relationship between the stress fluctuation amplitude and the position, the damage type is determined and attributed to the corresponding position to generate the damage classification result;

[0096] Based on the damage classification results, the coordinate data of each damage point is extracted. Through the correlation analysis between the coordinates and the damage type, the damage data is assigned to each coordinate, the damage coordinate relationship is generated and recorded, and the damage analysis results are generated;

[0097] Based on the defect location results, the stress fluctuation comparison algorithm is used to extract the coordinates and stress fluctuation data of each defect. Specifically, the three-dimensional coordinate information and stress fluctuation amplitude value are read from the defect location results. The stress fluctuation amplitude in the command parameters is set to 1-10 Pascals, and the coordinate information format is a three-dimensional format. By comparing the stress fluctuation amplitude differences at different coordinate positions, the stress changes of each defect point are classified, the fluctuation difference information is recorded, and the stress fluctuation classification results are generated;

[0098] Based on the stress fluctuation classification results, the damage type classification algorithm is used to extract the stress amplitude data of the defect points. Specifically, the stress amplitude value and position of each defect point are read from the stress fluctuation classification results. The damage type classification command is used. The stress amplitude value unit in the command parameter is Pascal, and the upper and lower limits are set to 0-100 Pascal. By comparing the relationship between the stress fluctuation amplitude and the position, the damage type of each defect point is classified and its coordinate information is recorded to generate the damage classification result.

[0099] Based on the damage classification results, a damage coordinate relationship analysis algorithm is used to extract the coordinate data of each damage point. Specifically, the three-dimensional coordinates and damage type information are read from the damage classification results. The coordinate and damage type association analysis command is used. The command parameters are set to the three-dimensional coordinate format and the coordinate unit is meter. Through the association analysis of the coordinates and damage types, the damage data is assigned to the corresponding coordinates, the relationship between the damage type and the coordinates is recorded, and the damage analysis results are generated.

[0100] The specific steps to generate the spatial distribution model are:

[0101] Based on the damage analysis results, the three-dimensional coordinates of each defect point are extracted, and the coordinates are assigned to each other. By comparing the spatial positions of each defect point, the three-dimensional coordinates of each defect point are assigned to generate the defect coordinate results;

[0102] Based on the defect coordinate results, a three-dimensional space model is constructed to extract the coordinate information of the defect point. The three-dimensional distribution data is generated through the correspondence between the coordinate data of the defect point and the stress fluctuation.

[0103] Based on the three-dimensional distribution data, the corresponding relationship between the three-dimensional coordinates of the defect point and the stress area is attributed. By comparing the coordinate distribution of the defect point and the stress concentration area, the defect and stress distribution in the three-dimensional space is generated, and the spatial distribution model is generated;

[0104] Based on the damage analysis results, a three-dimensional coordinate attribution algorithm is used to extract the three-dimensional coordinates of each defect point. Specifically, the three-dimensional coordinate data is read from the damage analysis results, and a coordinate attribution command is used. The coordinate format in the command parameter is a three-dimensional coordinate format, and the unit is meter. The relative distance of the three-dimensional coordinates is calculated by comparing the spatial positions of each defect point, and the attribution range is set to 0.01-1 meter. The three-dimensional coordinates of each defect point are attributed, and the coordinate data after attribution is recorded to generate the defect coordinate results;

[0105] Based on the defect coordinate results, a three-dimensional space model construction algorithm is used to extract the coordinate information of the defect point. Specifically, the three-dimensional coordinate data and stress fluctuation data are read from the defect coordinate results. A three-dimensional model construction command is used. The coordinate unit in the command parameter is meter, and the stress fluctuation unit is Pascal. The data of each defect point is classified according to the correspondence between the coordinate and the stress fluctuation data. The boundary conditions of the three-dimensional model are set as a minimum boundary of 1 meter and a maximum boundary of 100 meters. The distribution data of each defect point in the three-dimensional model is recorded to generate three-dimensional distribution data.

[0106] Based on the three-dimensional distribution data, a three-dimensional spatial distribution analysis algorithm is used to extract the correspondence between the three-dimensional coordinates of the defect point and the stress area. Specifically, the coordinates of the defect point and the coordinates of the stress concentration area in the three-dimensional distribution data are read, and the spatial distribution analysis command is used. The three-dimensional coordinate format is set in the command parameters, and the coordinate unit is meter. The coordinate range of the stress concentration area is set to 0.1-10 meters. By comparing the coordinate distribution of the defect point and the stress concentration area, the correspondence between the defect point and the stress concentration area is classified and recorded to generate a spatial distribution model.

[0107] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A geophysical data analysis method for building foundation piles, characterized in that: The following steps are involved: Step 1: Based on the pile foundation multipath signal, the path time difference and signal strength are extracted, the direct path signal is screened and the propagation characteristics of the path signal are recorded to obtain the direct path signal result; Step 2: Based on the direct path signal result, extract the phase and amplitude of each path signal, perform amplitude difference comparison and phase adjustment operations, correct its phase amplitude according to the signal propagation law, adjust the propagation parameters by matching the phase and amplitude, and generate a signal correction result; Step 3: Based on the signal correction result, a blind signal separation algorithm is used to separate the reflected signal path and the direct signal, and the propagation time of the abnormal reflection path is extracted. The amplitude of the abnormal signal of the path is compared by time difference, the propagation characteristics of the abnormal signal path are identified, and the abnormal signal analysis result is generated; Step 4: Based on the abnormal signal analysis results, a graph neural network is used to extract stress data in the pile foundation structure model, and the concentrated area is determined according to the stress value fluctuation. The corresponding relationship between the reflection signal path and the stress fluctuation is attributed, and the defect is located and the defect coordinates and type are recorded to generate the defect location result; Step 5: Based on the defect location results, the defect position coordinates and stress fluctuation amplitude are extracted, and the damage types are compared and classified. The damage data is assigned according to the coordinates of the defect area to generate a damage analysis result; Step 6: Based on the damage analysis results, the defect position and coordinate information are extracted, and a defect spatial distribution model is established according to the three-dimensional coordinates. The position of each defect point and the coordinates of the stress concentration area are assigned to generate a spatial distribution model.

2. The geophysical data analysis method for building foundation piles according to claim 1, characterized in that: The direct path signal results include time difference, intensity and path length, the signal correction results include corrected signal phase, signal amplitude and frequency response, the abnormal signal analysis results include time difference of the reflection path, abnormal signal amplitude and path propagation angle, the defect location results include stress concentration position, defect coordinates and defect type, the damage analysis results include damage position, damage type and structural integrity level, and the spatial distribution model includes defect point coordinates, stress concentration area position and three-dimensional distribution information.

3. The geophysical data analysis method for building foundation piles according to claim 1, characterized in that: The specific steps of generating the direct path signal result are: Based on the multi-path signal of the pile foundation, the time difference and signal strength of each path are extracted, and the path time difference is calculated. By comparing the time difference of each signal, the direct path signal is distinguished from other signals, and the signals are recorded to generate the path time difference result. Based on the path time difference result, extract the strength value of each path signal, classify the signal strength, filter out the signal strength data of the direct path by comparing the signal strength of each path, record the strength interval, and generate the signal strength classification result; Based on the signal strength classification result, the propagation speed of the direct path signal is recorded, the signal propagation path is verified using signal strength and time difference, the propagation speed data of the direct path is obtained and archived, and the direct path signal result is generated.

4. The geophysical data analysis method for building foundation piles according to claim 1, characterized in that: The specific steps of generating the signal correction result are: Based on the direct path signal result, extract the phase information of the direct path signal, classify the time difference and propagation speed of the signal, record the phase characteristics of each signal, classify the phase according to the path characteristics, and obtain the signal phase result; Based on the signal phase result, extract the signal amplitude data of each path, compare the amplitude difference, correct the amplitude characteristics of the signal by comparing the upper and lower limits of the amplitude, record the adjusted amplitude value, and obtain the signal amplitude result; Based on the signal amplitude result, a phase and amplitude matching operation of the signal is performed, and the phase and amplitude are corrected in combination with the propagation speed. The corrected phase and amplitude data are synchronously recorded to generate a signal correction result.

5. The geophysical data analysis method for building foundation piles according to claim 1, characterized in that: The specific steps of generating the abnormal signal analysis result are: Based on the signal correction result, a blind signal separation algorithm is used to extract the propagation time of each signal, calculate the time difference between the direct signal and the reflected signal, distinguish the reflected path and the direct path by the time difference, record the time difference and signal type of each path, and generate a signal path separation result; Based on the signal path separation result, the signal amplitude of each reflection path is extracted, and the amplitude change analysis is performed. By comparing the amplitude differences of different paths, the abnormal path is identified and the amplitude information is recorded to generate the abnormal signal amplitude result; Based on the abnormal signal amplitude result, the propagation speed and angle data of the abnormal path are extracted to perform propagation characteristic analysis. By comparing the speed and angle changes, the characteristics of the abnormal signal path are attributed and recorded to generate the abnormal signal analysis result.

6. The geophysical data analysis method for building foundation piles according to claim 1, characterized in that: The blind signal separation algorithm is based on the formula: Where: σ is stress, E is the elastic modulus of the material, ΔL is the change in the length of the reflected signal path, L is the initial signal path length, α is the temperature variation coefficient, T is the temperature variation, β is the adjustment coefficient of the internal structure characteristics of the material, and γ is the signal attenuation rate.

7. The geophysical data analysis method for building foundation piles according to claim 1, characterized in that: The specific steps of generating the defect location result are: Based on the abnormal signal analysis results, the stress data in the pile foundation structure model is extracted to extract stress fluctuations. By analyzing the stress changes in each area, the stress concentration area is identified and recorded to generate stress fluctuation results; Based on the stress fluctuation results, a graph neural network is used to extract stress data corresponding to the reflection signal path, compare the signal path with the stress change, and through the corresponding relationship analysis, the association between the attribute signal path and the stress area is generated to generate the path and stress association result; Based on the path and stress association results, the coordinate information of the defect position is extracted to locate the defect area. By comparing the stress data and the coordinate information, the type and position of the defect are classified, the relevant information is recorded, and the defect location result is generated.

8. The geophysical data analysis method for building foundation piles according to claim 1, characterized in that: The graph neural network, according to the formula: σ=(E·ò+C·v+k1·α·T+k2·δ) Where: σ is stress, E is Young's modulus, is strain, C is the signal propagation velocity, v is the acceleration of the signal path node, k1 is the thermal expansion weight coefficient, α is the thermal expansion coefficient of the material, T is the ambient temperature, k2 is the signal attenuation weight coefficient, and δ is the signal attenuation coefficient.

9. The geophysical data analysis method for building foundation piles according to claim 1, characterized in that: The specific steps of generating the damage analysis results are: Based on the defect location results, the coordinates and stress fluctuation data of each defect are extracted, the coordinates are compared with the stress fluctuations, and the stress changes of the defect points are classified by calculating the differences in stress fluctuation amplitudes at different positions to generate stress fluctuation classification results; Based on the stress fluctuation classification result, the stress amplitude data of the defect point is extracted to classify the damage type. By comparing the relationship between the stress fluctuation amplitude and the position, the damage type is determined and attributed to the corresponding position, and the damage classification result is generated; Based on the damage classification result, the coordinate data of each damage point is extracted, and the damage data is attributed to each coordinate through the correlation analysis between the coordinates and the damage type, and the damage coordinate relationship is generated and recorded to generate the damage analysis result.

10. The geophysical data analysis method for building foundation piles according to claim 1, characterized in that: The specific steps of generating the spatial distribution model are: Based on the damage analysis results, the three-dimensional coordinates of each defect point are extracted, and the coordinates are assigned to each other. By comparing the spatial positions of the defect points, the three-dimensional coordinates of each defect point are assigned to generate defect coordinate results; Based on the defect coordinate results, a three-dimensional space model is constructed to extract the coordinate information of the defect point, and three-dimensional distribution data is generated through the correspondence between the coordinate data of the defect point and the stress fluctuation; Based on the three-dimensional distribution data, the corresponding relationship between the three-dimensional coordinates of the defect point and the stress area, by comparing the coordinate distribution of the defect point and the stress concentration area, the defect and stress distribution in the three-dimensional space is generated to generate a spatial distribution model.

Images