Building foundation pile defect detection method and system based on intelligent induction

Abstract

The invention discloses a building foundation pile defect detection method and system based on intelligent sensing, and relates to the technical field of sensing. The method comprises the following steps: training a model for predicting foundation pile defects; clustering the foundation pile environment characteristic data to obtain a plurality of environment clusters; calculating defect responsivity and defect influence factors for each environment cluster; collecting sensor data of a to-be-detected foundation pile, a foundation pile environment data sequence and a foundation pile defect criterion table; inputting sensor data of an actual foundation pile into the defect prediction model to obtain a foundation pile defect prediction value; calculating an environment cluster to which each foundation pile environment data sequence belongs, and collecting corresponding defect responsivity and defect influence factors; based on the sensor data of the foundation pile and the defect responsivity and the influence factor of the environment clustering cluster, calculating the influence range and the severity of the foundation pile defect; and based on the foundation pile defect criterion table, the foundation pile defect prediction value and the defect influence range and severity, the bearing capacity and integrity of the foundation pile are judged.

Description

Technical Field

[0001] The present invention belongs to the field of sensor technology, and in particular relates to a method and system for detecting building foundation pile defects based on intelligent sensing. Background Art

[0002] The building pile defect detection method based on intelligent sensing is a new detection technology that integrates sensing technology, signal processing technology, artificial intelligence and Internet of Things technology. This technology aims to achieve accurate and efficient detection of building pile defects through intelligent means, and provide a scientific basis for the safety assessment and maintenance of engineering structures.

[0003] Traditional detection methods include static load test and core drilling method. Traditional detection methods have limitations in detection accuracy and efficiency, and it is difficult to meet the high requirements of modern engineering for pile foundation quality detection; although existing intelligent detection technology has made certain progress, it is often limited to a single data processing or analysis link and lacks systematic intelligent solutions; some detection methods such as core drilling method are destructive and costly; static load test has high requirements for the site and may have a certain impact on the environment; existing technology still has deficiencies in the identification of defect types, making it difficult to accurately classify and locate pile foundation defects. Summary of the invention

[0004] In order to overcome the shortcomings and deficiencies of the above-mentioned prior art, the first purpose of the present invention is to provide a building foundation pile defect detection method based on intelligent sensing; the second purpose of the present invention is to provide a building foundation pile defect detection system based on intelligent sensing.

[0005] The first object of the present invention adopts the following technical solution:

[0006] A method for detecting building foundation pile defects based on intelligent sensing comprises the following steps:

[0007] Step 1: Conduct X sample collection experiments. During each sample collection experiment, collect sensor data of pile foundations, pile foundation defect label data, pile foundation environmental feature data, and pile foundation defect regression training data; where X is the number of sample collection experiments selected;

[0008] Step 2: Using the sensor data of the pile as input and the label data of the pile defects as output, a defect prediction model for predicting pile defects is trained;

[0009] Step 3: Based on the pile environment characteristic data, use a clustering algorithm to cluster the pile environments of the X sample collection experiments to obtain N environmental clusters; N is the number of preset environmental clusters;

[0010] Step 4: Based on the pile defect regression training data, the corresponding defect response and defect impact factor are calculated for each environmental cluster;

[0011] Step 5: Collect sensor data of the pile to be tested, pile environment data sequence and pile defect criterion table;

[0012] Step 6: Input the sensor data of the actual pile foundation into the defect prediction model to obtain the pile foundation defect prediction value;

[0013] Step 7: Calculate and obtain the environmental cluster to which each pile environmental data sequence belongs, and collect the defect response and defect influencing factor corresponding to the environmental cluster;

[0014] Step 8: Based on the sensor data of the piles and the defect responsiveness and defect impact factor of each pile environment data sequence, calculate the impact range and severity of the pile defects;

[0015] Step 9: Determine the bearing capacity and integrity of the piles based on the pile defect criterion table, pile defect prediction values, and the impact range and severity of the pile defects.

[0016] Preferably, the experimental method of the sample collection experiment is:

[0017] Some foundation piles are selected as experimental foundation piles, and loading tests are carried out using different loading parameters; the loading parameters include the loaded load weight and the distance between the loading point and the center of the foundation pile.

[0018] Preferably, the method of collecting the sensor data of the pile foundation, the label data of the pile foundation defect, the characteristic data of the pile foundation environment and the regression training data of the pile foundation defect is:

[0019] In each sample collection experiment, the sensor data of the piles, which consist of the loading parameters, pile material, pile type, pile length, pile width, and pile slope, are collected;

[0020] In each sample collection experiment, the defect types and locations of the piles generated when the piles are loaded are collected as pile defect label data;

[0021] In each sample collection experiment, p groups of monitoring point sets are selected within a range of M meters around the foundation pile, each group of monitoring point sets includes q monitoring points, and the distances between the monitoring points in each group of monitoring point sets and the foundation pile are different; where M, p and q are preset coefficients respectively;

[0022] During the sample collection experiment, the vibration speed at the location of each monitoring point when the monitoring point receives vibration is collected;

[0023] The vibration velocity of each monitoring point, the distance between the monitoring point and the loading position, and the load weight used for loading are taken as a set of regression training samples, and all regression training samples are used as the regression training data of the pile defects in this sample collection experiment;

[0024] The pile foundation environment characteristic data within a range of M meters around the pile foundation is taken as the pile foundation environment characteristic data.

[0025] Preferably, the method of training a defect prediction model for predicting pile defects is:

[0026] The pile sensor data of each sample collection experiment is used as the input of the defect prediction model. The defect prediction model takes the pile defect prediction value corresponding to each sample collection experiment as the output, the pile defect label data of each sample collection experiment as the prediction target, the difference between the pile defect prediction value and the pile defect label data as the prediction error, and minimizing the sum of prediction errors as the training goal; the defect prediction model is trained until the sum of prediction errors reaches convergence and the training is stopped.

[0027] Preferably, a clustering algorithm is used to cluster the foundation pile environments of the X sample collection experiments to obtain N environment clusters:

[0028] The environmental characteristic data of the foundation pile in each sample collection experiment is taken as a data point, and each dimensional coordinate of the data point corresponds to an environmental parameter in the environmental data;

[0029] The data points of all sample collection experiments are divided into N environmental clusters using a clustering algorithm.

[0030] Preferably, the method of calculating the corresponding defect responsiveness and defect impact factor for each environmental cluster is as follows:

[0031] The pile defect regression training data of the sample collection experiment are grouped according to the clustering results of the environmental clustering clusters to obtain N groups of pile defect regression training data groups;

[0032] For each environment cluster:

[0033] A set of monitoring point sets is randomly selected from the p sets of monitoring point sets in each sample collection experiment, and all regression training samples in all the selected monitoring point sets are combined into a set of regression training data. This is collected p times in total to generate p sets of regression training data;

[0034] The least square method is used to perform regression analysis on each set of regression training data to obtain the corresponding defect responsiveness and defect influencing factor;

[0035] The average value of the defect responsiveness corresponding to all the group regression training data of the environmental cluster cluster is taken as the defect responsiveness of the environmental cluster cluster;

[0036] The average value of the defect impact factors corresponding to all the group regression training data of the environmental cluster cluster is taken as the defect impact factor of the environmental cluster cluster.

[0037] Preferably, the method of collecting the sensor data of the pile to be detected and the pile environment data sequence is:

[0038] Collecting the sensor data of the piles including the loading parameters planned for the piles to be tested and the pile material, pile type, pile length, pile width and pile slope of the piles to be tested;

[0039] Connect the planned loading position and the protection object position, collect the order of different pile environment types from the planned loading position to the protection object position, combine the environmental characteristic data and vibration transmission distance of the area where each pile environment type is located to form a group of pile environment data, arrange the pile environment data in the order of pile environment types, and obtain a pile environment data sequence; the vibration transmission distance is the length of the line connecting the planned loading position and the protection object position in the area where the pile environment type is located.

[0040] Preferably, the environmental cluster to which each pile environment data sequence belongs is calculated, and the defect responsiveness and defect influencing factor corresponding to the environmental cluster are collected in the following manner:

[0041] The pile environment data in the pile environment data sequence are numbered in sequence order, and the number is marked as i;

[0042] The distance between the environmental characteristic data of the i-th group of foundation pile environmental data and the center point of each environmental cluster is calculated, and the environmental cluster with the closest distance is taken as the environmental cluster to which the foundation pile environmental data sequence belongs, and the defect responsiveness and defect influencing factor corresponding to the environmental cluster are marked as Si and Fi respectively.

[0043] Preferably, the scope and severity of the pile foundation defects are calculated as follows:

[0044] The vibration transmission distance of the i-th group of pile environment data is marked as Ri;

[0045] The severity of the pile defect when it reaches the location of the i-th pile environment type is marked as Di;

[0046] Mark the planned load weight as Q;

[0047] For i=1, i.e. the first pile environment type, the defect severity D1 of the first pile environment type is calculated based on the defect responsiveness and defect impact factor; the calculation formula of D1 is as follows:

[0048]

[0049] Among them, S1 is the defect responsiveness of the first pile environment type; Q is the planned load weight; R1 is the vibration transmission distance of the first group of pile environment data; α is the defect influence factor;

[0050] Then for any i>1, calculate the defect severity Di of the i-th pile environment type; the calculation formula of Di is as follows:

[0051]

[0052] Wherein, Si is the defect response of the i-th pile environment type; D(i-1) is the defect severity of the previous pile environment type; Ri is the vibration transmission distance of the i-th group of pile environment data; α is the defect influence factor;

[0053] Then the final severity DI of the pile defect is the defect severity of the last pile environment type; DI=Di; wherein I is the number of elements in the pile environment data sequence.

[0054] The second object of the present invention adopts the following technical solution:

[0055] The building foundation pile defect detection system based on intelligent sensing is used to implement a building foundation pile defect detection method based on intelligent sensing, including:

[0056] Sample collection module: used to conduct multiple sample collection experiments and collect various data of pile foundations; defect prediction model training module: based on the collected sample data, a model for predicting pile foundation defects is trained; environmental clustering module: the environmental characteristic data of pile foundations are clustered to obtain multiple environmental clustering clusters; defect responsiveness and influencing factor calculation module: the defect responsiveness and defect influencing factor are calculated for each environmental clustering cluster; data collection module: the sensor data of the pile foundation to be detected, the pile foundation environmental data sequence and the pile foundation defect criterion table are collected; defect prediction module: the sensor data of the actual pile foundation is input into the defect prediction model to obtain the pile foundation defect prediction value; environmental cluster matching module: the environmental clustering cluster to which each pile foundation environmental data sequence belongs is calculated, and the corresponding defect responsiveness and defect influencing factor are collected; defect impact range and severity calculation module: the impact range and severity of pile foundation defects are calculated based on the sensor data of the pile foundation and the defect responsiveness and influencing factor of the environmental clustering cluster; pile foundation state assessment module: the bearing capacity and integrity of the pile foundation are judged based on the pile foundation defect criterion table, the pile foundation defect prediction value and the defect impact range and severity.

[0057] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0058] 1. The present invention uses a clustering algorithm to cluster the environmental characteristic data of pile foundations to obtain multiple environmental clustering clusters, which is helpful to identify the defect patterns of pile foundations under different environmental conditions and improve the adaptability and generalization ability of the model; based on the pile foundation defect regression training data, the least squares method is used for regression analysis to calculate the defect responsiveness and defect influencing factor of each environmental clustering cluster, quantify the impact of pile foundation defects under different environmental conditions, and provide a scientific basis for subsequent evaluation and prediction.

[0059] 2. The present invention arranges the foundation pile environmental data in order to form a foundation pile environmental data sequence. The orderly data processing can more accurately reflect the development of foundation pile defects under different environmental conditions, thereby improving the accuracy of the assessment. Based on the foundation pile sensor data and the defect responsiveness and influencing factors of the environmental clustering cluster, the influence range and severity of the foundation pile defects are calculated. By quantifying the influence range and severity of the defects, the health status of the foundation pile can be more intuitively assessed.

[0060] 3. The present invention combines the pile defect criterion table, the pile defect prediction value, the defect impact range and severity to comprehensively evaluate the bearing capacity and integrity of the pile. The comprehensive evaluation method provides comprehensive pile status information and improves the safety of the project. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0062] Figure 1 A flow chart of a building foundation pile defect detection method based on intelligent sensing of the present invention is shown;

[0063] Figure 2 The module diagram of the building foundation pile defect detection system based on intelligent sensing of the present invention is shown;

[0064] Figure 3 A flow chart of calculating defect responsiveness and defect impact factor according to the present invention is shown. DETAILED DESCRIPTION

[0065] 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.

[0066] In addition, the described features, structures or characteristics may be combined in one or more example embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the example embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced while omitting one or more of the specific details, or other methods, components, steps, etc. may be adopted. In other cases, well-known structures, methods, implementations or operations are not shown or described in detail to avoid obscuring various aspects of the present disclosure.

[0067] Embodiment 1:

[0068] See also Figure 1 As shown, the building foundation pile defect detection method based on intelligent sensing in this embodiment comprises the following steps:

[0069] Step 1: Conduct A sample collection experiments. During each sample collection experiment, collect sensor data of pile foundations, pile foundation defect label data, pile foundation environmental feature data, and pile foundation defect regression training data; where A is the number of sample collection experiments selected.

[0070] In each sample collection experiment, the sensor data of the piles, which consist of the loading parameters, pile material, pile type, pile length, pile width, and pile slope, are collected;

[0071] In each sample collection experiment, the defect types and locations of the piles generated when the piles are loaded are collected as pile defect label data;

[0072] In each sample collection experiment, p sets of monitoring points are selected within a range of X meters around the foundation pile, each set of monitoring points includes q monitoring points, and the distances between the monitoring points and the foundation pile in each set of monitoring points are different; where X, p and q are preset coefficients respectively;

[0073] During the sample collection experiment, the vibration speed at the location of each monitoring point when the monitoring point receives vibration is collected;

[0074] The vibration velocity of each monitoring point, the distance between the monitoring point and the loading position, and the load weight used for loading are taken as a set of regression training samples, and all regression training samples are used as the regression training data of the pile defects in this sample collection experiment;

[0075] The pile foundation environment characteristic data within a range of X meters around the pile foundation is used as the pile foundation environment characteristic data.

[0076] Some foundation piles are selected as experimental foundation piles, and loading tests are performed using different loading parameters; the loading parameters include the loaded load weight and the distance between the loading point and the center of the foundation pile.

[0077] Step 2: Using the sensor data of the pile foundation as input and the pile foundation defect label data as output, a defect prediction model for predicting pile foundation defects is trained.

[0078] The method of training a defect prediction model for predicting pile defects is as follows:

[0079] The pile sensor data of each sample collection experiment is used as the input of the defect prediction model. The defect prediction model takes the pile defect prediction value corresponding to each sample collection experiment as the output, takes the pile defect label data of each sample collection experiment as the prediction target, takes the difference between the pile defect prediction value and the pile defect label data as the prediction error, and takes minimizing the sum of prediction errors as the training goal; the defect prediction model is trained until the sum of prediction errors reaches convergence, and the training is stopped.

[0080] Step 3: Based on the pile environment characteristic data, use a clustering algorithm to cluster the pile environments of the A-time sample collection experiment to obtain N environmental clusters; N is the number of preset environmental clusters;

[0081] The clustering algorithm is used to cluster the foundation pile environment of the A-time sample collection experiment, and the method of obtaining N environment clusters is as follows:

[0082] The environmental characteristic data of the foundation pile in each sample collection experiment is taken as a data point, and each dimensional coordinate of the data point corresponds to an environmental parameter in the environmental data;

[0083] The data points of all sample collection experiments are divided into N environmental clusters using a clustering algorithm.

[0084] Step 4: Based on the pile defect regression training data, the corresponding defect response and defect impact factor are calculated for each environmental cluster;

[0085] See also Figure 3 As shown, the method of calculating the corresponding defect responsiveness and defect impact factor for each environmental cluster is as follows:

[0086] The pile defect regression training data of the sample collection experiment are grouped according to the clustering results of the environmental clustering clusters to obtain N groups of pile defect regression training data groups;

[0087] For each environment cluster:

[0088] A set of monitoring point sets is randomly selected from the p sets of monitoring point sets in each sample collection experiment, and all regression training samples in all the selected monitoring point sets are combined into a set of regression training data. This is collected p times in total to generate p sets of regression training data;

[0089] The least square method is used to perform regression analysis on each set of regression training data to obtain the corresponding defect responsiveness and defect influencing factor;

[0090] The average value of the defect responsiveness corresponding to all the group regression training data of the environmental cluster cluster is taken as the defect responsiveness of the environmental cluster cluster;

[0091] The average value of the defect impact factors corresponding to all the group regression training data of the environmental cluster cluster is taken as the defect impact factor of the environmental cluster cluster.

[0092] Step 5: Collect sensor data of the pile to be tested, pile environment data sequence and pile defect criterion table;

[0093] The method of collecting the sensor data of the pile to be tested and the pile environment data sequence is as follows:

[0094] Collecting the sensor data of the piles including the loading parameters planned for the piles to be tested and the pile material, pile type, pile length, pile width and pile slope of the piles to be tested;

[0095] Connect the planned loading position and the protection object position, collect the order of different pile environment types from the planned loading position to the protection object position, combine the environmental characteristic data and vibration transmission distance of the area where each pile environment type is located to form a group of pile environment data, arrange the pile environment data in the order of pile environment types, and obtain a pile environment data sequence; the vibration transmission distance is the length of the line connecting the planned loading position and the protection object position in the area where the pile environment type is located.

[0096] Step 6: Input the sensor data of the actual pile foundation into the defect prediction model to obtain the pile foundation defect prediction value;

[0097] Step 7: Calculate and obtain the environmental cluster to which each pile environmental data sequence belongs, and collect the defect response and defect influencing factor corresponding to the environmental cluster;

[0098] The environmental cluster to which each pile environmental data sequence belongs is calculated, and the defect response and defect influencing factor corresponding to the environmental cluster are collected in the following manner:

[0099] The pile environment data in the pile environment data sequence are numbered in sequence order, and the numbers are marked as i, i=1, 2, 3, ... 3, i.

[0100] The distance between the environmental characteristic data of the i-th group of foundation pile environmental data and the center point of each environmental cluster is calculated, and the environmental cluster with the closest distance is taken as the environmental cluster to which the foundation pile environmental data sequence belongs, and the defect responsiveness and defect influencing factor corresponding to the environmental cluster are marked as Si and Fi respectively.

[0101] Step 8: Based on the sensor data of the piles and the defect responsiveness and defect impact factor of each pile environment data sequence, calculate the impact range and severity of the pile defects;

[0102] The method for calculating the scope and severity of pile defects is:

[0103] The vibration transmission distance of the i-th group of pile environment data is marked as Ri;

[0104] The severity of the pile defect when it reaches the location of the i-th pile environment type is marked as Di;

[0105] Mark the planned load weight as Q;

[0106] For i=1, i.e. the first pile environment type, the defect severity D1 of the first pile environment type is calculated based on the defect responsiveness and defect impact factor; the calculation formula of D1 is as follows:

[0107]

[0108] Among them, S1 is the defect responsiveness of the first pile environment type; Q is the planned load weight; R1 is the vibration transmission distance of the first group of pile environment data; α is the defect influence factor;

[0109] Then for any i>1, calculate the defect severity Di of the i-th pile environment type; the calculation formula of Di is as follows:

[0110]

[0111] Among them, Si is the defect responsiveness of the i-th pile environment type; D(i-1) is the defect severity of the previous pile environment type; Ri is the vibration transmission distance of the i-th group of pile environment data; α is the defect influence factor.

[0112] Then the final severity DI of the pile defect is the defect severity of the last pile environment type; DI=Di; wherein I is the number of elements in the pile environment data sequence.

[0113] Step 9: Determine the bearing capacity and integrity of the piles based on the pile defect criterion table, pile defect prediction values, and the impact range and severity of the pile defects.

[0114] Based on the test results, a detailed pile defect report is generated, including information such as defect type, location, severity, etc., providing visual test results, analyzing the causes of defects, providing maintenance suggestions, and recording and tracking the maintenance history of the piles.

[0115] The beneficial effects of this embodiment are: improving detection accuracy and efficiency, realizing intelligent prediction and evaluation of pile foundation defects; adapting to detection needs under different environmental conditions through cluster analysis; providing detailed defect reports and maintenance suggestions, and providing a scientific basis for the safe use and maintenance of pile foundations.

[0116] Embodiment 2:

[0117] See also Figure 2 As shown, the building foundation pile defect detection system based on intelligent sensing of this embodiment includes a sample collection module, a defect prediction model training module, an environmental clustering module, a defect responsiveness and influencing factor calculation module, a data collection module, a defect prediction module, an environmental clustering matching module, a defect impact range and severity calculation module, and a foundation pile status evaluation module.

[0118] Sample collection module: used to conduct multiple sample collection experiments and collect various data of pile foundations.

[0119] Defect prediction model training module: Based on the collected sample data, a model for predicting pile defects is trained.

[0120] Environmental clustering module: clusters the environmental feature data of the pile foundation to obtain multiple environmental clusters.

[0121] Defect responsiveness and impact factor calculation module: Calculates defect responsiveness and defect impact factor for each environmental cluster.

[0122] Data collection module: collects sensor data of the piles to be tested, pile environment data sequence and pile defect criterion table.

[0123] Defect prediction module: Input the sensor data of the actual pile foundation into the defect prediction model to obtain the predicted value of the pile foundation defect.

[0124] Environmental cluster matching module: calculates the environmental cluster to which each pile environmental data sequence belongs, and collects the corresponding defect responsiveness and defect influencing factors.

[0125] Defect impact scope and severity calculation module: Based on the sensor data of the pile and the defect responsiveness and impact factor of the environmental clustering cluster, the impact scope and severity of the pile defects are calculated.

[0126] Pile foundation condition assessment module: Determine the bearing capacity and integrity of the pile foundation based on the pile foundation defect criterion table, pile foundation defect prediction value, defect impact range and severity.

[0127] The beneficial effects of this embodiment are: realizing intelligent detection and evaluation of pile foundation defects, improving detection accuracy and efficiency; facilitating system maintenance and upgrading through modular design; providing a scientific basis for the safe use and maintenance of pile foundations, reducing safety risks, and extending the service life of pile foundations.

[0128] The impact factor of the present invention is used to measure the degree of influence of different factors or variables on a certain result or decision. The definition of the impact factor refers to the numerical value assigned to each factor when comparing and evaluating multiple factors to reflect its importance or priority. These impact factors can be determined according to specific circumstances and needs, and are usually jointly formulated and confirmed by professionals or relevant stakeholders. By reasonably setting the impact factor, it can help the program or system to make decisions or predictions more accurately.

[0129] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

[0130] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A building foundation pile defect detection method based on intelligent sensing, characterized in that: The following steps are involved: Step 1: Conduct X sample collection experiments. During each sample collection experiment, collect sensor data of pile foundations, pile foundation defect label data, pile foundation environmental feature data, and pile foundation defect regression training data; where X is the number of sample collection experiments selected; Step 2: Using the sensor data of the pile as input and the label data of the pile defects as output, a defect prediction model for predicting pile defects is trained; Step 3: Based on the pile environment characteristic data, use a clustering algorithm to cluster the pile environments of the X sample collection experiments to obtain N environmental clusters; N is the number of preset environmental clusters; Step 4: Based on the pile defect regression training data, the corresponding defect response and defect impact factor are calculated for each environmental cluster; Step 5: Collect sensor data of the pile to be tested, pile environment data sequence and pile defect criterion table; Step 6: Input the sensor data of the actual pile foundation into the defect prediction model to obtain the pile foundation defect prediction value; Step 7: Calculate and obtain the environmental cluster to which each pile environmental data sequence belongs, and collect the defect response and defect influencing factor corresponding to the environmental cluster; Step 8: Based on the sensor data of the piles and the defect responsiveness and defect impact factor of each pile environment data sequence, calculate the impact range and severity of the pile defects; Step 9: Determine the bearing capacity and integrity of the piles based on the pile defect criterion table, pile defect prediction values, and the impact range and severity of the pile defects.

2. A building foundation pile defect detection method based on intelligent sensing according to claim 1, characterized in that: The experimental method of the sample collection experiment is: Some foundation piles are selected as experimental foundation piles, and loading tests are performed using different loading parameters; the loading parameters include the loaded load weight and the distance between the loading point and the center of the foundation pile.

3. A building foundation pile defect detection method based on intelligent sensing according to claim 1, characterized in that: The method of collecting the sensor data of the pile foundation, the label data of the pile foundation defect, the characteristic data of the pile foundation environment and the regression training data of the pile foundation defect is as follows: In each sample collection experiment, the sensor data of the piles, which consist of the loading parameters, pile material, pile type, pile length, pile width, and pile slope, are collected; In each sample collection experiment, the defect types and locations of the piles generated when the piles are loaded are collected as pile defect label data; In each sample collection experiment, p groups of monitoring points are selected within the range of M meters around the foundation pile. Each group of monitoring points includes q monitoring points, and the distances between the monitoring points in each group of monitoring points and the foundation pile are different. During the sample collection experiment, the vibration speed at the location of each monitoring point when the monitoring point receives vibration is collected; The vibration velocity of each monitoring point, the distance between the monitoring point and the loading position, and the load weight used for loading are taken as a set of regression training samples, and all regression training samples are used as the regression training data of the pile defects in this sample collection experiment; The pile foundation environment characteristic data within a range of M meters around the pile foundation is taken as the pile foundation environment characteristic data.

4. The method for detecting building foundation pile defects based on intelligent sensing according to claim 1, characterized in that: The method of training the defect prediction model for predicting pile defects is as follows: The pile sensor data of each sample collection experiment is used as the input of the defect prediction model, the defect prediction model uses the pile defect prediction value corresponding to each sample collection experiment as the output, the pile defect label data of each sample collection experiment as the prediction target, the difference between the pile defect prediction value and the pile defect label data as the prediction error, and minimizing the sum of the prediction errors as the training target; The defect prediction model is trained until the sum of the prediction errors reaches convergence and then the training is stopped.

5. The method for detecting building foundation pile defects based on intelligent sensing according to claim 1, characterized in that: The method of clustering the pile environment of the X-time sample collection experiment using the clustering algorithm to obtain N environment clusters is as follows: The environmental characteristic data of the foundation pile in each sample collection experiment is taken as a data point, and each dimensional coordinate of the data point corresponds to an environmental parameter in the environmental data; the data points of all sample collection experiments are divided into N environmental clusters using a clustering algorithm.

6. The method for detecting building foundation pile defects based on intelligent sensing according to claim 1, characterized in that: The method for calculating the corresponding defect responsiveness and defect impact factor for each environmental cluster is as follows: The pile defect regression training data of the sample collection experiment are grouped according to the clustering results of the environmental clustering clusters to obtain N groups of pile defect regression training data groups; For each environment cluster: A set of monitoring point sets is randomly selected from the p sets of monitoring point sets in each sample collection experiment, and all regression training samples in all the selected monitoring point sets are combined into a set of regression training data. This is collected p times in total to generate p sets of regression training data; The least squares method is used to perform regression analysis on each group of regression training data to obtain the corresponding defect responsiveness and defect impact factor; the average value of the defect responsiveness corresponding to all groups of regression training data of the environmental cluster cluster is taken as the defect responsiveness of the environmental cluster cluster; the average value of the defect impact factor corresponding to all groups of regression training data of the environmental cluster cluster is taken as the defect impact factor of the environmental cluster cluster.

7. The method for detecting building foundation pile defects based on intelligent sensing according to claim 1, characterized in that: The method of collecting the sensor data of the pile to be detected and the pile environment data sequence is: Collecting the sensor data of the piles including the loading parameters planned for the piles to be tested and the pile material, pile type, pile length, pile width and pile slope of the piles to be tested; Connect the planned loading position and the protection object position, collect the order of different pile environment types from the planned loading position to the protection object position, combine the environmental characteristic data and vibration transmission distance of the area where each pile environment type is located to form a group of pile environment data, arrange the pile environment data in the order of pile environment types, and obtain a pile environment data sequence; the vibration transmission distance is the length of the line connecting the planned loading position and the protection object position in the area where the pile environment type is located.

8. The method for detecting building foundation pile defects based on intelligent sensing according to claim 7, characterized in that: The method of calculating and obtaining the environment cluster to which each pile environment data sequence belongs and collecting the defect responsiveness and defect impact factor corresponding to the environment cluster is as follows: The pile environment data in the pile environment data sequence are numbered in sequence order, and the number is marked as i; The distance between the environmental characteristic data of the i-th group of foundation pile environmental data and the center point of each environmental cluster is calculated, and the environmental cluster with the closest distance is taken as the environmental cluster to which the foundation pile environmental data sequence belongs, and the defect responsiveness and defect influencing factor corresponding to the environmental cluster are marked as Si and Fi respectively.

9. A building foundation pile defect detection method based on intelligent sensing according to claim 8, characterized in that: The method for calculating the scope and severity of the pile foundation defects is as follows: The vibration transmission distance of the i-th group of pile environment data is marked as Ri; The severity of the pile defect when it reaches the location of the i-th pile environment type is marked as Di; Mark the planned load weight as Q; For i=1, i.e. the first pile environment type, the defect severity D1 of the first pile environment type is calculated based on the defect responsiveness and defect impact factor; the calculation formula of D1 is as follows: Among them, S1 is the defect responsiveness of the first pile environment type; Q is the planned load weight; R1 is the vibration transmission distance of the first group of pile environment data; α is the defect influence factor; For any i>1, calculate the defect severity Di of the i-th pile environment type; the calculation formula of Di is as follows: Wherein, Si is the defect response of the i-th pile environment type; D(i-1) is the defect severity of the previous pile environment type; Ri is the vibration transmission distance of the i-th group of pile environment data; α is the defect influence factor; Then the final severity DI of the pile defect is the defect severity of the last pile environment type; DI=Di; wherein I is the number of elements in the pile environment data sequence.

10. A building foundation pile defect detection system based on intelligent sensing, used to implement a building foundation pile defect detection method based on intelligent sensing as claimed in claim 1, characterized in that: System includes Sample collection module: used to conduct multiple sample collection experiments and collect various data of pile foundations; defect prediction model training module: based on the collected sample data, a model for predicting pile foundation defects is trained; environmental clustering module: the environmental characteristic data of pile foundations are clustered to obtain multiple environmental clustering clusters; defect responsiveness and influencing factor calculation module: the defect responsiveness and defect influencing factor are calculated for each environmental clustering cluster; data collection module: the sensor data of the pile foundation to be detected, the pile foundation environmental data sequence and the pile foundation defect criterion table are collected; defect prediction module: the sensor data of the actual pile foundation is input into the defect prediction model to obtain the pile foundation defect prediction value; environmental cluster matching module: the environmental clustering cluster to which each pile foundation environmental data sequence belongs is calculated, and the corresponding defect responsiveness and defect influencing factor are collected; defect impact range and severity calculation module: the impact range and severity of pile foundation defects are calculated based on the sensor data of the pile foundation and the defect responsiveness and influencing factor of the environmental clustering cluster; pile foundation state assessment module: the bearing capacity and integrity of the pile foundation are judged based on the pile foundation defect criterion table, the pile foundation defect prediction value and the defect impact range and severity.

Images