Method and system for measuring and identifying high crack of house wall

Abstract

The invention discloses a method and system for measuring and identifying a high crack of a house wall, and belongs to the technical field of buildings, and the method comprises the steps: collecting complete wall crack point cloud data, carrying out the geometric feature extraction of the wall crack point cloud data, and dynamically adjusting the deployment density of an environment and a strain sensor. And completing the prediction of the future width of the crack through the prediction model, completing the classification of the crack risk level based on the predicted value and the structure standard, and generating a final maintenance strategy according to the crack dynamic expansion condition, the risk level and the construction resource configuration rule. According to the scheme, by introducing dynamic data fusion and prediction analysis, through continuous data collection and environment change analysis, pre-judgment and intelligent decision making of the future development trend of the crack are achieved, the damage level of the crack is evaluated through an automatic algorithm, the system can generate personalized repair plans under different crack conditions, and the repair efficiency of the crack is improved. The repairing efficiency is improved.

Description

Technical Field

[0001] The invention belongs to the field of building technology, and in particular relates to a method and system for measuring and identifying cracks at high places on a wall of a house. Background Art

[0002] With the rapid development of the construction industry and the widespread use of high-rise buildings and complex structures, the detection and identification of cracks in building walls, especially those at high altitudes, has become an increasingly important task in building safety assessments. Cracks not only affect a building's appearance but also directly impact its safety. If cracks are not promptly and accurately detected and repaired, they can lead to further damage to the building and even pose potential safety hazards. Existing wall crack detection technologies primarily rely on manual inspection, traditional measurement tools, and static image recognition methods. However, these technologies exhibit numerous shortcomings when dealing with high-altitude cracks, complex structures, and dynamic environmental changes.

[0003] First, traditional manual inspection methods require workers to climb or use tools such as scaffolding to conduct high-altitude inspections, which is not only time-consuming and labor-intensive, but also poses a high safety hazard. Manual inspection is even more unrealistic and dangerous in high-rise buildings or inaccessible areas. Secondly, existing traditional measurement tools, such as rulers and crack measurement calipers, cannot provide accurate crack data, especially when measuring cracks at heights, where accuracy is difficult to guarantee. These tools can only provide limited information on crack width and depth, but cannot fully reflect the spatial structure, expansion trend, and risk level of the cracks. Third, existing image recognition technologies mostly rely on simple two-dimensional image processing methods, which cannot obtain three-dimensional information about cracks and have low recognition accuracy in complex environments (such as uneven lighting, diverse wall materials, complex crack morphology, etc.). In addition, these technologies often only provide static crack detection results and lack tracking and analysis of the dynamic evolution of cracks, resulting in an inability to accurately predict the future development trend of cracks and their potential threats to building structures.

[0004] Therefore, existing crack detection technologies have many limitations in terms of high-altitude crack detection, accurate measurement, dynamic analysis, and intelligent decision-making. To this end, we propose a method and system for measuring and identifying high-altitude cracks in house walls to address the above problems. Summary of the Invention

[0005] The purpose of the present invention is to solve the limitations of high-position crack detection in the prior art and to propose a method and system for measuring and identifying high-position cracks in house walls.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A method for measuring and identifying cracks at high places on a house wall, comprising:

[0008] Perform a global coarse-resolution scan of the wall to be tested to generate coarse point cloud data. Calculate the crack edge point density of the wall sub-area through local crack edge density detection. Perform a local high-resolution supplementary scan of the key crack area to generate high-resolution point cloud data. Finally, fuse the coarse point cloud data with the high-resolution point cloud data to form complete wall crack point cloud data.

[0009] Extract geometric features from wall crack point cloud data. Through crack structure connectivity modeling, edge density analysis, and crack fitting algorithm, extract the spatial features of the cracks and ultimately output the geometric feature set of the cracks.

[0010] The geometric feature set of the crack is input to calculate the distribution of key monitoring points of the crack, and the deployment density of environmental and strain sensors is dynamically adjusted. Dynamic environmental parameters are collected by deploying a strain sensor network, and a multi-source dynamic data set indexed by crack features is output for key crack areas.

[0011] A time series prediction model for crack expansion trends is constructed based on multi-source dynamic data sets. An environmental-stress sensitivity adjustment factor is introduced into the prediction model. The future width of cracks is predicted through the time series prediction model. The crack risk level is then graded based on the predicted value and structural standards.

[0012] A maintenance strategy is generated based on the predicted future crack width and risk level. The maintenance strategies are stratified according to the risk level, and the construction plan selection is further refined based on the predicted future crack width and average depth to generate the final maintenance strategy.

[0013] Preferably, in the global coarse resolution scan, the wall contour, potential crack direction and initial crack edge features are captured, and the key crack area is determined based on the crack edge point density of the wall area.

[0014] Preferably, geometric features of the crack point cloud data are extracted by considering the density of crack edge points and the connectivity of the crack area, and the crack structure regularization term is used to constrain the crack point cloud clustering process. After clustering is completed, the main axis direction fitting and edge feature expansion algorithm are used based on the selected crack area to extract the crack main axis and quantify the crack length and width.

[0015] Preferably, the local complexity factor of the crack is introduced in the process of fitting the main axis direction, and the average deviation of the crack point to the main axis is measured by the local complexity factor of the crack. If the crack deviates most from the straight line and the branches are most complex, the value of the local complexity factor of the crack is the largest, and then the local multi-end fitting is switched.

[0016] Preferably, the sensor network collects three dynamic environmental parameters in real time within the key crack area: current local temperature, current local humidity, and local strain of the wall where the crack is located.

[0017] Preferably, the environmental-stress sensitive adjustment factor simulates the relationship between environmental dynamics and crack propagation by the following calculation:

[0018]

[0019] Where T is the current temperature, T ref is the design reference temperature, S is the local strain of the wall, S ref is the design reference strain, λ is the system adjustment parameter. The environment-stress sensitive adjustment factor α env It plays a regulatory role in the dynamic prediction model. When the environment fluctuates greatly or the strain is high, it amplifies the sensitivity of the crack growth prediction.

[0020] Preferably, the final maintenance strategy includes a maintenance schedule, a recommended construction plan, construction team and equipment recommendations, and a re-inspection and observation plan.

[0021] A system for measuring and identifying cracks at high places on a house wall, comprising:

[0022] The data acquisition module is configured to perform a global coarse-resolution scan of the wall to be tested to generate coarse point cloud data, perform a local high-resolution supplementary scan of the key crack area to generate high-density point cloud data, and finally fuse the global low-resolution point cloud data with the local high-resolution point cloud data to form complete wall crack point cloud data;

[0023] The data preprocessing module is set to extract geometric features from the wall crack point cloud data. Through crack structure connectivity modeling, edge density analysis and crack fitting algorithm, the spatial features of the cracks are extracted and the geometric feature set of the cracks is output.

[0024] The dynamic feedback module is configured to collect three dynamic environmental parameters in real time, calculate the distribution of key monitoring points of the cracks, and dynamically adjust the deployment density of sensors in the data acquisition module;

[0025] The prediction module is configured to construct a time series prediction model for crack expansion trends based on multi-source dynamic data sets. This model is used to predict the future width of cracks and to classify crack risk levels based on the predicted values ​​and structural criteria.

[0026] The strategy generation module is configured to generate a hierarchical maintenance strategy based on the risk level-crack dynamic prediction dual-factor mechanism according to the dynamic expansion of cracks, risk level, and construction resource allocation rules. The maintenance strategy is layered according to the risk level, and the construction plan selection is further refined based on the future crack width prediction and average depth prediction to generate the final maintenance strategy.

[0027] To sum up, the technical effects and advantages of the present invention are as follows: This solution introduces dynamic data fusion and predictive analysis, and through continuous data collection and environmental change analysis, it realizes the prediction and intelligent decision-making of the future development trend of cracks. The hazard level of cracks is evaluated through automated algorithms, so that the system can generate personalized repair plans under different crack conditions, improve repair efficiency, and reduce errors and risks of manual operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a structural schematic diagram of the present invention;

[0029] Figure 2 It is a structural schematic diagram of the present invention. DETAILED DESCRIPTION

[0030] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0031] like Figure 1 As shown, a method for measuring and identifying cracks at high places on a house wall comprises:

[0032] Perform a global coarse-resolution scan of the wall to be tested to generate coarse point cloud data. Calculate the crack edge point density of the wall sub-area through local crack edge density detection. Perform a local high-resolution supplementary scan of the key crack area to generate high-density point cloud data. Finally, fuse the global low-resolution point cloud data with the local high-resolution point cloud data to form complete wall crack point cloud data.

[0033] Extract geometric features from wall crack point cloud data. Through crack structure connectivity modeling, edge density analysis, and crack fitting algorithm, extract the spatial features of the cracks and ultimately output the geometric feature set of the cracks.

[0034] Calculate the distribution of key monitoring points of cracks, dynamically adjust the deployment density of environmental and strain sensors, collect dynamic environmental parameters through the deployment of sensor networks, and output a multi-source dynamic dataset indexed by crack characteristics for crack sub-regions;

[0035] A time series prediction model for crack expansion trends is constructed based on multi-source dynamic data sets. An environmental-stress sensitivity adjustment factor is introduced into the prediction model to predict the future width of cracks. The crack risk level is then graded based on the predicted values ​​and structural standards.

[0036] Based on the dynamic expansion of cracks, risk levels, and construction resource allocation rules, a graded maintenance strategy is generated based on a dual-factor mechanism of risk level and dynamic crack prediction. Maintenance strategies are stratified according to risk level, and construction options are further refined based on future crack width predictions and average depth predictions to generate the final maintenance strategy.

[0037] Step 1

[0038] Regarding the 3D spatial data collection step for high-altitude crack detection, cracks are often located in hard-to-reach areas high above the wall. Traditional manual inspection or single-resolution laser scanning faces the challenges of numerous blind spots and difficulty capturing crack details. To address this issue, this step proposes an adaptive multi-scale laser radar scanning method for high-altitude cracks (AMLS-LiDAR). This method improves the 3D data integrity and spatial resolution of key crack areas while maintaining detection efficiency.

[0039] The system first performs a global coarse resolution scan on the high-altitude area of ​​the wall to be measured to generate preliminary point cloud data. Capturing the wall's contours, potential crack directions, and initial crack edge features. The LiDAR sensor uniformly scans the entire high-altitude wall surface at a base resolution of r0. For example, for a 30-meter-high exterior wall, an initial coarse-resolution scan samples at 0.05-meter intervals, quickly generating gridded data for the entire surface.

[0040] Then, the system calculates the crack edge point density s of the wall sub-area i by detecting the local crack edge density. i , and introduce the following resolution improvement factors:

[0041] r i =r0·1+δ·s i )

[0042] Among them, r i is the local scanning resolution of region i, r0 is the basic scanning resolution (e.g. 0.05 m), δ is the system adaptive adjustment coefficient (e.g. δ = 2), s i is the ratio of the number of crack edge points in the i-th region to the total number of regional points (e.g. s i = 0.2 means 20% of the points are crack edge points). i It determines the actual scanning accuracy of the sub-area. The denser the crack edge points are, the higher the scanning resolution will be.

[0043] The system then crack The sub-area i within is scanned with high resolution to generate a high-density point cloud The system automatically improves the resolution to fill in the spatial information missed during the coarse resolution phase, especially at key locations such as crack intersections, cracks near wall corners, or high-altitude structural nodes (such as beam-column connections), reducing data blind spots.

[0044] For example, in a corner area on the north facade of the building, a rough scan detected a crack edge density s i =0.3. After the additional scan, the final crack edge point cloud density increased by 1.6 times, completely restoring the characteristics of high-altitude fine cracks and crack depth changes. It successfully captured crack micro-branches with a depth of only 1 cm and a width of less than 2 mm, which are details that are difficult to capture with traditional low-resolution scanning.

[0045] Finally, the system fuses the global low-resolution point cloud P coarse With the local high-resolution point cloud P fine , forming a complete three-dimensional data of high cracks P=P coarse ∪P fine , providing high-quality input for subsequent crack feature extraction (step 2). The point cloud data P retains the actual sampling resolution r of each sub-region. i As a point cloud attribute, it provides resolution reference information for the subsequent feature extraction stage, further assisting in the accurate extraction of geometric features.

[0046] The adaptive multi-scale scanning strategy of this solution breaks through the limitations of "global fixed-resolution scanning" in complex high-altitude crack scenarios, improves the data density and edge integrity of key crack areas (such as crack-dense areas, corners, nodes, etc.), ensures the accuracy of crack length, width, depth, spatial direction and other features extracted in subsequent steps, and builds a highly reliable spatial data foundation.

[0047] Step 2

[0048] The input of this step is the three-dimensional point cloud data of the high crack output in the previous step P = P coarse ∪P fine , where P contains the complete three-dimensional coordinates (x, y, z) of the crack and the scanning resolution r of each sub-area i The core task of this step is to perform in-depth geometric feature extraction and quantitative analysis on the crack point cloud data, and output standardized feature parameters {L,W max ,D avg,θ}, which is used to support subsequent dynamic prediction and risk assessment steps. To address the complex structures unique to high-altitude cracks (such as multi-branch cracks, nonlinear orientation, and blurred edges), the system proposes a "fracture connectivity-density joint feature extraction method (CDJ-Extraction)". Through crack structure connectivity modeling, edge density analysis, and crack fitting algorithm, it accurately extracts the spatial characteristics of cracks and enhances the adaptability to abnormal crack structures (such as cross cracks and fracture cracks).

[0049] The system first extracts the main crack area C in P through the improved density-connected clustering algorithm (DL-Cluster). DL-Cluster considers the density s of the crack edge points i The crack structure regularization term Ω(C) is used to constrain the crack point cloud clustering process and improve the integrity of the crack main area. It is defined as follows:

[0050]

[0051] in, For point p k To the edge of the crack The shortest distance, E is the set of adjacent points inside the crack, ||p m -p n || is the distance between adjacent points, ∈ is a small constant to avoid division by zero, and λ1+λ2=1 is the regularization term coefficient. The first term measures the density of the edge distribution of the crack points, and the second term measures the structural connectivity between the crack points to ensure that the crack area is not mis-segmented due to the sparseness of the local point cloud. After clustering is completed, the system uses the main axis direction fitting and edge feature expansion algorithm based on the extracted crack area C to extract the main axis of the crack and quantify the crack length L and width W. max Taking the main axis direction of the crack as the reference, the projection length along the main axis direction in the C area is calculated as the crack length L, and the maximum boundary spacing in the short axis direction is W. max .

[0052] In order to cope with the "multi-branch, multi-level crack structure" often associated with high-altitude cracks, the system introduces a "crack local complexity factor" φ in the fitting process. i , used to automatically adjust the crack fitting accuracy and improve the quantitative stability under multi-branch cracks. i Defined as:

[0053]

[0054] Among them, p j is the crack point, l(p j ) is p j The projection points on the fitting axis, n is the total number of points in the crack area, φ iMeasures the average deviation of the crack point from the main axis and reflects the nonlinearity of the crack. i The larger the value, the more the crack deviates from the straight line and the more complex the branching is. The system will adaptively switch to the local multi-segment fitting strategy to improve L and W. max The extraction accuracy is improved to avoid distortion of crack features due to “forced uniaxial fitting”.

[0055] Based on the above analysis of the crack main axis and local structure, the system further uses high-resolution data area i The crack edge features are refined in the high-resolution area, and the average crack depth D is quantified. avg And the main crack strike angle θ. θ is the angle between the main axis of the crack and the horizontal plane. Especially for the corner area, the system will combine r i The strike factor of the high-resolution sub-area is fine-tuned, and multiple segments of θ are extracted segmentally for subsequent dynamic evolution analysis.

[0056] This step focuses on the core scenario of "complex structure of high-altitude cracks" and integrates multiple factors such as crack edge tightness, structural connectivity, and nonlinear complexity to ensure the accuracy, robustness, and reproducibility of crack geometric feature extraction. It is especially suitable for complex scenarios such as high-altitude corners, staggered cracks, or micro-branched cracks. The final output feature parameters {L, W max ,D avg ,θ} will be directly used as the input of the next step, providing high-quality and structured basic feature input for environmental dynamic data fusion and crack trend prediction.

[0057] Step 3

[0058] This step uses the crack geometric feature set {L,W max ,D avg ,θ} is the input. Since high-altitude cracks are not only affected by their own structural characteristics, but also closely related to the surrounding environment (temperature, humidity) and local stress of the wall, it is necessary to deploy an environment and structure dynamic perception network in the crack area to collect time-varying data of the crack's surrounding environment and structure. In view of the fact that high-altitude cracks are difficult to access and have complex structures, the system uses {L,W max ,D avg The system calculates the distribution of key monitoring points along the crack and dynamically adjusts the deployment density of environmental and strain sensors. For example, when the crack length L is greater than 3m and the crack strike angle θ is greater than 60° (close to vertical), the system automatically increases the density of strain sensors above and at both ends of the crack to focus on changes in structural loads under "high-risk crack strikes."

[0059] After the sensor network is deployed, the system collects three dynamic environmental parameters in real time within the key sub-area of ​​the crack: the current local temperature T, the current local humidity H, and the local strain S of the wall where the crack is located.

[0060] To maintain the consistency of dynamic data and the coupling of fracture characteristics, the system directly outputs a multi-source dynamic dataset indexed by fracture characteristics for all fracture sub-regions, standardized into the following structured output:

[0061] X={L,W max ,D avg ,θ,T,H,S}

[0062] Ultimately, the output X of this step will become the standardized input of the “dynamic prediction model” in step 4, directly serving the time series modeling of future crack expansion.

[0063] Step 4

[0064] The input of this step is the multi-source fusion dataset X={L,W max ,D avg ,θ,T,H,S}, the data contains both the spatial characteristics of the cracks and the dynamic environmental stress information. This step focuses on building a time series prediction model for the crack expansion trend based on X, and completing the assessment of the crack risk level under the same model framework. The system proposes a "Dynamic Crack Prediction-Risk Assessment Integrated Model (DRI-Model)" that directly realizes the future expansion trend of the cracks through a single model structure. The prediction is carried out and the automatic classification of crack risk level R is completed based on the predicted value and structural standards.

[0065] The model first performs dynamic prediction on the input X, focusing on predicting the crack width at the next time t+1 and average depth In terms of model design, in order to improve the response ability of crack prediction to dynamic environmental changes, DRI-Model innovatively introduces an "environment-stress sensitivity adjustment factor" α env , modeling the relationship between environmental dynamics and crack propagation:

[0066]

[0067] Where T is the current temperature, T ref is the design reference temperature, S is the local strain of the wall, S ref is the design reference strain, and λ is the system adjustment parameter. env It plays a regulatory role in the dynamic prediction model. When the environment fluctuates greatly or the strain is high, the system automatically amplifies the sensitivity in the crack growth prediction.

[0068] Based on this factor, the system predicts the future width of the crack:

[0069]

[0070] Among them, W max is the maximum width of the current crack, and η is the fitting coefficient of the prediction model. This prediction mechanism integrates environmental and structural dynamic factors, achieving a strong coupling between crack expansion trends and environmental dynamics, and can better reflect the dynamic behavior of high-altitude cracks in real building environments.

[0071] In the risk intelligence assessment module, the system directly And related standards to complete the risk level classification. The system adopts the following mapping relationship based on the building structure safety specifications:

[0072] like Then R = low risk;

[0073] like Then R = medium risk;

[0074] like Then R = high risk.

[0075] DRI-Model strictly follows the integrated design logic of "dynamic prediction + risk classification", avoiding the modeling fragmentation problem caused by the separation of prediction and risk analysis in traditional systems, and achieving the simplification of modeling process and the unification of evaluation chain. and the risk level R, for direct use by the next strategy generation module.

[0076] Step 5

[0077] The input of this step is the output of step 4 This step focuses on achieving "intelligent maintenance strategy generation based on dynamic risks". After completing the risk assessment, it automatically generates graded maintenance strategies, construction plans and auxiliary decision support based on the dynamic expansion of cracks, risk levels and construction resource allocation rules.

[0078] Based on the dual-factor mechanism of "risk level + dynamic crack prediction", the system proposes an "intelligent decision-making hierarchical generation method" to achieve step-by-step output of the maintenance strategy M:

[0079] In the first step, the system stratifies maintenance strategies based on risk level R (low / medium / high):

[0080] R = Low risk: The system marks it as "observation crack" and prioritizes "regular inspection + observation";

[0081] R = Medium Risk: The system marks it as a "routine maintenance crack" and recommends "planned maintenance";

[0082] R = High risk: The system marks it as an "emergency repair crack" and recommends "emergency treatment + special reinforcement."

[0083] In the second step, the system is further based on and Refine the construction plan selection:

[0084] like The system recommends "structural grouting + high-strength crack sealing + carbon fiber reinforcement";

[0085] like The system recommends "chemical anchor bolts + epoxy resin injection + crack surface waterproof coating";

[0086] like and The recommended solution is “flexible sealant + conventional crack sealing + surface coating reinforcement”.

[0087] The system introduces a "dynamic matching mechanism for construction resources" to automatically recommend work teams and equipment based on R and crack characteristics:

[0088] When R = high risk or When the system automatically recommends "high-altitude professional construction team", matching "suspended hanging basket + high-altitude safety rope + crack emergency sealing equipment";

[0089] When R = medium risk and cracks are concentrated from the ground to the middle of the wall, the "conventional building maintenance team + mobile scaffolding" combination is recommended;

[0090] For "high-risk + difficult-to-reach areas" (such as cracks high up on walls or in blind corners of structures), the system recommends the use of "drone crack review system + high-altitude auxiliary optical monitoring" equipment to improve construction accessibility and inspection efficiency.

[0091] In order to meet the "closed-loop guarantee requirements" of the engineering site, the system has designed a "re-inspection-observation plan linkage module". The specific operations are as follows:

[0092] For R = high risk and or For cracks, the system will forcibly output a "special re-inspection task" and plan to re-inspect the cracks "one month after the emergency repair is completed";

[0093] The system will simultaneously generate a "dynamic observation plan" and arrange two re-inspections within 3-6 months after construction completion, focusing on monitoring new crack expansion or stress recovery;

[0094] If R = medium risk and the expansion rate is low, the system only arranges "semi-annual regular inspection + observation" to avoid wasting resources.

[0095] The system's "RT-Maintainer mechanism" also automatically generates maintenance schedules for different priorities:

[0096] R = low risk, recommended to conduct an inspection once every six months;

[0097] R = Medium risk, recommended to be completed within 1-3 months;

[0098] R = High risk, recommendation: "Immediate repair, reinforcement or sealing within 48 hours."

[0099] The system maintenance strategy M ultimately includes the following complete content:

[0100] Maintenance schedule (e.g., "immediate execution," "within 1-3 months," "inspection within six months");

[0101] Recommended construction plans (such as "grouting + carbon fiber reinforcement", "chemical anchor bolts + plugging", etc.);

[0102] Recommendations for construction teams and equipment (e.g., "high-altitude work team + hanging basket," "conventional construction team + scaffolding," or "drone-assisted observation," etc.);

[0103] Re-inspection and observation plan (such as "special re-inspection after 1 month" or "6-month periodic inspection").

[0104] By coupling risk levels, dynamic predictions, and construction resources, RT-Maintainer achieves a fully automated transition from "risk assessment" to "maintenance action plan," ensuring the scientific nature of the strategy and the feasibility of construction.

[0105] The technical solutions in the above-mentioned embodiments of the present application have at least the following technical effects or advantages: by combining advanced three-dimensional measurement technology with dynamic prediction models, it can provide unprecedented high-precision data during the measurement of high-altitude cracks, breaking through the measurement limitations of traditional methods, especially in high-altitude or difficult-to-reach crack areas, ensuring the comprehensiveness and accuracy of data collection, and by introducing dynamic data fusion and predictive analysis, realizing the prediction and intelligent decision-making of future development trends of cracks.

[0106] The present application also provides a method and system for measuring and identifying cracks at high places on a wall of a house. Figure 2 As shown, the data acquisition module is configured to perform a global coarse resolution scan of the wall to be tested to generate coarse point cloud data, perform a local high-resolution supplementary scan of the key crack area to generate high-density point cloud data, and finally fuse the global low-resolution point cloud data with the local high-resolution point cloud data to form complete wall crack point cloud data;

[0107] The data preprocessing module is set to extract geometric features from the wall crack point cloud data. Through crack structure connectivity modeling, edge density analysis and crack fitting algorithm, the spatial features of the cracks are extracted and the geometric feature set of the cracks is output.

[0108] The dynamic feedback module is configured to collect three dynamic environmental parameters in real time, calculate the distribution of key monitoring points of the cracks, and dynamically adjust the deployment density of sensors in the data acquisition module;

[0109] The prediction module is configured to construct a time series prediction model for crack expansion trends based on multi-source dynamic data sets. This model is used to predict the future width of cracks and to classify crack risk levels based on the predicted values ​​and structural criteria.

[0110] The strategy generation module is configured to generate a graded maintenance strategy based on a dual-factor mechanism of risk level and dynamic crack prediction, according to the dynamic expansion of cracks, risk level, and construction resource allocation rules. The maintenance strategy is stratified according to risk level, and the construction plan selection is further refined based on the future crack width prediction and average depth prediction to generate the final maintenance strategy.

[0111] The technical solutions in the embodiments of the present application described above have at least the following technical effects or advantages: the system can assess the dynamic changes of cracks based on real-time building data and provide data support for repair decisions, ensuring that cracks can be accurately repaired or reinforced at an early stage, preventing crack expansion from posing a threat to building structural safety.

[0112] The system works as follows: By integrating dynamic data fusion and predictive analysis, as well as continuous data collection and environmental change analysis, it enables prediction and intelligent decision-making regarding future crack development trends. Based on real-time building data, the system assesses the dynamic changes in cracks and provides data support for repair decisions, ensuring that cracks are accurately repaired or reinforced at an early stage, preventing their expansion from posing a threat to the building's structural safety. Furthermore, the system uses an automated algorithm to assess the hazard level of cracks, enabling the system to generate personalized repair plans for different crack scenarios, improving repair efficiency and reducing the errors and risks associated with manual operations.

[0113] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A method for measuring and identifying cracks at high places on a building wall, characterized in that: include: Perform a global coarse-resolution scan of the wall to be tested to generate coarse point cloud data. Calculate the crack edge point density of the wall sub-area through local crack edge density detection. Perform a local high-resolution supplementary scan of the key crack area to generate high-resolution point cloud data. Finally, fuse the coarse point cloud data with the high-resolution point cloud data to form complete wall crack point cloud data. Extract geometric features from wall crack point cloud data. Through crack structure connectivity modeling, edge density analysis, and crack fitting algorithm, extract the spatial features of the cracks and ultimately output the geometric feature set of the cracks. The geometric feature set of the crack is input to calculate the distribution of key monitoring points of the crack, and the deployment density of environmental and strain sensors is dynamically adjusted. Dynamic environmental parameters are collected by deploying a strain sensor network, and a multi-source dynamic data set indexed by crack features is output for key crack areas. A time series prediction model for crack expansion trends is constructed based on multi-source dynamic data sets. An environmental-stress sensitivity adjustment factor is introduced into the prediction model. The future width of cracks is predicted through the time series prediction model. The crack risk level is then graded based on the predicted value and structural standards. A maintenance strategy is generated based on the predicted future crack width and risk level. The maintenance strategies are stratified according to the risk level, and the construction plan selection is further refined based on the predicted future crack width and average depth to generate the final maintenance strategy.

2. The method for measuring and identifying cracks at high places on a house wall according to claim 1, characterized in that: In the global coarse resolution scan, the wall contour, potential crack direction and initial crack edge features are captured, and the key crack area is determined based on the crack edge point density of the wall area.

3. The method for measuring and identifying cracks at high places on a house wall according to claim 1, characterized in that: The geometric feature extraction of crack point cloud data is performed by considering the density of crack edge points and the connectivity of crack areas, and using crack structure regularization terms to constrain the crack point cloud clustering process. After clustering is completed, the main axis direction fitting and edge feature expansion algorithm are used based on the selected crack area to extract the crack main axis and quantify the crack length and width.

4. The method for measuring and identifying cracks at high places on a house wall according to claim 3, characterized in that: The local complexity factor of the crack is introduced in the main axis direction fitting process, and the average deviation from the crack point to the main axis is measured by the local complexity factor of the crack. When the crack deviates most from the straight line and the branches are most complex, the value of the local complexity factor of the crack is the largest, and then the local multi-terminal fitting is switched.

5. The method for measuring and identifying cracks at high places on a house wall according to claim 1, characterized in that: The sensor network collects three dynamic environmental parameters in real time in the key crack area: current local temperature, current local humidity and local strain of the wall where the crack is located.

6. The method for measuring and identifying cracks at high places on a house wall according to claim 1, characterized in that: The environmental-stress sensitive adjustment factor simulates the relationship between environmental dynamics and crack propagation by the following calculation: Where T is the current temperature, T ref is the design reference temperature, S is the local strain of the wall, S ref is the design reference strain, λ is the system adjustment parameter; the environment-stress sensitive adjustment factor α env It plays a regulatory role in the dynamic prediction model. When the environmental fluctuations are large or the strain is high, the sensitivity of the crack growth prediction is amplified.

7. The method for measuring and identifying cracks at high places on a house wall according to claim 1, characterized in that: The final maintenance strategy includes a maintenance schedule, recommended construction plans, construction team and equipment recommendations, and a re-inspection and observation plan.

8. A system for measuring and identifying cracks at high places on house walls, characterized in that: include: The data acquisition module is configured to perform a global coarse-resolution scan of the wall to be tested to generate coarse point cloud data, perform a local high-resolution supplementary scan of the key crack area to generate high-density point cloud data, and finally fuse the global coarse point cloud data with the local high-density point cloud data to form complete wall crack point cloud data; The data preprocessing module is set to extract geometric features from the wall crack point cloud data. Through crack structure connectivity modeling, edge density analysis and crack fitting algorithm, the spatial features of the cracks are extracted and the geometric feature set of the cracks is output. The dynamic feedback module is set to collect three dynamic environmental parameters in real time, calculate the distribution of key monitoring points of the crack by inputting the geometric feature set of the crack, and dynamically adjust the deployment density of sensors in the data acquisition module; The prediction module is configured to construct a time series prediction model for crack expansion trends based on multi-source dynamic data sets. This model is used to predict the future width of cracks and to classify crack risk levels based on the predicted values ​​and structural criteria. The strategy generation module is configured to generate a maintenance strategy based on the prediction of the future width of the crack and the risk level; The maintenance strategies are stratified according to the risk level, and the construction plan selection is further refined based on the future crack width prediction and average depth prediction to generate the final maintenance strategy.

Images