A method and system for intelligent foundation investigation and evaluation

Abstract

The application discloses a kind of intelligent foundation detection and evaluation method and system, the method includes: based on the situation of foundation and the target of detection, setting ground penetrating radar and micro electro mechanical sensor array;Data acquisition is carried out based on ground penetrating radar and micro electro mechanical sensor array, and ground penetrating radar data and micro electro mechanical sensor data are obtained;Ground penetrating radar data and micro electro mechanical sensor data are analyzed and processed;After analyzing, the ground penetrating radar data and micro electro mechanical sensor data are weighted and fused by space-time synchronization mechanism, and the fusion weight is dynamically adjusted according to soil type;Dynamic risk assessment model is established, and the risk assessment grade of the foundation to be monitored is obtained by inputting the fused multivariate data into the dynamic risk assessment model.The application obtains foundation information from multiple dimensions, can more accurately identify small cracks, cavities and other defects, and accurately monitor the small vibration, pressure and humidity change of foundation.

Description

Technical Field

[0001] This invention relates to the field of foundation detection technology, and in particular to a method and system for intelligent foundation detection and assessment. Background Technology

[0002] Foundation detection technologies are used to assess the health and stability of building foundations, as well as the characteristics of the surrounding soil environment. These detection methods are widely used in building safety assessments, maintenance inspections, and reinforcement design. Common foundation detection technologies include geological exploration methods, surface and subsurface physical exploration methods, drilling and sampling methods, ground penetrating radar (GPR) technology, and acoustic reflection methods.

[0003] While existing technologies for detecting building foundations have been widely applied and achieved significant results, they also have some problems and shortcomings. These problems mainly lie in detection accuracy, cost, applicability, and operational complexity. The following are some of the main issues:

[0004] 1. Detection Accuracy Issues: 1) Insufficient Accuracy of Detection Results: Existing foundation detection technologies (such as ground-penetrating radar, acoustic reflection method, and static cone penetration test) may not be able to provide high-precision assessments of foundation conditions in certain situations. For example, ground-penetrating radar signal propagation is limited when encountering complex soil layers or strong interference conditions, which may lead to a decrease in the detection accuracy of underground structures. 2) Depth Limitations: Some detection methods (such as ultrasonic testing and ground-penetrating radar) have low detection accuracy for areas with greater underground depth (such as deep or ultra-deep foundations), and cannot effectively reflect the actual condition of deep foundations.

[0005] 2. Depth and Coverage Issues: 1) Limited Detection Depth: While drilling methods can obtain accurate soil samples, their depth is limited and requires a considerable amount of time for on-site operations. For projects such as deep foundation pits and large buildings, existing technologies struggle to quickly and comprehensively cover the entire subgrade. 2) Insufficient Spatial Coverage: Existing detection methods mostly involve point sampling, making it difficult to comprehensively cover all areas of the foundation. Since the foundation may experience localized or uneven dynamic problems, a single detection method may miss certain factors.

[0006] Therefore, proposing a method and system for intelligent ground detection and assessment to solve the problems existing in the current technology is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0007] The purpose of this invention is to provide a method and system for intelligent foundation detection and assessment, which can acquire foundation information from multiple dimensions, more accurately identify defects such as tiny cracks and voids, and accurately monitor dynamic changes in the foundation such as minute vibrations, pressure, and humidity.

[0008] To achieve the above objectives, the present invention provides the following solution:

[0009] A method for intelligent ground-based detection and assessment includes the following steps:

[0010] S1. Based on the foundation conditions and the target to be detected, set up ground penetrating radar and microelectromechanical sensor array;

[0011] S2. Data acquisition is performed based on ground-penetrating radar and microelectromechanical sensor array to obtain radar data and microelectromechanical sensor data.

[0012] S3. For ground-penetrating radar data, a signal processing algorithm combining wavelet transform and time-frequency analysis is used for analysis and processing; for microelectromechanical sensor data, a signal processing algorithm combining Kalman filtering algorithm and multiple linear regression model is used for analysis and processing.

[0013] S4. The analyzed ground-penetrating radar data and microelectromechanical sensor data are weighted and fused through a spatiotemporal synchronization mechanism, and the fusion weight is dynamically adjusted according to the soil type.

[0014] S5. Establish a dynamic risk assessment model. Input the fused multivariate data into the dynamic risk assessment model to obtain the risk assessment level of the foundation to be monitored.

[0015] Preferably, in S1, a ground-penetrating radar with a multi-antenna array is used to transmit high-frequency electromagnetic waves into the ground and acquire underground information from multiple angles at the same time; the microelectromechanical sensor array includes an accelerometer, a pressure sensor and a humidity sensor, which are distributed and installed on the ground surface and in the shallow soil to form a three-dimensional monitoring network.

[0016] Preferably, step S2 further includes storing the multiple collected detection radar data and microelectromechanical sensor data into a preset historical ground database.

[0017] Preferably, in S3, for ground-penetrating radar data, a signal processing algorithm combining wavelet transform and time-frequency analysis is used for analysis and processing, specifically including:

[0018] The original time-domain signal is decomposed and denoised using wavelet transform, and then the energy distribution of the radar signal at different times and frequencies is obtained through time-frequency analysis. The wavelet transform formula is as follows:

[0019]

[0020] Among them, W f (a, b) are wavelet transform coefficients, f(t) is the original signal, a is the scaling factor, b is the translation factor, and ψ * () represents the complex conjugate of the wavelet basis functions.

[0021] Preferably, in S3, for microelectromechanical sensor data, a signal processing algorithm combining Kalman filtering and multiple linear regression models is used for analysis and processing, specifically including:

[0022] Accelerometer data was fused and state estimated using a Kalman filter algorithm; pressure and humidity sensor data were analyzed using a multiple linear regression model to establish their relationship with foundation stability; the formula for the multiple linear regression model is as follows:

[0023] y = β0 + β1x1 + β2x2 + ... + β n x n +∈

[0024] Where y is the foundation stability index, x1, x2, ..., x n β is the independent variable for pressure or humidity. i is the regression coefficient, and ∈ is the error term.

[0025] Preferably, in S5, the dynamic risk assessment model adopts a CNN-optimized multi-parameter model, which inputs the fused multi-data and historical foundation data stored in S2 into the dynamic risk assessment model to obtain the risk assessment level of the foundation to be monitored; the formula of the CNN-optimized multi-parameter model is as follows:

[0026] Risk_Score=W1*Crack_dersity+W2*Vibration_amp+W3*Pressure_dev

[0027] Where W1, W2, and W3 are weights, Risk_Score is the risk score, Crack_density is the crack density, Vibration_amp is the vibration amplitude, and Pressure_dev is the pressure deviation.

[0028] A system for intelligent foundation detection and assessment, employing any one of the intelligent foundation detection and assessment methods described above, includes:

[0029] The equipment installation module is used to set up ground-penetrating radar and microelectromechanical sensor arrays based on the foundation conditions and the target to be detected.

[0030] The data acquisition module is used to acquire data based on ground penetrating radar and microelectromechanical sensor array to obtain radar data and microelectromechanical sensor data.

[0031] The data analysis and processing module uses a signal processing algorithm combining wavelet transform and time-frequency analysis to analyze and process ground-penetrating radar data; and a signal processing algorithm combining Kalman filtering algorithm and multiple linear regression model to analyze and process microelectromechanical sensor data.

[0032] The data fusion module is used to perform weighted fusion of the analyzed ground-penetrating radar data and microelectromechanical sensor data through a spatiotemporal synchronization mechanism. The fusion weight is dynamically adjusted according to the soil type.

[0033] The risk assessment module is used to establish a dynamic risk assessment model. The fused multivariate data is input into the dynamic risk assessment model to obtain the risk assessment level of the foundation to be monitored.

[0034] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements a method for intelligent ground detection and assessment as described above.

[0035] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0036] (1) The system of this invention combines a ground-penetrating radar multi-antenna array and a microelectromechanical sensor array to acquire foundation information from multiple dimensions, enabling more accurate identification of minute defects such as cracks and cavities, and precise monitoring of minute vibrations, pressure, and humidity changes in the foundation. In practical applications, for cavities with a depth of 5-10 meters and a diameter greater than 0.1 meters, the detection accuracy can reach over 90%, which is a significant improvement compared to traditional single detection methods.

[0037] (2) The present invention uses a multi-antenna ground penetrating radar in conjunction with a distributed microelectromechanical sensor array to achieve comprehensive coverage from shallow to deep layers and from local to overall, making up for the shortcomings of traditional detection methods in terms of depth and spatial coverage.

[0038] (3) The method of the present invention reduces the detection costs of a large amount of manpower, equipment and sample collection required by traditional drilling methods. The automated data acquisition and processing system greatly shortens the detection cycle. For medium-sized building foundation detection projects, traditional methods may take several weeks, while the present invention can complete a comprehensive detection and evaluation in only 3-5 days. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 A schematic diagram of the method flow for intelligent ground-based detection and assessment based on multi-technology fusion provided by the present invention;

[0041] Figure 2 A schematic diagram of the system architecture for intelligent ground-based detection and assessment based on multi-technology fusion provided by the present invention. Detailed Implementation

[0042] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0043] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0044] This invention relates to multiple technical fields, including geophysical exploration technology, signal processing, machine learning, image processing, and engineering geology.

[0045] Geophysical Exploration (GE):

[0046] Ground penetrating radar (GPR) is a geophysical exploration technology widely used for detecting underground structures. GPR technology helps identify underground objects and structures by emitting electromagnetic waves and receiving their reflected waves. This technology is used in fields such as geological exploration, environmental monitoring, and infrastructure inspection.

[0047] Signal Processing and Analysis (SP):

[0048] This invention relates to a wide range of signal processing techniques, including time-domain signal analysis, reflected wave feature extraction, noise removal, and gain adjustment. Signal processing is the core component for acquiring and analyzing useful information from GPR data, employing techniques such as filtering, gain compensation, and peak detection.

[0049] Machine Learning and Data Analysis (MLDA):

[0050] The model training and optimization phases involve the automated analysis of GPR data using machine learning methods (such as support vector machines and convolutional neural networks). This falls under data analysis and intelligent recognition technologies; machine learning methods can help extract features from complex data and perform pattern recognition and classification.

[0051] Computer Vision and Image Processing (CVIP):

[0052] Visualization techniques (such as depth profiles, heatmaps, and contour maps) are used to graphically present processed GPR data, helping to intuitively understand underground structures. This part involves computer vision and image processing technologies.

[0053] Engineering Geology and Infrastructure Inspection (EGI):

[0054] GPR (Geological Profiling) is widely used in the field of engineering geology, especially in the inspection and assessment of infrastructure such as building foundations, roads, and bridges, to check foundation conditions and structural safety. The technical applications involved here mainly serve the exploration and risk assessment of engineering projects.

[0055] Reference Figure 1 As shown, the present invention provides a method for intelligent ground-based detection and assessment, comprising the following steps:

[0056] S1. Based on the foundation conditions and the target to be detected, set up ground penetrating radar and microelectromechanical sensor array;

[0057] S2. Data acquisition is performed based on ground-penetrating radar and microelectromechanical sensor array to obtain radar data and microelectromechanical sensor data.

[0058] S3. For ground-penetrating radar data, a signal processing algorithm combining wavelet transform and time-frequency analysis is used for analysis and processing; for microelectromechanical sensor data, a signal processing algorithm combining Kalman filtering algorithm and multiple linear regression model is used for analysis and processing.

[0059] S4. The analyzed ground-penetrating radar data and microelectromechanical sensor data are weighted and fused through a spatiotemporal synchronization mechanism, and the fusion weight is dynamically adjusted according to the soil type.

[0060] S5. Establish a dynamic risk assessment model. Input the fused multivariate data into the dynamic risk assessment model to obtain the risk assessment level of the foundation to be monitored.

[0061] Furthermore, in S1, the detection method combines ground-penetrating radar (GPR) with a microelectromechanical system (MEMS) sensor array. The GPR emits high-frequency electromagnetic waves underground. A multi-antenna GPR array is employed, comprising a transmitting antenna array, a receiving antenna array, a signal transmitting unit, and a signal receiving unit. The transmitting antenna array emits electromagnetic waves under the control of the signal transmitting unit, and the receiving antenna array transmits the received reflected waves to the signal receiving unit, enabling simultaneous acquisition of underground information from multiple angles. For example, in a city building foundation detection project, a 500MHz multi-antenna GPR with an antenna spacing of 0.2m effectively detects structural information within a depth range of 5-10m. Simultaneously, a MEMS sensor array consisting of accelerometers, pressure sensors, and humidity sensors is deployed, distributed across the foundation surface and shallow soil, forming a three-dimensional monitoring network. Sensor nodes are arranged every 2-3m at different corners and key support points of the building. Each node integrates multiple sensors, converting the collected physical quantities into electrical signals and transmitting them wirelessly. The sensor nodes transmit data to the data analysis and processing module via a wireless communication module.

[0062] Specifically, the present invention employs a dual-frequency staggered arrangement of GPR antenna arrays (500MHz deep detection + 800MHz shallow high resolution), with an antenna spacing of 0.2m and a depth coverage of 5-10m; MEMS nodes: integrated temperature drift self-calibration circuit (error <0.1%), and the wireless networking protocol adopts ZigBee+LoRa dual-mode anti-interference transmission.

[0063] Furthermore, S2 also includes storing the multiple collected detection radar data and microelectromechanical sensor data into a preset historical ground database.

[0064] Furthermore, in S3, ground-penetrating radar data is analyzed and processed using a signal processing algorithm that combines wavelet transform and time-frequency analysis, specifically including:

[0065] The original time-domain signal is decomposed and denoised using wavelet transform, and then the energy distribution of the radar signal at different times and frequencies is obtained through time-frequency analysis. The wavelet transform formula is as follows:

[0066]

[0067] Among them, W f (a, b) are wavelet transform coefficients, f(t) is the original signal, a is the scaling factor, b is the translation factor, and ψ * () represents the complex conjugate of the wavelet basis functions.

[0068] The time-frequency analysis formula is as follows:

[0069]

[0070] Where S(t, f) is the time-frequency distribution function, s(τ) is the signal, and g * () represents the complex conjugate of the window function.

[0071] Furthermore, in S3, for microelectromechanical sensor data, a signal processing algorithm combining Kalman filtering and multiple linear regression models is used for analysis and processing, specifically including:

[0072] Accelerometer data was fused and state estimated using a Kalman filter algorithm; pressure and humidity sensor data were analyzed using a multiple linear regression model to establish their relationship with foundation stability; the formula for the multiple linear regression model is as follows:

[0073] y = β0 + β1x1 + β2x2 + ... + β n x n +∈

[0074] Where y is the foundation stability index, x1, x2, ..., x n β is the independent variable for pressure or humidity. i is the regression coefficient, and ∈ is the error term.

[0075] Furthermore, in S5, the dynamic risk assessment model employs a CNN-optimized multi-parameter model, inputting the fused multi-data elements and the historical foundation data stored in S2 into the dynamic risk assessment model to obtain the risk assessment level of the foundation to be monitored; the formula for the CNN-optimized multi-parameter model is as follows:

[0076] Risk_Score=W1*Crack_density+W2*Vibration_amp+W3*Pressure_dev

[0077] W1, W2, and W3 are weights, parameters optimized by the CNN model, representing the contribution weight of each indicator to the overall risk. Risk_Score is the risk score, a comprehensive quantitative measure of foundation safety risk; a higher value indicates a higher risk level. Crack_denxity is the crack density, the length / number of cracks per unit area of ​​the foundation surface or interior, reflecting the degree of structural damage. Vibration_amp is the vibration amplitude, the peak displacement / acceleration of the foundation caused by construction or environmental vibration, characterizing the impact of external disturbances on stability. Pressure_dev is the pressure deviation, the deviation between the actual soil pressure at key foundation points and the safety threshold, reflecting abnormal load distribution or soil failure risk.

[0078] Specifically, the safety and stability of building foundations are classified into three levels: safe, warning, and dangerous. When ground-penetrating radar (GPR) does not detect obvious defects such as cavities or cracks, and the vibration, pressure, and humidity parameters monitored by microelectromechanical systems (MEMS) sensors are all within normal ranges, the foundation is determined to be at the safe level. If GPR detects a small number of tiny cracks or cavities, or if MEMS sensor data shows slight abnormal fluctuations but does not exceed the set threshold, it is determined to be at the warning level. When GPR detects large-scale cavities or cracks, or if MEMS sensor data significantly exceeds the normal range, indicating a potentially significant safety hazard, it is determined to be at the dangerous level. For foundations at the safe level, regular routine monitoring is recommended. For foundations at the warning level, it is recommended to further employ high-precision local detection technology for detailed inspection to determine the specific details of the defects and develop corresponding maintenance plans, such as repairing tiny cracks and draining areas with abnormal humidity. For foundations at the dangerous level, all building activities should be immediately suspended, and professional personnel should be organized to conduct a comprehensive assessment and reinforcement design, taking effective reinforcement measures such as grouting reinforcement and foundation replacement.

[0079] This invention utilizes a ground-penetrating radar multi-antenna array to emit high-frequency electromagnetic waves. These waves propagate through the underground medium and reflect upon encountering interfaces between different media, with the receiving antenna acquiring the reflected wave data. Simultaneously, a microelectromechanical system (MEMS) sensor array collects real-time data on ground vibration, pressure, and humidity. The collected data is transmitted wirelessly to a data processing center. At the data processing center, algorithms such as wavelet transform, time-frequency analysis, Kalman filtering, and multiple linear regression models are used to process and analyze the data. Based on the analysis results, combined with pre-set thresholds and evaluation models, the safety and stability level of the foundation is determined, and finally, an evaluation report is generated, providing corresponding recommendations.

[0080] like Figure 2 As shown, a system for intelligent foundation detection and assessment, applying any one of the above-mentioned methods for intelligent foundation detection and assessment, includes:

[0081] The equipment installation module is used to set up ground-penetrating radar and microelectromechanical sensor arrays based on the foundation conditions and the target to be detected.

[0082] The data acquisition module is used to acquire data based on ground penetrating radar and microelectromechanical sensor array to obtain radar data and microelectromechanical sensor data.

[0083] The data analysis and processing module uses a signal processing algorithm combining wavelet transform and time-frequency analysis to analyze and process ground-penetrating radar data; and a signal processing algorithm combining Kalman filtering algorithm and multiple linear regression model to analyze and process microelectromechanical sensor data.

[0084] The data fusion module is used to perform weighted fusion of the analyzed ground-penetrating radar data and microelectromechanical sensor data through a spatiotemporal synchronization mechanism. The fusion weight is dynamically adjusted according to the soil type.

[0085] The risk assessment module is used to establish a dynamic risk assessment model. The fused multivariate data is input into the dynamic risk assessment model to obtain the risk assessment level of the foundation to be monitored.

[0086] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements a method for intelligent ground detection and assessment as described above.

[0087] The adaptive data fusion implementation program of this invention is as follows:

[0088]

[0089]

[0090] In a specific embodiment, this invention was applied to the foundation testing project of a newly built residential community. The community is located in a suburban area with a high groundwater level and some backfilled areas. During foundation construction, the intelligent foundation detection and assessment system of this invention was used for real-time monitoring. First, microelectromechanical (MEMS) sensor arrays were deployed on the foundation surface and in the shallow soil according to the design plan, while a multi-antenna ground-penetrating radar (GPR) scanned the foundation. During construction, the GPR detected an abnormal reflection signal in an area approximately 6 meters underground. After signal processing and analysis, it was determined that a small cavity might exist. Simultaneously, nearby MEMS sensors detected abnormal fluctuations in pressure and humidity data in the area. Based on this data, the system quickly determined that the foundation in that area was in an early warning state and generated a detailed report. Following the report's recommendations, the construction team confirmed the existence of the cavity through local drilling and promptly carried out grouting reinforcement. In subsequent construction, the system was continuously used for monitoring, ensuring the stability of the foundation, avoiding potential building safety hazards caused by foundation problems, and guaranteeing the construction quality of the entire community and the safety of residents' lives and property.

[0091] Example 1: Detection of voids in clay foundations

[0092] Deployment: The GPR array is scanned in a cross grid pattern (speed 0.5m / s) to arrange 32 MEMS nodes (spacing 3m);

[0093] Data anomaly: GPR detected a strong reflection signal (amplitude > threshold 150%) at coordinates (12,35), MEMS node #15 pressure dropped sharply by 20 kPa, and humidity increased by 15%;

[0094] Dynamic response: The fusion algorithm automatically increases the GPR weight (α = 0.7), the CNN model identifies it as a hole (probability 92.3%), the system triggers an early warning and generates a reinforcement plan (grouting volume 1.2m). 3 );

[0095] Example 2: Electromagnetic Interference Environment Test

[0096] Scenario: Foundation next to a high-voltage substation (background noise 60dB)

[0097] Anti-interference measures: Activate frequency domain notch filtering (suppress 50Hz power frequency interference), switch LoRa transmission mode (reducing packet loss rate from 18% to 2%);

[0098] Results: The accuracy rate of hole identification remained above 90%.

[0099] This invention's method for foundation detection and assessment employs adaptive filtering and interference suppression techniques to address various environmental disturbances. Even in high humidity or electromagnetic interference environments, it effectively reduces the impact of environmental factors on the detection results, ensuring data accuracy. Utilizing advanced machine learning and signal processing algorithms, it achieves automated data processing and analysis, reducing reliance on the experience of professional personnel and enabling a more objective and accurate assessment of foundation safety and stability, thus improving the reliability of result interpretation.

[0100] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, and optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0101] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for intelligent foundation detection and assessment, characterized in that, Includes the following steps: S1. Based on the foundation conditions and the target to be detected, set up ground penetrating radar and microelectromechanical sensor array; S2. Data acquisition is performed based on ground-penetrating radar and microelectromechanical sensor array to obtain radar data and microelectromechanical sensor data. S3. For ground-penetrating radar data, a signal processing algorithm combining wavelet transform and time-frequency analysis is used for analysis and processing; for microelectromechanical sensor data, a signal processing algorithm combining Kalman filtering algorithm and multiple linear regression model is used for analysis and processing. S4. The analyzed ground-penetrating radar data and microelectromechanical sensor data are weighted and fused through a spatiotemporal synchronization mechanism, and the fusion weight is dynamically adjusted according to the soil type. S5. Establish a dynamic risk assessment model, input the fused multivariate data into the dynamic risk assessment model, and obtain the risk assessment level of the foundation to be monitored. In step S3, ground-penetrating radar data is analyzed and processed using a signal processing algorithm that combines wavelet transform and time-frequency analysis, specifically including: The original time-domain signal is decomposed and denoised using wavelet transform, and then the energy distribution of the detection radar signal at different times and frequencies is obtained through time-frequency analysis; the wavelet transform formula is as follows: ; in, These are the wavelet transform coefficients. The original signal, As a scale factor, The translation factor is... For the complex conjugate of wavelet basis functions; In step S3, for the microelectromechanical sensor data, a signal processing algorithm combining the Kalman filter algorithm and the multiple linear regression model is used for analysis and processing, specifically including: Accelerometer data was fused and state estimated using a Kalman filter algorithm; pressure and humidity sensor data were analyzed using a multiple linear regression model to establish their relationship with foundation stability; the formula for the multiple linear regression model is as follows: ; in, As an indicator of foundation stability, Pressure or humidity is the independent variable. For regression coefficients, This is the error term.

2. The method for intelligent foundation detection and assessment according to claim 1, characterized in that, In S1, a ground-penetrating radar with a multi-antenna array is used to transmit high-frequency electromagnetic waves into the ground and acquire underground information from multiple angles at the same time; the microelectromechanical sensor array includes an accelerometer, a pressure sensor and a humidity sensor, which are distributed and installed on the foundation surface and in the shallow soil to form a three-dimensional monitoring network.

3. The method for intelligent foundation detection and assessment according to claim 1, characterized in that, S2 also includes storing the multiple collected detection radar data and microelectromechanical sensor data into a preset historical ground database.

4. The method for intelligent foundation detection and assessment according to claim 1, characterized in that, In step S5, the dynamic risk assessment model employs a CNN-optimized multi-parameter model. The fused multi-data elements and the historical foundation data stored in step S2 are input into the dynamic risk assessment model to obtain the risk assessment level of the foundation to be monitored. The formula for the CNN-optimized multi-parameter model is as follows: ； in, , , As weight, To score the risk, Crack density, The amplitude of vibration. This is for pressure deviation.

5. A system for intelligent foundation detection and assessment, employing the intelligent foundation detection and assessment method according to any one of claims 1-4, characterized in that, include: The equipment installation module is used to set up ground-penetrating radar and microelectromechanical sensor arrays based on the foundation conditions and the target to be detected. The data acquisition module is used to acquire data based on ground penetrating radar and microelectromechanical sensor array to obtain radar data and microelectromechanical sensor data. The data analysis and processing module uses a signal processing algorithm combining wavelet transform and time-frequency analysis to analyze and process ground-penetrating radar data; and a signal processing algorithm combining Kalman filtering algorithm and multiple linear regression model to analyze and process microelectromechanical sensor data. The data fusion module is used to perform weighted fusion of the analyzed ground-penetrating radar data and microelectromechanical sensor data through a spatiotemporal synchronization mechanism. The fusion weight is dynamically adjusted according to the soil type. The risk assessment module is used to establish a dynamic risk assessment model. The fused multivariate data is input into the dynamic risk assessment model to obtain the risk assessment level of the foundation to be monitored.

6. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements a method for intelligent ground detection and assessment as described in any one of claims 1 to 4.

Images