Foundation pit deformation visual monitoring and early warning system and method

Abstract

The invention discloses a foundation pit deformation visual monitoring and early warning system and method, and relates to the technical field of foundation pit visual monitoring. The surface of a to-be-monitored foundation pit is covered with electronic skin; the electronic skin comprises a composite film layer, a strain sensing layer, a pressure signal layer, an environment sensing layer, a power supply unit and a surface function layer. The multispectral camera array is arranged on the periphery of the foundation pit to be monitored; fluorescent two-dimensional code mark points are printed on the surface functional layer. The electronic skin can directly and continuously monitor tiny strain and soil pressure change on the surface of a foundation pit through the strain sensing layer and the pressure signal layer, and high-precision mechanical data are provided. The multispectral camera array provides visible light images for identifying mark points for displacement measurement, and also provides thermal images for monitoring temperature field changes and capturing potential anomalies which are difficult to find by naked eyes or a single sensor. The surface of the foundation pit is covered with the electronic skin, continuous and distributed sensing is conducted on the whole monitoring area, and possible local deformation of traditional discrete measuring points is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of foundation pit visual monitoring, and in particular to a foundation pit deformation visual monitoring and early warning system and method. Background Art

[0002] As an indispensable component of modern construction, foundation pit engineering safety is directly linked to the quality and stability of the entire project. Deformation monitoring, a key method for ensuring pit safety, has long relied on traditional measurement methods and sensor technologies. However, with the continuous expansion of construction projects, foundation pit deformation monitoring faces multiple challenges, including increasing accuracy, improving monitoring efficiency, and enhancing real-time performance.

[0003] The current industry faces numerous challenges in deformation monitoring, particularly in foundation pits and tunnels, where the available monitoring methods are relatively limited. Common, general-purpose monitoring methods first create a 3D model of the structure being monitored and then output the test results using this 3D information. This process consumes large amounts of data, is slow to process, and offers relatively low accuracy. Furthermore, the complex algorithms hinder real-time monitoring.

[0004] Specifically, traditional monitoring methods have the following limitations:

[0005] Low efficiency of manual monitoring: Traditional deformation monitoring mainly relies on manual work, which is not only time-consuming and labor-intensive, but also has limited monitoring frequency, making it difficult to detect and address deformation problems in a timely manner.

[0006] Insufficient data accuracy: Some monitoring methods and instruments lack accuracy and find it difficult to capture tiny deformation changes, resulting in inaccurate monitoring results and affecting subsequent analysis and decision-making.

[0007] Limited monitoring range: For large or complex civil engineering structures, traditional monitoring methods often find it difficult to cover the entire area, resulting in monitoring blind spots and increasing safety risks.

[0008] Complex data processing: The large amount of data generated by deformation monitoring needs to be analyzed and processed, but traditional data processing methods are cumbersome and inefficient, and cannot meet the needs of real-time monitoring. Summary of the Invention

[0009] In order to solve the above technical problems, the present invention provides a foundation pit deformation visual monitoring and early warning system and method. The following technical solutions are adopted:

[0010] A method for visually monitoring and early warning of foundation pit deformation comprises the following steps:

[0011] Step 1: Cover the surface of the pit to be monitored with an electronic skin; the electronic skin includes a composite film layer, a strain sensing layer, a pressure signal layer, an environmental sensing layer, a power supply unit, and a surface functional layer; deploy a multispectral camera array around the pit to be monitored; and print fluorescent QR code markers on the surface functional layer.

[0012] Step 2: The strain sensing layer and pressure signal layer on the electronic skin monitor the strain and soil pressure changes on the foundation pit surface, generating resistance change signals and capacitance change signals. The environmental sensing layer collects ambient temperature and humidity data. The multispectral camera array synchronously captures foundation pit images, thermal images, and marker point images at a set frequency.

[0013] Step 3: The computer monitoring server collects electronic skin data at set intervals. The electronic skin data includes resistance change signal data, capacitance change signal data, and ambient temperature and humidity data. It also simultaneously collects multispectral camera array visual data. The multispectral camera array visual data includes foundation pit images, thermal images, and marker point images.

[0014] Step 4: The computer monitoring server synchronizes the frame rate of the electronic skin data timestamp with the multispectral camera array visual data;

[0015] Step 5: Calculate the local strain value based on the resistance change signal data; calculate the pressure distribution on the foundation pit surface based on the capacitance change signal data; identify the fluorescent QR code marker points through image processing algorithms, establish a spatial coordinate system, calculate the displacement field of the foundation pit surface through image matching algorithms, and analyze the temperature distribution and abnormal areas in the thermal infrared image;

[0016] Step 6: Fusing the strain data monitored by the electronic skin with the displacement data monitored by the visual system;

[0017] Step 7: Use a risk assessment algorithm based on fusion data to output the deformation risk level of the fluorescent QR code marker.

[0018] By adopting the above technical solution, the electronic skin can directly and continuously monitor the tiny strains and soil pressure changes on the foundation pit surface through the strain sensing layer and pressure signal layer, providing high-precision mechanical data.

[0019] The multispectral camera array not only provides visible light images for identifying markers for displacement measurement, but also provides thermal images for monitoring temperature field changes, capturing potential anomalies that are difficult to detect with the naked eye or a single sensor (such as leakage and temperature anomalies caused by stress concentration).

[0020] The environmental perception layer supplements temperature and humidity information, helping to correct the impact of environmental factors on sensor readings and material properties, and improve the accuracy of monitoring data.

[0021] The electronic skin covers the surface of the foundation pit, achieving continuous and distributed perception of the entire monitoring area and avoiding local deformation that may be missed by traditional discrete measurement points.

[0022] Combined with a multispectral camera array that shoots at high frequencies, the system can achieve near real-time data collection, enabling timely detection of deformation trends and sudden anomalies.

[0023] By fusing the intrinsic mechanical data (strain, pressure) provided by the electronic skin with the external geometric and physical data (displacement, temperature) provided by the multispectral camera, the state of the foundation pit can be cross-verified from multiple dimensions.

[0024] For example, an increase in strain in a certain area should correspond to a change in displacement or pressure in that area. Data fusion can eliminate the limitations or false positives of a single sensor, providing a more comprehensive and reliable assessment of the foundation pit condition.

[0025] The system realizes the automation of the entire process from data collection, processing to analysis and early warning, reducing manual intervention and improving monitoring efficiency and consistency.

[0026] The risk assessment algorithm based on fused data can automatically calculate the deformation risk level of each marked point (representing the area), provide a quantitative basis for engineering decision-making, and achieve an intelligent leap from monitoring to early warning.

[0027] The electronic skin is made of flexible materials, which can better fit the surface of the foundation pit and adapt to certain deformations.

[0028] Fluorescent QR code marking points are visible under specific lighting. Combined with image processing algorithms, they can be reliably identified even in complex backgrounds. The QR code itself is unique, making it easy to track long-term changes in specific points.

[0029] Timestamp and frame rate synchronization ensures accurate temporal correspondence of data from different sources, providing a basis for subsequent accurate analysis and event backtracking.

[0030] Multispectral data (visible light, thermal) combined with strain, pressure and other data can generate richer visualization products, such as displacement cloud maps, strain cloud maps, pressure distribution maps, temperature anomaly maps, etc., enabling engineers to intuitively understand the complex state of the foundation pit.

[0031] By integrating advanced electronic skin sensing technology and multispectral vision technology, comprehensive, accurate, real-time and intelligent monitoring of foundation pit deformation is achieved, which significantly improves the safety level of foundation pit projects and has important engineering application value.

[0032] Optionally, the method further includes step 8 of setting a deformation risk level threshold, and generating an alarm when the deformation risk level exceeds the deformation risk level threshold.

[0033] By implementing this technical solution, the alarm mechanism ensures that pit deformation can be identified and addressed before it reaches a critical state that could pose a safety risk. Compared to traditional methods that can suffer from delayed monitoring data analysis and manual judgment delays, this automated threshold-triggered alarm significantly shortens the time from detection to response, improving the efficiency of risk management.

[0034] Optionally, the composite film layer is a polyimide silicone composite film, the strain sensing layer is a silver nanowire resistor grid, the pressure signal layer is a flexible capacitive pressure sensor array, the environmental sensing layer is a flexible temperature and humidity sensor array, the power supply units supply power to the strain sensing layer, the pressure signal layer and the environmental sensing layer respectively, the surface functional layer is a nano-antireflection film, and the fluorescent QR code marking points are printed by inkjet printing and UV curing process.

[0035] By employing this technical solution, both polyimide (PI) and silicone possess excellent flexibility, allowing them to adhere tightly to the foundation pit surface. This ensures good contact between the sensor layer and the monitored object, even with surface irregularities, thereby reducing measurement errors. PI exhibits excellent resistance to high and low temperatures, as well as chemical corrosion, while silicone offers excellent water and moisture resistance. This combined application enhances the long-term stability and reliability of the electronic skin in the complex and harsh outdoor foundation pit environment, such as exposure to sunlight, rain, and temperature fluctuations.

[0036] PI provides good mechanical strength and rigid support, while silicone provides softness and sealing. While ensuring flexibility, the composite structure also has certain tensile and tear resistance, and is not easily damaged during installation or slight deformation of the foundation pit.

[0037] The silver nanowires (AgNWs) used in the silver nanowire resistor grid have extremely low resistivity and excellent conductivity. The grid structure formed is very sensitive to tiny deformations and can accurately capture the tensile and compressive strains on the foundation pit surface (reflected by resistance changes), achieving high-precision strain measurement.

[0038] The grid structure enables large-area, uniform strain sensing, helping to obtain continuous strain field distribution information. AgNWs are easy to fabricate on flexible substrates, perfectly matching the flexible properties of the composite film layer.

[0039] The sensor array structure can provide spatial pressure distribution information, not just single-point pressure information. It can identify concentrated areas of pressure or changing trends, providing richer data for analyzing soil stress states. Capacitive sensors typically offer excellent linear output and long-term stability. The flexible substrate allows the sensor array to conform to curved surfaces, adapting to changes in the shape of the foundation pit surface, and is less susceptible to additional stress interference from its own rigidity.

[0040] Simultaneously monitoring temperature and humidity can more comprehensively capture the impact of environmental changes on the performance of the foundation pit and the electronic skin itself.

[0041] Nano-AR coatings can reduce the reflection and scattering of light on the surface of electronic skin, improve the clarity and contrast of images taken by multispectral cameras (especially visible light images), and facilitate subsequent marker recognition and image analysis.

[0042] Optionally, the formula for calculating the local strain value in step 5 is:

[0043] ;

[0044] in is the local strain value, is the relative change of phase resistance, and k is the material constant.

[0045] Optional,

[0046] The formula for the pressure distribution on the foundation pit surface in step 5 is:

[0047] ;

[0048] Where P is the calculated pressure value, is the base pressure value, is the measured pressure value, is the initial capacitance value.

[0049] Optionally, the specific steps of calculating the displacement field of the foundation pit surface using the image matching algorithm are as follows:

[0050] Step 51, extracting SIFT feature points from images of continuous time phases;

[0051] Step 52: Establishing feature point correspondence through nearest neighbor matching and consistency check;

[0052] Step 53: Evaluate the matching quality using a random sampling consistency algorithm and remove abnormal matching points;

[0053] Step 54: Calculate the displacement field of the foundation pit surface based on the matching point pairs.

[0054] Optionally, in step 6, the step of fusing the strain data monitored by the electronic skin with the displacement data monitored by the visual system is:

[0055] Step 61: Use a weighted fusion method to fuse the electronic skin strain data and the visual displacement data:

[0056] Step 62: spatially align the electronic skin monitoring points with the visual monitoring points to establish a corresponding relationship;

[0057] Step 63, respectively evaluating the uncertainty of the electronic skin data and the visual data;

[0058] Step 64: assign weights to the two data sources based on the uncertainty index and calculate the fusion result.

[0059] Optionally, the formula for calculating the fusion result in step 64 is:

[0060] ;

[0061] in It is electronic skin data, is the visual displacement data, is the weight of the electronic skin data, is the weight of the visual displacement data, It is the result of fusion.

[0062] Optionally, the specific steps of the risk assessment algorithm based on fusion data are as follows:

[0063] Based on the fusion results of the electronic skin strain data and visual displacement data corresponding to the fluorescent QR code markers, the fusion results of the strain, displacement, and displacement rate indicators are calculated respectively, and the risk level is determined based on the risk level threshold range;

[0064] Adopt fuzzy comprehensive evaluation method and integrate multi-index risk assessment results;

[0065] Based on the comprehensive evaluation results, the deformation risk level of the corresponding fluorescent QR code marking point is determined.

[0066] A foundation pit deformation visual monitoring and early warning system is used to implement a foundation pit deformation visual monitoring and early warning method. The visual monitoring and early warning system includes electronic skin, a multispectral camera array, a computer monitoring server and an alarm 4. The electronic skin includes a composite film layer, a strain sensing layer, a pressure signal layer, an environmental perception layer, a power supply unit and a surface functional layer; the bottom of the composite film layer is attached to the surface of the foundation pit to be detected, the strain sensing layer is laid on the top surface of the composite film layer, the pressure signal layer is laid on the strain sensing layer based on the isolation layer, the environmental perception layer is laid on the pressure signal layer based on the isolation layer, and the surface functional layer is laid on the environmental perception layer; the power supply unit supplies power to the strain sensing layer, the pressure signal layer and the environmental perception layer respectively; the multispectral camera array is arranged around the foundation pit to be monitored, the computer monitoring server is communicatively connected to the strain sensing layer, the pressure signal layer, the environmental perception layer and the multispectral camera array respectively, and outputs the deformation risk level of the corresponding fluorescent two-dimensional code mark point based on the electronic skin data and the visual displacement data. If the deformation risk level exceeds the deformation risk level threshold, the alarm 4 is controlled to alarm.

[0067] In summary, the present invention includes at least one of the following beneficial technical effects:

[0068] This invention provides a visual monitoring and early warning system and method for foundation pit deformation. The electronic skin, through a strain sensing layer and a pressure signal layer, can directly and continuously monitor minute strains on the foundation pit surface and soil pressure changes, providing high-precision mechanical data. A multispectral camera array not only provides visible light images for identifying markers and measuring displacement, but also thermal images for monitoring temperature field changes, capturing potential anomalies that are difficult to detect with the naked eye or with a single sensor.

[0069] The electronic skin covers the surface of the foundation pit, enabling continuous, distributed sensing of the entire monitoring area, avoiding local deformations that might be missed by traditional discrete measurement points. Combined with a high-frequency multispectral camera array, the system enables near-real-time data acquisition, enabling timely detection of deformation trends and sudden anomalies.

[0070] By fusing the intrinsic mechanical data provided by the electronic skin with the extrinsic geometric and physical data provided by the multispectral camera, the state of the foundation pit can be cross-validated from multiple dimensions. A risk assessment algorithm based on fused data automatically calculates the deformation risk level for each marker point. The alarm mechanism ensures that foundation pit deformation can be identified and addressed before it reaches a critical state that could cause safety issues. Compared to traditional methods, which can involve delayed analysis of monitoring data and manual judgment delays, this automated threshold-triggered alarm significantly shortens the time from monitoring to response, improving the efficiency of risk management. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 It is a flow chart of a method for visually monitoring and early warning of foundation pit deformation according to the present invention;

[0072] Figure 2 This is a schematic diagram of the electrical component connection principle of a foundation pit deformation visual monitoring and early warning system of the present invention;

[0073] Figure 3 This is a schematic diagram of the cross-sectional structure of the electronic skin of a foundation pit deformation visual monitoring and early warning system of the present invention;

[0074] Explanation of the accompanying symbols: 11. Composite film layer; 12. Strain sensing layer; 13. Pressure signal layer; 14. Environmental sensing layer; 15. Power supply unit; 16. Surface functional layer; 2. Multispectral camera array; 3. Computer monitoring server; 4. Alarm. DETAILED DESCRIPTION

[0075] The present invention will be further described in detail below with reference to the accompanying drawings.

[0076] The embodiment of the present invention discloses a foundation pit deformation visual monitoring and early warning system and method.

[0077] Reference Figure 1-Figure 3,Example 1, a foundation pit deformation visual monitoring and early warning method, comprising the following steps:

[0078] Step 1: Cover the surface of the foundation pit to be monitored with an electronic skin; the electronic skin includes a composite film layer 11, a strain sensing layer 12, a pressure signal layer 13, an environmental sensing layer 14, a power supply unit 15, and a surface functional layer 16; a multispectral camera array 2 is deployed around the foundation pit to be monitored; and fluorescent QR code markers are printed on the surface functional layer 16;

[0079] Step 2: The strain sensing layer 12 and pressure signal layer 13 on the electronic skin monitor the strain on the foundation pit surface and the changes in soil pressure, generating resistance change signals and capacitance change signals. The environmental sensing layer 14 collects ambient temperature and humidity data. The multispectral camera array 2 synchronously captures foundation pit images, thermal images, and marker point images at a set frequency.

[0080] Step 3: The computer monitoring server 3 collects electronic skin data at set intervals. The electronic skin data includes resistance change signal data, capacitance change signal data, and ambient temperature and humidity data. The computer monitoring server 3 also collects visual data from the multispectral camera array 2. The visual data from the multispectral camera array 2 includes foundation pit images, thermal images, and marker point images.

[0081] Step 4: The computer monitoring server 3 synchronizes the frame rate of the electronic skin data timestamp with the visual data of the multispectral camera array 2;

[0082] Step 5: Calculate the local strain value based on the resistance change signal data; calculate the pressure distribution on the foundation pit surface based on the capacitance change signal data; identify the fluorescent QR code marker points through image processing algorithms, establish a spatial coordinate system, calculate the displacement field of the foundation pit surface through image matching algorithms, and analyze the temperature distribution and abnormal areas in the thermal infrared image;

[0083] Step 6: Fusing the strain data monitored by the electronic skin with the displacement data monitored by the visual system;

[0084] Step 7: Use a risk assessment algorithm based on fusion data to output the deformation risk level of the fluorescent QR code marker.

[0085] Through the strain sensing layer 12 and the pressure signal layer 13, the electronic skin can directly and continuously monitor the tiny strains on the foundation pit surface and the changes in soil pressure, providing high-precision mechanical data.

[0086] The multispectral camera array 2 not only provides visible light images for identifying markers for displacement measurement, but also provides thermal images for monitoring temperature field changes, capturing potential anomalies that are difficult to detect with the naked eye or a single sensor (such as leakage and temperature anomalies caused by stress concentration).

[0087] The environmental sensing layer 14 supplements the temperature and humidity information, which helps to correct the impact of environmental factors on sensor readings and material properties, and improve the accuracy of monitoring data.

[0088] The electronic skin covers the surface of the foundation pit, achieving continuous and distributed perception of the entire monitoring area and avoiding local deformation that may be missed by traditional discrete measurement points.

[0089] Combined with a multispectral camera array that shoots at high frequencies, the system can achieve near real-time data collection, enabling timely detection of deformation trends and sudden anomalies.

[0090] By fusing the intrinsic mechanical data (strain, pressure) provided by the electronic skin with the external geometric and physical data (displacement, temperature) provided by the multispectral camera, the state of the foundation pit can be cross-verified from multiple dimensions.

[0091] For example, an increase in strain in a certain area should correspond to a change in displacement or pressure in that area. Data fusion can eliminate the limitations or false positives of a single sensor, providing a more comprehensive and reliable assessment of the foundation pit condition.

[0092] The system realizes the automation of the entire process from data collection, processing to analysis and early warning, reducing manual intervention and improving monitoring efficiency and consistency.

[0093] The risk assessment algorithm based on fused data can automatically calculate the deformation risk level of each marked point (representing the area), provide a quantitative basis for engineering decision-making, and achieve an intelligent leap from monitoring to early warning.

[0094] The electronic skin is made of flexible materials, which can better fit the surface of the foundation pit and adapt to certain deformations.

[0095] Fluorescent QR code marking points are visible under specific lighting. Combined with image processing algorithms, they can be reliably identified even in complex backgrounds. The QR code itself is unique, making it easy to track long-term changes in specific points.

[0096] Timestamp and frame rate synchronization ensures accurate temporal correspondence of data from different sources, providing a basis for subsequent accurate analysis and event backtracking.

[0097] Multispectral data (visible light, thermal) combined with strain, pressure and other data can generate richer visualization products, such as displacement cloud maps, strain cloud maps, pressure distribution maps, temperature anomaly maps, etc., enabling engineers to intuitively understand the complex state of the foundation pit.

[0098] By integrating advanced electronic skin sensing technology and multispectral vision technology, comprehensive, accurate, real-time and intelligent monitoring of foundation pit deformation is achieved, which significantly improves the safety level of foundation pit projects and has important engineering application value.

[0099] Example 2 further includes step 8 of setting a deformation risk level threshold, and issuing an alarm when the deformation risk level exceeds the deformation risk level threshold.

[0100] This alarm mechanism ensures that pit deformation is identified and addressed before it reaches a critical level that could pose a safety risk. Compared to traditional methods, which can involve delayed analysis of monitoring data and manual judgment, this automated threshold-triggered alarm significantly shortens the time from detection to response, improving the efficiency of risk management.

[0101] In Example 3, the composite film layer 11 is a polyimide silicone composite film, the strain sensing layer 12 is a silver nanowire resistor grid, the pressure signal layer 13 is a flexible capacitive pressure sensor array, and the environmental sensing layer 14 is a flexible temperature and humidity sensor array. The power supply unit 15 supplies power to the strain sensing layer 12, the pressure signal layer 13, and the environmental sensing layer 14 respectively. The surface functional layer 16 is a nano-antireflection film, and the fluorescent QR code marking points are printed by inkjet printing and UV curing process.

[0102] Both polyimide (PI) and silicone offer excellent flexibility, allowing them to adhere tightly to the surface of the foundation pit. Even with surface irregularities, this ensures good contact between the sensor layer and the monitored object, reducing measurement errors. PI offers excellent resistance to high and low temperatures, as well as chemical corrosion, while silicone offers excellent water and moisture resistance. This combined use enhances the long-term stability and reliability of the electronic skin in the complex and harsh outdoor foundation pit environment, such as exposure to sunlight, rain, and temperature fluctuations.

[0103] PI provides good mechanical strength and rigid support, while silicone provides softness and sealing. While ensuring flexibility, the composite structure also has certain tensile and tear resistance, and is not easily damaged during installation or slight deformation of the foundation pit.

[0104] The silver nanowires (AgNWs) used in the silver nanowire resistor grid have extremely low resistivity and excellent conductivity. The grid structure formed is very sensitive to tiny deformations and can accurately capture the tensile and compressive strains on the foundation pit surface (reflected by resistance changes), achieving high-precision strain measurement.

[0105] The grid structure enables large-area, uniform strain sensing, helping to obtain continuous strain field distribution information. AgNWs are easy to fabricate on flexible substrates, perfectly matching the flexible properties of the composite film layer.

[0106] The sensor array structure can provide spatial pressure distribution information, not just single-point pressure information. It can identify concentrated areas of pressure or changing trends, providing richer data for analyzing soil stress states. Capacitive sensors typically offer excellent linear output and long-term stability. The flexible substrate allows the sensor array to conform to curved surfaces, adapting to changes in the shape of the foundation pit surface, and is less susceptible to additional stress interference from its own rigidity.

[0107] Simultaneously monitoring temperature and humidity can more comprehensively capture the impact of environmental changes on the performance of the foundation pit and the electronic skin itself.

[0108] Nano-AR coatings can reduce the reflection and scattering of light on the surface of electronic skin, improve the clarity and contrast of images taken by multispectral cameras (especially visible light images), and facilitate subsequent marker recognition and image analysis.

[0109] In Example 4, the formula for calculating the local strain value in step 5 is:

[0110] ;

[0111] in is the local strain value, is the relative change of phase resistance, and k is the material constant.

[0112] In Example 5, the formula for the pressure distribution on the foundation pit surface in step 5 is:

[0113] ;

[0114] Where P is the calculated pressure value, is the base pressure value, is the measured pressure value, is the initial capacitance value.

[0115] Example 6: The specific steps of calculating the displacement field of the foundation pit surface using the image matching algorithm are as follows:

[0116] Step 51, extracting SIFT feature points from images of continuous time phases;

[0117] Step 52: Establishing feature point correspondence through nearest neighbor matching and consistency check;

[0118] Step 53: Evaluate the matching quality using a random sampling consistency algorithm and remove abnormal matching points;

[0119] Step 54: Calculate the displacement field of the foundation pit surface based on the matching point pairs.

[0120] In Example 7, in step 6, the step of fusing the strain data monitored by the electronic skin with the displacement data monitored by the visual system is:

[0121] Step 61: Use a weighted fusion method to fuse the electronic skin strain data and the visual displacement data:

[0122] Step 62: spatially align the electronic skin monitoring points with the visual monitoring points to establish a corresponding relationship;

[0123] Step 63, respectively evaluating the uncertainty of the electronic skin data and the visual data;

[0124] Step 64: assign weights to the two data sources based on the uncertainty index and calculate the fusion result.

[0125] In Example 8, the formula for calculating the fusion result in step 64 is:

[0126] ;

[0127] in It is electronic skin data, is the visual displacement data, is the weight of the electronic skin data, is the weight of the visual displacement data, It is the result of fusion.

[0128] Example 9, the specific steps of the risk assessment algorithm based on fusion data are as follows:

[0129] Based on the fusion results of the electronic skin strain data and visual displacement data corresponding to the fluorescent QR code markers, the fusion results of the strain, displacement, and displacement rate indicators are calculated respectively, and the risk level is determined based on the risk level threshold range;

[0130] Adopt fuzzy comprehensive evaluation method and integrate multi-index risk assessment results;

[0131] Based on the comprehensive evaluation results, the deformation risk level of the corresponding fluorescent QR code marking point is determined.

[0132] Example 10, a foundation pit deformation visual monitoring and early warning system, used to implement a foundation pit deformation visual monitoring and early warning method, the visual monitoring and early warning system includes an electronic skin, a multispectral camera array 2, a computer monitoring server 3 and an alarm 4, the electronic skin includes a composite film layer 11, a strain sensing layer 12, a pressure signal layer 13, an environmental perception layer 14, a power supply unit 15 and a surface functional layer 16; the bottom of the composite film layer 11 is attached to the surface of the foundation pit to be detected, the strain sensing layer 12 is laid on the top surface of the composite film layer 11, the pressure signal layer 13 is laid on the strain sensing layer 12 based on the isolation layer, and the environmental perception layer 14 is laid on the pressure signal layer 13 based on the isolation layer, and the surface functional layer 16 is laid on the environmental perception layer 14. The power supply unit 15 supplies power to the strain sensing layer 12, the pressure signal layer 13 and the environmental perception layer 14 respectively. The multispectral camera array 2 is arranged around the foundation pit to be monitored. The computer monitoring server 3 is communicated with the strain sensing layer 12, the pressure signal layer 13, the environmental perception layer 14 and the multispectral camera array 2 respectively. The deformation risk level of the corresponding fluorescent two-dimensional code mark point is output based on the electronic skin data and the visual displacement data. If the deformation risk level exceeds the deformation risk level threshold, the alarm 4 is controlled to alarm.

[0133] The following specific embodiments are used to illustrate the implementation principle of the present invention:

[0134] A new subway station project in a city center has a foundation pit depth of approximately 25 meters. The surrounding environment is complex, with existing buildings and roads nearby. To ensure construction safety and environmental stability, the project decided to use the foundation pit deformation visual monitoring and early warning system proposed in this paper.

[0135] Electronic skin installation: After cleaning the slopes and floor surfaces formed by the excavation pit, a polyimide-silicone composite film is laid as a base (composite film layer 11). A resistor grid composed of silver nanowires (strain sensing layer 12) is precisely laid on top to sense surface strain. A flexible capacitive pressure sensor array (pressure signal layer 13) is laid above the strain sensing layer, through a layer of isolation material, to monitor soil pressure. A flexible temperature and humidity sensor array (environmental sensing layer 14) is then laid through the isolation layer to monitor ambient temperature and humidity. Finally, a nano-antireflection film is applied to the top layer (surface functional layer 16). All sensing layers are powered by an integrated flexible power supply unit (power supply unit 15). Dozens of evenly distributed fluorescent QR code markers are printed on the surface functional layer using inkjet printing and UV curing.

[0136] Multispectral Camera Array Deployment: An array of four multispectral cameras (Multispectral Camera Array 2) was installed at a safe distance around the foundation pit. Each camera simultaneously captures visible light and thermal infrared images, specifically focusing on fluorescent QR code markers. The cameras synchronize their capture every five minutes to ensure consistent data timing.

[0137] Computer Monitoring Server Connection: A high-performance computer monitoring server (computer monitoring server 3) is located in a monitoring center near the construction site. The server is connected to the data acquisition module in the electronic skin via a wired network, regularly receiving resistance change signals, capacitance change signals, and ambient temperature and humidity data. Simultaneously, the server also receives image data (excavation pit images, thermal images, and marker point images) from the multispectral camera array via the network.

[0138] Data Collection: The electronic skin continuously monitors and sends data to a server. The multispectral camera array captures images at a set frequency and transmits them to the server.

[0139] Data synchronization: The electronic skin data and camera image data received by the server are both time-stamped. The server first performs time stamp calibration to ensure that the data from different devices are synchronized in time for subsequent correlation analysis.

[0140] Mechanical Data Analysis: The server processes the resistance change signal and calculates the local strain value of each area based on a known formula, reflecting the tension or compression of the foundation pit surface. Simultaneously, the capacitance change signal is processed to calculate the pressure distribution map of the foundation pit surface, identifying the concentrated areas of soil pressure.

[0141] Visual Data Analysis: The server processes the image data. First, an image processing algorithm is used to identify and decode fluorescent QR code markers in the visible light image. Based on the positions of these markers, a spatial coordinate system corresponding to the foundation pit surface is established. Then, image matching algorithms such as SIFT feature point extraction and matching, and RANSAC to eliminate mismatches are used to calculate the displacement of each marker in the continuous image, thereby obtaining the displacement field of the entire foundation pit surface. In addition, the server also analyzes thermal infrared images to identify areas of abnormal temperature, which may be related to leakage or stress concentration.

[0142] Ambient temperature and humidity data are used to correct strain and pressure data to eliminate or reduce measurement deviations caused by environmental factors.

[0143] The server fuses the strain data calculated by the electronic skin with the displacement data calculated by the vision system. First, spatial registration is performed to match the sensor points on the electronic skin with the nearest visual landmarks. Uncertainty in both data sources is assessed (for example, based on sensor accuracy and image matching confidence), and the strain and displacement data are weighted accordingly. Finally, a weighted fusion method is used to obtain comprehensive deformation information for each corresponding region.

[0144] Risk Assessment and Early Warning: Based on the fused data, the server runs a risk assessment algorithm. This algorithm first calculates indicators such as strain, displacement, and rate of change for the area represented by each marker point. These indicators are then compared against pre-set risk thresholds to preliminarily determine the risk level. A fuzzy comprehensive evaluation method is then used to combine the risk assessment results of these individual indicators to produce a more comprehensive and objective overall risk level. For example, if both strain and displacement in a given area exceed warning values ​​and the displacement rate accelerates, the overall risk level for that area will be assessed as "high."

[0145] Alarm and Visualization: The server displays the deformation risk level of each marked point in real time on a monitoring interface, typically color-coded (e.g., green for low risk, yellow for medium risk, and red for high risk) on a digital model or image of the foundation pit. When the risk level of one or more marked points exceeds a preset alarm threshold (e.g., reaching "high" risk), the server immediately triggers an alarm (alarm 4), emitting an audible and visual alarm. The server also sends an alert message to project managers and relevant safety officers, prompting them to take immediate inspection or reinforcement measures.

[0146] Through this system, the subway foundation pit project has achieved continuous, automated, and multi-dimensional monitoring of the entire excavation surface, which can timely detect potential deformation risks and ensure the safe and smooth progress of the project.

[0147] The above are all preferred embodiments of the present invention, and are not intended to limit the scope of protection of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for visual monitoring and early warning of foundation pit deformation, characterized by: The following steps are involved: Step 1: Covering the surface of the foundation pit to be monitored with an electronic skin; the electronic skin comprises a composite film layer (11), a strain sensing layer (12), a pressure signal layer (13), an environmental sensing layer (14), a power supply unit (15), and a surface functional layer (16); deploying a multispectral camera array (2) around the foundation pit to be monitored; and printing fluorescent two-dimensional code marking points on the surface functional layer (16); Step 2: The strain sensing layer (12) and the pressure signal layer (13) on the electronic skin monitor the strain on the foundation pit surface and the soil pressure change, generating a resistance change signal and a capacitance change signal; the environmental sensing layer (14) collects environmental temperature and humidity data; the multispectral camera array (2) synchronously captures the foundation pit image, thermal image, and marker point image at a set frequency; Step 3, the computer monitoring server (3) collects electronic skin data at set intervals, the electronic skin data including resistance change signal data, capacitance change signal data, and ambient temperature and humidity data, and simultaneously collects visual data from the multispectral camera array (2), the visual data from the multispectral camera array (2) including foundation pit images, thermal images, and marker point images; Step 4, the computer monitoring server (3) synchronizes the frame rate of the electronic skin data timestamp with the visual data of the multispectral camera array (2); Step 5: Calculate the local strain value based on the resistance change signal data; calculate the pressure distribution on the foundation pit surface based on the capacitance change signal data; identify the fluorescent QR code marker points through image processing algorithms, establish a spatial coordinate system, calculate the displacement field of the foundation pit surface through image matching algorithms, and analyze the temperature distribution and abnormal areas in the thermal infrared image; Step 6: Fusing the strain data monitored by the electronic skin with the displacement data monitored by the visual system; Step 7: Use a risk assessment algorithm based on fusion data to output the deformation risk level of the fluorescent QR code marker.

2. The method for visual monitoring and early warning of foundation pit deformation according to claim 1, characterized in that: The method further includes step 8 of setting a deformation risk level threshold, and generating an alarm when the deformation risk level exceeds the deformation risk level threshold.

3. The method for visual monitoring and early warning of foundation pit deformation according to claim 2, characterized in that: The composite film layer (11) is a polyimide silicone composite film, the strain sensing layer (12) is a silver nanowire resistor grid, the pressure signal layer (13) is a flexible capacitive pressure sensor array, the environmental sensing layer (14) is a flexible temperature and humidity sensor array, the power supply unit (15) supplies power to the strain sensing layer (12), the pressure signal layer (13) and the environmental sensing layer (14), respectively, the surface functional layer (16) is a nano-antireflection film, and the fluorescent two-dimensional code marking points are printed by inkjet printing and ultraviolet curing process.

4. The method for visual monitoring and early warning of foundation pit deformation according to claim 3, characterized in that: The formula for calculating the local strain value in step 5 is: ； in is the local strain value, is the relative change of phase resistance, and k is the material constant.

5. The method for visual monitoring and early warning of foundation pit deformation according to claim 4, characterized in that: The formula for pressure distribution on the foundation pit surface in step 5) is: ； Where P is the calculated pressure value, is the base pressure value, is the measured pressure value, is the initial capacitance value.

6. The method for visual monitoring and early warning of foundation pit deformation according to claim 5, characterized in that: The specific steps of calculating the displacement field of the foundation pit surface using the image matching algorithm are as follows: Step 51, extracting SIFT feature points from images of continuous time phases; Step 52: Establishing feature point correspondence through nearest neighbor matching and consistency check; Step 53: Evaluate the matching quality using a random sampling consistency algorithm and remove abnormal matching points; Step 54: Calculate the displacement field of the foundation pit surface based on the matching point pairs.

7. The method for visual monitoring and early warning of foundation pit deformation according to claim 6, characterized in that: In step 6, the steps for fusing the strain data monitored by the electronic skin with the displacement data monitored by the visual system are: Step 61: Use a weighted fusion method to fuse the electronic skin strain data and the visual displacement data: Step 62: spatially align the electronic skin monitoring points with the visual monitoring points to establish a corresponding relationship; Step 63, respectively evaluating the uncertainty of the electronic skin data and the visual data; Step 64: assign weights to the two data sources based on the uncertainty index and calculate the fusion result.

8. The method for visual monitoring and early warning of foundation pit deformation according to claim 7, characterized in that: The formula for calculating the fusion result in step 64) is: ； in It is electronic skin data, is the visual displacement data, is the weight of the electronic skin data, is the weight of the visual displacement data, It is the result of fusion.

9. The method for visual monitoring and early warning of foundation pit deformation according to claim 8, characterized in that: The specific steps of the risk assessment algorithm based on fusion data are as follows: Based on the fusion results of the electronic skin strain data and visual displacement data corresponding to the fluorescent QR code markers, the fusion results of the strain, displacement, and displacement rate indicators are calculated respectively, and the risk level is determined based on the risk level threshold range; Adopt fuzzy comprehensive evaluation method and integrate multi-index risk assessment results; Based on the comprehensive evaluation results, the deformation risk level of the corresponding fluorescent QR code marking point is determined.

10. A foundation pit deformation visual monitoring and early warning system, characterized by: A method for visual monitoring and early warning of foundation pit deformation according to claim 9 is used to implement the visual monitoring and early warning system, comprising an electronic skin, a multispectral camera array (2), a computer monitoring server (3) and an alarm (4), wherein the electronic skin comprises a composite film layer (11), a strain sensing layer (12), a pressure signal layer (13), an environmental perception layer (14), a power supply unit (15) and a surface functional layer (16); the bottom of the composite film layer (11) is attached to the surface of the foundation pit to be detected, the strain sensing layer (12) is laid on the top surface of the composite film layer (11), the pressure signal layer (13) is laid on the strain sensing layer (12) based on the isolation layer, and the environmental perception layer (14) is laid on the isolation layer. The layer is laid on the pressure signal layer (13), the surface functional layer (16) is laid on the environmental perception layer (14), the power supply unit (15) supplies power to the strain sensing layer (12), the pressure signal layer (13) and the environmental perception layer (14), respectively, the multispectral camera array (2) is arranged around the foundation pit to be monitored, the computer monitoring server (3) is respectively connected to the strain sensing layer (12), the pressure signal layer (13), the environmental perception layer (14) and the multispectral camera array (2), and outputs the deformation risk level of the corresponding fluorescent two-dimensional code mark point based on the electronic skin data and the visual displacement data, and controls the alarm (4) to alarm if the deformation risk level exceeds the deformation risk level threshold.

Images