Intelligent precast beam field transportation and erection state real-time monitoring system based on BIM (Building Information Modeling)

Abstract

The invention relates to the technical field of building construction, and discloses an intelligent precast beam field transportation and erection state real-time monitoring system based on BIM modeling, and the system integrates a data collection module, a BIM model construction module, a state gradient calculation module, a data visualization module, a risk prediction module, an alarm notification module and other multifunctional modules. By collecting multi-dimensional state data in real time and combining the BIM technology, three-dimensional visualization display of the data is achieved, and a user can visually monitor the real-time state of prefabricated beam transportation and erection. According to the system, a state gradient calculation module is particularly introduced to monitor the rapid change of the state, and meanwhile, a risk prediction model is constructed by using a long short-term memory (LSTM) network to accurately predict future states and potential risks. A prediction result visually displays a risk area through a BIM model, and early warning information is automatically issued. According to the method, the limitation of a traditional monitoring method is effectively solved, the comprehensiveness, the real-time performance and the predictability of monitoring are improved, and a powerful technical guarantee is provided for the transporting and erecting process of the precast beam.

Description

Technical Field

[0001] The present invention relates to the technical field of building construction, and particularly to a real-time monitoring system for the transportation and erection status of an intelligent precast beam yard based on BIM modeling. Background Technique

[0002] In the field of building construction, especially during the transportation and erection process of a precast beam yard, ensuring the safety and efficiency of the construction status is of utmost importance. Traditional management methods often rely on manual monitoring and empirical judgment. However, it is difficult to comprehensively and real-time capture the multi-dimensional status data and its change gradients during the construction process, resulting in a lag in the discovery and response to potential risks, which not only affects the construction efficiency but also may pose threats to project quality and personnel safety. With the rapid development of Building Information Modeling (BIM) technology, its application in building construction management is becoming increasingly widespread. BIM technology can realize the three-dimensional visualization modeling of building projects, providing a powerful tool for the simulation, monitoring, and management of the construction process. However, during the transportation and erection process of a precast beam yard, how to combine BIM technology with real-time monitoring technology to effectively capture and visualize multi-dimensional status data (such as displacement, stress, temperature, etc.) and its change gradients, and then realize the timely discovery and prediction of risks, is still an urgent problem to be solved.

[0003] Existing risk prediction methods often rely on simple statistical models and are difficult to fully consider the complex relationships among historical status data, status gradient data, and environmental data, resulting in limited accuracy of prediction results. Therefore, constructing an intelligent risk prediction model that can comprehensively consider multiple data sources is of great significance for improving the safety and efficiency of the transportation and erection process of a precast beam yard. Summary of the Invention

[0004] The purpose of the present invention is to provide a real-time monitoring system for the transportation and erection status of an intelligent precast beam yard based on BIM modeling to solve the problems raised in the above background technique.

[0005] To achieve the above purpose, the present invention provides the following technical solution: A real-time monitoring system for the transportation and erection status of an intelligent precast beam yard based on BIM modeling, the system includes: A data acquisition module for real-time collecting multi-dimensional status data during the transportation and erection of precast beams, including position, speed, acceleration, stress, vibration, and environmental temperature; A BIM model construction module for constructing a three-dimensional building information model according to the actual layout and dimensions of the precast beam yard and transportation and erection; A status gradient calculation module for calculating the change gradients of the collected multi-dimensional status data to monitor the rapid changes in the status; A data visualization module, which is used to map real-time status data and its change gradient into a BIM model to achieve three-dimensional visualization display of the data; A model construction module, which is used to construct a risk prediction model based on the long short-term memory network (LSTM) to predict future status and potential risks; A model input data processing module, which is used to prepare the input data of the LSTM risk prediction model, including historical status data, status gradient data, and environmental data; A model output parsing module, which is used to parse the output of the LSTM risk prediction model to identify potential risks and their levels; A risk three-dimensional display module, which is used to mark the predicted risk areas in the BIM model; An alarm and notification module, which automatically publishes corresponding early warning information according to the risk levels identified by the model output parsing module.

[0006] Preferably, the formula for the state gradient calculation module to calculate the change gradient is: Δy(t) = y(t) - y(t-1); where y(t) is the state data value at time t, y(t-1) is the state data value at time t-1, and Δy(t) represents the change gradient of the state data at time t.

[0007] Preferably, the mapping of the real-time status data and its change gradient into the BIM model to achieve three-dimensional visualization display of the data includes: S101: Receive the real-time status data transmitted by the data acquisition module; S102: Receive the change gradient of the state data transmitted by the state gradient calculation module; S103: Construct a three-dimensional grid model according to the geometric information and location information of the BIM model; the three-dimensional grid model includes geometric representations of precast beams, transportation equipment, erection structures, and the surrounding environment; S104: Map the real-time status data and its change gradient to the corresponding nodes of the three-dimensional grid, and represent the magnitude and change gradient of the data value through color, transparency, and texture; S105: Use a rasterization algorithm to generate a three-dimensional image of the BIM model containing the real-time status data and its change gradient.

[0008] Preferably, the method of representing the magnitude and change gradient of the data value through color and transparency is: For the real-time status data, use a linear mapping algorithm to convert it into a color value, and the color mapping formula is: ; where is the state data value, and are the minimum and maximum values of the status data respectively, is the color value after mapping, and are the minimum and maximum color values in the color mapping table respectively; For the change gradient of the status data, an alpha transparency adjustment algorithm is used to represent its magnitude, and the alpha transparency mapping formula is: ;

[0009] Wherein, is the change gradient of the status data, is the maximum value of the change gradient, is the alpha transparency value after mapping, and are the maximum and minimum values of the alpha transparency respectively.

[0010] Preferably, the method for representing the magnitude and change gradient of data values through textures is as follows: Define the mapping relationship between the status data categories and texture patterns, wherein each status data or change gradient range corresponds to a specific texture pattern; For the received real-time status data, according to its category or change gradient, determine the texture pattern to be used by looking up the mapping relationship table; When mapping the real-time status data and its change gradient to the corresponding nodes of the 3D mesh model, apply the corresponding texture to the nodes according to the determined texture pattern; the application of the texture pattern adopts a texture mapping algorithm, specifically: i: Set the texture coordinate system (u, v), where u and v represent the horizontal and vertical coordinates on the texture image respectively; ii: For each node on the 3D mesh model, calculate its coordinates (u, v) in the texture coordinate system according to its position information and normal direction; iii: Sample the corresponding color value from the texture pattern according to the (u, v) coordinates; iv: Blend the sampled color value with the original color value of the node to obtain the final color value for rendering the node.

[0011] Preferably, the steps for constructing the risk prediction model include: S201: Collect historical status data, status gradient data and environmental data as the training data set; S202: Define the LSTM network structure, including an input layer, a hidden layer and an output layer, where the hidden layer contains at least one LSTM unit; S203: Determine the dimensions and formats of the model input data, including the time step and the number of features; S204: Train the LSTM model using the training dataset, adjust the network weights through the backpropagation algorithm and the time backpropagation algorithm, and minimize the prediction error; S205: After the training is completed, verify the prediction performance of the model, and adjust the model structure or training parameters as needed.

[0012] Preferably, the input data prepared by the model input data processing module includes: A sequence of state data for the past N time steps, where N is a preset time step; The corresponding gradient sequence of state data changes; The current and past M time steps of environmental data, where M is the time range considering the influence of environmental factors.

[0013] Preferably, the output of the risk prediction model includes: The predicted values of the state data for the next K time steps, where K is the prediction time range; The corresponding predicted gradient of state data changes; The level and type of potential risks, determined by a classifier or a threshold determination method.

[0014] Preferably, the steps for the model output parsing module to parse the model output include: S301: Receive the output data of the LSTM risk prediction model; S302: Analyze the predicted state data and its change gradient to identify abnormal changes and trends; S303: Evaluate the specific impact of potential risks on the precast beam transportation and erection process according to the predicted risk level and type; S304: Generate a risk report, including risk description, level, location, and recommended measures.

[0015] Preferably, the steps for the risk three-dimensional display module to mark risks in the BIM model include: S401: Receive the risk report generated by the model output parsing module; S402: Find the corresponding location in the BIM model according to the location information in the risk report; S403: Use eye-catching colors, marks, or animation effects to highlight the risk area in the BIM model; S404: Add risk description and level information near the risk area.

[0016] Compared with the prior art, the beneficial effects of the present invention are: The system collects multi-dimensional state data during the transportation and erection of precast beams in real time through the data acquisition module, including position, speed, acceleration, stress, vibration, and environmental temperature, etc., achieving a full-range monitoring of the state of precast beams. Combining with the three-dimensional building information model established by the BIM model construction module, the system can display the actual state of precast beams in real time, enabling the monitoring personnel to intuitively and comprehensively grasp the dynamic situation during the transportation and erection process, greatly improving the comprehensiveness and real-time nature of monitoring.

[0017] The state gradient calculation module effectively captures the rapid changes in the state by calculating the change gradients of multi-dimensional state data, providing an important basis for timely discovering potential risks. The data visualization module maps the real-time state data and its change gradients into the BIM model, realizing the three-dimensional visualization display of data, enabling the monitoring personnel to more intuitively observe the data change trend, and enhancing the accuracy and predictability of state monitoring.

[0018] The system constructs a risk prediction model based on the long short-term memory network (LSTM), which can fully consider historical state data, state gradient data, and environmental data to accurately predict the future state and potential risks. The model output analysis module can identify potential risks and their levels, and the risk three-dimensional display module marks the predicted risk areas in the BIM model, enabling the monitoring personnel to take measures in advance to effectively avoid or reduce risks, improving the accuracy and timeliness of risk prediction.

[0019] The alarm and notification module automatically issues corresponding early warning information according to the identified risk levels, providing timely and accurate decision-making support for the decision-making layer, enabling the management layer to respond quickly and make effective decisions, optimizing resource allocation, and improving management efficiency. The implementation of this system promotes the development of construction management towards intelligence and informatization, sets a new technical benchmark for the construction industry, and helps to enhance the competitiveness and sustainable development ability of the entire industry. Brief Description of the Drawings

[0020] Figure 1 It is the working principle diagram of the real-time monitoring system for the transportation and erection state of the intelligent precast beam yard based on BIM modeling described in the present invention; Figure 2 It is the flow chart of the present invention from receiving data to generating the three-dimensional image of the BIM model; Figure 3 It is the flow chart for constructing the risk prediction model. Detailed Embodiment

[0021] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0022] Please refer to Figures 1 - 3 , the present invention provides a technical solution: an intelligent precast beam yard transportation and erection status real-time monitoring system based on BIM modeling, and the system includes: Data acquisition module: A high-precision sensor network is adopted, including a GPS locator, an accelerometer, a stress sensor, a vibration sensor, and an environmental temperature sensor, which are respectively responsible for real-time collection of multi-dimensional status data such as the position, speed, acceleration, stress, vibration, and environmental temperature of the precast beam during transportation and erection. The sensor data is transmitted to the central data processing unit by wireless or wired means to ensure the real-time and integrity of the data.

[0023] BIM model construction module: Using BIM software (such as Autodesk Revit, Graphisoft ArchiCAD, etc.), a high-precision three-dimensional building information model is constructed according to the actual layout, dimensions, and construction details of the precast beam yard and transportation and erection. The model contains the structural information, spatial position relationship of the precast beam, and interaction information with other building components, providing a basis for subsequent data mapping and risk display.

[0024] State gradient calculation module: Preprocess the collected multi-dimensional status data, including data cleaning, denoising, and normalization. Calculate the time series change gradient of each status parameter, and use numerical differentiation methods (such as the finite difference method) to calculate the instantaneous change rate of data points to capture the rapid changes in the state.

[0025] Data visualization module: Develop a data mapping algorithm to associate the real-time status data and its change gradient with the corresponding elements in the BIM model. Using graphics rendering technologies (such as WebGL, DirectX, etc.), dynamically display the real-time data in the BIM model to achieve three-dimensional visualization of the data.

[0026] Model construction module: Construct a risk prediction model based on the long short-term memory network (LSTM). Utilize the powerful processing ability of LSTM for time series data to learn the state change patterns during the transportation and erection of the precast beam. The model is trained using historical status data, state gradient data, and environmental data as inputs, and the output is the predicted future status and potential risks.

[0027] Model Input Data Processing Module: Organize historical data to form a training set and a validation set for the training and validation of the LSTM model. The real-time collected data, after preprocessing, serves as the input data for model prediction.

[0028] Model Output Parsing Module: Develop a parsing algorithm to post-process the output of the LSTM model, identify potential risk points and their risk levels. According to the preset risk thresholds, risks are classified into different levels, such as low risk, medium risk, and high risk.

[0029] Risk Three-Dimensional Display Module: Map the parsed risk areas and level information back into the BIM model, and visually display the risk areas by means of color coding or icon annotation, etc. Realize the dynamic update of the risk area, and automatically adjust the display content as the prediction results change.

[0030] Alarm and Notification Module: According to the risk level, automatically trigger corresponding early warning mechanisms, such as sending text messages, emails, or activating on-site alarms. The early warning information includes risk descriptions, locations, levels, and recommended countermeasures to ensure that relevant personnel can respond quickly and take measures to reduce risks.

[0031] The present invention will be further described below in conjunction with Embodiments 1 to 4: Embodiment 1:

[0032] This embodiment details the implementation steps of the state gradient calculation module and the data visualization module, specifically including: State Gradient Calculation Module: Responsible for calculating the change gradient of the collected multi-dimensional state data to monitor the rapid changes in the state. The calculation formula for the change gradient is: Δy(t) = y(t) - y(t-1); where y(t) represents the state data value at time t, which may be position, speed, acceleration, stress, vibration, or environmental temperature, etc.; y(t-1) represents the state data value at time t-1; and Δy(t) represents the change gradient of the state data at time t, reflecting the change amount and change rate of the state data between two adjacent times. By calculating the change gradient, the system can more sensitively capture the rapid changes in the state data, providing an important basis for subsequent risk prediction.

[0033] The data visualization module is responsible for mapping the real-time state data and its change gradient into the BIM model to achieve three-dimensional visualization display of the data. The specific steps are as follows: S101: Receive the real-time state data transmitted by the data acquisition module, including position, speed, acceleration, stress, vibration, and environmental temperature, etc.

[0034] S102: Receive the change gradient of the state data transmitted by the state gradient calculation module.

[0035] S103: Construct a three-dimensional grid model according to the actual layout and dimensions of the precast beam yard and transportation and erection. This model includes geometric representations of precast beams, transportation equipment, erection structures, and the surrounding environment, such as bridge piers, brackets, cranes, transportation vehicles, etc.

[0036] S104: Map the real-time status data and its change gradients to the corresponding nodes of the three-dimensional grid. For example, map the stress data on the precast beam to the corresponding positions of the beam body, and map the speed data of the transportation equipment to the vehicle model. Represent the magnitude and change gradients of the data values through visual attributes such as color, transparency, and texture. For example, different colors can be used to represent different stress ranges, transparency can be used to represent the change gradient of the data, or texture can be used to represent the distribution of the data.

[0037] S105: Adopt a rasterization algorithm to convert the three-dimensional grid model containing real-time status data and its change gradients into a two-dimensional image, that is, the three-dimensional image of the BIM model. This image can be updated in real time to reflect the dynamic changes in the process of transporting and erecting precast beams.

[0038] Embodiment 2: This embodiment further elaborates in detail how to represent real-time status data and its change gradients through color, transparency, and texture, and map them to the three-dimensional grid model. The following are the specific implementation methods.

[0039] ① Color and transparency representation method a. Color mapping: For real-time status data, adopt a linear mapping algorithm to convert it into color values for intuitively displaying the magnitude of the data. The color mapping formula is as follows: ;

[0040] Where, is the status data value, and are respectively the minimum and maximum values of the status data, is the color value after mapping, and are respectively the minimum and maximum color values in the color mapping table; b. Transparency mapping: For the change gradient of the status data, adopt a transparency adjustment algorithm to represent its magnitude in order to highlight the rapidity of data changes. The transparency mapping formula is as follows: ;

[0041] Where, is the change gradient of the status data, is the maximum value of the change gradient, is the transparency value after mapping, and are the maximum and minimum values of transparency respectively.

[0042] ② Texture representation method a. Definition of texture mapping relationship: Define the mapping relationship between the state data category and the texture pattern. According to actual requirements, map each state data or change gradient range to a specific texture pattern. For example, a high-stress area can use a red texture, a low-stress area can use a blue texture, and an area with a large change gradient can use a striped or gradient texture.

[0043] b. Determination of texture pattern: For the received real-time state data, determine the texture pattern to be used by looking up the pre-defined mapping relationship table according to its category or change gradient.

[0044] c. Texture mapping algorithm: When mapping the real-time state data and its change gradient to the corresponding nodes of the three-dimensional grid model, the following texture mapping algorithm is adopted: i. Set the texture coordinate system (u, v), where u and v represent the horizontal and vertical coordinates on the texture image respectively. ii. For each node on the three-dimensional grid model, calculate its coordinates (u, v) in the texture coordinate system according to its position information and normal direction. This usually involves the process of converting the three-dimensional coordinates of the node into two-dimensional texture coordinates, and may need to consider the geometric shape of the model and the size of the texture image. iii. Sample the corresponding color value in the texture pattern according to the calculated (u, v) coordinates. This is usually achieved by looking up the pixel value at the corresponding position in the texture image. iv. Blend the sampled color value with the original color value of the node to obtain the final color value for rendering the node. The blending method can be defined according to actual requirements, such as weighted average, linear interpolation or other algorithms.

[0045] Example 3: This example details the specific steps for constructing a risk prediction model, which aims to predict potential risks during the transportation and erection of precast beams using historical state data, state gradient data, and environmental data. The following is the detailed implementation method: S201. Data collection and preprocessing: Historical state data: Collect historical state data from the monitoring system of precast beam transportation and erection, including but not limited to the position, speed, acceleration, stress, vibration, etc. of the precast beam.

[0046] State gradient data: The change gradient of state data calculated based on historical state data, reflecting the change amount and change rate of state data between adjacent time steps.

[0047] Environmental data: Collect environmental data related to the transportation and erection of precast beams, such as weather conditions (temperature, humidity, wind speed, etc.), geological conditions (soil type, foundation stability, etc.), and other relevant factors at the construction site.

[0048] Organize the above data into a training data set to ensure the time series integrity and consistency of the data.

[0049] S202. Definition of LSTM network structure: Input layer: Receive the preprocessed input data, including the state data sequence, the state data change gradient sequence, and the environmental data.

[0050] Hidden layer: Contain at least one LSTM (Long Short-Term Memory) unit to capture long-term dependencies in time series data. The number of hidden layers and the number of LSTM units in each layer can be adjusted according to the complexity of the problem and the scale of the data.

[0051] Output layer: Design the output layer according to the prediction target. For example, it can predict the state data value at a future time step, the change trend of the state data, or the potential risk level.

[0052] S203. Determine the input data dimension and format: Time step N: Determine the state data sequence of the past N time steps as part of the input. The selection of N should be based on the temporal characteristics of the data and the requirements of the prediction target.

[0053] Number of features: Determine the number of features in each time step, including the dimension of the state data, the dimension of the state data change gradient, and the dimension of the environmental data. Organize the input data into a two-dimensional array form of (N, number of features) to meet the input requirements of the LSTM network.

[0054] S204. Model training: Use the training data set to train the LSTM model. Adjust the network weights through the backpropagation algorithm and the Backpropagation Through Time (BPTT) algorithm to minimize the prediction error. Select an appropriate loss function, such as the Mean Squared Error (MSE) or cross-entropy loss, to measure the difference between the predicted value and the actual value. Adopt an optimization algorithm, such as Adam or SGD, to update the network weights and improve the convergence speed and prediction performance of the model.

[0055] S205. Model validation and adjustment: After training is completed, use the validation dataset to verify the prediction performance of the model, and evaluate the accuracy, stability, and generalization ability of the model. According to the validation results, adjust the model structure (such as adding hidden layers, adjusting the number of LSTM units) or training parameters (such as learning rate, batch size) to further improve the prediction performance of the model. Methods such as cross-validation and grid search can be used to optimize the model parameters to ensure that the model performs stably and excellently on different datasets.

[0056] Organize the processed input data above into a two-dimensional array form according to the input requirements of the LSTM network, and input it into the LSTM model for prediction. Specifically, it includes: a. State data sequence: Extract the state data of the past N time steps to form a state data sequence. For example, if N is set to 10, then for each prediction time point, the state data of the 10 time steps before that time point will be extracted as input.

[0057] b. State data change gradient sequence: The state data change gradient calculated based on the state data sequence reflects the change trend of the state data in the time series.

[0058] c. Environmental data: Extract the current and past M time steps of environmental data, where M is the time range considering the influence of environmental factors. For example, the weather conditions within the past 24 hours can be considered as part of the environmental data.

[0059] Example 4: In this example, through the output and subsequent processing module of the risk prediction model, a comprehensive prediction and intuitive display of potential risks during the transportation and erection of precast beams are achieved. The specific steps are as follows: The output content of the risk prediction model described in the present invention is rich and specific, aiming to provide comprehensive risk prediction information. The following is the detailed implementation method of the model output and its subsequent processing: The model output content includes: Predicted values of state data for the next K time steps: The model predicts the state data values within the next K time steps based on the input historical data and environmental factors. For example, during the transportation and erection of precast beams, key state data such as the stress and displacement of the beam body in the future several time steps can be predicted.

[0060] Corresponding predicted state data change gradients: In addition to directly predicting the state data values, the model also calculates and outputs the change gradients of these state data in the future time steps, that is, the change rate and trend of the state data over time. This helps to identify potential risk points and change trends.

[0061] Levels and Types of Potential Risks: The model analyzes the predicted status data and change gradients through built-in classifiers or threshold determination methods to determine the levels (such as low risk, medium risk, high risk) and types (such as mechanical failures, structural instability, etc.) of potential risks.

[0062] The model output parsing module is responsible for parsing the data output by the model and converting it into information that is easy to understand and apply. The specific steps are as follows: S301: Receive the output data of the LSTM risk prediction model: The module first receives the predicted values of future status data, change gradients, and risk level and type information output by the model.

[0063] S302: Analyze the predicted status data and its change gradients: The module conducts in-depth analysis on the received predicted data to identify abnormal changes or trends. For example, by comparing the current status data with the predicted data, or analyzing the mutation points of the change gradients, potential risk points are discovered.

[0064] S303: Evaluate the specific impacts of potential risks: Based on the predicted risk levels and types, the module further evaluates the specific impacts of these risks on the process of precast beam transportation and erection. This includes the possible consequences of the risks, the scope and degree of the impacts, etc.

[0065] S304: Generate a risk report: Based on the above analysis, the module generates a detailed risk report. The report content includes risk descriptions, levels, locations (such as specific coordinates or components in the BIM model), and recommended measures to be taken. These measures may include adjusting the transportation speed, strengthening structural monitoring, conducting emergency repairs, etc.

[0066] The risk three-dimensional display module uses the BIM model to visually display risk information to users. The specific steps are as follows: S401: Receive the risk report: The module first receives the risk report generated by the model output parsing module to obtain the specific information of the risks.

[0067] S402: Locate the risk positions: Based on the location information in the risk report (such as coordinates or component identifiers in the BIM model), the module finds the corresponding risk positions in the BIM model.

[0068] S403: Highlight the risk areas: To draw the user's attention, the module uses eye-catching colors, markings, or animation effects to highlight the risk areas in the BIM model. For example, high-risk areas can be marked in red and a flashing effect can be added; medium-risk areas can be marked in yellow and a border can be added, etc.

[0069] S404: Add risk description and level information: Near the highlighted risk area, the module adds risk description and level information. Such information may include the type of risk, level, possible impacts, and recommended measures, etc. In this way, users can intuitively understand the specific situation of the risk and take corresponding countermeasures.

[0070] It should be noted that in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device.

[0071] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. The real-time monitoring system for the transportation and erection status of intelligent prefabricated beam yard based on BIM modeling is characterized by: The system comprises: Data acquisition module, used to collect multi-dimensional status data in real time during the transportation and erection of precast beams, including position, speed, acceleration, stress, vibration and ambient temperature; BIM model building module, used to build a three-dimensional building information model based on the actual layout and size of the prefabricated beam yard and transportation and erection; The state gradient calculation module is used to calculate the change gradient of the collected multi-dimensional state data to monitor the rapid change of the state; Data visualization module, used to map real-time status data and its change gradient into the BIM model to achieve three-dimensional visualization of data; Model building module, used to build a risk prediction model based on long short-term memory network (LSTM) to predict future status and potential risks; Model input data processing module, used to prepare input data for LSTM risk prediction model, including historical state data, state gradient data and environmental data; Model output parsing module, used to parse the output of LSTM risk prediction model and identify potential risks and their levels; The risk 3D display module is used to mark the predicted risk areas in the BIM model; The alarm and notification module automatically issues corresponding warning information based on the risk level identified by the model output analysis module.

2. According to the BIM modeling-based real-time monitoring system for the transportation and erection status of an intelligent prefabricated beam yard according to claim 1, it is characterized in that: The formula for calculating the change gradient by the state gradient calculation module is: Δy(t) = y(t) - y(t-1); Among them, y(t) is the state data value at time t, y(t-1) is the state data value at time t-1, and Δy(t) represents the change gradient of the state data at time t.

3. The real-time monitoring system for transportation and erection status of intelligent prefabricated beam yard based on BIM modeling according to claim 2 is characterized in that: The real-time status data and its change gradient are mapped to the BIM model to realize the three-dimensional visualization of the data, including: S101: receiving real-time status data transmitted by a data acquisition module; S102: receiving a state data change gradient transmitted by a state gradient calculation module; S103: constructing a three-dimensional grid model according to the geometric information and position information of the BIM model; the three-dimensional grid model includes geometric representations of prefabricated beams, transportation equipment, erection structures, and surrounding environments; S104: Mapping the real-time status data and its change gradient to corresponding nodes of the three-dimensional grid, and representing the size and change gradient of the data value through color, transparency and texture; S105: Using a rasterization algorithm, a three-dimensional image of the BIM model including real-time status data and its change gradient is generated.

4. The real-time monitoring system for transportation and erection status of intelligent prefabricated beam yard based on BIM modeling according to claim 3 is characterized in that: The method of expressing the size and change gradient of data values ​​by color and transparency is as follows: For real-time status data, a linear mapping algorithm is used to convert it into a color value. The color mapping formula is: ； in, is the state data value, and are the minimum and maximum values ​​of the state data, respectively. is the mapped color value, and They are the minimum and maximum color values ​​in the color map, respectively; For the state data change gradient, the transparency adjustment algorithm is used to represent its size. The transparency mapping formula is: ； in, is the state data change gradient, is the maximum value of the gradient, is the transparency value after mapping, and are the maximum and minimum values ​​of transparency respectively.

5. The real-time monitoring system for transportation and erection status of intelligent prefabricated beam yard based on BIM modeling according to claim 3 is characterized in that: The method of expressing the size and gradient of data values ​​through texture is: Defining a mapping relationship between state data categories and texture patterns, wherein each state data or change gradient range corresponds to a specific texture pattern; For the received real-time status data, according to its category or change gradient, the texture pattern to be used is determined by looking up the mapping relationship table; When real-time state data and its change gradient are mapped to corresponding nodes of the three-dimensional mesh model, corresponding textures are applied to the nodes according to the determined texture patterns; the texture pattern is applied using a texture mapping algorithm, specifically: i: Set the texture coordinate system (u, v), where u and v represent the horizontal and vertical coordinates on the texture image respectively; ii: For each node on the 3D mesh model, calculate its coordinates (u, v) in the texture coordinate system based on its position information and normal direction; iii: According to the (u, v) coordinate, sample the corresponding color value in the texture pattern; iv: Merge the sampled color value with the original color value of the node to get the final color value, which is used to render the node.

6. The real-time monitoring system for transportation and erection status of intelligent prefabricated beam yard based on BIM modeling according to claim 1 is characterized in that: The steps to build a risk prediction model include: S201: Collect historical state data, state gradient data and environment data as training data sets; S202: define an LSTM network structure, including an input layer, a hidden layer and an output layer, wherein the hidden layer includes at least one LSTM unit; S203: Determine the dimension and format of the model input data, including the time step and the number of features; S204: train the LSTM model using the training data set, adjust the network weights through the back propagation algorithm and the time back propagation algorithm, and minimize the prediction error; S205: After the training is completed, verify the prediction performance of the model and adjust the model structure or training parameters as needed.

7. The real-time monitoring system for transportation and erection status of intelligent prefabricated beam yard based on BIM modeling according to claim 6 is characterized in that: The input data prepared by the model input data processing module includes: The state data sequence of the past N time steps, where N is the preset time step; The corresponding state data change gradient sequence; Environmental data for the current and past M time steps, where M is the time range for considering the impact of environmental factors.

8. The real-time monitoring system for transportation and erection status of intelligent prefabricated beam yard based on BIM modeling according to claim 7 is characterized in that: The outputs of the risk prediction model include: The predicted value of the state data for the next K time steps, where K is the predicted time range; The corresponding predicted state data change gradient; The level and type of potential risk are determined by classifiers or threshold judgment methods.

9. The real-time monitoring system for transportation and erection status of intelligent prefabricated beam yard based on BIM modeling according to claim 8 is characterized in that: The step of the model output parsing module parsing the model output comprises: S301: Receive output data of LSTM risk prediction model; S302: Analyze the predicted state data and its change gradient to identify abnormal changes and trends; S303: Evaluate the specific impact of potential risks on the precast beam transportation and erection process based on the predicted risk level and type; S304: Generate a risk report, including risk description, level, location and recommended measures.

10. The real-time monitoring system for transportation and erection status of intelligent prefabricated beam yard based on BIM modeling according to claim 9 is characterized in that: The steps of marking risks in the BIM model by the risk three-dimensional display module include: S401: receiving a risk report generated by a model output parsing module; S402: Find the corresponding location in the BIM model according to the location information in the risk report; S403: Use eye-catching colors, markers or animation effects to highlight risk areas in the BIM model; S404: Add risk description and level information near the risk area.

Images