Method for detecting construction quality of pile foundation rich in karst geology

Abstract

The invention discloses a karst-rich geological pile foundation construction quality detection method, and belongs to the technical field of engineering construction quality detection. The method comprises the following steps before construction: constructing a digital three-dimensional model of a site and a geologic structure; creating an ideal construction model based on the constructed digital three-dimensional model; a special tracing material is used in the construction process and is used for tracing the flowing state and the distribution condition of the pouring material in real time; a pressure sensor, a displacement sensor, a tilt angle sensor and a detector specially used for detecting a tracing material are uniformly arranged along a construction area to form an omnibearing monitoring network; in the construction process, monitoring data are analyzed in real time to recognize an abnormal mode, and construction parameters are dynamically adjusted according to the abnormal mode; the construction process is visually presented, and potential problems are predicted; the construction quality is comprehensively evaluated by utilizing a tracing material in combination with a multi-source detection technology, so that the existing defect problem is accurately positioned.

Description

Technical Field

[0001] The present application relates to the technical field of engineering construction quality inspection, and more specifically, to a method for inspecting the construction quality of pile foundations in karst-rich geology. Background Art

[0002] In modern construction, pile foundations, as an important part of infrastructure, are widely used in engineering projects under complex geological conditions such as karst geology. The quality of pile foundation construction directly affects the safety and service life of the building. Therefore, it is particularly important to ensure the quality of pile foundation construction.

[0003] Karst-rich geology is a special geological environment with complex geological structure, high degree of karst development, and a large number of cavities and cracks. This geological condition places higher demands on pile foundation construction, especially in terms of the driving depth of the pile foundation, the integrity of the pile body, and the interaction between the pile foundation and the surrounding soil, which can easily lead to substandard construction quality. Traditional pile foundation quality inspection methods mostly rely on static load tests, acoustic wave inspection and other technologies. These methods may be difficult to fully reflect the actual situation of the pile foundation under complex geological conditions, and may even miss potential quality problems.

[0004] In recent years, with the advancement of science and technology, modern detection technologies such as electromagnetic detection, ultrasonic detection, and acoustic emission monitoring have gradually been applied to the monitoring and evaluation of pile foundation construction quality. Although the application of these new technologies has improved the accuracy and efficiency of detection to a certain extent, under karst-rich geological conditions, due to the particularity of karst characteristics, existing detection methods still face many challenges. For example, how to accurately judge the impact of karst on pile foundations during the detection process, and how to monitor quality problems that may occur during the construction process in real time, are technical problems that need to be solved urgently.

[0005] In summary, how to effectively carry out pile foundation construction quality inspection under karst-rich geological conditions to ensure the reliability and effectiveness of construction quality has become a technical problem that needs to be solved urgently. Summary of the invention

[0006] In order to overcome a series of defects in the prior art, the purpose of this application is to provide a method for detecting the construction quality of pile foundations in karst-rich geology, including the following steps.

[0007] The following preparations are carried out before construction: high-frequency ultrasonic scanning technology is used to conduct a comprehensive inspection of the construction area, and a digital three-dimensional model of the site and geological structure is constructed; an ideal construction model is created based on the constructed digital three-dimensional model; special tracer materials are used during the construction process to track the flow state and distribution of the injection material in real time; pressure sensors, displacement sensors, inclination sensors and detectors specifically used to detect tracer materials are evenly arranged along the construction area to form a comprehensive monitoring network.

[0008] During the construction process, monitoring data is analyzed in real time to identify abnormal patterns, and construction parameters are dynamically adjusted accordingly to effectively respond to various changes and abnormal situations.

[0009] Compare and analyze real-time monitoring data with digital 3D models to visualize construction progress and predict potential problems.

[0010] When serious anomalies are detected or predicted, graded warnings are automatically issued and corresponding treatment suggestions are provided.

[0011] By using tracer materials and combining multi-source detection technology, a comprehensive assessment of construction quality can be conducted to accurately locate existing defects, thereby achieving comprehensive quality control and optimization of the entire construction process.

[0012] Furthermore, high-frequency ultrasonic scanning technology was used to conduct a comprehensive inspection of the construction area and construct a digital three-dimensional model of the site and geological structure. The specific steps are as follows.

[0013] According to the geological conditions, structural characteristics and expected detection depth, an ultrasonic probe with a suitable frequency is selected to obtain high-resolution images while ensuring sufficient penetration depth.

[0014] Remove impurities on the surface of the detection area that affect the scanning quality to ensure effective transmission of ultrasonic signals.

[0015] According to the preset scanning grid, the ultrasonic probe is used to perform a comprehensive scan and collect ultrasonic signals in real time, including waveform, amplitude and time parameters.

[0016] The collected ultrasonic signals are preprocessed to extract geological structure characteristic information.

[0017] Based on the pre-processed ultrasonic data, a 3D reconstruction algorithm is used to construct a digital model of the site and geological structure.

[0018] The extracted structural feature information is integrated into the digital model, and the key features are annotated in a visual way. At the same time, the detailed position coordinates, size information and related quantitative analysis results of each feature are provided.

[0019] Furthermore, the collected ultrasonic signal is preprocessed, including the following steps.

[0020] The ultrasonic data were cleaned to remove obvious outliers and interference signals.

[0021] Wavelet transform is used to reduce background noise and improve the signal-to-noise ratio.

[0022] Time alignment is performed on the ultrasonic signals after cleaning to ensure that the signals collected at different positions have the same time reference.

[0023] The signal is calibrated according to the known acoustic velocity and geometric parameters of the formation material to ensure the accuracy of the measurement results.

[0024] Furthermore, the geological structure characteristic information is extracted through the following steps.

[0025] Signal processing techniques are used to extract key features from preprocessed ultrasonic signals.

[0026] Based on the extracted key features, the time inversion algorithm is applied to reconstruct the propagation path of the sound wave in the formation by backpropagating the ultrasonic signal, so as to accurately locate the karst development position and structural characteristics.

[0027] The results of the time inversion algorithm are combined with a pre-established feature library to identify and classify geological structural features.

[0028] The identified structural feature information is combined with the geometric data of the construction area to generate detailed structural feature information.

[0029] Furthermore, based on the preprocessed ultrasonic data, a 3D reconstruction algorithm is used to construct a digital model of the site and geological structure, including the following steps.

[0030] The preprocessed ultrasonic data are segmented according to the scanning grid and position information, and spatial registration is performed using an iterative closest point algorithm to accurately match the segmented ultrasonic data segments to the actual geological structure locations.

[0031] Based on the registered ultrasonic data, the coordinates of each sampling point in three-dimensional space are calculated to generate a dense point cloud representing the site and structure, and preliminary noise removal and smoothing are performed at the same time.

[0032] Using the moving least squares method, point cloud data is converted into a continuous surface model to ensure accurate reflection of geological and structural features, and segmented reconstruction and splicing are performed as needed.

[0033] The underground structural features are reconstructed based on the depth information and reflection characteristics in the ultrasonic data. At the same time, the implicit function representation method is used to capture the karst development characteristics or other structural details and integrate them into the surface model to form a three-dimensional geometric model.

[0034] Optimize the 3D geometry model by using mesh simplification algorithms to reduce unnecessary details while retaining key features; apply smoothing algorithms to reduce surface noise; and use edge-preserving filters to enhance important geometric features.

[0035] Material density, thickness variation or defect location information is mapped as attributes onto the optimized 3D geometric model and semantic labels are added to generate the final digital 3D model.

[0036] Furthermore, the segmented ultrasonic data segments are accurately mapped to the actual geological structure positions by optimizing the following objective function: Minimize(∑ i=1 n ‖p i -T(q i )‖ 2 +λ∑ j=1 m ‖r j -T(s j )‖ 2 ), where T represents the transformation matrix used to align the ultrasound data segment to the three-dimensional geometric model; n represents the total number of points in the ultrasound data segment that need to be registered; p i Represents points in the actual geological structure, which are used to align with points in the ultrasonic data segment; i represents a point in the ultrasonic data segment; T(q i ) represents point q i The position after transformation by the transformation matrix T; λ is the balance factor used to balance the influence of spatial registration and geometric constraints; m represents the number of additional geometric constraint points; r j Represents the target geometric constraint point; s j Represents the geometric constraint point; T(s j ) means s j The position after transformation by transformation matrix T.

[0037] Further, the point cloud data is converted into a continuous surface model using a moving least squares method, which includes the following steps.

[0038] For each point p in the point cloud d , defining its local neighborhood Nd based on a fixed radius.

[0039] For each neighborhood point p g ∈Nd calculate the weight w(p d ,p g ), the specific formula is: w(p d ,p g )=exp(-‖p g -p d ‖2 / (2σ 2 )), where σ 2 is the variance, which determines the distance point p d The influence of the distance on the weight; ‖p g -p d ‖ is point p d With p g The distance between.

[0040] Fit p by minimizing the following objective function d Local plane: min a,b,c Σ g∈Nd w(p d ,p g )·(z g -(ax g +by g +c)) 2 , where (x g ,y g ) is point p g The coordinates in the local coordinate system; a, b, c are the plane parameters to be fitted; z g Represents point p g height.

[0041] For each point p d The plane is fitted to the points in the neighborhood of to generate a continuous surface model of the point cloud.

[0042] All points p d The fitting plane parameters (a, b, c) are applied to the entire point cloud data to construct a smooth surface model.

[0043] Furthermore, creating an ideal construction model based on the constructed digital three-dimensional model includes the following steps.

[0044] Import the constructed digital 3D model into professional engineering analysis software, combine it with engineering specifications and construction requirements, and identify the key factors affecting construction quality.

[0045] The construction parameters of pile foundation in karst-rich geology are set based on the identified key factors, including pile diameter, pile length, pouring material ratio and pouring pressure.

[0046] Iterative optimization is performed through professional analysis software to achieve theoretically optimal construction results while meeting design specifications.

[0047] Use the optimized parameters to perform virtual construction simulation on the digital 3D model, analyze possible abnormal situations, determine their causes, impact range and corresponding response strategies.

[0048] Based on engineering specifications, construction process requirements and simulation results, a comprehensive quality control indicator system is established, and an allowable error range is set for each indicator.

[0049] Integrate optimized construction parameters, virtual construction simulation results, abnormal situation response strategies and quality control indicator system to build a complete ideal construction model.

[0050] Furthermore, during the construction process, the monitoring data is analyzed in real time to identify abnormal patterns, and the construction parameters are dynamically adjusted accordingly, including the following steps.

[0051] Based on project characteristics, geological conditions and historical experience, normal operating parameter ranges and abnormal alarm thresholds are set for each monitoring indicator.

[0052] The real-time data collected by the monitoring network is quickly analyzed, including trend analysis and correlation analysis, and the analysis results are presented in an intuitive way to facilitate operators to quickly understand the current construction status.

[0053] Cluster analysis algorithms are used to conduct in-depth analysis of real-time data to identify possible abnormal patterns. When an anomaly is detected, an alarm is immediately issued and the specific location, type and severity of the anomaly are provided.

[0054] Based on the identified abnormal patterns and the preset response plan library, corresponding adjustment strategies can be quickly formulated and executed.

[0055] Furthermore, the real-time monitoring data is compared and analyzed with the digital 3D model to visually present the construction progress and predict potential problems, including the following steps.

[0056] Dynamically map real-time monitoring data to the digital 3D model to achieve instant update of model parameters and ensure that the model accurately reflects the current construction status.

[0057] The updated digital 3D model is compared with the expected construction progress in real time, and the differences between the actual construction status and the plan are clearly displayed through color coding.

[0058] Predict construction status changes over a period of time in the future and dynamically display the prediction results in the 3D model.

[0059] Based on the comparison analysis and prediction results, potential problem areas are automatically identified and marked in the 3D model, and a detailed problem description is generated.

[0060] Generates comprehensive reports with 3D visualizations, data charts, and text descriptions, while providing decision recommendations based on potential issues identified.

[0061] Furthermore, the following formula is used to comprehensively predict the changes in construction status in the future: Y^ t+h =α·F t+h ETS +(1-α)·(β0+β1·(t+h)), where, F t+h ETS is the predicted value of the exponential smoothing model; α is the weighting coefficient; Y^ t+h is the predicted value at time t+h; t is the current observation time point; h is the predicted time span; β0 and β1 are regression coefficients.

[0062] Furthermore, when a serious anomaly is detected or predicted, a graded warning is automatically issued and corresponding processing suggestions are provided, including the following steps.

[0063] All sensor data including pressure, flow, temperature and electromagnetic wave feedback from nano-tracers are continuously monitored.

[0064] Identify data points that deviate from normal ranges and, using a combination of historical data and expert knowledge, assess the severity of detected anomalies.

[0065] Based on currently detected abnormal patterns and historical data, predictive analytical models are applied to predict potential problems and assess their severity, probability of occurrence, and economic and safety impact, identifying risks in advance for intervention.

[0066] Classification is performed based on the anomaly detection and prediction results combined with the preset classification standards.

[0067] Detailed warning information is automatically generated based on the grading results, including anomaly description, cause analysis, potential impact assessment and warning level, and combined with the real-time updated three-dimensional digital model to intuitively display the location and impact range of the anomaly or predicted problem.

[0068] Based on the detected anomaly type and warning level, corresponding processing suggestions are extracted from the pre-established knowledge base, including immediate measures, resources to be mobilized, potential solutions, and security issues that need attention.

[0069] Send early warning information and handling suggestions to relevant personnel through preset communication channels.

[0070] Furthermore, the k-nearest neighbor method is used to identify data points that deviate from the normal range. The specific steps include: given a data point X u , calculate the average distance D to its K nearest neighbors u , the specific formula is: D u =(1 / K)·∑ p=1 K ‖X u -X p‖, where D u is the data point X u The average distance of X p is a given data point X u The pth nearest neighbor of ; ‖·‖ is the Euclidean distance; set a judgment threshold; if D u > judgment threshold, then X u Data points that are judged to be outside the normal range.

[0071] Furthermore, the construction quality is comprehensively evaluated by using tracer materials in combination with multi-source detection technology to accurately locate existing defects, thereby achieving comprehensive quality control and optimization of the entire construction process, including the following steps.

[0072] The tracer material signal feedback is collected and the collected data is denoised and enhanced, and a three-dimensional image is reconstructed through a tomography algorithm.

[0073] Apply computer vision and deep learning algorithms to automatically detect and locate anomalies in 3D images and provide precise 3D coordinates and ranges.

[0074] Compare the actual construction situation with the ideal construction model, quantify key quality indicators, and give a comprehensive score based on quality standards.

[0075] Use mechanical models to dynamically simulate the construction process, predict problem areas and optimization strategies, and evaluate the effects of different construction solutions.

[0076] Automatically generate an evaluation report containing 3D visualization, anomaly information, quality scores, and simulation results, and provide optimization suggestions to improve construction quality and provide lessons learned.

[0077] Compared with the prior art, the present application has the following beneficial effects.

[0078] This application combines high-frequency ultrasonic scanning technology and a multi-sensor monitoring network to track the flow of injection materials through tracer materials, and adjust construction parameters in real time to ensure quality control and optimization during the construction of karst-rich geological pile foundations. At the same time, a three-dimensional digital model is used to intuitively present the construction progress and predict potential problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 The present invention is a flowchart of a method for detecting the construction quality of pile foundations in karst-rich geology disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0080] In order to make the purpose, technical scheme and advantages of the implementation of the present invention clearer, the technical scheme in the embodiment of the present invention will be described in more detail below in conjunction with the drawings in the embodiment of the present invention. In the drawings, the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The described embodiments are part of the embodiments of the present invention, not all of the embodiments.

[0081] Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in the field without making any creative work shall fall within the scope of protection of the present invention.

[0082] The embodiments and directional terms described below with reference to the accompanying drawings are exemplary and intended to be used to explain the present invention, but should not be construed as limiting the present invention.

[0083] like Figure 1 As shown, a method for detecting the construction quality of pile foundation in karst-rich geology includes the following steps.

[0084] The following preparations are carried out before construction: high-frequency ultrasonic scanning technology is used to conduct a comprehensive inspection of the construction area, and a digital three-dimensional model of the site and geological structure is constructed; an ideal construction model is created based on the constructed digital three-dimensional model; special tracer materials are used during the construction of pile foundations in karst-rich geology to track the flow state and distribution of the injection material in real time; pressure sensors, displacement sensors, inclination sensors and detectors specifically used to detect tracer materials are evenly arranged along the construction area to form a comprehensive monitoring network.

[0085] During the construction process, monitoring data is analyzed in real time to identify abnormal patterns, and construction parameters are dynamically adjusted accordingly to effectively respond to various changes and abnormal situations.

[0086] Compare and analyze real-time monitoring data with digital 3D models to visualize construction progress and predict potential problems.

[0087] When serious anomalies are detected or predicted, graded warnings are automatically issued and corresponding treatment suggestions are provided.

[0088] By using tracer materials and combining multi-source detection technology, a comprehensive assessment of construction quality can be conducted to accurately locate existing defects, thereby achieving comprehensive quality control and optimization of the entire construction process.

[0089] Step 1 is to carry out comprehensive preparations before construction. By using high-frequency ultrasonic scanning technology, the geological structure of the construction area can be accurately detected and a detailed digital three-dimensional model can be constructed. The model not only includes the surface structure, but also covers the distribution of underground rock formations, which greatly enhances the understanding of the geological environment. Such a digital three-dimensional model provides a reliable foundation for subsequent construction, so that an ideal construction model can be created on the premise of fully understanding the site conditions. This process effectively reduces the construction risks caused by unknown geological conditions and ensures the safety and effectiveness of construction activities. In addition, this high-precision digital model also provides an important basis for the design and planning of construction, so that each step in the construction process can be accurately connected to the actual geological conditions, thereby achieving effective control of construction quality. In the construction process of karst-rich geological pile foundations, special tracer materials are used to track the flow state and distribution of the injection material in real time, further enhancing the visualization and monitoring capabilities of the construction. The tracer material can provide direct flow feedback during the injection process, helping construction personnel to understand the distribution of the material in the pile in a timely manner. Through real-time monitoring of the flow state, the construction team can better judge the injection effect of the material and ensure its uniformity and integrity. This process helps to promptly detect and resolve possible blockages or poor flow problems, thereby ensuring stable construction quality. In addition, the use of tracer materials combined with the deployment of other sensors forms a comprehensive monitoring network, greatly improving the transparency and controllability of the construction process and laying a solid foundation for subsequent quality assessment.

[0090] Step 2: Through advanced data analysis technology, construction units can quickly identify abnormal fluctuations in monitoring data and take necessary adjustment measures in time to deal with emergencies. This dynamic adjustment capability significantly improves the adaptability of the construction process and reduces quality risks caused by environmental changes or improper construction operations. By combining real-time monitoring with data analysis, the construction team can flexibly adjust construction parameters in the face of various complex situations, thereby achieving refined management and operation to ensure the quality and progress of construction.

[0091] The comparative analysis provided in step 3 not only helps the construction team to clearly understand the current construction status, but also provides data support for the prediction of potential problems. Through intuitive data display, construction personnel can better identify deviations and potential risks in construction, providing a strong basis for subsequent decision-making. The advantage of this process is that by establishing a relationship between digital models and real-time data, more scientific construction management can be achieved to ensure the continuous optimization and improvement of construction quality.

[0092] In step 4, when a serious anomaly is detected or predicted, this method also has the function of automatically issuing a graded warning and providing corresponding treatment suggestions in a timely manner. This mechanism can form a complete emergency response system during the construction process. When unforeseen problems arise, it can respond quickly and intervene to avoid more serious consequences. Through the intelligent early warning mechanism, construction units can effectively reduce quality risks and ensure construction safety. At the same time, the existence of a graded early warning mechanism also enhances the construction team's ability to respond, making the handling of emergencies more efficient and professional.

[0093] In step 5, the construction quality is comprehensively evaluated using tracer materials combined with multi-source detection technology, which makes it possible to accurately locate defects. This process not only achieves comprehensive quality control and optimization of the entire construction process, but also provides valuable reference data for subsequent maintenance and management. Through the fusion analysis of multi-source data, the construction team can fully grasp the construction quality, discover and solve potential defects in a timely manner, and thus improve the overall quality level of the construction. This comprehensive quality assessment mechanism based on high-tech means not only improves construction efficiency, but also provides experience and technical support for future engineering construction.

[0094] In summary, this method of pile foundation construction quality inspection in karst-rich geology has formed a complete construction quality control system through high-frequency ultrasonic scanning, application of tracer materials, real-time monitoring data analysis, digital model comparison and intelligent early warning mechanism. This not only improves the safety and effectiveness of the construction process, but also realizes comprehensive monitoring and optimization of construction quality, providing an important technical reference for future construction management. Through the comprehensive application of these technical means, construction units can maintain the stability and reliability of construction quality in a dynamic and complex construction environment, and promote the efficiency and safety of engineering construction.

[0095] Furthermore, high-frequency ultrasonic scanning technology was used to conduct a comprehensive inspection of the construction area and construct a digital three-dimensional model of the site and geological structure. The specific steps are as follows.

[0096] According to geological conditions, structural characteristics and expected detection depth, select an ultrasonic probe with a suitable frequency to obtain high-resolution images while ensuring sufficient penetration depth. The choice of frequency directly affects the penetration depth and resolution of ultrasound. Lower-frequency ultrasound has better penetration ability and is suitable for deep detection; while higher frequencies can provide clearer images and are suitable for detailed analysis of shallow structures. Through reasonable frequency selection, the advantages of ultrasonic technology can be fully utilized in different construction environments to obtain high-quality detection data.

[0097] Remove impurities on the surface of the inspection area that affect the scanning quality to ensure the effective transmission of the ultrasonic signal. At the construction site, there are often surface impurities, oil stains or other foreign matter, which will affect the transmission efficiency of the ultrasonic signal. For this reason, the inspection area must be cleaned before the inspection to remove impurities that affect the scanning quality. This pre-processing step ensures that the ultrasonic signal can effectively penetrate the material and be transmitted to the probe, thereby ensuring the accuracy and reliability of the data.

[0098] According to the preset scanning grid, the ultrasonic probe is used to perform a comprehensive scan and collect ultrasonic signals in real time, including waveform, amplitude and time parameters. This step provides a full range of detection of the construction area. Through the continuous monitoring of the collected data, it can be ensured that no important geological features are missed. In this process, the collection and storage of real-time data not only improves the detection efficiency, but also provides a sufficient basis for subsequent data processing.

[0099] The collected ultrasonic signals are preprocessed to extract geological structure feature information. The preprocessing process includes steps such as denoising, signal smoothing, and feature extraction, which are crucial to improving the quality of the data. By removing environmental noise and interference, effective geological information is extracted to ensure the accuracy of subsequent analysis. This process not only reduces the error rate in subsequent analysis, but also improves the accuracy of the final model.

[0100] Based on the pre-processed ultrasonic data, a 3D reconstruction algorithm is used to construct a digital model of the site and geological structure. The 3D reconstruction algorithm can convert the 2D ultrasonic signal into a 3D visualization model, and present the overall shape of the geological structure through point cloud data, voxel model, etc. This digital model can intuitively reflect the geological characteristics and help the construction team to comprehensively evaluate and analyze the site conditions.

[0101] The extracted structural feature information is integrated into the digital model, key features are visually annotated, and detailed location coordinates, size information and related quantitative analysis results of each feature are provided. Through this detailed annotation and analysis, construction personnel can quickly identify the characteristics of geological structures and make accurate judgments in subsequent construction. This process not only improves the readability of the data, but also provides rich information support for the construction team, enabling them to effectively avoid potential risks during the actual construction process and ensure construction quality and safety.

[0102] In summary, the process of using high-frequency ultrasonic scanning technology to conduct a comprehensive inspection of the construction area and build a digital 3D model involves several key steps. Each step reflects the potential for the application of high-tech means in modern construction quality control. Through the reasonable selection of ultrasonic frequency, cleaning of impurities, comprehensive scanning, data preprocessing, 3D reconstruction and information integration, this method not only improves the accuracy and efficiency of geological detection, but also provides strong support for risk assessment and decision-making at the construction site.

[0103] Furthermore, the collected ultrasonic signal is preprocessed, including the following steps.

[0104] Clean the ultrasonic data to remove obvious outliers and interference signals. This step is a key link to ensure the accuracy and reliability of subsequent data analysis. First of all, cleaning the ultrasonic data is the first step of preprocessing, which aims to remove obvious outliers and interference signals to ensure the quality of the data. Due to the complex environment of the construction site, ultrasonic signals may be interfered by various factors during propagation, such as equipment failure, environmental noise or other external influences. By identifying and deleting these abnormal data, the preprocessing process can significantly improve the effectiveness of subsequent analysis. The cleaned data set can not only reflect a more realistic geological condition, but also reduce the risk of errors in the data analysis process, providing a reliable basis for subsequent processing.

[0105] Wavelet transform is used to reduce background noise and improve the signal-to-noise ratio. Wavelet transform technology is widely used in the field of signal processing due to its good time-frequency localization ability. In ultrasonic signal processing, wavelet transform can effectively decompose the signal into different frequency components, thereby achieving accurate noise suppression. By removing unnecessary high-frequency noise and retaining the main signal components, wavelet transform significantly improves the signal-to-noise ratio of the signal and enhances the clarity and recognizability of the signal. This step ensures that subsequent analysis can more accurately extract geological features and further improve the quality of digital 3D model construction.

[0106] Time alignment is performed on the cleaned ultrasonic signals to ensure that the signals collected at different locations have the same time reference. The purpose of time alignment is to ensure that the signals collected at different locations have the same time reference so that the signals can be effectively compared and analyzed. Since there may be differences in signal propagation time when ultrasonic waves are collected at different locations, it is necessary to correct each signal using a corresponding algorithm. Time alignment not only improves the comparability between signals, but also provides a unified time frame for subsequent feature extraction and analysis, ensuring the integrity and consistency of the data. This process is crucial for the accurate identification of geological features and can avoid analysis errors caused by time differences.

[0107] The signal is calibrated according to the known velocity and geometry of the formation material to ensure the accuracy of the measurement results. In ultrasonic testing, the velocity of sound varies from formation material to formation material, so when analyzing the signal, it must be calibrated to obtain accurate measurement results under different geological conditions. By combining the known velocity of sound and material geometry parameters, the signal is adjusted and corrected accordingly, which can significantly improve the accuracy of the test data. This calibration process helps to convert the ultrasonic signal into a true value corresponding to the actual geological features, thereby enhancing the credibility and application value of the digital 3D model.

[0108] In summary, the preprocessing step of ultrasonic signals is crucial to ensure the accuracy and reliability of subsequent data analysis. Through a series of operations such as data cleaning, wavelet transform to reduce noise, time alignment and signal calibration, the preprocessing process can effectively improve the quality and availability of signals. This not only provides the necessary foundation for the accurate identification of geological structures, but also provides a scientific basis for the subsequent management and decision-making of construction. In the context of the current complex construction environment and the requirement for high efficiency, this efficient ultrasonic signal preprocessing method is particularly important and will help promote the further development and application of construction engineering detection technology.

[0109] Furthermore, the geological structure characteristic information is extracted through the following steps.

[0110] Signal processing technology is used to extract key features from the preprocessed ultrasonic signal. Common signal processing techniques include Fourier transform, wavelet transform and time domain analysis. These technologies can effectively identify important components in the signal, such as reflected waves, refracted waves and scattered waves, so as to extract key features representing geological structures. By efficiently processing the signal, the extracted feature information not only includes the depth, thickness and boundary position of the stratum, but also reflects the physical properties of the material (such as density and elastic modulus), providing an important basis for subsequent analysis.

[0111] Based on the extracted key features, the time inversion algorithm is applied to reconstruct the propagation path of the sound wave in the stratum by back-propagating the ultrasonic signal, so as to accurately locate the karst development location and structural characteristics. This process utilizes the time delay characteristics of the sound wave, allowing us to trace back the propagation path of the sound wave based on the detected signal. Through in-depth analysis of the propagation characteristics of the sound wave, the development location and scale of the karst can be accurately identified, providing valuable information support for construction. The advantage of this algorithm lies in its efficiency and accuracy, which enables the complex geological environment to be effectively interpreted and promotes the development of geological exploration technology.

[0112] The results of the time inversion algorithm are combined with the pre-established feature library to identify and classify geological structure features. The feature library contains rich information about known geological conditions and related features. Using this database, the collected signals can be compared and analyzed. By matching with the feature library, different types of geological structures, such as karst, rock layers, and faults, can be identified and classified. This combined technology not only improves the efficiency and accuracy of geological analysis, but also provides a scientific basis for subsequent engineering design and construction.

[0113] The identified structural feature information is combined with the geometric data of the construction area to generate detailed structural feature information. In this process, the geometric data of the construction area includes information such as the shape, size, depth and material properties of the site. Combining the extracted geological structure feature information with these geometric data can generate a more detailed geological structure map, which intuitively shows the overall situation of the construction site. This comprehensive analysis not only helps the construction team better understand the site conditions, but also provides strong support for subsequent construction decisions, thereby effectively reducing construction risks and improving construction efficiency.

[0114] In summary, the process of extracting geological structure feature information includes several key steps, such as applying signal processing technology, using time inversion algorithm, combining feature library for identification and classification, and integrating with geometric data of construction area. These steps not only complement each other to form a systematic analysis process, but also significantly improve the accuracy and reliability of geological detection. Through this method, the construction team can fully understand the geological conditions of the construction area and make scientific decisions, thereby promoting the smooth progress of the construction project. In modern construction projects, in-depth geological analysis and accurate feature extraction are important guarantees to ensure construction safety and efficiency, which also provides new ideas and directions for more efficient engineering management in the future.

[0115] Furthermore, based on the preprocessed ultrasonic data, a 3D reconstruction algorithm is used to construct a digital model of the site and geological structure, including the following steps.

[0116] The pre-processed ultrasonic data is segmented according to the scanning grid and position information, and the iterative closest point algorithm is used for spatial registration to accurately match the segmented ultrasonic data segments to the actual geological structure location. The iterative closest point (ICP) algorithm gradually adjusts the position and direction between the data segments to closely match the actual geological location, thereby accurately mapping the segmented ultrasonic data to the geological structure in three-dimensional space. This operation can greatly improve the matching degree between the data and the actual site conditions, providing high-quality input data for subsequent modeling.

[0117] Based on the registered ultrasonic data, the coordinates of each sampling point in three-dimensional space are calculated to generate a dense point cloud representing the site and structure, and preliminary noise removal and smoothing are performed at the same time. The generation of the point cloud not only includes surface sampling, but also goes deep into the internal structure of the stratum, reflecting the specific position of each ultrasonic sampling point in space. The construction of this dense point cloud enables the geological structure information of the construction site to be visualized in three dimensions, and the preliminary noise removal and smoothing ensure the quality of the data, thereby reducing errors and unnecessary complexity. The processing at this stage provides high-precision basic data for the subsequent generation of three-dimensional surface models.

[0118] Use the moving least squares method to convert the point cloud data into a continuous surface model to ensure that the geological and structural characteristics are accurately reflected, and segmented reconstruction and splicing are performed as needed. The moving least squares method (MLS) generates a smooth and continuous surface model by calculating the least squares error between adjacent points to ensure that the model can accurately reflect the geological and structural characteristics. In this process, if the site geology is relatively complex, segmented reconstruction and splicing are required to ensure that the final generated surface model has high integrity and continuity. This step is important for accurately capturing key features in the geological environment, such as karst development or underground cavities.

[0119] The underground structural features are reconstructed based on the depth information and reflection characteristics in the ultrasonic data, and the implicit function representation is used to capture the karst development characteristics or other structural details and integrate them into the surface model to form a three-dimensional geometric model. The implicit function representation method can flexibly cope with complex geometric forms, such as irregular karst forms and cavities, to ensure the accuracy of the model. This process is particularly important for the quality inspection of karst pile construction, because it can clearly present the hidden structural characteristics of the construction site and help the construction team discover potential problems in advance.

[0120] Optimizing the three-dimensional geometric model includes: using a mesh simplification algorithm to reduce unnecessary details while retaining key features, which can effectively reduce the computational complexity of the model and improve processing efficiency; applying a smoothing algorithm to reduce surface noise, which can make the model smoother and reduce errors; and using an edge-preserving filter to enhance important geometric features, which ensures that key structural information is not lost while simplifying the model, which is especially important for the fine characterization of karst or defect edges.

[0121] The material density, thickness variation or defect location information is mapped as attributes to the optimized 3D geometric model, and semantic labels are added to generate the final digital 3D model. This operation closely combines the actual physical properties of the construction site with the geometric model, so that the model not only has geometric information, but also expresses key attributes related to construction quality. The integration of these attribute information provides the construction team with a complete display of geological features, and can further enhance the interpretability of the model through semantic labels to ensure its application effect in actual construction.

[0122] In summary, through the above series of steps, ultrasonic data from preprocessing, point cloud generation, 3D surface modeling to final optimization and attribute mapping, a complete digital 3D model is formed. Each step is closely linked, which not only improves the accuracy of the data, but also ensures the operability and practicality of the final model. The 3D model provides an accurate display of the construction site and its geological characteristics, which can help the construction team monitor and evaluate the construction quality in real time during the construction process, greatly reducing construction risks and improving construction efficiency.

[0123] Furthermore, the segmented ultrasonic data segments are accurately mapped to the actual geological structure positions by optimizing the following objective function: Minimize(∑ i=1 n ‖p i -T(q i )‖ 2 +λ∑ j=1 m ‖r j -T(s j )‖ 2 ), where T represents the transformation matrix used to align the ultrasound data segment to the three-dimensional geometric model; n represents the total number of points in the ultrasound data segment that need to be registered; p i Represents points in the actual geological structure, which are used to align with points in the ultrasonic data segment; i represents a point in the ultrasonic data segment; T(q i ) represents point q i The position after transformation by the transformation matrix T; λ is the balance factor used to balance the influence of spatial registration and geometric constraints; m represents the number of additional geometric constraint points; r j Represents the target geometric constraint point; s j Represents the geometric constraint point; T(s j ) means s jThe position after transformation by the transformation matrix T. The above optimization objective function ensures the accurate correspondence between the ultrasonic data segment and the actual geological structure by combining spatial registration and geometric constraints. In the case of high data noise or complex geological structure, the introduction of geometric constraints provides additional robustness, which can not only achieve high registration accuracy but also retain key structural features. In addition, the dynamic adjustment of the balance factor λ provides flexibility and can balance the priorities between different objectives according to actual needs. This optimization objective function provides strong mathematical support for three-dimensional reconstruction and can significantly improve the accuracy and efficiency of quality inspection during karst pile construction.

[0124] Further, the point cloud data is converted into a continuous surface model using a moving least squares method, which includes the following steps.

[0125] For each point p in the point cloud d , and define its local neighborhood Nd based on a fixed radius. This step is to ensure that the local area of ​​each point in the point cloud can be described and fitted by the information of other points in the neighborhood. By selecting an appropriate neighborhood radius, it can be ensured that the number of points in the neighborhood is large enough to provide reliable geometric features, while maintaining locality so that the fitted local plane can accurately reflect the micro-geometric features. This way of dividing the local neighborhood improves the algorithm's adaptability and can be flexibly adjusted according to the density and distribution characteristics of the point cloud, avoiding the problems of over-smoothing or local detail loss that may exist in the global processing method.

[0126] For each neighborhood point p g ∈Nd calculate the weight w(p d ,p g ), the specific formula is: w(p d ,p g )=exp(-‖p g -p d ‖ 2 / (2σ 2 )), where σ 2 is the variance, which determines the distance point p d The influence of the distance on the weight; ‖p g -p d ‖ is point p d With p g This process ensures that the closer the point is, the greater the impact on the fitting, while the farther the point is, the less impact it has on the fitting, thus avoiding the introduction of noise or irrelevant information by distant points. In this way, the ability to capture local features in the point cloud can be enhanced, making the fitting results more consistent with the details of the actual geological or structural surface. In addition, the variance parameter σ 2 The setting allows adjusting the speed of weight decay, which further improves the robustness to different point cloud densities and noise levels.

[0127] Fit p by minimizing the following objective function d Local plane: min a,b,c Σ g∈Nd w(p d ,p g )·(z g -(ax g +by g +c)) 2 , where (x g ,y g ) is point p g The coordinates in the local coordinate system; a, b, c are the plane parameters to be fitted; z g Represents point p g This step ensures that the points in each local neighborhood can be approximately described by a plane, so that a continuous geometric surface can be constructed. Since the objective function includes weights, the fitting result more accurately reflects the local structure and filters out the negative impact of distant points on the fitting. Through this local plane fitting, point clouds with complex geometric features, such as the geological structure of a construction site, can be effectively processed, avoiding the over-smoothing problem that may occur in global fitting.

[0128] For each point p d The plane is fitted to the points in the neighborhood to generate a continuous surface model of the point cloud. Since each point is fitted with a plane based on its local neighborhood information, this ensures that the point cloud surface model has overall smoothness and continuity while maintaining local details. The surface model generated in this way can accurately reflect the local characteristics of the actual geological structure, while avoiding the discreteness caused by noise or irregular sampling in the point cloud data. This result provides a smooth and realistic geometric surface for digital modeling and 3D reconstruction.

[0129] All points p d The fitting plane parameters (a, b, c) are applied to the entire point cloud data to construct a smooth surface model. This step not only integrates the results of local neighborhood fitting, but also achieves the continuity and consistency of the entire point cloud through global stitching. In this way, a smooth surface model can be generated that accurately describes complex three-dimensional structures while retaining key geometric features. The geometric surface after plane fitting can not only reflect the overall trend, but also capture local geometric changes and details, making the model highly practical in digital applications.

[0130] In summary, the moving least squares method ensures the accuracy, smoothness and geometric consistency of the point cloud model in the construction of the continuous surface of point cloud data through the steps of local neighborhood definition, weight calculation, local plane fitting, and continuous surface generation. Each step effectively controls the noise and data discreteness in the processing process, so that the final three-dimensional geometric model not only has the ability to accurately describe local details, but also maintains the smoothness and continuity of the overall structure. This process is particularly suitable for ultrasonic scanning data applied to construction sites, and can help build high-resolution three-dimensional geological structure models and promote digital modeling and structural detection in engineering applications.

[0131] Furthermore, creating an ideal construction model based on the constructed digital three-dimensional model includes the following steps.

[0132] Import the constructed digital 3D model into professional engineering analysis software, and identify the key factors affecting construction quality in combination with engineering specifications and construction requirements. The core of this step is to automatically or semi-automatically identify key factors that may affect construction quality, such as geological structure and karst development, in combination with the specifications of specific engineering projects. These factors include the unevenness of the foundation, the depth and size of the karst, the groundwater level, etc. Identifying them helps determine risk areas during the construction process and provides a basis for subsequent construction parameter setting and optimization. The innovation of this step is to digitize and systematize complex geological conditions information, thereby improving the scientificity and accuracy of construction planning.

[0133] Based on the identified key factors, the construction parameters of the pile foundation in karst-rich geology are set, including pile diameter, pile length, ratio of grouting materials and grouting pressure. Parameters such as pile diameter, pile length, ratio of grouting materials and grouting pressure directly affect the stability and bearing capacity during the construction process. When setting these parameters, it is necessary to balance the construction quality and engineering cost, and consider the uncertainties in karst-rich geology. Through reasonable parameter setting, the anti-settlement ability of the pile foundation can be effectively enhanced, the bonding effect of the grouting material can be improved, and the disturbance of the foundation during construction can be reduced, thereby ensuring the construction quality. The innovation of this process is that it provides precise construction parameters in complex geological environments, minimizing construction risks.

[0134] Iterative optimization is performed through professional analysis software to achieve the theoretically optimal construction effect while meeting the design specifications. This step uses simulation and calculation technology to simulate the construction effects under different parameter combinations to ensure that the optimal parameter combination is found while meeting the design specifications. Through such optimization, material waste and unnecessary costs can be reduced, while improving the durability and stability of the pile foundation. The innovation of iterative optimization is that through large-scale simulation, the accuracy of the design and the feasibility of the construction are significantly improved, ensuring the realization of the theoretically optimal construction plan.

[0135] Use the optimized parameters to perform virtual construction simulation on the digital 3D model, analyze possible abnormal situations, determine their causes, impact range and corresponding response strategies. Through virtual simulation, abnormal situations that may occur during the construction process can be discovered in advance, such as uneven pile foundation pouring, material settlement that does not meet expectations, etc. Combined with the results of virtual simulation, the root cause and impact range of the problem can be analyzed in detail, and response strategies can be formulated in advance. The innovation of this step is to use virtual construction to provide early warning and solutions for actual construction, thereby reducing the impact of emergencies during construction on the progress and quality of the project.

[0136] According to the engineering specifications, construction process requirements and simulation results, a comprehensive quality control indicator system is established, and an allowable error range is set for each indicator. This system covers all key indicators in pile foundation construction, such as pile foundation verticality, uniformity of grouting materials, pressure control, etc. The setting of the allowable error range for each indicator ensures that the construction quality control has a basis to follow, which is helpful for real-time monitoring and problem feedback at the construction site. The innovation of this step is to provide a systematic and standardized quality assurance mechanism for construction, which can maximize the stability and consistency of the construction process.

[0137] The optimized construction parameters, virtual construction simulation results, abnormal situation response strategies and quality control indicator system are integrated to build a complete ideal construction model. This ideal construction model can provide data support for each link in the construction process, ensuring that the construction team can accurately execute and monitor in real time to ensure construction quality. The innovation of this step is that it provides all-round guidance for construction under complex geological conditions, making the construction process more intelligent and efficient.

[0138] In summary, by constructing the various steps of an ideal construction model, the accuracy and safety of pile foundation construction under karst-rich geological conditions can be significantly improved. From the construction of a digital three-dimensional model, to the optimization of construction parameters, virtual construction simulation, and then to the establishment of a quality control system, the entire process forms a closely linked closed-loop mechanism. Such a model not only improves the scientificity and efficiency of construction, but also greatly reduces construction risks and quality issues through prior virtual simulation and optimization, providing solid technical support for the successful implementation of the project.

[0139] Furthermore, during the construction process, the monitoring data is analyzed in real time to identify abnormal patterns, and the construction parameters are dynamically adjusted accordingly, including the following steps.

[0140] Based on the project characteristics, geological conditions and historical experience, normal operating parameter ranges and abnormal alarm thresholds are set for each monitoring indicator. This step combines the previous geological exploration and engineering design data, and through scientific modeling, sets clear warning lines for abnormal situations that may occur during the construction process. The innovation of this process is to ensure that every link in the construction process is within the controllable range. At the same time, the setting of abnormal alarm thresholds helps the construction team to detect potential problems in a timely manner, thereby avoiding engineering quality problems caused by unnecessary errors.

[0141] The real-time data collected by the monitoring network is quickly analyzed, including trend analysis and correlation analysis, and the analysis results are presented in an intuitive way, so that operators can quickly understand the current construction status. Rapid trend analysis and correlation analysis of these data can help identify changing trends that may occur during the construction process. For example, a gradual change in injection pressure may indicate a change in material density, while material settlement may reflect changes in geological conditions. By presenting these analysis results to operators in an intuitive way such as charts or color coding, they can quickly understand the current construction status. The innovation of this step is that it greatly improves the readability and response speed of the data, ensuring that the construction team can make decisions in a timely manner, thereby improving the real-time control capabilities of the construction.

[0142] Cluster analysis algorithms are used to deeply analyze real-time data to identify possible abnormal patterns. When an abnormality is detected, an alarm is immediately sounded, and the specific location, type, and severity of the abnormality are provided. Cluster analysis can automatically group construction data according to different characteristics to identify possible abnormal patterns. The algorithm can identify potential construction anomalies, such as uneven distribution of cast materials, pile foundation deflection, or rapid material settlement. When an abnormality is detected, an alarm will be immediately sounded, detailing the specific location, type, and severity of the abnormality. The innovation of this process is that through the application of intelligent algorithms, the accuracy and response speed of anomaly detection are improved, and uncontrollable factors in the construction process are reduced.

[0143] According to the identified abnormal patterns, combined with the preset response solution library, the corresponding adjustment strategy is quickly formulated and executed. When the cluster analysis algorithm identifies the abnormal pattern, it will automatically combine the pre-established response solution library to quickly formulate the corresponding adjustment strategy. For example, if the density of the grouting material is detected to be too high, the grouting pressure or material ratio can be automatically adjusted, or if the pile foundation is skewed, the working angle of the construction equipment can be dynamically adjusted. These response strategies can be dynamically adjusted according to the changes in real-time data, thereby ensuring stability and accuracy during the construction process. The innovation of this step is that through automated and intelligent construction adjustments, the time delay of manual intervention is reduced, and the continuity of construction quality is ensured through rapid response.

[0144] In summary, during the construction process, by analyzing monitoring data in real time, quickly identifying abnormal patterns and dynamically adjusting construction parameters, the safety and quality control of construction can be significantly improved. Each step uses intelligent technology to improve the efficiency of data collection, processing and response, ensuring that every link in the construction process is in the best control state. This not only helps to reduce sudden risks during construction, but also optimizes construction efficiency by adjusting parameters in real time, making construction more flexible and controllable.

[0145] Furthermore, the real-time monitoring data is compared and analyzed with the digital 3D model to visually present the construction progress and predict potential problems, including the following steps.

[0146] Dynamically map real-time monitoring data to the digital 3D model to update model parameters instantly and ensure that the model accurately reflects the current construction status. This process seamlessly integrates data such as pile foundation settlement, material pressure, and temperature into the model through the monitoring network to ensure that construction personnel always have the latest situation on site. The innovation of this step is that it enhances the dynamics and accuracy of the model, avoids data lag, and enables the construction team to observe the construction progress in real time in a virtual environment and make adjustments based on real-time data.

[0147] The updated digital 3D model is compared with the expected construction progress in real time, and the difference between the actual construction situation and the plan is clearly displayed through color coding. The color coding intuitively visualizes the progress of the construction, such as marking the completed, delayed and upcoming parts with different colors. This comparison not only allows construction personnel to easily identify areas of progress deviation, but also provides immediate feedback for construction management. The innovation of this process is to improve the visualization and management efficiency of the construction progress, ensuring that the construction goals at each stage can be accurately achieved.

[0148] Predict changes in construction status over a period of time in the future and display the prediction results in a dynamic way in the 3D model. The innovation of this step is that it enhances the foresight and planning of construction. Construction personnel can identify possible delays or technical problems in advance and make timely adjustments to reduce the impact of emergencies on the construction period.

[0149] Based on the comparison analysis and prediction results, potential problem areas are automatically identified and marked in the 3D model, and a detailed description of the problem is generated. The innovation of this process lies in its automated and intelligent analysis capabilities, which can raise an alarm before the problem expands, avoiding the escalation of problems due to negligence during the construction process.

[0150] Generate a comprehensive report containing 3D visualization, data charts and text descriptions, and provide corresponding decision-making suggestions based on the potential problems identified. The innovation of this step is to transform complex construction problems into actionable suggestions through information integration and analysis, helping the construction team make more informed decisions and improving the scientificity and effectiveness of construction management.

[0151] In summary, comparing and analyzing real-time monitoring data with digital 3D models not only improves the visual management of the construction process, but also further enhances the forward-looking and risk control capabilities of construction through dynamic prediction and problem identification. Each step improves the efficiency of data processing through intelligent technology, allowing the construction team to understand the construction status, identify problems and make effective adjustments in a more intuitive, accurate and timely manner. This process significantly improves the transparency and responsiveness of construction management, making the construction process more controllable and reliable.

[0152] Furthermore, the following formula is used to comprehensively predict the changes in construction status in the future: Y^ t+h =α·F t+h ETS +(1-α)·(β0+β1·(t+h)), where, F t+h ETS is the predicted value of the exponential smoothing model; α is the weighting coefficient; Y^ t+h is the predicted value at time t+h; t is the current observation time point; h is the predicted time span; β0 and β1 are regression coefficients. This formula comprehensively predicts changes in construction status by combining the exponential smoothing model and the linear regression model. Exponential smoothing captures short-term trends, and linear regression handles long-term trends. The prediction results are not only suitable for short-term construction management, but also provide a reference for long-term planning. The technical effect of this combined model significantly improves the flexibility and accuracy of the prediction, providing an efficient decision support tool for the construction team.

[0153] Furthermore, when a serious anomaly is detected or predicted, a graded warning is automatically issued and corresponding processing suggestions are provided, including the following steps.

[0154] Continuously monitor all sensor data, including pressure, flow, temperature, and electromagnetic wave feedback from nanotracers. This step can fully capture subtle changes in the construction environment through high-frequency data collection and real-time updates, ensuring response speed and monitoring accuracy, and reducing delayed responses to abnormal phenomena. At the same time, the use of nanotracers also improves the ability to detect microscopic environments, thereby being sensitive to changes in the internal state of geology or materials.

[0155] Identify data points that deviate from the normal range, and at the same time, combine historical data and expert knowledge to evaluate the severity of the detected anomalies. This step can accurately identify potential hidden dangers in the construction process by setting reasonable monitoring thresholds and data anomaly models, especially early warning of data anomalies, which helps prevent further deterioration of abnormal problems. At the same time, the introduction of the expert knowledge base provides a reference for the severity assessment of anomalies, making anomaly judgment more scientific and accurate.

[0156] Based on currently detected abnormal patterns and historical data, predictive analysis models are applied to predict potential problems, assess their severity, probability of occurrence, and economic and safety impacts, and identify risks in advance for intervention. This step not only detects potential hazards early by predicting future risks, but also provides the construction team with sufficient time to take preventive measures to reduce the impact of sudden problems. This forward-looking assessment method improves the initiative of construction management and prevents delayed responses due to information lags.

[0157] According to the abnormal detection and prediction results, the classification is carried out in combination with the preset classification standards. The graded warning mechanism ensures that the urgency of the problem can be quickly identified, so that appropriate response measures can be taken for problems of different levels. In terms of the efficiency and flexibility of its response, the abnormalities are quickly classified through preset standards, making the matching of treatment plans more targeted, avoiding waste of resources or insufficient response. At the same time, the graded warning can also help the construction management team to reasonably arrange resources and personnel for corresponding processing, thereby improving the efficiency of problem response.

[0158] Based on the classification results, detailed warning information is automatically generated, including abnormal description, cause analysis, potential impact assessment and warning level, and combined with the real-time updated 3D digital model to intuitively display the location and impact range of abnormal or predicted problems. This step uses 3D visualization technology to enable the construction team to intuitively understand the specific location of the abnormality and its impact on the overall project, enhancing the visibility and comprehensibility of abnormal information and helping managers make more targeted decisions.

[0159] According to the detected abnormality type and warning level, corresponding processing suggestions are extracted from the pre-established knowledge base, including immediate measures, resources to be mobilized, potential solutions, and safety matters that need attention. The innovation of this step is that the knowledge base provides a solution library based on historical experience and expert knowledge, which can quickly provide response guidance to the construction team and reduce the decision-making time for personnel to deal with problems. The processing suggestions cover technical, resource and safety aspects, ensuring the comprehensiveness and operability of the processing, thereby greatly improving the efficiency of solving abnormal problems.

[0160] Early warning information and handling suggestions are sent to relevant personnel through preset communication channels. The innovation of this step is that the efficient information transmission is ensured through the automated communication mechanism, without manual intervention, and the automated management of the entire process is realized, which improves the emergency response speed and reduces the losses that may be caused by delayed information transmission.

[0161] In summary, through real-time monitoring, automatic identification and hierarchical management, efficient and intelligent support is provided for exception handling during the construction process. From exception detection and prediction to the automatic generation of response suggestions, the safety and quality of construction are guaranteed while improving the accuracy and efficiency of management. The visualization and automatic communication functions of the 3D model further optimize the information transmission and decision-making process, making the entire exception handling process more coherent and reliable, and effectively reducing the risks in construction.

[0162] Furthermore, the k-nearest neighbor method is used to identify data points that deviate from the normal range. The specific steps include: given a data point X u , calculate the average distance D to its K nearest neighbors u , the specific formula is: D u =(1 / K)·∑ p=1 K ‖X u -X p ‖, where D u is the data point X u The average distance of X p is a given data point X u The pth nearest neighbor of ; ‖·‖ is the Euclidean distance; set a judgment threshold; if D u > judgment threshold, then X u The K-Nearest Neighbors (KNN) method captures the local abnormality of data points through simple neighborhood distance calculation in anomaly detection, and can efficiently identify data points that deviate from the normal range in combination with the judgment threshold. The flexibility of this method makes it applicable to various types of data sets and application scenarios, especially when there is no clear distribution assumption.

[0163] Furthermore, the construction quality is comprehensively evaluated by using tracer materials in combination with multi-source detection technology to accurately locate existing defects, thereby achieving comprehensive quality control and optimization of the entire construction process, including the following steps.

[0164] Collect the signal feedback of the tracer material, denoise and enhance the collected data, and reconstruct the three-dimensional image through the tomography algorithm. During the construction process, the tracer material is used to obtain signal feedback from key positions. These signals will be affected by noise and other interference factors, so it is crucial to denoise and enhance the collected data. Through signal processing techniques such as filtering and denoising algorithms (such as wavelet transform and Kalman filtering), the signal-to-noise ratio of the data can be significantly improved. The tomography algorithm is used to reconstruct the processed signal into a three-dimensional image. The innovation of this process is that it can realize non-invasive analysis of the internal structure of the object and provide more comprehensive visual information. Through high-quality three-dimensional images, the assessment of construction quality will be more accurate, laying the foundation for subsequent defect detection.

[0165] Computer vision and deep learning algorithms are used to automatically detect and locate anomalies in 3D images and provide precise 3D coordinates and ranges. This step involves technologies such as target detection and image segmentation, using deep learning models such as convolutional neural networks (CNN) to identify defects in images. These models are trained to identify and classify various defects in construction and provide precise 3D coordinates and ranges. This automated detection method not only improves detection efficiency, but also reduces the error rate of manual operations, making defect location more accurate, which helps to implement precise subsequent processing.

[0166] Compare the actual construction situation with the ideal construction model, quantify key quality indicators, and give a comprehensive score based on the quality standards. By establishing an ideal construction model based on BIM (Building Information Model), a standard can be provided for the actual construction process. Using quality standards for comprehensive scoring can help identify deviations and deficiencies in the construction process. The innovation of this process is that it can clarify key quality indicators in construction, such as dimensional accuracy, material strength, and compliance of construction processes, thereby providing a clear basis for quality control.

[0167] Use mechanical models to dynamically simulate the construction process, predict problem areas and optimization strategies, and evaluate the effects of different construction solutions. This simulation can use methods such as finite element analysis (FEA) to deeply analyze problems such as stress concentration and displacement that may occur during the construction process. Through dynamic simulation, the effects of different construction solutions can be evaluated, possible defects can be identified in advance, and data-based decision support can be provided. The innovation of this process is that it can find the optimal construction solution through simulation analysis, reduce construction risks, and improve construction quality.

[0168] Automatically generate an evaluation report containing 3D visualization, anomaly information, quality score and simulation results, and provide optimization suggestions to improve construction quality and provide lessons learned. The evaluation report not only helps with quality control during the construction process, but also summarizes valuable lessons learned to guide the optimization of future construction projects. By combining the evaluation results with historical data, the construction quality can be further improved and provide reference for similar projects.

[0169] In summary, the use of tracer materials and multi-source detection technology for construction quality assessment can not only achieve comprehensive quality control of the construction process, but also accurately locate defects and provide a basis for subsequent optimization and improvement. Through the combination of signal processing, automatic detection, dynamic simulation and report generation, this process has technically achieved an efficient closed loop from data collection to quality assessment. Ultimately, the implementation of this mechanism will help improve construction quality, reduce risks and ensure the smooth completion of the project.

[0170] Finally, it should be pointed out that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the above embodiments, a person skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting the construction quality of pile foundation in karst geology. It is characterized in that The following steps are involved: The following preparations are carried out before construction: high-frequency ultrasonic scanning technology is used to conduct a comprehensive inspection of the construction area, and a digital 3D model of the site and geological structure is constructed; an ideal construction model is created based on the constructed digital 3D model; special tracer materials are used during the construction process to track the flow state and distribution of the injection material in real time; pressure sensors, displacement sensors, inclination sensors and detectors specifically used to detect tracer materials are evenly arranged along the construction area to form a comprehensive monitoring network; During the construction process, monitoring data is analyzed in real time to identify abnormal patterns and construction parameters are dynamically adjusted accordingly to effectively respond to various changes and abnormal situations; Compare and analyze real-time monitoring data with digital 3D models to visualize construction progress and predict potential problems; When a serious anomaly is detected or predicted, it automatically issues a graded warning and provides corresponding treatment suggestions; By using tracer materials and combining multi-source detection technology, a comprehensive assessment of construction quality can be conducted to accurately locate existing defects, thereby achieving comprehensive quality control and optimization of the entire construction process.

2. A method for detecting the construction quality of pile foundation in karst geology according to claim 1, It is characterized in that High-frequency ultrasonic scanning technology is used to conduct a comprehensive inspection of the construction area and build a digital three-dimensional model of the site and geological structure. The specific steps are as follows: According to geological conditions, structural characteristics and expected detection depth, select an ultrasonic probe with a suitable frequency to obtain high-resolution images while ensuring sufficient penetration depth; Remove impurities on the surface of the detection area that affect the scanning quality to ensure effective transmission of ultrasonic signals; According to the preset scanning grid, the ultrasonic probe is used to perform a comprehensive scan and collect ultrasonic signals in real time, including waveform, amplitude and time parameters; Pre-process the collected ultrasonic signals to extract geological structure characteristic information; Based on the pre-processed ultrasonic data, a 3D reconstruction algorithm is used to construct a digital model of the site and geological structure; The extracted structural feature information is integrated into the digital model, and the key features are annotated in a visual way. At the same time, the detailed position coordinates, size information and related quantitative analysis results of each feature are provided.

3. A method for detecting the construction quality of pile foundation in karst geology according to claim 2, It is characterized in that The collected ultrasonic signal is preprocessed, including the following steps: Clean the ultrasonic data to remove obvious outliers and interference signals; Apply wavelet transform to reduce background noise and improve signal-to-noise ratio; Time-align the ultrasonic signals after cleaning to ensure that the signals collected at different locations have the same time reference; The signal is calibrated according to the known acoustic velocity and geometric parameters of the formation material to ensure the accuracy of the measurement results.

4. A method for detecting the construction quality of pile foundation in karst geology according to claim 3, It is characterized in that The geological structure feature information is extracted through the following steps: Apply signal processing techniques to extract key features from pre-processed ultrasonic signals; Based on the extracted key features, the time inversion algorithm is applied to reconstruct the propagation path of the sound wave in the formation by back-propagating the ultrasonic signal, so as to accurately locate the karst development position and structural characteristics; Combine the results of the time inversion algorithm with a pre-established feature library to identify and classify geological structural features; The identified structural feature information is combined with the geometric data of the construction area to generate detailed structural feature information.

5. A method for detecting the construction quality of pile foundation in karst geology according to claim 2, It is characterized in that Based on the pre-processed ultrasonic data, a 3D reconstruction algorithm is used to construct a digital model of the site and geological structure, including the following steps: The pre-processed ultrasonic data is segmented according to the scanning grid and position information, and the iterative closest point algorithm is used for spatial registration to accurately correspond the segmented ultrasonic data segments to the actual geological structure positions; Based on the registered ultrasonic data, the coordinates of each sampling point in three-dimensional space are calculated to generate a dense point cloud representing the site and structure, and preliminary noise removal and smoothing are performed at the same time; Use the moving least squares method to convert point cloud data into a continuous surface model to ensure accurate reflection of geological and structural features, and perform segmented reconstruction and splicing as needed; Reconstruct underground structural features based on depth information and reflection characteristics in ultrasonic data, and use implicit function representation to capture karst development features or other structural details and integrate them into the surface model to form a three-dimensional geometric model; Optimize the 3D geometry model by using mesh simplification algorithms to reduce unnecessary details while retaining key features; apply smoothing algorithms to reduce surface noise; and use edge-preserving filters to enhance important geometric features. Material density, thickness variation or defect location information is mapped as attributes onto the optimized 3D geometric model and semantic labels are added to generate the final digital 3D model.

6. A method for detecting the construction quality of pile foundation in karst geology according to claim 5, It is characterized in that The segmented ultrasonic data segments are accurately mapped to the actual geological structure locations by optimizing the following objective function: Minimize(∑ i=1 n ‖p i -T(q i )‖ 2 +λ∑ j=1 m ‖r j -T(s j )‖ 2 ), where T represents the transformation matrix used to align the ultrasound data segment to the three-dimensional geometric model; n represents the total number of points in the ultrasound data segment that need to be registered; p i Represents points in the actual geological structure, which are used to align with points in the ultrasonic data segment; i represents a point in the ultrasonic data segment; T(q i ) represents point q i The position after transformation by the transformation matrix T; λ is the balance factor used to balance the influence of spatial registration and geometric constraints; m represents the number of additional geometric constraint points; r j Represents the target geometric constraint point; s j Represents the geometric constraint point; T(s j ) means s j The position after transformation by transformation matrix T.

7. A method for detecting the construction quality of pile foundation in karst geology according to claim 5, It is characterized in that Using the moving least squares method, the point cloud data is converted into a continuous surface model, including the following steps: For each point p in the point cloud d , define its local neighborhood Nd based on a fixed radius; For each neighborhood point p g ∈Nd calculate the weight w(p d ,p g ), the specific formula is: w(p d ,p g )=exp(-‖p g -p d ‖ 2 / (2σ 2 )), where σ 2 is the variance, which determines the distance point p d The influence of the distance on the weight; ‖p g -p d ‖ is point p d With p g The distance between Fit p by minimizing the following objective function d Local plane: min a,b,c Σ g∈Nd w(p d ,p g )·(z g -(ax g +by g +c)) 2 , where (x g ,y g ) is point p g The coordinates in the local coordinate system; a, b, c are the plane parameters to be fitted; z g Represents point p g Height; For each point p d Perform plane fitting on points in the neighborhood of to generate a continuous surface model of the point cloud; All points p d The fitting plane parameters (a, b, c) are applied to the entire point cloud data to construct a smooth surface model.

8. A method for detecting the construction quality of pile foundation in karst geology according to claim 1, It is characterized in that Creating an ideal construction model based on the constructed digital 3D model includes the following steps: Import the constructed digital 3D model into professional engineering analysis software, combine it with engineering specifications and construction requirements, and identify the key factors affecting construction quality; The construction parameters of pile foundation in karst-rich geology are set based on the identified key factors, including pile diameter, pile length, grouting material ratio and grouting pressure; Iterative optimization is performed through professional analysis software to achieve theoretically optimal construction results while meeting design specifications; Use optimized parameters to perform virtual construction simulation on the digital 3D model, analyze possible abnormal situations, determine their causes, impact range and corresponding response strategies; Establish a comprehensive quality control indicator system based on engineering specifications, construction process requirements and simulation results, and set an allowable error range for each indicator; Integrate optimized construction parameters, virtual construction simulation results, abnormal situation response strategies and quality control indicator system to build a complete ideal construction model.

9. A method for detecting the construction quality of pile foundation in karst geology according to claim 1, It is characterized in that During the construction process, monitoring data is analyzed in real time to identify abnormal patterns and dynamically adjust construction parameters accordingly, including the following steps: Based on project characteristics, geological conditions and historical experience, set normal operating parameter ranges and abnormal alarm thresholds for each monitoring indicator; Rapidly analyze the real-time data collected by the monitoring network, including trend analysis and correlation analysis, and present the analysis results in an intuitive way to help operators quickly understand the current construction status; Apply cluster analysis algorithms to deeply analyze real-time data to identify possible abnormal patterns. When an abnormality is detected, an alarm is immediately issued and the specific location, type and severity of the abnormality are provided; Based on the identified abnormal patterns and the preset response plan library, corresponding adjustment strategies can be quickly formulated and executed.

10. A method for detecting the construction quality of pile foundation in karst geology according to claim 1, It is characterized in that Compare and analyze real-time monitoring data with the digital 3D model to visualize construction progress and predict potential problems, including the following steps: Dynamically map real-time monitoring data to the digital 3D model to achieve instant updates of model parameters and ensure that the model accurately reflects the current construction status; Compare the updated digital 3D model with the expected construction progress in real time, and clearly display the differences between the actual construction status and the plan through color coding; Predict changes in construction status over a period of time in the future and dynamically display the prediction results in the 3D model; Based on the comparison analysis and prediction results, potential problem areas are automatically identified and marked in the 3D model, and a detailed problem description is generated; Generates comprehensive reports with 3D visualizations, data charts, and text descriptions, while providing decision recommendations based on potential issues identified.

11. A method for detecting the construction quality of pile foundation in karst geology according to claim 10, It is characterized in that The following formula is used to comprehensively predict the changes in construction status in the future: Y^ t+h =α·F t+h ETS +(1-α)·(β 0 +β 1 ·(t+h)), where F t+h ETS is the predicted value of the exponential smoothing model; α is the weighting coefficient; Y^ t+h is the predicted value at time t+h; t is the current observation time point; h is the predicted time span; β 0 and β 1 is the regression coefficient.

12. A method for detecting the construction quality of pile foundation in karst geology according to claim 1, It is characterized in that When a serious anomaly is detected or predicted, a graded warning is automatically issued and corresponding treatment suggestions are provided, including the following steps: Continuously monitor all sensor data, including pressure, flow, temperature, and electromagnetic wave feedback from nanotracers; Identify data points that deviate from normal ranges and, using a combination of historical data and expert knowledge, assess the severity of detected anomalies; Apply predictive analytics models to predict potential problems based on currently detected abnormal patterns and historical data, and assess their severity, probability of occurrence, and economic and safety impacts, identifying risks in advance for intervention; Classification is performed based on the abnormal detection and prediction results combined with the preset classification standards; Automatically generate detailed warning information based on the classification results, including abnormality description, cause analysis, potential impact assessment and warning level, and combine it with the real-time updated 3D digital model to intuitively display the location and impact range of abnormalities or predicted problems; Based on the detected anomaly type and warning level, corresponding processing suggestions are extracted from the pre-established knowledge base, including immediate measures, resources to be mobilized, potential solutions, and security matters that need attention; Send early warning information and handling suggestions to relevant personnel through preset communication channels.

13. A method for detecting the construction quality of pile foundation in karst geology according to claim 12, It is characterized in that Use the k nearest neighbor method to identify data points that deviate from the normal range. The specific steps include: given a data point X u , calculate the average distance D to its K nearest neighbors u , the specific formula is: D u =(1 / K)·∑ p=1 K ‖X u -X p ‖, where D u is the data point X u The average distance of X p is a given data point X u The pth nearest neighbor of ; ‖·‖ is the Euclidean distance; set a judgment threshold; if D u > judgment threshold, then X u Data points that are judged to be outside the normal range.

14. A method for detecting the construction quality of pile foundation in karst geology according to claim 1, It is characterized in that Using tracer materials and multi-source detection technology, the construction quality is comprehensively evaluated to accurately locate the existing defects, thereby achieving comprehensive quality control and optimization of the entire construction process, including the following steps: Collect tracer material signal feedback and perform denoising and enhancement processing on the collected data, and reconstruct a three-dimensional image through a tomography algorithm; Apply computer vision and deep learning algorithms to automatically detect and locate anomalies in 3D images and provide precise 3D coordinates and ranges; Compare the actual construction situation with the ideal construction model, quantify key quality indicators, and make a comprehensive score based on the quality standards; Use mechanical models to dynamically simulate the construction process, predict problem areas and optimization strategies, and evaluate the effects of different construction solutions; Automatically generate an evaluation report containing 3D visualization, anomaly information, quality scores, and simulation results, and provide optimization suggestions to improve construction quality and provide lessons learned.

Images