Intelligent analysis system for construction project management based on BIM technology

Abstract

The present invention belongs to the technical field of construction project management, and specifically discloses an intelligent analysis system for construction project management based on BIM technology. The system includes: a steel mesh initial data acquisition module, a steel mesh casting data acquisition module, a steel mesh morphological index analysis module, a steel mesh laying risk analysis module, and a steel mesh laying risk feedback terminal. The present invention solves the problem of the inability to accurately identify the spatial position deviation of steel bars in the prior art by collecting the overall image of the steel mesh before casting and building a BIM model. At the same time, by analyzing the risk of steel mesh laying from three dimensions: binding point characteristics, spacing characteristics, and deformation characteristics, the current binding point, spacing, and deformation can be accurately identified, achieving a detailed quantitative analysis of the spatial position deviation of steel bars and the quality of connections, thus filling the blind spot of the prior art in detecting the risk of steel bar connection failure. In addition, the deficiencies of the prior art in analyzing the state of steel bars are improved by introducing a risk compensation factor.

Description

Technical Field

[0001] The present invention belongs to the technical field of construction project management, and in particular relates to an intelligent analysis system for construction engineering project management based on BIM technology. Background Art

[0002] Concrete is the most widely used material in construction projects. Concrete pouring is the core link of building structure. Its quality directly determines the safety, durability and functionality of the building structure. In order to ensure the safety of the building structure, the quality of concrete pouring needs to be managed.

[0003] Existing technologies, such as the Chinese invention patent application with application number 202411854286.X, disclose a highway concrete layer construction quality management system. This system uses an ultrasonic detection module to acquire horizontal ultrasonic conduction detection data in real time, and combines this with the time-domain reflection feature analysis technology of a vibration result analysis module to accurately locate areas of insufficient vibration. A path optimization algorithm based on the vibration plan generation module dynamically generates supplementary vibration plans, and dual-mode communication technology in the background synchronous transmission module enables real-time interaction between detection data and control instructions. This helps improve the quality and efficiency of highway concrete layer construction, reduce engineering quality risks caused by insufficient vibration, and extend the service life of the highway.

[0004] With regard to the existing technical solutions, the existing focus is on the overall part of poured concrete, and steel bars are the main core structure of concrete pouring. Currently, there is no specific analysis of the status of the steel bars after pouring, and there are still deficiencies in the following aspects: 1. The spatial position deviation of the steel bars cannot be accurately identified, such as the degree of disorder in the arrangement of the steel bars and the degree of displacement deviation are not quantified in detail, resulting in deviations in the subsequent bearing capacity assessment of the concrete.

[0005] 2. Ignoring the hidden risks of steel bar connection quality. Concrete flow may wash away the binding wires, causing failure of the connection nodes and interruption of steel bar stress transmission, which in turn leads to instability of the steel bar skeleton. This makes the concrete pouring quality analysis relatively limited and fails to integrate the impact of potential risk situations on concrete pouring quality as much as possible. Summary of the Invention

[0006] In view of this, in order to solve the above problems, an intelligent analysis system for construction project management based on BIM technology is proposed.

[0007] The purpose of the present invention can be achieved through the following technical solutions: The present invention provides an intelligent analysis system for construction project management based on BIM technology, the system comprising: the system comprising: a steel mesh initial data acquisition module, which collects the overall image of the corresponding laid steel mesh before concrete pouring, builds a steel mesh BIM model based on the image, and marks the bundling points and steel bar laying spacing in the BIM model.

[0008] The steel mesh pouring data acquisition module uses an X-ray detector to perform multi-angle inspection on the concrete components after pouring, and generates a fused panoramic radiographic image after registration.

[0009] The steel mesh morphological index analysis module is composed of a bundling point recognition unit, a spacing recognition unit and a deformation recognition unit. Each unit sequentially recognizes the current bundling point feature, the current spacing feature and the current deformation feature based on the radiographic image.

[0010] The steel mesh laying risk analysis module calculates the bundling status consistency and spacing arrangement consistency based on the current bundling point characteristics and the current spacing characteristics in combination with the steel mesh BIM model, and sets a risk compensation factor based on the current deformation characteristics. The comprehensive risk index of the steel mesh laying is calculated by comprehensively considering the bundling status and spacing arrangement consistency as well as the risk compensation factor.

[0011] The steel mesh laying risk feedback terminal triggers the corresponding early warning instruction when the comprehensive risk index exceeds the preset threshold, and provides feedback on the comprehensive risk index of steel mesh laying.

[0012] Compared with the existing technology, the beneficial effects of the present invention are as follows: (1) The present invention digitizes the steel bar laying design parameters based on the collection of the overall image of the steel mesh before pouring and the construction of a BIM model, thereby solving the technical problem in the existing technology that the spatial position deviation of the steel bar cannot be accurately identified, resulting in deviations in the subsequent bearing capacity assessment of concrete, and provides an accurate benchmark for steel bar status analysis.

[0013] (2) The present invention analyzes the risk of steel mesh laying from three dimensions: binding point characteristics, spacing characteristics, and deformation characteristics. This solves the current problem of focusing only on the overall concrete part and ignoring the quality risk of steel bar connections. It can accurately identify the current binding points, spacing, and deformation conditions, and achieve a detailed quantitative analysis of the spatial position deviation and connection quality of the steel bars, thus filling the blind spot in the detection of steel bar connection failure risks in existing technologies.

[0014] (3) The present invention calculates the comprehensive risk index of steel mesh laying by combining the BIM model annotation data with the actual detection characteristics, statistically analyzing the bundling status and spacing arrangement, and introducing risk compensation factors. It fully integrates the potential risks and the hidden risks of steel bar connection quality and spatial position deviation, provides an accurate basis for concrete pouring quality assessment, and improves the shortcomings of existing technologies in steel bar status analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0016] Figure 1 This is a schematic diagram of the system module connection of the present invention.

[0017] Figure 2 This is a schematic diagram of the structure of the steel mesh morphology index analysis module of the present invention.

[0018] Figure 3 It is a schematic diagram of the overall implementation process of the present invention. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0020] See also Figure 1 and Figure 3 As shown, the present invention provides an intelligent analysis system for construction project management based on BIM technology, which includes: a steel mesh initial data acquisition module, a steel mesh casting data acquisition module, a steel mesh morphological index analysis module, a steel mesh laying risk analysis module and a steel mesh laying risk feedback terminal.

[0021] In the above, the steel mesh casting data acquisition module is connected to the steel mesh initial data acquisition module and the steel mesh shape index analysis module respectively, and the steel mesh laying risk analysis module is connected to the steel mesh shape index analysis module and the steel mesh shape index analysis module respectively.

[0022] The steel mesh initial data acquisition module collects the overall image of the corresponding laid steel mesh before concrete pouring, builds a steel mesh BIM model based on the image, and marks the binding points and steel bar laying spacing in the BIM model.

[0023] Specifically, the marking of the bundling points and the spacing between steel bars includes: marking the coordinates of the bundling points in an image coordinate system, and identifying the number of bundling circles and the maximum bundling gap area of ​​each bundling point through image recognition technology, and then marking the position coordinates, number of bundling circles and maximum bundling gap area of ​​each bundling point in the BIM model.

[0024] Import the transverse spacing of each transverse reinforcement and the longitudinal spacing of each longitudinal reinforcement in the laid steel mesh, and then mark the transverse and longitudinal spacing in the BIM model.

[0025] The embodiment of the present invention collects the overall image of the steel mesh before pouring and builds a BIM model to digitize the steel bar laying design parameters, thereby solving the technical problem in the existing technology that the spatial position deviation of the steel bars cannot be accurately identified, resulting in deviations in the subsequent bearing capacity assessment of concrete, and provides an accurate benchmark for steel bar status analysis.

[0026] The steel mesh casting data acquisition module performs multi-angle detection on the concrete components after casting through an X-ray detector, and generates a fused panoramic radiographic image after registration.

[0027] For example, the orthographic and 45-degree oblique images of the same detection area may be registered based on ORB feature point matching to generate a fused image, and the matching accuracy may be less than or equal to ≤2 pixels in order to further limit the matching accuracy.

[0028] See also Figure 2 As shown, the steel mesh morphological index analysis module is composed of a bundling point recognition unit, a spacing recognition unit and a deformation recognition unit. Each unit sequentially recognizes the current bundling point feature, the current spacing feature and the current deformation feature based on the radiographic image.

[0029] Specifically, the specific identification process of the current bundling point feature includes: A1, identifying the current number of bundling circles and the center point position of each bundling point from the fused panoramic radiographic image, and converting the center point position into the current position coordinates.

[0030] The number of binding circles can be identified by using the annular projection of the binding wire in the radiographic image to detect the annular contour through the Hough circle transform, and then counting the number of concentric circle layers. The number of circles corresponds to the number of concentric rings. For example, 2 layers indicates double binding. The grayscale gradient is combined to distinguish the effective number of circles. There is a grayscale difference between the binding wire and the steel bar, and the grayscale of the metal wire in the radiographic image is higher.

[0031] A2. Based on the coordinates of the binding points marked in the BIM model, the steel mesh area in the fused image is divided into pixel sub-areas, and image enhancement is performed on each pixel sub-area.

[0032] A3. After performing target detection on each enhanced pixel sub-region, the bounding box position of the bundling point is obtained. Non-maximum suppression is performed on the detection results and duplicate boxes are eliminated.

[0033] A4. Extract the mean value, variance, and circularity of the grayscale of the binding wire area within the binding point boundary box, and use the three as the actual grayscale features of the binding wire area.

[0034] A5. Use the Otsu threshold segmentation method to convert each pixel sub-region into a binary image. Perform a morphological opening operation on the binary image to remove noise points. The gap area between the binding wire and the steel bar is recorded as the connected gap area. Calculate the pixel area of ​​each connected gap area, and take the maximum value as the maximum gap area.

[0035] A6. The current position coordinates of each bundling point, the current number of bundling circles, the actual grayscale characteristics of the corresponding bundling wire area, and the maximum gap area are used as the current bundling point features.

[0036] In one specific embodiment, obtaining the bounding box of the binding points in step A3 requires a process involving data labeling, model training, inference detection, and geometric calibration. The binding point locations are manually labeled in the X-ray image using tools such as LabelMe to generate standard format data containing coordinates and categories. Subsequently, object detection models such as YOLOv6 are trained using CIoULoss to optimize bounding box regression, and mosaic data enhancement is combined to improve small object detection capabilities. During inference, the model outputs the original predicted box, which is then filtered for overlapping boxes using non-maximum suppression. Image distortion is then corrected using camera calibration parameters. Pixel coordinates are converted to actual physical coordinates using a scale factor. Ultimately, a bounding box accurately enclosing the binding wire image is obtained, providing a spatial reference for subsequent state determination.

[0037] It should be added that image enhancement, Otsu threshold segmentation method, target detection, morphological opening operation and non-maximum suppression all adopt existing technical means, and their specific execution process is not the focus of the present invention and will not be repeated here.

[0038] It can be understood that the specific calculation formula of the circularity is: ,in, Indicates circularity, represents the pixel contour area, represents the perimeter of the pixel outline, Represents pi.

[0039] Specifically, the specific recognition process of the current spacing feature includes: B1, after pre-processing the fused panoramic radiographic image, extracting the steel bar contour edge through Canny edge detection, using probabilistic Hough transform to detect long straight lines, and then extracting the horizontal and vertical steel bar center lines.

[0040] B2. Extract the Y coordinates of the center lines of all transverse reinforcements, sort them, calculate the differences between adjacent Y coordinates, eliminate outliers, and take the average of the remaining Y coordinate differences as the measured transverse spacing. Similarly, extract the X coordinates of the center lines of the longitudinal reinforcements to obtain the measured longitudinal spacing.

[0041] B3. Set the measured transverse spacing of all transverse reinforcements and the measured longitudinal spacing of all longitudinal reinforcements as the current spacing features.

[0042] It should be noted that the specific extraction of rebar outlines in step B1 is as follows: 1) Use median filtering or Gaussian filtering to remove radiographic noise. Histogram equalization or adaptive histogram equalization is used to enhance image contrast and highlight the grayscale difference between the rebar and the concrete background. To address potential radiographic artifacts or uneven illumination in radiographic images, a top-hat transform or homomorphic filtering can be used to further correct background brightness.

[0043] 2) After preprocessing, edges are extracted using the Canny edge detection algorithm. This involves applying a Gaussian smoothing kernel (e.g., 3×3 or 5×5) to the image for secondary denoising. Gradient magnitude and direction are calculated, and non-maximum suppression is used to refine the edges. Finally, a double threshold is applied to connect true edges and filter out false responses. The high threshold is typically 2-3 times the low threshold. After edge extraction, a contour search algorithm, such as OpenCV's findContours function, is used in RETR_EXTERNAL mode to extract the outer contour and obtain a continuous edge contour of the rebar. Morphological closing operations are then used to connect edge breakpoints caused by noise or ray attenuation, ultimately yielding a complete, clear rebar contour. The morphological closing operation can select circular or linear structuring elements.

[0044] It is understandable that the specific processes and functions involved in the filtering, histogram equalization enhancement, Canny edge detection algorithm, contour search algorithm and morphological closing operation described above are all existing technical means and will not be described in detail.

[0045] It should also be noted that the parameter settings for detecting long straight lines using the probabilistic Hough transform in step B1 may be: setting the minimum line segment length to 100 pixels and the line segment interval to ≤ 20 pixels.

[0046] In another specific embodiment, the specific identification process of the current deformation feature includes: C1, performing mask extraction on a single steel bar area through BIM model positioning, and applying a skeletonization algorithm to generate a steel bar skeleton with a single pixel width.

[0047] Among them, the skeletonization algorithm is an existing mature algorithm and will not be explained in detail.

[0048] C2. Perform quadratic polynomial fitting on the pixel coordinates of each point on the skeleton to calculate the fitting residuals, and define the line connecting the two end points of the steel bar as the target axis. Calculate the maximum distance and average distance from each point on the steel bar skeleton to the target axis, and calculate the difference between the two to obtain the maximum deviation distance.

[0049] Understandably, during concrete pouring, steel bars may experience local bending or non-uniform deformation due to vibration, collision, etc. The quadratic polynomial fitting residual quantifies the degree of deviation between the skeleton curve and the ideal curve. Larger residuals indicate more significant local deformation.

[0050] As you can understand, using the line connecting the two ends of the rebar as the ideal axis and calculating the difference between the maximum distance from each point to the axis and the average distance can highlight the point of maximum abnormal deformation, such as a single point bend, and avoid the average distance masking local extreme deformation. This is suitable for detecting dangerous shapes such as inflection points or sharp bends. Distance calculation can also be performed using the distance calculation formula between two points.

[0051] C3. Select each section of the steel skeleton according to the preset interval distance, extract the outline of the selected section, calculate the minimum circumscribed circle area of ​​each section, and calculate the ratio of the outline area of ​​each section to its minimum circumscribed circle area, and record the ratio as the outline rule ratio.

[0052] Understandably, the ideal cross-section of a steel bar is a regular circle. If it becomes oval or partially concave due to compression or collision, the ratio of its outline area to the minimum circumscribed circle area decreases. This metric is directly related to the geometric integrity of the cross-section; a smaller ratio indicates more severe cross-sectional distortion, such as irregular shapes caused by flattening or twisting.

[0053] C4. Take the standard deviation of the minimum circumscribed circle area corresponding to each section as the steel bar profile difference, and extract the minimum profile regularity ratio.

[0054] Understandably, the standard deviation of the minimum circumscribed circle area of ​​each section along the length of the steel bar reflects the fluctuation range of the cross-sectional dimensions. A larger standard deviation indicates more uneven deformation along the axis of the bar, such as local expansion or contraction.

[0055] C5. Taking the fitting residual, maximum deviation distance, steel bar profile difference and minimum profile regularity ratio as current deformation features.

[0056] Understandably, the specific process of solving the minimum circumscribed circle is as follows: perform a convex hull operation on the contour vertex set, remove the inner concave points, and retain the peripheral feature points.

[0057] Based on the convex hull vertices, all vertex pairs are traversed through the rotating calculus method to calculate the minimum circle enclosing all points. For non-convex contours, such as local concavities caused by necking, the distance from the concave point to the circumscribed circle is additionally calculated. If the radius exceeds the radius, the circle radius is expanded until all points are included.

[0058] Connected domains are marked on the binary contour image, the total number of white pixels is calculated, and the total number of white pixels is converted into actual area, which is the product of the total number of pixels and the image scale.

[0059] It should be added that by fitting numerical indicators such as residuals and maximum deviation distance, the steel bar deformation is transformed from qualitative observation to quantitative evaluation, which facilitates the setting of thresholds and provides an objective basis for construction quality acceptance. At the same time, by decomposing the steel bar deformation into quantifiable indicators such as local bending, cross-sectional distortion, and uniformity defects, it not only covers the detailed detection of a single dimension, but also realizes the comprehensive evaluation of the overall morphology. The quadratic polynomial fitting calculation is an existing technical means, and its specific calculation formula is no longer shown.

[0060] Furthermore, the specific setting process of the spacing distance in step C3 includes: C31, extracting the number of bending points based on the pixel coordinates of each point on the skeleton, and taking the ratio of the number of bending points to the total length of the steel bar as the bending density.

[0061] C32. If the bending density is less than or equal to the preset bending density, the default division interval distance is used as the preset interval distance. Otherwise, the steel bar is divided into several segments, the number of bending points of each steel bar segment is extracted, and its bending density is calculated.

[0062] C33. The steel bar segment with a bending density less than or equal to the preset bending density is recorded as a normal intercepted segment, the default division interval distance is used as the preset interval distance, and the steel bar segment with a bending density greater than the preset bending density is recorded as an encrypted intercepted bending segment.

[0063] C34, the bending density of the encrypted intercepted bending segment, the preset bending density and the default division interval distance are respectively recorded as 、 and , the preset spacing distance of the encrypted intercepted curved section is obtained by combining the three , , is a natural constant, The preset minimum separation distance.

[0064] In a specific embodiment, the number of bending points is extracted based on the skeleton pixel coordinates. The coordinates are first preprocessed such as denoising and smoothing, and then the curvature of each point is calculated by the three-point difference method or the quadratic difference method. The local peak points whose curvature is greater than the neighborhood are screened by setting an absolute or relative curvature threshold. The dense neighboring points are eliminated by combining the neighborhood non-maximum suppression, and the non-structural deformation points near the skeleton endpoints are filtered. Finally, the number of bending points that truly reflects the bending characteristics of the skeleton is obtained. The three-point difference method and the quadratic difference method are existing algorithms, and the calculation formulas are no longer specifically displayed.

[0065] Understandably, when it is not dense at all, that is, when the deformation of the steel skeleton is small, the fixed default interval distance is directly used to quickly complete the section selection. However, the bending point distribution of the steel skeleton is often uneven, such as in weak links of construction and stress concentration areas where bending is dense. At this time, the bending is relatively dense. By segmenting again, the bending density of each section is identified, and the detection interval of the sections with larger bending density is narrowed to ensure that the cross-section detection of key deformation areas, such as connection points and vibration collision areas, is more dense to avoid missing local serious bending.

[0066] It should be noted that the segmentation can be based on a fusion of deformation characteristics and engineering experience. The specific settings are as follows: 1) Utilize statistically significant changes in the spacing between bending points, such as when the standard deviation exceeds a threshold, to automatically identify the start and end locations of dense deformation zones. 2) Incorporating construction techniques, such as rebar connections, binding, and structural design, pre-mark high-risk sections, such as beam-column joints and post-cast strip overlaps, and forcibly treat them as independent segments to ensure inspection density in critical areas. 3) In sections without obvious deformation characteristics, divide them into equal intervals, such as 200mm, to ensure inspection coverage and avoid missing long, uniform deformations.

[0067] It should be noted that the preset minimum spacing distance is set based on a comprehensive consideration of equipment accuracy, engineering specifications, calculation efficiency, and sensitivity to steel bar deformation. For example, based on image resolution, where 1 pixel corresponds to 0.1 mm, the preset minimum spacing distance is at least 50 mm. The reference specification requires one inspection point every 100 mm, so the preset minimum spacing distance can be 50-100 mm to cover critical deformations. At the same time, historical data verification has determined that different steel bar types, such as main bars, have a preset minimum spacing distance of 50 mm, while structural bars have a preset minimum spacing distance of 80 mm.

[0068] It can be understood that the exponential decay model is used to smooth the compression interval, which not only ensures the detection accuracy in dense areas, but also avoids over-dense sampling by limiting the minimum interval. The introduction of the natural constant e can make the adjustment process continuous, adapt to the gradual changes in deformation in the project, and realize risk-oriented and dynamically optimized detection interval setting, balancing accuracy and efficiency.

[0069] The embodiment of the present invention analyzes the risk of steel mesh laying from three dimensions: binding point characteristics, spacing characteristics, and deformation characteristics. This solves the current problem of focusing only on the overall concrete part and ignoring the quality risks of steel bar connections. It can accurately identify the current binding points, spacing, and deformation conditions, and realize a detailed quantitative analysis of the spatial position deviation and connection quality of the steel bars, filling the blind spot in the detection of steel bar connection failure risks in existing technologies.

[0070] The steel mesh laying risk analysis module calculates the bundling state consistency and spacing arrangement consistency based on the current bundling point characteristics and the current spacing characteristics in combination with the steel mesh BIM model, and sets a risk compensation factor based on the current deformation characteristics. The comprehensive risk index of the steel mesh laying is calculated by comprehensively considering the bundling state and spacing arrangement consistency as well as the risk compensation factor.

[0071] Specifically, the specific statistical process of the bundling state consistency includes: D1, calculating the Euclidean distance between the current position coordinates of each bundling point and the marked position coordinates, and taking the maximum value as the maximum displacement distance of the bundling point.

[0072] D2. Calculate the absolute difference between the current number of strapping circles and the marked number of circles, as well as the absolute difference between the current maximum gap area and the marked maximum strapping gap area at each strapping point, and take the average of these values ​​to obtain the difference in the number of strapping circles and the difference in the strapping gap area at each strapping point.

[0073] D3. Compare the actual grayscale features of the binding wire area with the set reference grayscale feature interval, and count the proportion of the number of features exceeding the interval to the total number of features, which is recorded as the deviation feature ratio.

[0074] D4. For each over-limit feature, calculate its relative deviation from the corresponding reference feature value, perform weighted summation to obtain the comprehensive feature relative deviation value, input the deviation feature ratio and the comprehensive feature relative deviation value into the Sigmoid function, and then output the strapping status evaluation compensation factor.

[0075] D5. Normalize the maximum displacement distance of the bundling point, the difference in the number of bundling circles at the bundling point, and the difference in the bundling gap area. Combine the normalized indicators through a linear function to obtain the preliminary bundling state consistency.

[0076] D6. The initial bundling state consistency is corrected by the bundling state evaluation compensation factor to obtain the final bundling state consistency.

[0077] Understandably, the grayscale features of the tying wire are directly related to material properties and morphological regularity, such as contour distortion caused by rust, twisting, and loosening. While these feature anomalies do not directly reflect geometric displacement, they can implicitly reduce the mechanical properties of the tying wire, such as anchoring strength and anti-slip resistance. By converting grayscale deviations into compensation factors using a Sigmoid function, the accuracy of the fit can be dynamically adjusted. When grayscale features exhibit severe anomalies, such as excessive variance or circularity below a threshold, the compensation factor reduces the accuracy and highlights the impact of material or morphological defects. This overcomes the limitations of traditional geometric parameter analysis and makes the assessment more realistic for engineering applications. For example, if rust on the tying wire is located in the correct position, the actual anchoring strength may be insufficient. The compensation factor can effectively identify such hidden dangers.

[0078] In a specific embodiment, a full-factor evaluation system for bundling quality is constructed by integrating geometric parameters, circle number / gap difference and grayscale feature deviation. Geometric parameters, namely displacement, circle number and gap, reflect construction accuracy and mechanical tightness. Grayscale features such as mean, variance and circularity reveal the material uniformity, surface flatness and morphological regularity of the bundling wire. The fit results extend from spatial position and quantity compliance to material performance and morphological quality, thereby providing precise targeting for subsequent construction rectification, such as replacing bundling wire with grayscale abnormalities and adjusting the circle number / gap, to achieve quality control upgrade from surface compliance to inherent safety.

[0079] It should also be added that the normalization in step D5 can be processed by maximum-minimum normalization, and the index after normalization of the linear function combination is used to obtain the preliminary bundling state consistency, which refers to the weighted summation of the normalized results of the maximum displacement distance of the bundling point, the difference in the number of bundling circles at the bundling point, and the difference in the area of ​​the bundling gap.

[0080] It should be noted that when determining the weights of each indicator in the initial bundling state fit, it is necessary to set them in combination with engineering experience, etc., and give priority to considering the degree of influence of the indicator on the bundling quality. For example, if the difference in the area of ​​the bundling gap is too large, it is easy to cause the structure to loosen, and the weight should be higher, while taking into account the data fluctuation range and scenario requirements. Specifically, expert scoring method, inverse variance or correlation analysis, hierarchical analysis method, etc. can be used, and the rationality of the weights can be verified through sensitivity analysis and confusion matrix. Preferably, the recommended initial weights can be configured as follows: the weight of the maximum displacement distance of the bundling point is 0.45, which directly affects the structural stability; the weight of the difference in the number of bundling circles is 0.3, which allows a certain degree of fault tolerance but affects the strength; the weight of the difference in the area of ​​the bundling gap is 0.25, which affects the collaborative performance but is not the core. Finally, it is dynamically adjusted through actual data feedback so that the weighted sum of the fit indicators accurately reflects the bundling quality.

[0081] Specifically, the specific statistical process of the spacing arrangement consistency includes: E1, calculating the difference between the measured spacing and the marked spacing for each transverse and longitudinal steel bar, and comparing the number of steel bars whose difference exceeds a preset threshold with the total number of steel bars in the corresponding direction to obtain the transverse and longitudinal spacing deviation ratio.

[0082] E2. Take the maximum horizontal spacing difference and the maximum vertical spacing difference, perform normalization, and combine the horizontal spacing deviation ratio and the vertical spacing deviation ratio to obtain the spacing arrangement consistency through weighted summation.

[0083] Understandably, current conventional approaches to analyzing the conformity of building rebar spacing arrangements often focus on single-dimensional deviations, such as simply calculating the average difference between measured and marked rebar spacing and the overall variance. Alternatively, these approaches simply calculate the percentage of rebars with spacing deviations exceeding a threshold, using a single ratio as a measure of conformity. These methods fail to distinguish between directional differences in transverse and longitudinal rebars and rarely incorporate the impact of extreme deviations, insufficiently capturing the complex deviation characteristics of spacing arrangements. Calculating spacing differences and deviation ratios can intuitively reflect the number and proportion of rebar spacing deviations from design specifications, clarifying the overall distribution of deviations. Selecting the maximum spacing difference and normalizing it highlights the impact of extreme deviations on rebar arrangement, complementing the inadequacy of the deviation ratio, which only reflects the quantitative ratio. Finally, through weighted fusion of multiple parameters, a comprehensive assessment of the performance of rebar spacing in terms of both individual deviations and the percentage of deviations can be achieved. This provides a quantitative basis for accurately assessing whether rebar spacing arrangements meet design requirements and ensures the rationality of structural loads, thereby facilitating construction quality control and structural safety prediction.

[0084] It should be added that in order to determine the weights of the transverse and longitudinal spacing deviation ratios and the normalized maximum spacing difference, historical steel mesh construction inspection data can be collected and the transverse spacing deviation ratio, longitudinal spacing deviation ratio, normalized maximum transverse spacing difference, and normalized maximum longitudinal spacing difference can be used as inputs, and the conformity contribution value obtained through structural mechanics simulation is used as output. The weights are iteratively optimized using a multivariate linear regression or support vector machine regression training model. Among them, the spacing deviation ratio reflects the overall construction compliance, because the longitudinal reinforcement has a more direct impact on the structural bearing capacity, and the normalized maximum spacing difference reflects the local extreme deviation and stress concentration risk. The longitudinal extreme deviation has a more significant impact, so the weight of the longitudinal spacing deviation ratio is set to be greater than the weight of the transverse spacing deviation ratio, which is greater than the weight of the normalized maximum longitudinal spacing difference, which is greater than the weight of the normalized maximum transverse spacing difference, and the sum of the four is 1.

[0085] Specifically, the setting of the risk compensation factor includes: R1, taking the fitting residual, maximum deviation distance, steel bar profile difference and minimum profile rule ratio as each risk assessment item, and performing normalization processing on them respectively to obtain the risk value corresponding to each risk assessment item.

[0086] R2. Based on the preset risk weights corresponding to each risk assessment item, sum the risk weights of each risk assessment item whose risk value exceeds the preset threshold, and divide it by the total risk weights of the risk assessment items to obtain the risk weight ratio.

[0087] R3. The ratio of the number of risk assessment items whose statistical risk values ​​exceed the preset threshold to the total number of risk assessment items is used to obtain the risk coverage ratio, and the product of the risk weight ratio and the risk coverage ratio is used as the risk compensation factor.

[0088] It should be noted that the fitting residual reflects local bending fluctuations and can be normalized using a maximum-minimum method. The maximum deviation distance locates significant bending points, and the threshold can be inversely calculated based on the maximum bending radius allowed by the design. If the threshold is not exceeded, 0 is used as the normalization result. Otherwise, the relative deviation between the maximum deviation distance and the threshold is used as the normalization result. The profile regularity ratio is an indicator of cross-sectional distortion, with an ideal value of 1. The risk value is normalized by taking the inverse. For example, when the profile regularity ratio is 0.8, the risk value is 0.2, reflecting a risk of 20% cross-sectional area loss. The profile difference reflects the cross-sectional uniformity index and is normalized by the standard deviation.

[0089] It should also be noted that by normalizing each risk assessment item, the risk weight ratio is used to highlight the dominant role of high-risk items. For example, the maximum deviation distance is preset with the highest weight, and the compensation factor is directly increased when the limit is exceeded. At the same time, the risk coverage ratio is used to reflect the prevalence of the problem. The product of the two achieves a linear coupling between risk severity and impact range. In terms of parameter processing, normalization eliminates dimensional differences, and the preset weights are consistent with the standard risk classification. For example, the risk weight of main reinforcement bending is higher than that of cross-sectional differences. This avoids the complex parameter adjustment process of nonlinear functions and is more suitable for the needs of rapid on-site assessment. The preset risk weights can be comprehensively designed based on actual construction data and construction experience.

[0090] In another specific embodiment, the specific statistical process of the steel mesh laying comprehensive risk index is as follows: F1. If the bundling state consistency is lower than the preset threshold, set the secondary risk compensation factor to If the spacing arrangement fit is lower than the preset threshold, set the secondary risk compensation factor to , if both are lower than the corresponding preset threshold, set the secondary risk compensation factor to , and thus obtain the secondary risk compensation factor , The value is or or , , .

[0091] F2, the bundling state consistency, spacing arrangement consistency and risk compensation factor are recorded as 、 and .

[0092] F3. Calculate the comprehensive risk index of steel mesh laying , , and They represent the weights corresponding to the bundling state consistency and spacing arrangement consistency respectively.

[0093] It should be added that when the bundling or spacing consistency is lower than the threshold, the bundling risk and spacing risk are activated respectively, reflecting the amplification of the risk by a single defect. If both are not up to standard, the activation , reflecting the synergistic effect of dual risks, and the sum of weights is 1 to ensure the normalization effect.

[0094] Understandably, and The specific value of is designed based on engineering specifications, mechanical principles, data statistics, and scenario requirements. For example, the bundling and spacing risks are graded according to construction standards, and the bundling status is given a higher weight.

[0095] The embodiment of the present invention calculates the comprehensive risk index of steel mesh laying based on the BIM model annotation data and actual detection characteristics, statistics the bundling status consistency and spacing arrangement consistency, introduces a risk compensation factor, and fully integrates the potential risks and hidden risks of steel bar connection quality and spatial position deviation, providing an accurate basis for concrete pouring quality assessment, and improving the deficiencies of existing technology in steel bar status analysis.

[0096] The steel mesh laying risk feedback terminal triggers a corresponding early warning instruction when the comprehensive risk index exceeds a preset threshold, and provides feedback on the comprehensive risk index of the steel mesh laying.

[0097] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they should all fall within the scope of protection of the present invention.

Claims

1. The intelligent analysis system for construction project management based on BIM technology is characterized by: include: Collect an overall image of the corresponding steel mesh before concrete pouring, build a steel mesh BIM model based on the image, and mark the binding points and steel bar laying spacing in the BIM model; Use X-ray detectors to perform multi-angle inspections on the poured concrete components, and generate fused panoramic radiographic images after registration; It consists of a binding point recognition unit, a spacing recognition unit and a deformation recognition unit, each unit sequentially recognizes the current binding point feature, the current spacing feature and the current deformation feature based on the radiographic image; Based on the current bundling point characteristics and current spacing characteristics combined with the steel mesh BIM model, the bundling state consistency and spacing arrangement consistency are calculated in sequence. The risk compensation factor is set based on the current deformation characteristics. The comprehensive risk index of steel mesh laying is calculated by comprehensively considering the bundling state, spacing arrangement consistency and risk compensation factor. When the comprehensive risk index exceeds the preset threshold, the corresponding warning instruction is triggered, and the comprehensive risk index feedback of the steel mesh laying is carried out; The specific statistical process of the bundling state consistency includes: Calculate the Euclidean distance between the current position coordinates of each strapping point and the marked position coordinates, and take the maximum value as the maximum displacement distance of the strapping point; Calculate the absolute difference between the current number of strapping circles and the marked number of circles, as well as the current maximum gap area and the marked maximum strapping gap area at each strapping point, and take the average of these values ​​to obtain the difference in the number of strapping circles and the difference in the strapping gap area at each strapping point; Compare the actual grayscale features of the binding wire area with the set reference grayscale feature interval, and count the proportion of features exceeding the interval to the total number of features, which is recorded as the deviation feature ratio; For each over-limit feature, calculate its relative deviation from the corresponding reference feature value, perform weighted summation to obtain the comprehensive feature relative deviation value, input the deviation feature ratio and the comprehensive feature relative deviation value into the Sigmoid function to output the strapping status evaluation compensation factor; The maximum displacement distance of the strapping point, the difference in the number of strapping circles at the strapping point, and the difference in the area of ​​the strapping gap are normalized, and the normalized indicators are combined through a linear function to obtain the initial strapping state consistency; The final bundling state consistency is obtained by correcting the initial bundling state consistency through the bundling state evaluation compensation factor; The specific statistical process of the spacing arrangement consistency includes: For each transverse and longitudinal reinforcement, the difference between the measured spacing and the marked spacing is calculated. The number of reinforcements whose difference exceeds the preset threshold is compared with the total number of reinforcements in the corresponding direction to obtain the transverse and longitudinal spacing deviation ratio. The maximum horizontal spacing difference and the maximum vertical spacing difference are taken and normalized. The horizontal spacing deviation ratio and the vertical spacing deviation ratio are combined and weighted summed to obtain the spacing arrangement consistency. The specific statistical process of the steel mesh laying comprehensive risk index is as follows: If the degree of consistency of the binding state is lower than the preset threshold, the secondary risk compensation factor is set to If the spacing arrangement fit is lower than the preset threshold, set the secondary risk compensation factor to , if both are lower than the corresponding preset threshold, set the secondary risk compensation factor to , and thus obtain the secondary risk compensation factor , The value is or or , , ; The bundling state consistency, spacing arrangement consistency and risk compensation factor are recorded as 、 and ; Calculate the comprehensive risk index of steel mesh laying , , and They represent the weights corresponding to the bundling state consistency and spacing arrangement consistency respectively.

2. The construction project management intelligent analysis system based on BIM technology according to claim 1, characterized in that: The marking of the binding points and the spacing between steel bars includes: Mark the coordinates of the tying points in an image coordinate system, and identify the number of tying circles and the maximum tying gap area of ​​each tying point through image recognition technology, and then mark the position coordinates, number of tying circles and maximum tying gap area of ​​each tying point in the BIM model; Import the transverse spacing of each transverse reinforcement and the longitudinal spacing of each longitudinal reinforcement in the laid steel mesh, and then mark the transverse and longitudinal spacing in the BIM model.

3. The construction project management intelligent analysis system based on BIM technology according to claim 1 is characterized by: The specific identification process of the current bundling point feature includes: Identify the current number of strapping circles and the center point position of each strapping point from the fused panoramic radiographic image, and convert the center point position into the current position coordinates; Based on the coordinates of the binding points marked in the BIM model, the steel mesh area in the fused image is divided into pixel sub-areas, and image enhancement is performed on each pixel sub-area; After the enhanced pixel sub-regions are subjected to target detection, the bounding box positions of the binding points are obtained, and non-maximum suppression is performed on the detection results to eliminate duplicate boxes; The average value, variance and circularity of the grayscale of the binding wire area are extracted within the binding point boundary box, and the three are used as the actual grayscale features of the binding wire area. The Otsu threshold segmentation method was used to convert each pixel sub-region into a binary image. The morphological opening operation was performed on the binary image to remove noise points. The gap area between the binding wire and the steel bar was recorded as the connected gap area. The pixel area of ​​each connected gap area was calculated, and the maximum value was taken as the maximum gap area. The current position coordinates of each bundling point, the current number of bundling circles, the actual grayscale features of the corresponding bundling wire area and the maximum gap area are used as the current bundling point features.

4. The construction project management intelligent analysis system based on BIM technology according to claim 1 is characterized by: The specific recognition process of the current spacing feature includes: After preprocessing the fused panoramic radiographic image, the steel bar contour edges are extracted using Canny edge detection, and long straight lines are detected using probabilistic Hough transform, thereby extracting the horizontal and vertical steel bar center lines. Extract the Y coordinates of the center lines of all transverse reinforcements, sort them, calculate the differences between adjacent Y coordinates, remove outliers, and take the average of the remaining Y coordinate differences as the measured transverse spacing. Similarly, extract the X coordinates of the center lines of the longitudinal reinforcements to obtain the measured longitudinal spacing. Uses the measured transverse spacing of all transverse reinforcement and the measured longitudinal spacing of all longitudinal reinforcement as the current spacing characteristics.

5. The construction project management intelligent analysis system based on BIM technology according to claim 1 is characterized in that: The specific identification process of the current deformation feature includes: The single steel bar area is located by BIM model and mask extraction is performed, and a skeletonization algorithm is applied to generate a steel bar skeleton with a single pixel width; A quadratic polynomial is fitted to the pixel coordinates of each point on the skeleton to calculate the fitting residuals. The line connecting the two end points of the steel bar is defined as the target axis. The maximum distance and average distance from each point on the steel bar skeleton to the target axis are calculated, and the difference between the two is calculated to obtain the maximum deviation distance. Select each section of the steel skeleton according to a preset interval, extract the outline of the selected section, calculate the minimum circumscribed circle area of ​​each section, and calculate the ratio of the outline area of ​​each section to its minimum circumscribed circle area, and record the ratio as the outline rule ratio; The standard deviation of the minimum circumscribed circle area corresponding to each section is taken as the steel bar profile difference, and the minimum profile regularity ratio is extracted; The fitting residual, maximum deviation distance, steel bar profile difference and minimum profile regularity ratio are used as current deformation features.

6. The construction project management intelligent analysis system based on BIM technology according to claim 5, characterized in that: The specific process of setting the interval distance includes: The number of bending points is extracted based on the pixel coordinates of each point on the skeleton, and the ratio of the number of bending points to the total length of the steel bar is used as the bending density; If the bending density is less than or equal to the preset bending density, the default division interval is used as the preset interval; otherwise, the steel bar is divided into several segments, the number of bending points of each steel bar segment is extracted, and its bending density is calculated; The steel bar segments with a bending density less than or equal to the preset bending density are recorded as normal intercepted segments, the default division interval distance is used as the preset interval distance, and the steel bar segments with a bending density greater than the preset bending density are recorded as encrypted intercepted bending segments; The bending density of the encrypted intercepted bending segment, the preset bending density and the default division interval distance are respectively recorded as 、 and , the preset spacing distance of the encrypted intercepted curved section is obtained by combining the three , , is a natural constant, The preset minimum separation distance.

7. The construction project management intelligent analysis system based on BIM technology according to claim 5 is characterized by: The setting of the risk compensation factor includes: The fitting residual, maximum deviation distance, steel bar profile difference and minimum profile rule ratio are used as risk assessment items, and the corresponding risk values ​​of each risk assessment item are obtained after normalization processing. Based on the preset risk weights corresponding to each risk assessment item, sum the risk weights of each risk assessment item whose risk value exceeds the preset threshold, and divide the sum by the total risk weights of the risk assessment items to obtain the risk weight ratio; The risk coverage ratio is obtained by calculating the ratio of the number of risk assessment items whose risk values ​​exceed the preset threshold to the total number of risk assessment items, and the product of the risk weight ratio and the risk coverage ratio is used as the risk compensation factor.

Images