An intelligent detection and deviation correction system for steel structure construction

Abstract

This invention relates to the field of solid deformation measurement technology, and in particular provides an intelligent detection and deviation correction system for steel structure construction. It solves the problem that existing technologies have limited means of measuring solid deformation, often only able to detect deviations through static measurement or post-construction analysis, and cannot achieve real-time monitoring and dynamic correction during construction. The system includes a virtual model module integrating multiple sensors to collect data from steel structure components in real time; a steel deviation prediction module constructing a steel deviation prediction model and training it based on the analysis results of a construction physical model; inputting the construction physical model into the steel deviation prediction model to predict deviation trends; a correction suggestion generation module generating correction suggestions based on the predicted deviation trends; and triggering automatic adjustment commands based on the correction suggestions. This invention achieves intelligent monitoring and deviation correction during steel structure construction through the organic combination of virtual models, deviation prediction, and correction suggestions.

Description

Technical Field

[0001] This invention relates to the field of deformation technology for measuring solids, and particularly to an intelligent detection and deviation correction system for steel structure construction. Background Technology

[0002] Steel structure construction is widely used in modern architecture. Due to its high strength, light weight, and good seismic performance, it has become the preferred structural form for large buildings, bridges, and industrial plants. However, during steel structure construction, problems such as geometric dimensional deviations, weight deviations, and stress deformation can easily occur due to the complex dimensions of components, variable construction environments, and human operational errors. In particular, the measurement and control of solid deformation (such as elastic deformation, plastic deformation, or thermal deformation caused by temperature) is one of the difficulties in construction quality, as these deformations directly affect the installation accuracy and overall stability of the structure. Traditional steel structure construction inspection mainly relies on manual measurement and experience-based judgment, which suffers from low inspection efficiency, insufficient accuracy, and poor real-time performance. In addition, existing technologies have limited means of measuring solid deformation, often only able to detect deviations through static measurement or post-construction analysis, unable to achieve real-time monitoring and dynamic correction during construction. These problems not only increase construction costs and time but may also lead to structural safety hazards. Summary of the Invention

[0003] To achieve the above objectives, the present invention adopts the following technical solution:

[0004] In one aspect, the present invention provides an intelligent detection and deviation correction system for steel structure construction, comprising:

[0005] The virtual model module integrates multiple sensors to collect data such as the geometric dimensions, weight deviations, and stress deformation of steel structure components in real time; it constructs a construction physical model based on the physical construction scenario, compares the point cloud data with the construction physical model, and if the analysis result is greater than the preset value, it is output to the steel deviation prediction module.

[0006] The steel deviation prediction module is used to build a steel deviation prediction model based on historical and real-time steel structure data, train the steel deviation prediction model according to the analysis results of the construction physical model, and input the construction physical model into the steel deviation prediction model to predict the deviation trend.

[0007] The correction suggestion generation module generates correction suggestions based on the predicted deviation trend; triggers automatic adjustment commands based on the correction suggestions; and issues an early warning and locates the deviation problem area if the deviation exceeds a preset threshold.

[0008] In one optional implementation, the virtual model module includes:

[0009] The coordinate alignment submodule is used to align the measured point cloud data with the construction physical model in the coordinate system, select the position of the common steel structure node between the two, transform the measured point cloud from the scanning coordinate system to the design coordinate system of the construction physical model, and perform spatial positioning by generating point cloud lines.

[0010] An attribute comparison submodule is constructed to compare the geometric differences between the point cloud and the construction virtual model, and to calculate the deviation value between the actual cloud points and the construction physical model point by point; the point cloud is semantically segmented to identify basic components, and the corresponding construction attributes in the construction physical model are compared in an object-oriented manner.

[0011] The dataset generation submodule is used to construct key elements for identification, fit point cloud data to the elements of the construction physical model, and compare them with the design values; it generates a virtual point cloud dataset from the parameters in the construction physical model, and divides the virtual point cloud dataset into training set, validation set, and test set.

[0012] In one alternative implementation, an attribute comparison submodule is constructed, including:

[0013] The point cloud segmentation unit is used to extract hierarchical features from the point cloud and learn geometric features point by point; adaptively construct a local neighborhood graph and enhance the feature representation of steel beam nodes through graph attention mechanism; and perform planar segmentation of the point cloud to separate the building plan view from the component point cloud.

[0014] Independent cluster units are used to divide the point cloud into independent clusters, each cluster corresponding to a basic structure, and to refine the segmentation by combining silver information; for multi-source data, the 2D semantic segmentation results are mapped to the 3D point cloud to supplement semantic information;

[0015] The deviation calculation unit is used to compare the segmented construction point cloud with the construction attributes in the construction physical model object by object through feature matching to calculate the geometric deviation; if abnormal segmentation occurs, a feedback optimization mechanism is triggered to adjust the clustering threshold.

[0016] In one optional implementation, the steel deviation prediction module includes:

[0017] The variable distribution difference submodule is used to build a steel deviation prediction model based on historical steel structure data and real-time steel structure data. It merges the historical steel structure data and real-time steel structure data to obtain a dataset, and divides it into a training set and a test set. It introduces a control dataset and compares the variable distribution deviation between the original training set and the control dataset.

[0018] The optimal model submodule is used to train multiple base models using historical steel structure data, generate new training and test sets; train a meta-model using the new dataset generated from the base models, optimize the weights through backpropagation; and determine the optimal steel deviation prediction model through cross-validation.

[0019] The prediction deviation submodule is used to adjust the weights and retrain the steel deviation prediction model if the control dataset test shows a deviation in variable distribution; the construction physical model is input into the steel deviation prediction model to predict the deviation trend.

[0020] In one optional implementation, the variable distribution difference submodule includes:

[0021] The deviation calculation unit is used to calculate the distribution difference of each variable in the two datasets and generate a deviation probability value; by comparing the deviation probability values, it reflects the degree of deviation of the distribution of the variable in the control dataset from that in the original dataset; and checks whether the frequency of each category in the control dataset is consistent with that in the original dataset.

[0022] The target variable adjustment unit is used to readjust the target variable when the distribution of the target variable in the control dataset differs from the original training set by a greater than a preset standard value.

[0023] The offset determination unit is used to mark target variables with a deviation probability greater than a preset value. If the control dataset with a probability of 1 completely confirms the value of the variable, the unit will determine the degree of offset by visualizing the overlap of the distribution of the target variable in the original dataset and the control dataset. An automated offset report will be generated.

[0024] In one optional implementation, the optimal model submodule includes:

[0025] The prediction result generation unit is used to organize historical steel structure data according to features to form an initial training set; based on the historical steel structure data, multiple base models are generated, and multiple independent base models are trained using the initial training set to generate corresponding prediction results. The prediction results are the base models' capture of the trend or pattern of steel deviation in the historical steel structure data.

[0026] The feature combination processing unit is used to combine the prediction results of each base model with the features of the initial training set to form an expanded feature space; if each base model generates a prediction value, the number of features in the initial training set increases to be equal to the number of base models.

[0027] The standardization unit is used to apply the prediction results of the base model to the test set to generate a new test set; it also performs standardization on the newly generated training and test sets to eliminate the dimensional differences between different features.

[0028] In one optional implementation, the prediction result generation unit includes:

[0029] The parameter combination design subunit is used to design multiple unique parameter combinations based on the multi-dimensional features of the initial training set, and to design a base model that captures dynamic changing trends to identify trends or patterns in steel deviations.

[0030] The data deconstruction processing subunit is used to deconstruct historical steel structure data from multiple perspectives according to different representations of short-term deviation, long-term deviation, local deviation, and global deviation of steel deviation, and obtain the identification results corresponding to short-term deviation, long-term deviation, local deviation, and global deviation; the identification results are integrated into a comprehensive deviation prediction value using a weighted average.

[0031] The indicator adds a sub-unit to add a confidence index to the deviation prediction value, and uses Monte Carlo simulation to evaluate the reliability and uncertainty of the prediction results.

[0032] In one optional implementation, the prediction deviation submodule includes:

[0033] The adjustment mechanism trigger unit is used to calculate the bias probability, class bias, and target variable bias by comparing the variable distribution of the control dataset with that of the original training set. If a distribution bias greater than a preset value is detected, the weight adjustment mechanism is triggered. After the base model generates the secondary dataset, the meta-model optimizes the value selection process through backpropagation.

[0034] The prediction value generation unit is used to retrain the steel deviation prediction model based on the updated weights; the physical model parameters of the current construction are input into the adjusted steel deviation prediction model to generate the predicted value of future steel deviation.

[0035] The deviation judgment unit restores the standardized predicted values ​​to the actual physical dimensions; restores the long-term variation pattern of the data by fitting the trend term; calculates the absolute and relative deviations between the predicted values ​​and the actual measured values ​​to identify outliers; and issues an early warning if the deviation exceeds the threshold.

[0036] In one optional implementation, the correction suggestion generation module includes:

[0037] The first trend analysis module is used to analyze the trend of prediction deviation, to determine whether the prediction deviation shows a short-term local expansion or a long-term global accumulation; for short-term local expansion, a local correction plan is proposed; for long-term global accumulation, a comprehensive adjustment is made.

[0038] The second trend analysis module is used to obtain geometric dimension deviation, stress deformation deviation, or weight deviation from the predicted deviation trend.

[0039] The instruction conversion module is used to convert geometric deviations, stress deformation deviations, or weight deviations into executable instructions, which are linked with the construction equipment to automatically adjust welding parameters, displace component positions, or apply auxiliary supports based on correction suggestions; and to inspect the weld seams using ultrasonic waves.

[0040] In one alternative implementation, for geometric deviations in length, width, or height, adjustment schemes are generated using parametric design tools, such as suggesting partial cutting and welding or replacement of the entire component.

[0041] Optimize the support structure or adjust the distribution of external loads to address stress-deformation deviations; conduct mechanical property tests on bolts, nuts, and washers in the support structure.

[0042] Weight deviation provides solutions for counterweight adjustment or optimized material distribution.

[0043] The virtual model module of this invention integrates multiple sensors to collect key data such as geometric dimensions, weight deviations, and stress deformation of steel structure components in real time, and constructs a construction physical model in conjunction with the physical construction scenario. By comparing and analyzing point cloud data with the construction physical model, deviations during construction can be quickly identified. If the deviation exceeds a preset value, the result is output to the steel deviation prediction module. The steel deviation prediction module constructs a steel deviation prediction model based on historical steel structure data and real-time collected steel structure data, and trains the model using the analysis results of the construction physical model. By inputting the construction physical model into the prediction model, it can effectively predict the deviation trend of steel structure components, providing early warnings for potential problems during construction, thereby reducing the impact of deviations on construction quality. The correction suggestion generation module automatically generates targeted correction suggestions based on the predicted deviation trend and triggers corresponding automatic adjustment commands. If the deviation exceeds a preset threshold, an early warning signal is issued, and the deviation problem area is accurately located, facilitating timely corrective measures by construction personnel to ensure that construction quality meets requirements. Attached Figure Description

[0044] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0045] Figure 1 This is a block diagram of the intelligent detection and deviation correction system for steel structure construction provided in Embodiment 1 of the present invention;

[0046] Figure 2 This is a block diagram of the virtual model module provided in Embodiment 2 of the present invention;

[0047] Figure 3 This is a block diagram of the steel deviation prediction module provided in Embodiment 3 of the present invention;

[0048] Figure 4 A module block diagram is generated for the correction suggestions provided in Embodiment 4 of the present invention;

[0049] Figure 5 A block diagram of the electronic device provided by the present invention;

[0050] Figure 6 A block diagram of a computer-readable storage medium provided for this invention. Detailed Implementation

[0051] The technical solutions of the present invention will now be described with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0052] Hereinafter, the terms "first," "second," etc., are used for descriptive convenience only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0053] In this invention, unless otherwise explicitly specified and limited, the term "connection" should be interpreted broadly. For example, "connection" can be a fixed mechanical connection, a detachable mechanical connection, or an integral part; or, "connection" can be a direct connection or an indirect connection through an intermediate medium. Furthermore, unless otherwise explicitly specified and limited, the term "coupling" should be interpreted broadly. For example, "coupling" can be a direct electrical connection, such as physical contact and electrical conduction between two components; it can also be understood as an electrical connection between different components in a circuit structure through physical lines capable of transmitting electrical signals, such as copper foil or wires on a printed circuit board (PCB), to transmit electrical signals; or, "coupling" can be an indirect electrical connection between two components through an intermediate medium; or, "coupling" can be an electrical connection between two components in a non-contact manner, such as an electrical connection between two components using capacitive coupling to transmit electrical signals.

[0054] In this embodiment of the invention, directional terms such as "up," "down," "left," and "right" may be defined relative to the orientation of the components shown in the accompanying drawings. It should be understood that these directional terms can be relative concepts, used for relative description and clarification, and can change accordingly depending on the orientation of the components in the accompanying drawings.

[0055] Example 1:

[0056] like Figure 1 As shown, this embodiment of the invention provides an intelligent detection and deviation correction system for steel structure construction, comprising:

[0057] The virtual model module integrates multiple sensors to collect data such as the geometric dimensions, weight deviations, and stress deformation of steel structure components in real time; it constructs a construction physical model based on the physical construction scenario, compares the point cloud data with the construction physical model, and if the analysis result is greater than the preset value, it is output to the steel deviation prediction module.

[0058] The steel deviation prediction module is used to build a steel deviation prediction model based on historical and real-time steel structure data, train the steel deviation prediction model according to the analysis results of the construction physical model, and input the construction physical model into the steel deviation prediction model to predict the deviation trend.

[0059] The correction suggestion generation module generates correction suggestions based on the predicted deviation trend; triggers automatic adjustment commands based on the correction suggestions; and issues an early warning and locates the deviation problem area if the deviation exceeds a preset threshold.

[0060] In the above embodiments, the virtual model module integrates multiple sensors to collect key data such as the geometric dimensions, weight deviation, and stress deformation of steel structure components in real time, and constructs a construction physical model in conjunction with the physical construction scenario. By comparing and analyzing point cloud data with the construction physical model, deviations during construction can be quickly identified. If the deviation exceeds a preset value, the result is output to the steel deviation prediction module. The steel deviation prediction module constructs a steel deviation prediction model based on historical steel structure data and real-time collected steel structure data, and trains the model using the analysis results of the construction physical model. By inputting the construction physical model into the prediction model, the deviation trend of steel structure components can be effectively predicted, providing early warning for potential problems during construction, thereby reducing the impact of deviations on construction quality. The correction suggestion generation module automatically generates targeted correction suggestions based on the predicted deviation trend and triggers corresponding automatic adjustment commands. If the deviation exceeds a preset threshold, an early warning signal is issued, and the deviation problem area is accurately located, facilitating timely corrective measures by construction personnel to ensure that the construction quality meets requirements.

[0061] In summary, this embodiment achieves intelligent monitoring and deviation correction during steel structure construction by organically combining virtual models, deviation prediction, and correction suggestions. Its core value lies in real-time performance, accuracy, and automation: the combination of sensors and models enables timely detection of deviations during construction; the predictive model allows for advance prediction of deviation trends; and the correction suggestion module automatically generates solutions and triggers adjustment commands, significantly improving construction efficiency and quality while reducing construction risks and costs.

[0062] Example 2:

[0063] like Figure 2 As shown, based on Embodiment 1, the virtual model module provided in this embodiment of the invention specifically includes:

[0064] The coordinate alignment submodule is used to align the measured point cloud data with the construction physical model in the coordinate system, select the position of the common steel structure node between the two, transform the measured point cloud from the scanning coordinate system to the design coordinate system of the construction physical model, and perform spatial positioning by generating point cloud lines.

[0065] An attribute comparison submodule is constructed to compare the geometric differences between the point cloud and the construction virtual model, and to calculate the deviation value between the actual cloud points and the construction physical model point by point; the point cloud is semantically segmented to identify basic components, and the corresponding construction attributes in the construction physical model are compared in an object-oriented manner.

[0066] The construction attributes include design dimensions and material parameters, etc.

[0067] The dataset generation submodule is used to construct key elements for identification, fit point cloud data to the elements of the construction physical model, and compare them with the design values; it generates a virtual point cloud dataset from the parameters in the construction physical model, and divides the virtual point cloud dataset into training set, validation set, and test set.

[0068] In the above embodiments, the coordinate alignment submodule ensures that the measured point cloud data and the coordinate system of the construction physical model are aligned, allowing for comparison and analysis within the same coordinate system. By identifying common steel structure nodes, the point cloud is transformed from the scanning coordinate system to the design coordinate system, achieving precise spatial positioning. Coordinate alignment is fundamental for subsequent analysis, ensuring the accuracy and reliability of geometric and attribute comparisons. The attribute comparison submodule is primarily used for geometric difference analysis between the point cloud data and the construction virtual model. By calculating the deviation between the cloud points and the model point by point, the difference between actual construction and design is quantified. Simultaneously, semantic segmentation technology is used to identify basic components in the point cloud and compare them with the construction attributes in the construction physical model. Construction attributes include design dimensions, material parameters, etc., and the comparison results provide data support for construction quality assessment. The dataset generation submodule uses point cloud data to fit elements of the construction physical model for the identified key components and compares them with design values. Simultaneously, it generates a virtual point cloud dataset from the parameters in the construction physical model and divides it into training, testing, and evaluation sets, providing a standardized data foundation for model training, algorithm verification, and quality control.

[0069] In summary, this embodiment achieves efficient comparison and fusion of construction physical models and measured point cloud data through coordinate alignment, geometric difference analysis, and data generation; it provides standardized and data-driven technical support for construction accuracy detection, quality assessment, and optimization of construction processes, and is an important component of modern building information management.

[0070] Example 3:

[0071] Based on Example 2, the construction attribute comparison submodule provided in this embodiment of the invention specifically includes:

[0072] The point cloud segmentation unit is used to extract hierarchical features from the point cloud and learn geometric features point by point; adaptively construct a local neighborhood graph and enhance the feature representation of steel beam nodes through graph attention mechanism; and perform planar segmentation of the point cloud to separate the building plan view from the component point cloud.

[0073] Independent cluster units are used to divide the point cloud into independent clusters, each cluster corresponding to a basic structure, and to refine the segmentation by combining silver information; for multi-source data, the 2D semantic segmentation results are mapped to the 3D point cloud to supplement semantic information;

[0074] The deviation calculation unit is used to compare the segmented construction point cloud with the construction attributes in the construction physical model object by object through feature matching to calculate the geometric deviation; if abnormal segmentation occurs, a feedback optimization mechanism is triggered to adjust the clustering threshold.

[0075] In the above embodiments, the point cloud segmentation unit performs structured processing on the point cloud through hierarchical feature extraction and geometric feature learning; it adaptively constructs a local neighborhood graph and introduces a graph attention mechanism to enhance the feature representation of key parts such as steel beam nodes; simultaneously, it separates the building plan and component point cloud, providing a basis for clustering and comparative analysis. The independent clustering unit divides the point cloud into independent clusters, each cluster corresponding to a basic component; it refines the cluster segmentation by combining shape information (such as silver information) to ensure the accuracy of component identification; for multi-source data, this unit maps the 2D semantic segmentation results to the 3D point cloud to supplement semantic information and further improve segmentation accuracy. The deviation calculation unit uses feature matching technology to compare the segmented construction point cloud with the construction attributes in the construction physical model object by object to calculate the geometric deviation; in addition, for abnormal segmentation cases, this unit triggers a feedback optimization mechanism to dynamically adjust the clustering threshold to ensure the reliability of the segmentation results.

[0076] In summary, the attribute comparison submodule in this embodiment achieves refined comparative analysis of the construction physical model and measured point cloud data through point cloud segmentation, clustering, and deviation calculation. It provides standardized and automated technical means for construction quality inspection, geometric accuracy assessment, and anomaly identification, and is an important support for improving the efficiency and accuracy of construction management. It helps to detect construction errors in a timely manner, optimize the construction process, and ensure that the project quality meets the design requirements.

[0077] Example 4:

[0078] like Figure 3 As shown, based on Example 1, the steel deviation prediction module provided in this embodiment of the invention specifically includes:

[0079] The variable distribution difference submodule is used to build a steel deviation prediction model based on historical steel structure data and real-time steel structure data. It merges the historical steel structure data and real-time steel structure data to obtain a dataset, and divides it into a training set and a test set. It introduces a control dataset and compares the variable distribution deviation between the original training set and the control dataset.

[0080] The optimal model submodule is used to train multiple base models using historical steel structure data, generate new training and test sets; train a meta-model using the new dataset generated from the base models, optimize the weights through backpropagation; and determine the optimal steel deviation prediction model through cross-validation.

[0081] The prediction deviation submodule is used to adjust the weights and retrain the steel deviation prediction model if the control dataset test shows a deviation in variable distribution; the construction physical model is input into the steel deviation prediction model to predict the deviation trend.

[0082] In the above embodiments, the variable distribution difference submodule constructs a steel deviation prediction model based on historical and real-time steel structure data. It generates a dataset by merging historical and real-time data and divides it into training and testing sets. A control dataset is introduced, and the variable distribution deviation between the original training set and the control dataset is compared to provide data support for model optimization. The optimal model submodule trains multiple base models using historical steel structure data, generating new training and testing sets. It trains the new dataset using a meta-model and optimizes the model weights using backpropagation. Cross-validation is used to determine the optimal steel deviation prediction model, improving prediction accuracy. When the control dataset test shows a variable distribution shift, the prediction deviation submodule adjusts the weights and retrains the steel deviation prediction model. The construction physics model is input into the optimized prediction model, and the steel deviation trend prediction result is output.

[0083] In summary, the steel deviation prediction module in this embodiment achieves accurate prediction of steel deviation through the fusion of historical and real-time data, model training and optimization; it provides data-driven technical support for quality control, error warning and decision optimization during construction, which helps to improve construction accuracy and safety; it can effectively reduce material waste and ensure that the project quality meets design expectations.

[0084] Example 5:

[0085] Based on Example 4, the variable distribution difference submodule provided in this embodiment of the invention includes:

[0086] The deviation calculation unit is used to calculate the distribution difference of each variable in the two datasets and generate a deviation probability value; by comparing the deviation probability values, it reflects the degree of deviation of the distribution of the variable in the control dataset from that in the original dataset; and checks whether the frequency of each category in the control dataset is consistent with that in the original dataset.

[0087] The target variable adjustment unit is used to readjust the target variable when the distribution of the target variable in the control dataset differs from the original training set by a greater than a preset standard value.

[0088] The offset determination unit is used to mark target variables with a deviation probability greater than a preset value. If the control dataset with a probability of 1 completely confirms the value of the variable, the unit will determine the degree of offset by visualizing the overlap of the distribution of the target variable in the original dataset and the control dataset. An automated offset report will be generated.

[0089] The calculation formula for the deviation from the program calculation unit is as follows:

[0090]

[0091] In the formula, Representing variables Distribution deviation probability value Representing variables In the original dataset, the first k The frequency of each category; Representing variables In the control dataset, the first k The frequency of each category; Indicates a smoothing coefficient to avoid a denominator of zero; Representing variables In the k Standard deviation in each category; n Indicates the total number of categories of the variable;

[0092] The calculation formula for the target variable adjustment unit:

[0093]

[0094] In the formula, Represent the target variable Distribution difference index; Indicates the first [number] in the original training set j The target variable for each sample The value; Indicates the first in the control dataset j The target variable for each sample The value; Represents the target variable in the original training set. The average value; Indicates the target variable in the control dataset The average value; m Indicates the total number of samples;

[0095] The formula for calculating the degree of offset is as follows:

[0096]

[0097] In the formula, Represent the target variable The degree of deviation index; Represent the target variable The probability density function in the original dataset; Represent the target variable The probability density function in the control dataset; express The mean in the original dataset; Represent the target variable The mean in the control dataset; Represent the target variable Standard deviation in the original dataset; Represent the target variable The standard deviation in the control dataset. The deviation calculation unit formula measures the distribution deviation of a variable between the original and control datasets; the core idea is to compare the frequency differences of variables in the two datasets, while incorporating adjustments for standard deviation and natural logarithm to ensure robustness to outliers; by calculating the distribution deviation probability value, the system can quantify the distribution difference of each variable in the two datasets, thereby identifying those variables that exhibit significant fluctuations in the control dataset. The target variable adjustment unit is used to evaluate the distribution difference of the target variable in the two datasets; combining the concepts of mean squared error and standard deviation, it calculates the distribution difference index between the target variable and the control dataset by comparing their specific values; through the distribution difference index, the system can clearly define the degree of difference of the target variable in the two datasets, helping to adjust model parameters and ensure that the prediction results are closer to the actual values. This is particularly important in multivariate adjustment. The bias judgment unit calculates the difference in probability density functions of variables between the original and control datasets through integration, and combines the mean and standard deviation to quantify the degree of variable bias; it provides a comprehensive judgment of variable bias, helping the system identify those variables with significant differences in probability distribution, thereby providing a decision basis for the adjustment mechanism.

[0098] In the above embodiments, the variable distribution difference submodule combines a deviation calculation unit, a target variable adjustment unit, and a deviation degree judgment unit. It calculates the distribution difference of variables in the control dataset and the original dataset, generates deviation probability values, and precisely quantifies the degree of deviation in variable distribution. It checks whether the frequency of each category in the control dataset is consistent with that in the original dataset, ensuring data distribution consistency. When the difference between the distribution of the target variable in the control dataset and the original training set exceeds a preset standard, it automatically readjusts the target variable to correct the distribution deviation problem. It marks variables with deviation probability values ​​exceeding a preset threshold and uses visualization to compare the overlap of variable distributions in the original dataset and the control dataset, systematically evaluating the degree of deviation. It generates an automated deviation report, providing a clear basis for data quality assessment and subsequent optimization. This embodiment, through systematic calculation, adjustment, and evaluation functions, achieves precise control and optimization of dataset distribution differences, ensuring dataset consistency and reliability, and providing technical support for data quality control and analysis.

[0099] Example 6:

[0100] Based on Example 4, the optimal model submodule provided in this embodiment of the invention includes:

[0101] The prediction result generation unit is used to organize historical steel structure data according to features to form an initial training set; based on the historical steel structure data, multiple base models are generated, and multiple independent base models are trained using the initial training set to generate corresponding prediction results. The prediction results are the base models' capture of the trend or pattern of steel deviation in the historical steel structure data.

[0102] The feature combination processing unit is used to combine the prediction results of each base model with the features of the initial training set to form an expanded feature space; if each base model generates a prediction value, the number of features in the initial training set increases to be equal to the number of base models.

[0103] The standardization unit is used to apply the prediction results of the base model to the test set to generate a new test set; it also performs standardization on the newly generated training and test sets to eliminate the dimensional differences between different features.

[0104] In the above embodiments, the optimal model submodule can efficiently construct and optimize the steel deviation prediction model, thereby improving the model's accuracy and generalization ability. Specifically: the prediction result generation unit organizes historical steel structure data to form an initial training set, and then generates multiple independent base models based on these data; each base model captures the trend or pattern of steel deviation from different perspectives and generates corresponding prediction results; through the training of the base models, a preliminary understanding of the diversity and complexity of steel deviation is established. The feature combination processing unit combines the prediction results of each base model with the original features in the initial training set to form an expanded feature space; it not only retains the core information of the original data, but also incorporates the base model's capture results of deviation trends, making the feature space richer and more comprehensive, which helps to improve the model's expressive ability. The standardization processing unit applies the prediction results of the base models to the test set, generates a new test set, and performs standardization processing on the newly generated training set and test set; standardization eliminates the dimensional differences between different features, enabling the model to process each feature more fairly during training and prediction, thereby improving the model's stability and prediction accuracy.

[0105] In summary, the optimal model submodule in this embodiment can extract more comprehensive feature information from historical data and combine it with the prediction results of multiple base models to form a more robust and accurate steel deviation prediction model. This not only enhances the model's ability to handle data complexity but also ensures its stability and reliability through standardization, ultimately providing a scientific and efficient solution for predicting steel deviation.

[0106] Example 7:

[0107] Based on Example 6, the prediction result generation unit provided in this embodiment of the invention includes:

[0108] The parameter combination design subunit is used to design multiple unique parameter combinations based on the multi-dimensional features of the initial training set, and to design a base model that captures dynamic changing trends to identify trends or patterns in steel deviations.

[0109] The data deconstruction processing subunit is used to deconstruct historical steel structure data from multiple perspectives according to different representations of short-term deviation, long-term deviation, local deviation, and global deviation of steel deviation, and obtain the identification results corresponding to short-term deviation, long-term deviation, local deviation, and global deviation; the identification results are integrated into a comprehensive deviation prediction value using a weighted average.

[0110] The indicator adds a sub-unit to add a confidence index to the deviation prediction value, and uses Monte Carlo simulation to evaluate the reliability and uncertainty of the prediction results.

[0111] In the above embodiments, by organically combining the parameter combination design subunit, data deconstruction processing subunit, and index addition subunit, dynamic trends are captured. Through the design of unique parameter combinations, the base model can effectively identify the trends and patterns of steel deviations, providing a reliable foundation for prediction. A comprehensive deconstruction of deviation manifestations, analyzing historical data from four dimensions—short-term, long-term, local, and global—allows for a more comprehensive and detailed understanding of deviation characteristics, ensuring the comprehensiveness and accuracy of prediction results. The prediction results are integrated and optimized by using weighted averaging to combine the identification results of multi-dimensional deviations into a comprehensive deviation prediction value, ensuring the comprehensiveness and practicality of the prediction results. To assess prediction reliability, confidence indices and Monte Carlo simulations are introduced to quantify the reliability and uncertainty of prediction results, helping users better understand the applicability and potential risks of the prediction results. This approach provides comprehensive, accurate, and reliable prediction results in complex steel deviation scenarios, offering a scientific basis for decision-making.

[0112] Example 8:

[0113] Based on Example 4, the prediction deviation submodule provided in this embodiment of the invention includes:

[0114] The adjustment mechanism trigger unit is used to calculate the bias probability, class bias, and target variable bias by comparing the variable distribution of the control dataset with that of the original training set. If a distribution bias greater than a preset value is detected, the weight adjustment mechanism is triggered. After the base model generates the secondary dataset, the meta-model optimizes the value selection process through backpropagation.

[0115] The prediction value generation unit is used to retrain the steel deviation prediction model based on the updated weights; the physical model parameters of the current construction are input into the adjusted steel deviation prediction model to generate the predicted value of future steel deviation.

[0116] The deviation judgment unit restores the standardized predicted values ​​to the actual physical dimensions; restores the long-term variation pattern of the data by fitting the trend term; calculates the absolute and relative deviations between the predicted values ​​and the actual measured values ​​to identify outliers; and issues an early warning if the deviation exceeds the threshold.

[0117] The calculation formula for the adjustment mechanism trigger unit is as follows:

[0118]

[0119] In the formula, This indicates the trigger threshold for the weight adjustment mechanism; Representing variables The probability value of the distribution deviation (same as the deviation program calculation unit); Representing variables The degree of offset index (same as the degree of offset judgment unit); Represent the target variable Distribution difference index (same as target variable adjustment unit); N Indicates the total number of variables; m Indicates the total number of target variables;

[0120] Predicted value generation unit calculation formula:

[0121]

[0122] In the formula, This indicates the prediction deviation value; In the representation metamodel, the first l The weights of each base model; Indicates the activation function; Indicates the first r One construction physical model parameter; Indicates the first r Model coefficients for each parameter; Indicates the first s One standardized parameter; Indicates the first s The weights of each standardized parameter; The model error term is represented by L, R, and S, which represent the number of base models, the number of construction parameters, and the number of standardized parameters, respectively.

[0123] Deviation Judgment Unit Calculation Formula:

[0124]

[0125] In the formula, This indicates the degree of deviation between the predicted value and the actual value; Indicates the first t Predicted values ​​at each time point; Indicates the firstt Actual measured values ​​at each time point; This represents the standard deviation of the predicted values; This represents the standard deviation of the actual measured value; This represents the mean of the predicted values; This represents the mean of the actual measured values; T Indicates the total number of points in time;

[0126] The adjustment mechanism trigger unit formula integrates the degree of deviation and offset, and combines the distribution differences of the target variable to calculate the trigger threshold for the weight adjustment mechanism. This determines whether model adjustments are needed. When the trigger threshold exceeds a certain threshold, the adjustment mechanism is activated to ensure more accurate model predictions. The prediction value generation unit combines the weights and construction parameters of multiple base models, using activation functions and exponential weighting to generate the final prediction deviation value. This integrates the influence of multiple models and parameters, providing basic data for deviation assessment. The deviation assessment unit compares the predicted value with the actual measured value, using the proportional relationship between the exponential function and the standard deviation to calculate the degree of deviation between the predicted value and the actual value. This provides a quantitative assessment of prediction accuracy, helping the system understand the gap between the predicted results and the actual values, thereby further optimizing model parameters and the adjustment mechanism.

[0127] In the above embodiments, the prediction deviation submodule, through the combined use of multiple functional units, achieves the following overall: by comparing the variable distribution of the control dataset with that of the original training set, it calculates the deviation probability, category deviation, and target variable deviation. If the detected distribution deviation is greater than a preset value, a weight adjustment mechanism is triggered, and the meta-model optimizes the values ​​through backpropagation to ensure that the model parameters adapt to data changes; based on the updated weights, the steel deviation prediction model is retrained, and the physical model parameters of the current construction are input into the adjusted prediction model to generate predicted values ​​of future steel deviations, providing data support for decision-making; the standardized predicted values ​​are restored to actual physical dimensions, and the long-term change law of the data is restored by fitting the trend term; the absolute and relative deviations between the predicted values ​​and the actual measured values ​​are calculated to identify outliers; if the deviation exceeds a threshold, an early warning is issued to ensure the reliability of the prediction results. This embodiment, through a systematic deviation detection, weight adjustment, predicted value generation, and deviation judgment and early warning mechanism, achieves dynamic optimization and reliability assurance of the steel deviation prediction model, providing technical support for construction quality control and prediction.

[0128] Example 9:

[0129] like Figure 4 As shown, based on Embodiment 1, the correction suggestion generation module provided in this embodiment of the invention includes:

[0130] The first trend analysis module is used to analyze the trend of prediction deviation, to determine whether the prediction deviation shows a short-term local expansion or a long-term global accumulation; for short-term local expansion, a local correction plan is proposed; for long-term global accumulation, a comprehensive adjustment is made.

[0131] The second trend analysis module is used to obtain geometric dimension deviation, stress deformation deviation, or weight deviation from the predicted deviation trend.

[0132] Among them, for geometric dimension deviations such as length, width, or height deviations, adjustment schemes are generated through parametric design tools, such as suggesting local cutting and welding or replacing the entire component;

[0133] Optimize the support structure or adjust the distribution of external loads to compensate for stress and deformation deviations; conduct mechanical property tests on bolts, nuts, washers, etc. in the support structure;

[0134] Weight deviation provides solutions for counterweight adjustment or optimized material distribution;

[0135] The instruction conversion module is used to convert geometric deviations, stress deformation deviations, or weight deviations into executable instructions, which are linked with the construction equipment to automatically adjust welding parameters, displace component positions, or apply auxiliary supports based on correction suggestions; and to inspect the weld seams using ultrasonic waves.

[0136] In the above embodiments, the first trend analysis module distinguishes between short-term local expansion and long-term global accumulation deviation trends, providing local correction or comprehensive adjustment schemes respectively, ensuring the pertinence and efficiency of the correction measures. The second trend analysis module addresses deviations from three dimensions: geometric dimension deviation, stress deformation deviation, and weight deviation, proposing scientific and reasonable correction schemes respectively. For example, it uses parametric design tools to generate adjustment schemes, optimizes support structures, or adjusts load distribution, ensuring the coverage and operability of the correction strategies. Automated execution and linkage are achieved through an execution command conversion module that transforms correction suggestions into executable commands, linking with construction equipment to automatically adjust welding parameters, displace component positions, or apply auxiliary supports, reducing the need for manual intervention and improving construction efficiency and consistency. Through the collaborative work of each module, the system can comprehensively analyze, scientifically design, and automatically execute deviation problems, realizing closed-loop management from prediction to correction, improving construction quality, reducing costs, and providing strong technical support for the intelligent development of steel structure construction.

[0137] Figure 5 A block diagram of an exemplary electronic device suitable for implementing embodiments of the present invention is shown.

[0138] Electronic devices may include a central processing unit / microprocessor / main control chip, etc.; a storage medium coupled to the central processing unit / microprocessor / main control chip, etc., and storing computer-executable instructions therein for performing the steps of various methods of embodiments of the present invention when executed by a processor.

[0139] The central processing unit / microprocessor / main control chip, etc., may include, but are not limited to, one or more processors or microprocessors.

[0140] Storage media may include, but are not limited to, random access memory (RAM), read-only memory (ROM), flash memory, EPROM memory, EEPROM memory, registers, and computer storage media (such as hard disks, floppy disks, solid-state drives, removable disks, CD-ROMs, DVD-ROMs, Blu-ray discs, etc.).

[0141] In addition, the electronic device may include (but is not limited to) a data bus, an input / output bus / external bus / device bus, a display, and input / output devices (e.g., keyboard, mouse, speaker, etc.).

[0142] The central processing unit / microprocessor / main control chip, etc., can communicate with external devices via the I / O bus through a wired or wireless network (not shown).

[0143] The storage medium may also store at least one computer-executable instruction for performing the steps of various functions and / or methods in the embodiments described herein when run by a central processing unit / microprocessor / main control chip, etc.

[0144] In one embodiment, the at least one computer-executable instruction may also be compiled into or comprise a software product, wherein one or more computer-executable instructions are executed by a processor to perform the steps of the various functions and / or methods in the embodiments described herein.

[0145] Figure 6 A schematic diagram of a computer-readable storage medium according to an embodiment of the present invention is shown.

[0146] like Figure 6As shown, instructions, such as computer-readable instructions, are stored on a non-transitory computer-readable storage medium. When the computer-readable instructions are executed by a processor, the various methods described above can be performed. The non-transitory computer-readable storage medium includes, but is not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. Non-transitory non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. For example, the non-transitory computer-readable storage medium can be connected to a computing device such as a computer, and then, when the computing device executes the computer-readable instructions stored on the computer-readable storage medium, the various methods described above can be performed.

[0147] In the several embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0148] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0149] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0150] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product, stored in a storage medium, specifically includes several instructions for executing all or part of the steps of the methods of various embodiments of this invention through a computer device (which may be a personal computer, server, or network device, etc.). The aforementioned storage medium includes: USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, optical disks, and other media capable of storing program code.

[0151] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent detection and deviation correction system for steel structure construction, characterized in that, Include: The virtual model module integrates multiple sensors to collect real-time data on the geometric dimensions, weight deviations, and stress deformation of steel structure components. It constructs a construction physical model based on the physical construction scenario and compares the point cloud data with the construction physical model. If the analysis result is greater than the preset value, it is output to the steel deviation prediction module. The steel deviation prediction module is used to build a steel deviation prediction model based on historical and real-time steel structure data, train the steel deviation prediction model according to the analysis results of the construction physical model, and input the construction physical model into the steel deviation prediction model to predict the deviation trend. The correction suggestion generation module is used to generate correction suggestions based on the trend of prediction deviation. Based on the suggested corrections, an automatic adjustment command is triggered. If the deviation exceeds the preset threshold, an early warning will be issued and the area of ​​deviation problem will be located. The steel deviation prediction module includes: The variable distribution difference submodule is used to build a steel deviation prediction model based on historical steel structure data and real-time steel structure data. It merges the historical steel structure data and real-time steel structure data to obtain a dataset, and divides it into a training set and a test set. It introduces a control dataset and compares the variable distribution deviation between the original training set and the control dataset. The optimal model submodule is used to train multiple base models using historical steel structure data, generate new training and test sets; train a meta-model using the new dataset generated from the base models, optimize the weights through backpropagation; and determine the optimal steel deviation prediction model through cross-validation. The prediction deviation submodule is used to adjust the weights and retrain the steel deviation prediction model if the control dataset test shows a deviation in variable distribution; the construction physical model is input into the steel deviation prediction model to predict the deviation trend.

2. The intelligent detection and deviation correction system for steel structure construction as described in claim 1, characterized in that, The virtual model module includes: The coordinate alignment submodule is used to align the measured point cloud data with the construction physical model in the coordinate system, select the position of the common steel structure node between the two, transform the measured point cloud from the scanning coordinate system to the design coordinate system of the construction physical model, and perform spatial positioning by generating point cloud lines. An attribute comparison submodule is constructed to compare the geometric differences between the point cloud and the construction virtual model, and to calculate the deviation value between the actual cloud points and the construction physical model point by point; the point cloud is semantically segmented to identify basic components, and the corresponding construction attributes in the construction physical model are compared in an object-oriented manner. The dataset generation submodule is used to construct key elements for identification, fit point cloud data into construction physical model elements, and compare them with design values; it generates a virtual point cloud dataset from the parameters in the construction physical model, and divides the virtual point cloud dataset into training set, validation set, and test set.

3. The intelligent detection and deviation correction system for steel structure construction as described in claim 2, characterized in that, Construct an attribute comparison submodule, including: The point cloud segmentation unit is used to extract hierarchical features from the point cloud and learn geometric features point by point; adaptively construct a local neighborhood graph and enhance the feature representation of steel beam nodes through graph attention mechanism; and perform planar segmentation of the point cloud to separate the building plan view from the component point cloud. Independent cluster units are used to divide the point cloud into independent clusters, each cluster corresponding to a basic structure, and to refine the segmentation by combining silver information; for multi-source data, the 2D semantic segmentation results are mapped to the 3D point cloud to supplement semantic information; The deviation calculation unit is used to compare the segmented construction point cloud with the construction attributes in the construction physical model object by object through feature matching to calculate the geometric deviation; if abnormal segmentation occurs, a feedback optimization mechanism is triggered to adjust the clustering threshold.

4. The intelligent detection and deviation correction system for steel structure construction as described in claim 1, characterized in that, The variable distribution difference submodule includes: The deviation calculation unit is used to calculate the distribution difference of each variable in the two datasets and generate a deviation probability value; by comparing the deviation probability values, it reflects the degree of deviation of the distribution of the variable in the control dataset from that in the original dataset; and checks whether the frequency of each category in the control dataset is consistent with that in the original dataset. The target variable adjustment unit is used to readjust the target variable when the distribution of the target variable in the control dataset differs from the original training set by a greater than a preset standard value. The offset determination unit is used to mark target variables with a deviation probability greater than a preset value. If the target variable is completely missing in the control dataset with a probability of 1, the unit visualizes the overlap of the distribution of the target variable in the original dataset and the control dataset to determine the degree of offset. An automated offset report is generated.

5. The intelligent detection and deviation correction system for steel structure construction as described in claim 1, characterized in that, The optimal model submodule includes: The prediction result generation unit is used to organize historical steel structure data according to features to form an initial training set; based on the historical steel structure data, multiple base models are generated, and multiple independent base models are trained using the initial training set to generate corresponding prediction results. The prediction results are the base models' capture of the trend or pattern of steel deviation in the historical steel structure data. The feature combination processing unit is used to combine the prediction results of each base model with the features of the initial training set to form an expanded feature space; if each base model generates a prediction value, the number of features in the initial training set increases to be equal to the number of base models. The standardization unit is used to apply the prediction results of the base model to the test set to generate a new test set; the newly generated training set and test set are standardized to eliminate the dimensional differences between different features.

6. The intelligent detection and deviation correction system for steel structure construction as described in claim 5, characterized in that, The prediction result generation unit includes: The parameter combination design subunit is used to design multiple unique parameter combinations based on the multi-dimensional features of the initial training set, and to design a base model that captures dynamic changing trends to identify trends or patterns in steel deviations. The data deconstruction processing subunit is used to deconstruct historical steel structure data from multiple perspectives according to different representations of short-term deviation, long-term deviation, local deviation, and global deviation of steel deviation, and obtain the identification results corresponding to short-term deviation, long-term deviation, local deviation, and global deviation; the identification results are integrated into a comprehensive deviation prediction value using a weighted average. The indicator adds a sub-unit to add a confidence index to the deviation prediction value, and uses Monte Carlo simulation to evaluate the reliability and uncertainty of the prediction results.

7. The intelligent detection and deviation correction system for steel structure construction as described in claim 1, characterized in that, The prediction bias submodule includes: The adjustment mechanism trigger unit is used to calculate the bias probability, class bias, and target variable bias by comparing the variable distribution of the control dataset with that of the original training set. If a distribution bias greater than a preset value is detected, the weight adjustment mechanism is triggered. After the base model generates the secondary dataset, the meta-model optimizes the value selection process through backpropagation. The prediction value generation unit is used to retrain the steel deviation prediction model based on the updated weights; the physical model parameters of the current construction are input into the adjusted steel deviation prediction model to generate the predicted value of future steel deviation. The deviation judgment unit restores the standardized predicted values ​​to the actual physical dimensions; restores the long-term variation pattern of the data by fitting the trend term; calculates the absolute and relative deviations between the predicted values ​​and the actual measured values ​​to identify outliers; and issues an early warning if the deviation exceeds the threshold.

8. The intelligent detection and deviation correction system for steel structure construction as described in claim 1, characterized in that, The module for generating modification suggestions includes: The first trend analysis module is used to analyze the trend of prediction deviation, to determine whether the prediction deviation shows a short-term local expansion or a long-term global accumulation; for short-term local expansion, a local correction plan is proposed; for long-term global accumulation, a comprehensive adjustment is made. The second trend analysis module is used to obtain geometric dimension deviation, stress deformation deviation, or weight deviation from the predicted deviation trend. The instruction conversion module is used to convert geometric deviations, stress deformation deviations, or weight deviations into executable instructions, which are linked with the construction equipment to automatically adjust welding parameters, displace component positions, or apply auxiliary supports based on correction suggestions; and to inspect the weld seams using ultrasonic waves.

9. The intelligent detection and deviation correction system for steel structure construction as described in claim 8, characterized in that, in, For geometric dimensional deviations in length, width, or height, adjustment schemes can be generated using parametric design tools, with recommendations including local cutting and welding or replacement of the entire component. Optimize the support structure or adjust the distribution of external loads to address stress-deformation deviations; conduct mechanical property tests on bolts, nuts, and washers in the support structure. Weight deviation provides solutions for counterweight adjustment or optimized material distribution.

Images