Intelligent construction site model fusion intelligent management method and system based on BIM
By using local feature matching and comprehensive risk assessment, systematic cumulative deviations in construction are identified and warned of, which solves the problem of inaccurate construction plans, improves construction quality and efficiency, and reduces project delays and costs.
Patent Information
- Application Number
- CN202511847418.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-09
- Publication Date
- 2026-03-17
AI Technical Summary
In building construction, the automatic alignment process fails when key calibration reference points are obscured, causing the system to switch to local feature matching mode. The accumulation of errors leads to inaccurate construction plans, delays in the construction period, and additional cost expenditures. Existing technologies are unable to identify and provide early warnings of systematic cumulative deviations.
The matching confidence is calculated by matching local features, low confidence alignment events are recorded, the trend of systematic cumulative deviation is analyzed, and the comprehensive risk level is calculated by combining the component attributes of the BIM model and the construction schedule, and an alarm is triggered to suspend the downstream construction plan.
It enables early identification and warning of systematic cumulative deviations, avoids construction misalignment, improves construction quality and efficiency, and reduces project delays and cost expenditures.
Smart Images
Figure CN121684631A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of smart construction site management, in particular to a BIM-based smart construction site model fusion intelligent management method and system. BACKGROUND
[0002] In the field of modern construction, intelligent management systems based on building information modeling (BIM) are widely used. One of the core functions is to compare and align point cloud data collected on the construction site with the BIM design model to monitor the construction progress and check the installation accuracy of components in real time. However, in actual construction, due to the complex and variable site environment, such as the blocking of key calibration reference points by temporarily erected equipment or piled materials, the automatic alignment program cannot work according to the preset global calibration process.
[0003] When the automatic alignment program cannot run normally due to the absence of key calibration reference points, to avoid delays in the construction period, the site data management team may be forced to switch to a local feature matching mode and maintain system operation by manually reducing the global coordinate consistency check requirements. In this mode, the system log will continuously record low-priority prompts for "low local matching confidence" in multiple areas. In the local feature matching mode, the small and unavoidable errors in each local matching will gradually accumulate as the matching area expands, resulting in a continuous and gradually increasing systematic spatial deviation overall. Based on these point cloud data with accumulated deviations, the project management system will update the subsequent construction progress and material allocation plan, such as predicting the accurate hoisting position of large prefabricated components. Due to the deviations in the input data itself, these predictions and plans also inherit spatial inaccuracies. When large prefabricated components arrive at the site for hoisting, construction personnel may find that there is a visible spatial misalignment between the components and the foundation structure, making it impossible to smoothly position or safely install the components, thereby forcing a stoppage of work and the initiation of an emergency survey and correction process.
[0004] At this time, due to insufficient recording and reporting of initial human intervention, and the low-priority prompts from the system failing to effectively reveal the systematic risk of error accumulation, the project team has difficulty quickly locating the root cause of the problem. This information gap and cognitive blind spot seriously delays problem identification and solution development, resulting in delays in the construction period and additional costs. SUMMARY
[0005] To solve the problems in the prior art, the present application discloses a BIM-based smart construction site model fusion intelligent management method and system, which aims to solve the problem that in the local feature matching mode, systematic cumulative deviations are difficult to be effectively identified and warned, leading to inaccurate construction plans, delays in the construction period, and additional costs.
[0006] The technical solutions of the present application are as follows: In a first aspect, the present application discloses a BIM-based intelligent construction site model fusion intelligent management method, comprising: Local feature matching is performed on the point cloud data collected from the construction site and the BIM model, the matching confidence of the local feature matching is calculated, and when the matching confidence is lower than a preset threshold, a low-confidence alignment event is recorded according to the matching result of the local feature matching; When the low-confidence alignment event continuously occurs, the spatial deviation data in the continuously occurring low-confidence alignment event is obtained, the point cloud collection path of the point cloud data is combined, and whether there is a systematic cumulative deviation trend that is continuous and gradually increasing is analyzed, when the systematic cumulative deviation trend occurs, the predicted total spatial deviation value at the end of the current local feature matching region is calculated according to the systematic cumulative deviation trend and the coverage range of the point cloud collection path; According to the attribute information of the components in the BIM model and the construction progress plan, the semantic risk weight is comprehensively allocated to different construction regions and components; The technical risk factor is calculated in combination with the matching confidence and the predicted total spatial deviation value, the cross evaluation is performed according to the technical risk factor and the semantic risk weight, the comprehensive risk level is calculated, and when the comprehensive risk level reaches or exceeds a preset fuse threshold, an alarm is triggered and the related downstream construction plan is suspended.
[0007] Further, the low-confidence alignment event at least includes a timestamp of event occurrence, a unique identifier of the current local feature matching region in the BIM model affected, and spatial deviation data.
[0008] On the basis of the above, the step of performing local feature matching on the point cloud data collected from the construction site and the BIM model to calculate the matching confidence of the local feature matching comprises: The iterative closest point algorithm is used to perform local alignment on the point cloud data and the local region in the BIM model, and the root mean square error of the distance between the data points in the iteratively transformed point cloud data and the corresponding data points in the BIM model is calculated as the matching confidence.
[0009] Further, when the low-confidence alignment event continuously occurs, the step of obtaining the spatial deviation data in the continuously occurring low-confidence alignment event, combining the point cloud collection path of the point cloud data, and analyzing whether there is a systematic cumulative deviation trend that is continuous and gradually increasing comprises: The deviation vector of the spatial deviation data in the low-confidence alignment event continuously occurring along the point cloud collection path is analyzed in a preset length sliding window; If the deviation vector is continuously positive or negative in the same direction and the deviation absolute value shows an increasing trend, it is judged that there is a systematic cumulative deviation trend.
[0010] Furthermore, the step of calculating the expected total spatial deviation value at the end of the current local feature matching region based on the systematic cumulative deviation trend and the coverage of the point cloud acquisition path includes: Based on the systematic cumulative deviation trend and the coverage of the point cloud acquisition path, the total deviation of the point cloud data relative to the BIM model at the end of the current local matching area is predicted and calculated as the expected total spatial deviation value.
[0011] Specifically, the component's attribute information should include at least the component type, structural importance identifier, installation accuracy requirements, and whether it is a prefabricated component.
[0012] Based on this, the steps for comprehensively assigning semantic risk weights to different construction areas and components according to the attribute information of components in the BIM model and the construction schedule include: Based on the component attribute information in the BIM model and the construction schedule, the key time nodes of the component in the construction schedule and the dependency information of the component with adjacent components in the installation logic are obtained. Configure preset influence factors for the component's attribute information, key time nodes, and dependency information respectively; The semantic risk weight is obtained by weighting the weights based on preset impact factors.
[0013] More specifically, when the overall risk level reaches or exceeds the preset circuit breaker threshold, the steps to trigger an alarm and automatically suspend the relevant downstream construction plans include: When the overall risk level reaches or exceeds the preset circuit breaker threshold, the alarm response level is selected according to the overall risk level; Based on the alarm response level, an alarm is triggered, and relevant personnel are notified via voice broadcast and emergency notification. An emergency shutdown procedure was initiated, and downstream construction projects were suspended in accordance with the construction schedule, while power supply to the relevant areas was cut off. Based on the nature and scope of risk impact corresponding to the comprehensive risk level, risk management recommendations are generated and then pushed to relevant responsible persons.
[0014] As a technical improvement, risk management recommendations include at least one of the following: conducting on-site retesting, adjusting the component installation plan, and reinforcing the component structure.
[0015] Secondly, this application also discloses a BIM-based smart construction site model fusion intelligent management system for executing the aforementioned BIM-based smart construction site model fusion intelligent management method. The system includes: The event logging module is used to perform local feature matching between point cloud data collected from the construction site and the BIM model, calculate the matching confidence of the local feature matching, and record the low confidence alignment event based on the matching result of the local feature matching when the matching confidence is lower than a preset threshold. The deviation prediction module is used to acquire spatial deviation data in consecutive low-confidence alignment events when they occur. Combined with the point cloud acquisition path of the point cloud data, it analyzes whether there is a continuous and gradually increasing systematic cumulative deviation trend. When a systematic cumulative deviation trend occurs, it calculates the expected total spatial deviation value at the end of the current local feature matching region based on the systematic cumulative deviation trend and the coverage of the point cloud acquisition path. The risk weight allocation module is used to comprehensively allocate semantic risk weights to different construction areas and components based on the attribute information of components in the BIM model and the construction schedule. The risk calculation and control module is used to calculate the technical risk factor by combining the matching confidence level and the expected total spatial deviation value. Based on the technical risk factor and semantic risk weight, it performs cross-evaluation to calculate the comprehensive risk level. When the comprehensive risk level reaches or exceeds the preset circuit breaker threshold, it triggers an alarm and suspends the relevant downstream construction plan.
[0016] This technical solution provides a system for implementing the above method. Through modular design, it ensures the effective execution of the method and the stable operation of the system.
[0017] In summary, this application provides a BIM-based smart construction site model fusion intelligent management method and system. The method involves matching point cloud data with local features of the BIM model and calculating the matching confidence level. When the matching confidence level falls below a preset threshold, a low-confidence alignment event is recorded. Based on this, when consecutive low-confidence alignment events occur, spatial deviation data is acquired, and combined with the point cloud acquisition path, an analysis is performed to determine if a continuous and gradually increasing systematic cumulative deviation trend exists. Once a systematic cumulative deviation trend is identified, the expected total spatial deviation value at the end of the current local feature matching area is calculated based on this trend and the coverage of the point cloud acquisition path. Simultaneously, semantic risk weights are comprehensively assigned to different construction areas and components, incorporating the component attribute information and construction schedule in the BIM model. Finally, a technical risk factor is calculated by combining the matching confidence level and the expected total spatial deviation value. A cross-evaluation is then performed based on the technical risk factor and semantic risk weights to calculate the comprehensive risk level. When the comprehensive risk level reaches or exceeds a preset circuit breaker threshold, an alarm is triggered, and the relevant downstream construction plan is suspended. Attached Figure Description
[0018] Figure 1This is a flowchart illustrating a BIM-based smart construction site model fusion intelligent management method provided in an embodiment of this application.
[0019] Figure 2 This is a schematic diagram of the structure of a BIM-based smart construction site model fusion intelligent management system provided in an embodiment of this application.
[0020] Labeling Explanation: 210 Event Recording Module; 220 Deviation Prediction Module; 230 Risk Weight Allocation Module; 240 Risk Calculation and Control Module. Detailed Implementation
[0021] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.
[0022] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.
[0023] In the field of modern building construction, a typical and profound risk chain exists in the actual application of BIM intelligent management: the local matching mode triggered by the lack of key reference points will lead to systematic spatial deviation through error accumulation, and the failure of the system's early warning mechanism and manual intervention process will eventually lead to serious construction misalignment problems.
[0024] Specifically, when automatic global alignment fails due to occlusion of the reference point (e.g., by temporary equipment or materials), the system degrades to a local feature matching mode. In this mode, the small errors generated by each local match (e.g., unavoidable errors when matching using algorithms such as Iterative Nearest Point (ICP)) gradually accumulate and propagate as the matching area expands, potentially leading to systemic spatial deviations at the millimeter or even centimeter level. However, system logs typically only continuously record low-priority alerts such as "low confidence in local matching," failing to effectively reveal the systemic risk of this accumulated deviation. Furthermore, the initial manual intervention to reduce global coordinate check requirements to maintain operation may not be adequately recorded. This results in subsequent decisions based on biased data (such as predicting the hoisting position of prefabricated components) being inherently inaccurate. When misalignment manifests in the physical world (e.g., components cannot be installed), the root cause is difficult to trace, leading to project delays and cost overruns. This is not merely a technical issue, but a systemic problem stemming from the failure of technical processes, human-computer interaction, and project management collaboration.
[0025] Firstly, please see Figure 1 This application proposes a BIM-based smart construction site model fusion intelligent management method, which includes: S1. Perform local feature matching between the point cloud data collected from the construction site and the BIM model, calculate the matching confidence of the local feature matching, and record the low confidence alignment event based on the matching result of the local feature matching when the matching confidence is lower than the preset threshold. S2. When low confidence alignment events occur consecutively, obtain the spatial deviation data in the consecutive low confidence alignment events, combine it with the point cloud acquisition path of the point cloud data, analyze whether there is a continuous and gradually increasing systematic cumulative deviation trend, and when a systematic cumulative deviation trend occurs, calculate the expected total spatial deviation value at the end of the current local feature matching area based on the systematic cumulative deviation trend and the coverage of the point cloud acquisition path. S3. Based on the attribute information of components in the BIM model and the construction schedule, comprehensively assign semantic risk weights to different construction areas and components. S4. Calculate the technical risk factor by combining the matching confidence level and the expected total spatial deviation value. Perform cross-evaluation based on the technical risk factor and semantic risk weight to calculate the comprehensive risk level. When the comprehensive risk level reaches or exceeds the preset circuit breaker threshold, trigger an alarm and suspend the relevant downstream construction plan.
[0026] The method proposed in this application first performs local feature matching between point cloud data collected from the construction site and the BIM model, and then calculates the matching confidence. Specifically, various local feature matching algorithms can be used. For example, a feature point-based matching method extracts features such as corners and edges from the point cloud data and the BIM model, then matches these features and calculates the distance error of the matching points to evaluate the matching confidence. Another approach is a region-growing matching method, which starts from an initial matching point and gradually expands to the surrounding area to find the optimal matching region, and calculates the matching confidence based on the average distance or overlap of points within the matching region. When the matching confidence is lower than a preset threshold, the system records a low-confidence alignment event based on the matching results. These events can be recorded in a database, containing the time of the event, the identifier of the affected BIM model region, and specific spatial deviation data. For example, when scanning a newly poured concrete slab, if the local matching degree between the point cloud data and the BIM model is low, the system will record this event, including the slab ID, the time of occurrence, and the specific deviation vector.
[0027] When consecutive low-confidence alignment events occur, the system acquires spatial deviation data from these events and, combined with the point cloud data acquisition path, analyzes whether a continuous and gradually increasing systematic cumulative deviation trend exists. For example, statistical analysis of spatial deviation data in consecutive low-confidence alignment events can be performed to observe changes in the direction and magnitude of the deviation vector. If the deviation vector is consistently positive or negative in the same direction, and its absolute value shows a gradually increasing trend, a systematic cumulative deviation can be identified. For instance, when scanning a row of consecutive precast wall panels, if the local matching of the first few wall panels shows a small eastward deviation, and this deviation gradually increases in subsequent wall panels, then a systematic cumulative deviation trend can be identified. When a systematic cumulative deviation trend appears, the system calculates the expected total spatial deviation value at the end of the current local feature matching area based on this trend and the coverage of the point cloud acquisition path. This can be predicted by linear or nonlinear extrapolation of the cumulative deviation trend. For example, if the deviation is found to accumulate in a certain direction at a rate of 1 millimeter per meter, and the length of the current local matching area is 10 meters, then the expected total spatial deviation value will be 10 millimeters.
[0028] Furthermore, this application comprehensively assigns semantic risk weights to different construction areas and components based on the component attribute information and construction schedule in the BIM model. Component attribute information may include component type (e.g., beam, column, slab, wall), structural importance identifier (e.g., main structural component, secondary structural component), installation accuracy requirements (e.g., high accuracy, general accuracy), and whether it is a precast component. The construction schedule provides the construction sequence of components, key time nodes, and dependencies with other components. For example, a main structural precast beam with high installation accuracy requirements and a critical path for subsequent floor slab installation will be assigned a higher semantic risk weight. Conversely, a minor non-structural component with low installation accuracy requirements and minimal impact on the overall construction schedule will be assigned a lower semantic risk weight.
[0029] Finally, this application calculates a technical risk factor by combining the matching confidence level and the expected total spatial deviation value, and then performs a cross-evaluation based on the technical risk factor and semantic risk weights to calculate the comprehensive risk level. The technical risk factor can be calculated by weighting the reciprocal of the matching confidence level (the lower the matching confidence level, the higher the technical risk) and the magnitude of the expected total spatial deviation value. For example, if the matching confidence level is 0.8 and the expected total spatial deviation value is 5 mm, the technical risk factor might be 0.2 * 5 = 1. Then, the technical risk factor is cross-evaluated with the semantic risk weights. For example, if the technical risk factor is high and the semantic risk weight of the component is also high (e.g., a critical structural component), then the comprehensive risk level will be rated as high risk. Conversely, if the technical risk factor is low and the semantic risk weight of the component is also low, then the comprehensive risk level is low risk. When the comprehensive risk level reaches or exceeds the preset circuit breaker threshold, the system will trigger an alarm and suspend the relevant downstream construction plan. For example, if the overall risk level reaches the "red alert" level, the system will automatically send an emergency notification to the project manager and construction supervisor, and issue a warning through on-site broadcasts. At the same time, it will automatically suspend subsequent construction tasks related to the affected area, such as stopping the transportation of prefabricated components to the area or stopping the power supply to related equipment.
[0030] The core innovation of this application lies in the introduction of "systematic cumulative deviation trend" analysis and "semantic risk weight" comprehensive evaluation. By introducing systematic cumulative deviation trend analysis, semantic risk weight allocation, and intelligent circuit breaker mechanism, this application constructs a more comprehensive, intelligent, and proactive smart construction site model fusion management method, significantly improving the ability to identify, warn, and control risks during the construction process. It effectively solves the pain point of difficulty in detecting and timely intervening in error accumulation in existing technologies, providing strong guarantees for the quality, safety, and efficiency of building construction.
[0031] Specifically, in order to more comprehensively record and manage low-confidence alignment events, this application further defines the composition of low-confidence alignment events in the above method.
[0032] Low-confidence alignment events include at least the timestamp of the event, a unique identifier of the current local feature matching area in the affected BIM model, and spatial deviation data.
[0033] The timestamp of an event refers to the exact moment when a low-confidence alignment event is detected and recorded by the system. It is usually recorded in a format accurate to the second or millisecond, such as "YYYY-MM-DD HH:MM:SS.ms". This timestamp is used to accurately track the order and duration of events, providing a time-dimensional basis for subsequent deviation trend analysis.
[0034] The unique identifier for the currently matched local feature area in the affected BIM model refers to the unique code or name of the BIM model area where low-confidence alignment occurs during local feature matching between point cloud data and the BIM model. For example, this identifier could be the ID of a component in the BIM model, the name of a floor area, or the number of a construction section. This unique identifier allows for quick location of the specific affected area in the BIM model, facilitating subsequent inspection, analysis, and correction.
[0035] Spatial deviation data refers to the geometric positional differences between point cloud data and the BIM model during local feature matching. This data can include 3D coordinate deviations (e.g., Δx, Δy, Δz) or rotational deviations (e.g., rotation angles around the X, Y, and Z axes). Spatial deviation data quantifies the degree of inconsistency between point cloud data and the BIM model and is a core basis for judging alignment quality and analyzing deviation trends.
[0036] This application's solution explicitly includes the event timestamp, the unique identifier of the current local feature matching area in the affected BIM model, and spatial deviation data in the record of low-confidence alignment events. This ensures that each low-confidence alignment event carries rich and structured information. This information not only clearly describes the event's time, location, and specific deviation amount, but also provides the necessary data foundation for subsequent systematic cumulative deviation trend analysis. For example, the timestamp allows for chronological tracking of events, the unique identifier defines the analysis scope, and the spatial deviation data is a key input for analyzing changes in the direction and magnitude of deviations. This ensures that the monitoring data of the actual construction site conditions has higher traceability and analytical value.
[0037] In some embodiments described above, this application proposes performing local feature matching between point cloud data collected from the construction site and the BIM model, and calculating the matching confidence level of the local feature matching. However, in practical applications, how to efficiently and accurately perform local alignment between point cloud data and the BIM model, and quantify the reliability of its matching, is crucial to ensuring the accuracy of subsequent deviation analysis and risk assessment. If the matching method is not precise enough or the confidence level calculation method is unreasonable, it may lead to misjudgment of alignment deviation, thereby affecting the decision-making of the entire smart construction site management system.
[0038] In response, this application further proposes a step to perform local feature matching between point cloud data collected from the construction site and the BIM model, and to calculate the matching confidence of the local feature matching. The steps include: using an iterative nearest point algorithm to locally align the point cloud data and the local area in the BIM model, and calculating the root mean square error of the distance between the data points in the iteratively transformed point cloud data and the corresponding data points in the BIM model as the matching confidence.
[0039] Specifically, the Iterative Closest Point (ICP) algorithm is a widely used algorithm for point cloud registration. Its core idea is to find the optimal transformation (rotation and translation) between two point sets through iterative optimization, minimizing their distance. In this application, the ICP algorithm is used to accurately align point cloud data collected from the construction site with corresponding local areas in the BIM model. This algorithm continuously searches for nearest point pairs, calculates the transformation matrix, and applies the transformation until convergence or a preset number of iterations is reached, thereby achieving accurate matching between the point cloud data and the BIM model.
[0040] The matching confidence score can be understood as an indicator measuring the quality of local alignment between point cloud data and the BIM model. In this application, it is specifically obtained by calculating the root mean square error (RMSE) of the distance between data points in the iteratively transformed point cloud data and corresponding data points in the BIM model. The RMSE effectively reflects the overall deviation between the matched point sets; the smaller the value, the higher the alignment accuracy between the point cloud data and the BIM model, and the higher the matching confidence score. Conversely, a larger RMSE indicates a significant alignment deviation and a lower matching confidence score.
[0041] The proposed solution employs an iterative nearest-point algorithm for local alignment, enabling high-precision matching of actually acquired point cloud data with the theoretical BIM model, thus overcoming the accuracy limitations of traditional matching methods. The iterative optimization characteristics of the iterative nearest-point algorithm allow for the precise establishment of the spatial correspondence between point cloud data and the BIM model. Furthermore, by calculating the root mean square error (RMSE) as the matching confidence level, a quantitative and intuitive indicator is provided to evaluate the matching quality. This quantitative evaluation method enables the system to accurately determine the reliability of the current local feature matching, avoiding misjudgments in subsequent deviation analysis and risk assessment due to insufficient matching accuracy or inaccurate confidence level evaluation.
[0042] In some embodiments of this application, the step of acquiring spatial deviation data from consecutively occurring low-confidence alignment events, and analyzing whether there is a continuous and gradually increasing systematic cumulative deviation trend in conjunction with the point cloud acquisition path of the point cloud data, may specifically include the following: By analyzing the deviation vector of spatial deviation data in low-confidence alignment events that occur continuously along the point cloud acquisition path within a sliding window of preset length; If the deviation vector is consistently positive or negative in the same direction and the absolute value of the deviation shows an increasing trend, then it is determined that there is a systematic cumulative deviation trend.
[0043] Specifically, a sliding window can be understood as an analysis interval that moves across a time or spatial sequence. Its preset length is designed to ensure that a sufficient number of consecutive low-confidence alignment events are captured for effective trend analysis, while avoiding the introduction of excessive irrelevant data or delayed responses due to an overly long window. For example, this preset length can be dynamically configured based on the complexity of the actual construction project, the frequency of point cloud acquisition, and the real-time requirements for deviation response. Spatial deviation data is recorded in each low-confidence alignment event and can be represented as a deviation vector in three-dimensional space. The deviation vector refers to the actual direction and magnitude of the deviation between the point cloud data and the BIM model in the local feature matching area. By analyzing these continuously occurring deviation vectors within the sliding window, the evolution pattern of the deviation can be revealed. In practical applications, if the deviation vector is consistently positive or negative in the same direction—for example, multiple consecutive positive deviations in the X-axis direction or multiple consecutive negative deviations in the Z-axis direction—and the absolute value of the deviation shows a gradually increasing trend, then a systematic cumulative deviation trend can be identified. This trend indicates that the deviation is not a random error, but is caused by some persistent factors (such as problems with the calibration of measuring equipment, systematic errors in construction processes, etc.), and its impact is accumulating.
[0044] This application's solution introduces a sliding window of preset length to dynamically analyze spatial deviation data in low-confidence alignment events occurring continuously along the point cloud acquisition path. This method effectively filters out sporadic, random matching errors, focusing instead on identifying persistent and directional deviation patterns. By judging the trends in the direction and absolute value of the deviation vector, early signs of systematic accumulated deviations can be accurately captured. For example, when a point cloud acquisition device experiences slight systematic drift during movement, the deviation in each local match may be small, but through continuous analysis within the sliding window, it can be observed that the deviation vector continuously accumulates in a specific direction and its absolute value gradually increases, thus timely identifying this potential systematic problem.
[0045] Specifically, in the above method, the step of calculating the expected total spatial deviation value at the end of the current local feature matching region based on the systematic cumulative deviation trend and the coverage of the point cloud acquisition path can be further refined.
[0046] According to the above method, the step of calculating the expected total spatial deviation value at the end of the current local feature matching area based on the systematic cumulative deviation trend and the coverage of the point cloud acquisition path includes: predicting and calculating the total deviation of the point cloud data relative to the BIM model at the end of the current local matching area based on the systematic cumulative deviation trend and the coverage of the point cloud acquisition path, and using this as the expected total spatial deviation value.
[0047] This application's solution, through in-depth analysis of identified systematic cumulative deviation trends and combined with the actual coverage of the point cloud acquisition path, enables proactive prediction of potential future deviations. Specifically, when the system detects a continuous and gradually increasing deviation trend, it goes beyond simply identifying the deviation itself. Instead, it utilizes the evolution of this trend and the spatial information of the acquisition path to calculate the final cumulative deviation upon completion of the current construction or matching area. This predictive mechanism allows the management system to anticipate potential risks, providing more accurate data support for subsequent risk assessment and decision-making. For example, through linear regression, exponential smoothing, or other time series analysis methods, the total deviation at the path's endpoint can be predicted based on historical deviation data and the acquisition path length.
[0048] In some embodiments of this application, in order to more accurately allocate semantic risk weights to different construction areas and components, it is necessary to define the attribute information of the components in detail.
[0049] Specifically, the attribute information of the aforementioned components includes at least the component type, structural importance indicator, installation accuracy requirements, and whether it is a precast component. Component type refers to the component's classification in the BIM model, such as beams, columns, slabs, walls, pipes, or equipment, aiming to distinguish the function, material properties, and construction techniques of different components. Structural importance indicator indicates the component's criticality within the overall building structure, such as primary load-bearing, secondary load-bearing, or non-load-bearing, aiming to assess the structural risks that component failure may pose. Installation accuracy requirements refer to the allowable range of geometric deviations during installation; higher accuracy requirements indicate greater construction difficulty and potential deviation risks. Whether it is a precast component distinguishes the component's production method. Precast components are typically prefabricated in a factory and then hoisted and connected on-site; their construction risk management differs significantly from that of components cast or assembled on-site.
[0050] The solution proposed in this application clarifies the attribute information of the aforementioned components, enabling the allocation of semantic risk weights to fully consider the inherent characteristics and construction complexity of the components. For example, a beam with high structural importance, strict installation precision requirements, and is a precast component may face higher risks during construction and should therefore be assigned a higher semantic risk weight. Thus, the allocation of semantic risk weights is no longer based on a single or limited dimension, but rather integrates multiple key attributes of the components, thereby more comprehensively and accurately reflecting the potential risks of the components.
[0051] In some of the embodiments described above in this application, semantic risk weights are comprehensively assigned to different construction areas and components based on the attribute information of components in the BIM model and the construction schedule. However, in its implementation, if the weights are simply assigned based on the static attribute information of the components and the construction schedule, it may not be able to fully capture the dynamic relationships between components and the importance of time nodes during the construction process. This results in the semantic risk weights being assessed in a less refined and comprehensive manner, making it difficult to accurately reflect the complexity of actual construction risks.
[0052] In response, this application further proposes the following steps for comprehensively assigning semantic risk weights to different construction areas and components based on the attribute information of components in the BIM model and the construction schedule: Based on the component attribute information in the BIM model and the construction schedule, the key time nodes of the component in the construction schedule and the dependency information of the component with adjacent components in the installation logic are obtained. Configure preset influence factors for the component's attribute information, key time nodes, and dependency information respectively; The semantic risk weight is obtained by weighting the weights based on preset impact factors.
[0053] Specifically, component attribute information can be understood as data describing the component's own characteristics, such as component type, structural importance, installation accuracy requirements, and whether it is a prefabricated component. This information directly reflects the inherent risk characteristics of the component. Critical time nodes refer to the points in the construction schedule that have a decisive impact on the installation or completion of a component, such as the planned start date, planned completion date, and milestones. Their purpose is to identify time-sensitive risks. Dependency information refers to the logical sequence or interdependent relationships between a component and other components during installation or construction. For example, the installation of one component can only proceed after the completion of another component. Its purpose is to reveal potential bottlenecks and cascading risks in the construction process.
[0054] Specifically, configuring preset influence factors for component attribute information, key time nodes, and dependency information means assigning different weight coefficients to these information based on their importance to the overall risk assessment. For example, the influence factor for the attribute information of components with high structural importance can be set higher; the influence factor for time nodes on the critical path should also be increased accordingly; and the influence factor for the dependency information of components with complex dependencies should also be amplified. These influence factors can be set and adjusted based on engineering experience, expert knowledge, or historical data.
[0055] This application's solution, when allocating semantic risk weights, considers not only the static attribute information of components but also key time nodes in the construction schedule and installation logical dependencies between components. This allows for a deeper understanding and quantification of the potential risks of components throughout the entire construction lifecycle. By configuring preset influence factors for these different dimensions of information and performing weighted calculations, the final semantic risk weights comprehensively reflect the risks arising from the component's own characteristics, time sensitivity, and its correlation with other components. It is precisely this multi-dimensional and dynamic risk consideration that transforms the assessment of semantic risk weights from a simple static allocation into a more accurate capture of complex risk scenarios that may arise during construction.
[0056] In some of the embodiments described above in this application, when the overall risk level reaches or exceeds a preset circuit breaker threshold, an alarm is triggered and the relevant downstream construction plan is suspended. However, in actual smart construction site management, simply triggering an alarm and suspending construction may not be sufficient to comprehensively address complex construction risks, especially in scenarios requiring refined responses, multi-party collaboration, and the provision of specific handling suggestions. The response mechanism of the basic solution may lack sufficient flexibility and depth, potentially leading to untimely or incomplete risk handling, or even triggering secondary risks.
[0057] In response, this application further proposes the following steps for triggering an alarm and automatically suspending relevant downstream construction plans when the aforementioned comprehensive risk level reaches or exceeds a preset circuit breaker threshold: When the overall risk level reaches or exceeds the preset circuit breaker threshold, the alarm response level is selected according to the overall risk level; Based on the alarm response level, an alarm is triggered, and relevant personnel are notified via voice broadcast and emergency notification. An emergency shutdown procedure was initiated, and downstream construction projects were suspended in accordance with the construction schedule, while power supply to the relevant areas was cut off. Based on the nature and scope of risk impact corresponding to the comprehensive risk level, risk management recommendations are generated and then pushed to relevant responsible persons.
[0058] Specifically, when the overall risk level reaches or exceeds the preset circuit breaker threshold, the system will first intelligently select an appropriate alarm response level based on the severity and urgency of the overall risk level. For example, alarm response levels can be divided into "general alarm," "critical alarm," and "emergency shutdown alarm," each corresponding to different response intensities and scopes. The alarm response level can be understood as a series of pre-set action plans for different levels of risk, aiming to ensure the accuracy and effectiveness of risk response.
[0059] Furthermore, once the alarm response level is determined, the system will trigger the corresponding alarm mechanism based on that level. This includes not only traditional audible and visual alarms, but also emergency notifications broadcast throughout the entire construction site via voice broadcasting, and rapid dissemination of alarm information to all relevant personnel, such as project managers, construction team leaders, safety officers, and related technical personnel, through SMS, app push notifications, and other emergency notification methods. The aim is to ensure the timeliness and coverage of information transmission, avoiding delays in risk response due to information lag.
[0060] Based on this, in order to effectively control the spread of risk and avoid further losses, the system will initiate an emergency shutdown procedure. Specifically, this procedure will intelligently identify and suspend downstream construction plans directly related to or affected by the risk area based on the current construction schedule. Simultaneously, to prevent electrical accidents or further equipment damage, it will also cut off the power supply to the relevant areas. The aim is to immediately isolate the source of risk, prevent the accident from escalating, and buy time for subsequent risk assessment and handling.
[0061] Furthermore, to provide targeted solutions, the system automatically generates detailed risk mitigation recommendations based on the nature of the risk corresponding to the comprehensive risk level (e.g., structural deviations, material defects, schedule delays, etc.) and the scope of the risk's impact (e.g., individual components, local areas, the entire project, etc.). These recommendations may include at least one of the following: conducting on-site re-measurement, adjusting component installation plans, and reinforcing the component structure. These risk mitigation recommendations are then promptly pushed to the relevant responsible parties so that they can take swift action. The aim is to provide professional guidance for the effective resolution of risks, improving the efficiency and accuracy of risk mitigation.
[0062] This application's solution effectively addresses the shortcomings of basic solutions in terms of the depth and breadth of risk response by introducing a multi-layered alarm response mechanism, comprehensive notification methods, decisive work stoppage measures, and intelligent handling suggestions. The ability to select a matching response level based on risk severity ensures rational resource allocation, avoiding over- or under-response. Simultaneously, voice broadcasts and emergency notifications ensure that all key personnel are informed of the risk immediately, enabling rapid and coordinated responses. Initiating emergency work stoppage procedures and cutting off power physically isolates the risk, minimizing potential harm. More importantly, by generating and pushing specific risk handling suggestions, this solution elevates risk management from a reactive to a proactive approach, providing decision-makers with a scientific basis for quickly and accurately resolving problems.
[0063] Specifically, when the overall risk level reaches or exceeds the preset circuit breaker threshold, the generated risk management recommendations aim to provide relevant personnel with clear action guidelines to effectively address and resolve the identified risks.
[0064] Risk management recommendations include at least one of the following: conducting on-site re-measurement, adjusting the component installation plan, and strengthening the component structure. On-site re-measurement refers to conducting actual measurements and inspections of the construction area or component where deviations have occurred to verify the actual situation, scope, and nature of the deviations and to obtain the latest accurate data. Its purpose is to provide a reliable basis for subsequent risk assessment and management. Adjusting the component installation plan involves modifying and optimizing the original component installation plan, method, or sequence based on the on-site re-measurement results and identified deviations. For example, the installation sequence can be adjusted, connection methods changed, or new installation tools and techniques used to correct deviations and ensure correct component installation. Its purpose is to eliminate or mitigate the negative impact of deviations through plan adjustments. Strengthening the component structure refers to taking engineering measures to reinforce the component when its structural integrity or load-bearing capacity is affected by deviations. For example, supports can be added, connection points reinforced, or other structural reinforcement methods used to restore or improve the structural safety of the component. Its purpose is to ensure construction quality and structural safety, and to avoid potential structural risks.
[0065] This application's solution provides specific and actionable risk management recommendations, enabling relevant personnel to quickly obtain targeted solutions once the overall risk level reaches a preset trigger threshold and an alarm is activated. These recommendations directly address the root causes that may lead to systemic cumulative deviations or semantic risks. Examples include confirming the actual extent of deviations through on-site retesting, correcting errors during construction by adjusting component installation plans, and compensating for existing structural defects through component reinforcement. Therefore, this solution not only identifies problems promptly but also provides effective means to resolve them, thus forming a closed-loop risk management process.
[0066] The aforementioned BIM-based smart construction site model integration and intelligent management method provides an effective approach for integrating and managing on-site construction data with BIM models and conducting risk assessments. However, in real-world smart construction site environments, the effective implementation of this method often faces challenges. For example, relying on manual labor or fragmented tools to perform each step may lead to low data processing efficiency, insufficient real-time performance, increased human error rates, and difficulty in automating and rapidly responding to construction risks. These limitations may prevent this method from achieving its maximum effectiveness in complex and ever-changing construction sites.
[0067] Secondly, see Figure 2 This application proposes a BIM-based smart construction site model fusion intelligent management system for executing the aforementioned BIM-based smart construction site model fusion intelligent management method. Through modular design, the system functionally divides the key steps of the method, thereby achieving automated and integrated management. Specifically, the system includes: The event logging module 210 is used to perform local feature matching between point cloud data collected from the construction site and the BIM model, calculate the matching confidence of the local feature matching, and record the low confidence alignment event based on the matching result of the local feature matching when the matching confidence is lower than a preset threshold. The deviation prediction module 220 is used to acquire spatial deviation data in consecutive low-confidence alignment events when low-confidence alignment events occur consecutively, and analyze whether there is a continuous and gradually increasing systematic cumulative deviation trend in combination with the point cloud acquisition path of the point cloud data. When a systematic cumulative deviation trend occurs, the module calculates the expected total spatial deviation value at the end of the current local feature matching region based on the systematic cumulative deviation trend and the coverage of the point cloud acquisition path. The risk weight allocation module 230 is used to comprehensively allocate semantic risk weights to different construction areas and components based on the attribute information of components in the BIM model and the construction schedule. The risk calculation and control module 240 is used to calculate the technical risk factor by combining the matching confidence level and the expected total spatial deviation value, perform cross-evaluation based on the technical risk factor and semantic risk weight, calculate the comprehensive risk level, and trigger an alarm and suspend the relevant downstream construction plan when the comprehensive risk level reaches or exceeds the preset circuit breaker threshold.
[0068] The proposed solution decomposes the aforementioned BIM-based smart construction site model and intelligent management method into multiple functional modules, which work collaboratively to achieve automated and intelligent management of the construction process. The event recording module 210 serves as the data input portal, ensuring real-time comparison and preliminary anomaly detection between the site point cloud data and the BIM model. Based on this, the deviation prediction module 220 transforms isolated deviation information into predictive cumulative deviation trends through trend analysis of continuous abnormal events, thus shifting from passive problem detection to proactive early warning. The risk weight allocation module 230 integrates the rich semantic information of the BIM model and key nodes of the construction plan into the risk assessment system, ensuring that risk assessment is no longer limited to geometric deviations but comprehensively considers the importance of components, installation difficulty, and impact on the overall schedule. Finally, the risk calculation and control module 240 integrates technical-level deviation risks with management-level semantic risks to form a comprehensive risk level. Based on a preset circuit breaker mechanism, it can automatically trigger alarms and take control measures such as work stoppages, effectively avoiding the lag and subjectivity of manual judgment and ensuring the safety, accuracy, and efficiency of the construction process.
[0069] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.
Claims
1. A BIM-based intelligent construction site model fusion intelligent management method, characterized in that, The method comprises the following steps: Local feature matching is performed between point cloud data collected from a construction site and a BIM model, a matching confidence of the local feature matching is calculated, when the matching confidence is lower than a preset threshold, a low-confidence alignment event is recorded according to a matching result of the local feature matching; When a low-confidence alignment event occurs continuously, spatial deviation data in the continuously occurring low-confidence alignment events is obtained, whether there is a systematic cumulative deviation trend that is continuous and gradually increasing is analyzed in combination with a point cloud collection path of the point cloud data, when the systematic cumulative deviation trend occurs, a predicted total spatial deviation value at an end of a current local feature matching region is calculated according to the systematic cumulative deviation trend and a coverage range of the point cloud collection path; According to the attribute information of the components in the BIM model and the construction progress plan, semantic risk weights are comprehensively assigned to different construction regions and components; A technical risk factor is calculated in combination with the matching confidence and the predicted total spatial deviation value, cross-evaluation is performed according to the technical risk factor and the semantic risk weights, a comprehensive risk level is calculated, and when the comprehensive risk level reaches or exceeds a preset fuse threshold, an alarm is triggered and related downstream construction plans are suspended. 2.The BIM-based intelligent construction site model fusion intelligent management method according to claim 1, characterized in that, The low-confidence alignment event at least includes a timestamp of event occurrence, a unique identifier of the current local feature matching region in the BIM model affected, and the spatial deviation data. 3.The BIM-based intelligent construction site model fusion intelligent management method of claim 1, wherein, The step of performing local feature matching between point cloud data collected from a construction site and a BIM model and calculating a matching confidence of the local feature matching comprises the following steps: An iterative closest point algorithm is used to perform local alignment between the point cloud data and a local region in the BIM model, and a root mean square error of distances between data points in the point cloud data after iterative transformation and corresponding data points in the BIM model is calculated as the matching confidence.
4. The BIM-based intelligent construction site model fusion intelligent management method according to claim 1, characterized in that, The step of obtaining spatial deviation data in continuously occurring low-confidence alignment events when a low-confidence alignment event occurs continuously, and analyzing whether there is a systematic cumulative deviation trend that is continuous and gradually increasing in combination with a point cloud collection path of the point cloud data comprises the following steps: A deviation vector of the spatial deviation data in the low-confidence alignment events that occur continuously along the point cloud collection path is analyzed within a preset length of a sliding window; If the deviation vector is continuously positive or negative in the same direction and the deviation absolute value shows an increasing trend, it is determined that the systematic cumulative deviation trend exists.
5. The BIM-based intelligent construction site model fusion intelligent management method according to claim 1, characterized in that, The step of calculating a predicted total spatial deviation value at an end of a current local feature matching region according to the systematic cumulative deviation trend and a coverage range of the point cloud collection path comprises the following steps: According to the systematic cumulative deviation trend and the coverage range of the point cloud collection path, a total deviation amount of the point cloud data relative to the BIM model at the end of the current local matching region is predicted and calculated as the predicted total spatial deviation value.
6. The BIM-based intelligent construction site model fusion intelligent management method according to claim 1, characterized in that, The attribute information of the components at least includes component type, structural importance identifier, installation precision requirement, and whether it is a prefabricated component.
7. The BIM-based intelligent construction site model fusion intelligent management method according to claim 6, characterized in that, The step of comprehensively assigning semantic risk weights to different construction regions and components according to attribute information of the components in the BIM model and a construction progress plan comprises: According to the attribute information of the components in the BIM model and the construction progress plan, the key time nodes of the components in the construction progress plan and the dependency relationship information of the components in the installation logic with adjacent components are obtained in association; The attribute information of the components, the key time nodes and the dependency relationship information are respectively configured with preset influence factors; Based on the preset influence factors, the semantic risk weights are obtained by weighted calculation. 8.The BIM-based intelligent construction site model fusion intelligent management method of claim 1, wherein, The step of triggering an alarm and automatically suspending related downstream construction plans when the comprehensive risk level reaches or exceeds a preset fuse threshold value comprises: When the comprehensive risk level reaches or exceeds a preset fuse threshold value, an alarm response level is selected according to the comprehensive risk level; According to the alarm response level, an alarm is triggered, and relevant personnel are notified through voice broadcast and emergency notification; An emergency shutdown program is started, and related downstream construction plans are suspended according to the construction progress plan, and the power supply of the related area is cut off; According to the risk nature and risk influence range corresponding to the comprehensive risk level, a risk disposal suggestion is generated, and the risk disposal suggestion is pushed to the relevant person in charge. 9.The BIM-based intelligent construction site model fusion intelligent management method of claim 8, wherein, The risk disposal suggestion comprises at least one of on-site re-measurement, adjustment of component installation scheme and component structure reinforcement. 10.A BIM-based intelligent construction site model fusion intelligent management system, configured to perform the BIM-based intelligent construction site model fusion intelligent management method according to any one of claims 1 to 9. The system comprises: An event recording module is configured to perform local feature matching between point cloud data collected from a construction site and a BIM model, calculate a matching confidence of the local feature matching, and record a low-confidence alignment event according to a matching result of the local feature matching when the matching confidence is lower than a preset threshold. A deviation prediction module is configured to obtain spatial deviation data in continuously occurring low-confidence alignment events when the low-confidence alignment events continuously occur, analyze whether there is a systematic cumulative deviation trend that is continuous and gradually increasing in combination with a point cloud collection path of the point cloud data, and calculate a predicted total spatial deviation value at an end of a current local feature matching area according to the systematic cumulative deviation trend and a coverage range of the point cloud collection path when the systematic cumulative deviation trend occurs. A risk weight allocation module is configured to comprehensively assign semantic risk weights to different construction regions and components according to attribute information of the components in the BIM model and a construction progress plan. A risk calculation and control module is configured to calculate a technical risk factor in combination with the matching confidence and the predicted total spatial deviation value, cross-evaluate the technical risk factor and the semantic risk weights, calculate a comprehensive risk level, and trigger an alarm and suspend related downstream construction plans when the comprehensive risk level reaches or exceeds a preset fuse threshold value.