Construction dynamic monitoring method based on BIM and multi-source data

Abstract

The invention provides a construction dynamic monitoring method based on BIM and multi-source data, and the method comprises the steps: deploying a sensor at a key part of a construction building, so as to monitor a construction process and obtain key construction data; the method comprises the steps of collecting construction data and uploading the construction data to a back-end computer in real time, preprocessing the construction data and normalizing the construction data to serve as an input vector, analyzing the construction process condition in real time by utilizing a trained neural network model, predicting whether an abnormal condition exists or not, comparing the construction data of the abnormal condition with BIM model data, finding out corresponding improper construction, and determining whether the abnormal condition exists or not. And the information is fed back to on-site constructors. The method has the beneficial effects that real-time monitoring and accurate prediction in the construction process are realized, so that the safety and the operation efficiency of a BIM-based construction project are remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of construction monitoring, and in particular relates to a construction dynamic monitoring method based on BIM and multi-source data. Background Art

[0002] In the field of construction engineering, construction based on building information modeling (BIM) has become increasingly popular. BIM provides a multi-dimensional digital representation of construction projects, supporting all kinds of information from design, construction to maintenance stages in the decision-making process. Using BIM technology, engineers and architects can efficiently plan, design, build and manage buildings and infrastructure on an integrated platform.

[0003] Existing technologies integrate BIM with on-site monitoring systems to monitor key parameters during construction, such as position errors, structural stability, and environmental conditions. This integration usually involves combining sensor data with BIM models to achieve real-time monitoring and data-driven decision support. However, these systems often face limitations and challenges, such as difficulties in data integration, insufficient ability to handle high-frequency time series data, and lack of ability to effectively predict and process multi-source data. Summary of the invention

[0004] In view of this, the present invention aims to propose a construction dynamic monitoring method based on BIM and multi-source data, so as to realize real-time monitoring and accurate prediction during the construction process, thereby significantly improving the safety and operation efficiency of BIM-based construction projects.

[0005] To achieve the above object, the technical solution of the present invention is achieved as follows: Construction dynamic monitoring method based on BIM and multi-source data.

[0006] Furthermore, sensors are deployed at key locations of the construction building to monitor the construction process and obtain key construction data; Collect construction data and upload it to the back-end computer in real time, pre-process the construction data and normalize it as an input vector, use the trained neural network model to analyze the construction process in real time, and predict whether there are abnormal situations, compare the construction data of abnormal situations with the BIM model data, find out the corresponding improper construction, and feedback to the on-site construction personnel; deploy sensors at key parts of the construction building, including: Install temperature and humidity sensors after concrete pouring to monitor temperature and humidity during the curing process; Install stress and strain sensors at key connection points of steel structures to monitor changes in structural bearing capacity; Use laser scanners to regularly scan the installation positions of prefabricated components and compare them with the BIM design model to monitor installation position deviations.

[0007] Furthermore, an automated script is used to identify outliers in the collected construction data, and then all construction data is normalized as input data, including: ; ; ; ; ; In the formula, , , , , They are temperature data, humidity data, stress data, strain data, and position deviation data. At 20°C, At 35°C, is 40%, is 90%, is 10MPa, and Select according to the specific environment.

[0008] Furthermore, the neural network model includes: A feature extraction layer receiving input data, extracting features of multi-source input data; The fully connected layer receives the results of the feature extraction layer and performs weighted fusion on the extracted features; The output layer connected to the fully connected layer predicts the probability of various results based on the fusion results of the fully connected layer.

[0009] Furthermore, the feature extraction layer includes using an LSTM layer to extract long-term dependency and periodic features of temperature and humidity data and stress and strain data.

[0010] Furthermore, the fully connected layer includes receiving the time series features of the temperature and humidity data and stress and strain data output by the LSTM layer and the directly input position deviation data for weighted fusion, as follows: ; In the formula, is the fusion function, , , , They are the temperature, humidity, stress and strain features after LSTM layer processing. , , , , The weights correspond to each data respectively, and the weights of each data are allocated according to the construction focus of the current stage.

[0011] Furthermore, whenever the construction environment changes significantly, the weights of temperature, humidity, stress, strain, and position deviation need to be readjusted using the following formula: ; ; In the formula, are the weights of the adjusted temperature, humidity, stress, strain, and position deviation, is the adjustment factor, is the current characteristic value of each construction data, is the historical mean of each construction data.

[0012] Furthermore, the output layer includes output categories such as temperature and humidity exceeding the standard, stress exceeding the limit, position deviation being too large, and everything being normal. The sensitivity and threshold of the output category are dynamically adjusted according to the real-time data and the construction stage.

[0013] Furthermore, a change threshold of the characteristic of each construction data is set according to the historical data to determine whether the construction environment has changed significantly; The adjusted weights of temperature, humidity, stress, strain, and position deviation need to be normalized so that the sum of all weights is 1.

[0014] Furthermore, the dynamic weight adjustment adopts a smooth adjustment mechanism to make a gradual transition to the weight adjustment. The following formula is used for smoothing: ; In the formula, represent The smoothed weights, is the weight adjustment caused by changes in the construction environment, is the smoothing factor.

[0015] Furthermore, a computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the construction dynamic monitoring method based on BIM and multi-source data is implemented.

[0016] Compared with the prior art, the construction dynamic monitoring method based on BIM and multi-source data described in the present invention has the following beneficial effects: (1) The construction dynamic monitoring method based on BIM and multi-source data described in the present invention can capture long-term trends and periodic changes in the data by using LSTM to process time series data such as temperature, humidity and stress and strain data, so that the system can predict and respond to potential risks in a timely manner; (2) The construction dynamic monitoring method based on BIM and multi-source data described in the present invention effectively fuses data from different sensors, including real-time and historical data, by integrating the MLP fully connected layer, and optimizes the information flow. This dynamic data fusion strategy enables each type of data to be weighted according to its importance in a specific construction stage, thereby enhancing the accuracy of the decision-making process, especially in a complex construction environment, and enabling better management and adjustment of construction strategies; (3) The construction dynamic monitoring method based on BIM and multi-source data described in the present invention combines the multi-dimensional data of the BIM model, which not only improves structural safety, but also optimizes resource allocation and construction scheduling, reduces unnecessary delays and cost overruns, and improves the efficiency and quality assurance of the entire construction project. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the accompanying drawings: Figure 1 It is a schematic diagram of a flow chart of a construction monitoring method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the neural network model architecture described in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0019] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0020] Construction dynamic monitoring method based on BIM and multi-source data, such as Figure 1 As shown, sensors are deployed at key locations of the construction building to monitor the construction process and obtain key construction data; construction data is collected and uploaded to the back-end computer in real time, the construction data is preprocessed and normalized as an input vector, and the trained neural network model is used to analyze the construction process in real time and predict whether there are abnormal situations. The construction data of the abnormal situation is compared with the BIM model data to find out the corresponding improper construction and provide feedback to the on-site construction personnel.

[0021] Specifically, the deployment of sensors at key locations of construction buildings includes: installing temperature and humidity sensors after concrete pouring to monitor the temperature and humidity during the curing process, preferably collecting data every 30 minutes, and the range is usually 20°C to 35°C and 40% to 90% relative humidity; installing stress and strain sensors at key connection points of the steel structure to monitor changes in the structural bearing capacity, preferably recording data every 10 minutes; using a laser scanner to regularly scan the installation position of prefabricated components and compare them with the BIM design model to monitor installation position deviations.

[0022] Furthermore, an automated script is used to identify outliers in the collected construction data, and then all construction data are normalized as input data as follows: ; ; ; ; ; In the formula, , , , , They are temperature data, humidity data, stress data, strain data, and position deviation data. At 20°C, At 35°C, is 40%, is 90%, is 10MPa, and Select according to the specific environment.

[0023] Furthermore, the neural network model is a hybrid model that combines a multi-layer perceptron (MLP) and a long short-term memory network (LSTM), and uses LSTM to process input data containing time series and multiple sources, such as Figure 2 As shown, it includes: a feature extraction layer that receives input data and extracts features of multi-source input data; a fully connected layer that receives the results of the feature extraction layer and performs weighted fusion on the extracted features; an output layer connected to the fully connected layer and predicts the probability of various results according to the fusion results of the fully connected layer.

[0024] Specifically, the feature extraction layer is used to receive multi-source input data from the construction site and perform preliminary feature extraction, including using the LSTM layer to extract long-term dependency and periodicity features of temperature and humidity data and stress and strain data.

[0025] Specifically, the fully connected layer (MLP) is used to receive the results processed by the feature extraction layer and fuse them with the directly input position deviation data, including receiving the time series features of the temperature and humidity data and stress and strain data output by the LSTM layer and the directly input position deviation data for weighted fusion, as follows: ; In the formula, is the fusion function, , , , They are the temperature, humidity, stress and strain features after LSTM layer processing. , , , , The weights correspond to the data respectively, and the weights of the data are allocated according to the construction focus of the current stage. For example, during the concrete pouring, more attention may be paid to the temperature and humidity data, while during the structure assembly stage, more attention may be paid to the stress, strain and position deviation.

[0026] Furthermore, due to the large changes in the on-site construction environment, the errors in the daily measurements of each sensor are large. Coupled with the limitations of the multi-layer perceptron (MLP) and the long short-term memory network (LSTM), the temperature and humidity have always exceeded the standard, the stress has exceeded the limit, or the position deviation is too large during the monitoring process, causing the output of the neural network model to be trapped. Therefore, this application needs to adaptively adjust the weights. Whenever the construction environment changes significantly, the weights of temperature, humidity, stress, strain, and position deviation need to be readjusted using the following formula: ; ; In the formula, are the weights of the adjusted temperature, humidity, stress, strain, and position deviation, is the adjustment factor, is the current characteristic value of each construction data, is the historical mean of each construction data.

[0027] Specifically, the threshold of change of the characteristics of each construction data is set according to the historical data to determine whether the construction environment has changed significantly; the weights of the adjusted temperature, humidity, stress, strain, and position deviation need to be normalized so that the sum of all weights is 1. The purpose of dynamically adjusting weights is to adapt to changes in the construction environment and to improve the robustness of the neural network model in complex construction environments. Because sudden environmental changes may cause drastic fluctuations in monitoring data, directly adjusting weights will introduce instability and affect model output. Therefore, the dynamic adjustment of weights uses a smooth adjustment mechanism to make a gradual transition to weight adjustment, and all the weights are adjusted. The following formula is used for smoothing: ; In the formula, represent The smoothed weights, is the weight adjustment caused by changes in the construction environment, is the smoothing factor. For example: .

[0028] The corresponding smoothed weights are: Specifically, the output layer uses the Softmax activation function to convert the output of the fully connected layer into the probability of predicting various potential construction problems. These problems include but are not limited to excessive temperature and humidity, excessive stress, excessive position deviation, etc., including the output categories of excessive temperature and humidity, excessive stress, excessive position deviation, and everything is normal. The sensitivity and threshold of the output category are dynamically adjusted according to the real-time data and the construction stage.

[0029] Those of ordinary skill in the art will appreciate that the units and method steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0030] In the several embodiments provided in the present application, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the division of the units described above is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. The above-mentioned units may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present invention.

[0031] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

[0032] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A construction dynamic monitoring method based on BIM and multi-source data, characterized by: Deploy sensors at key locations of the construction building to monitor the construction process and obtain key construction data; collect construction data and upload it to the back-end computer in real time, pre-process the construction data and normalize it as an input vector, use the trained neural network model to analyze the construction process in real time, and predict whether there are any abnormalities, compare the construction data of abnormalities with the BIM model data, find out the corresponding improper construction, and provide feedback to the on-site construction personnel; Deployment of sensors in key parts of construction buildings includes: Install temperature and humidity sensors after concrete pouring to monitor temperature and humidity during the curing process; Install stress and strain sensors at key connection points of steel structures to monitor changes in structural bearing capacity; Use laser scanners to regularly scan the installation positions of prefabricated components and compare them with the BIM design model to monitor installation position deviations; Among them, whenever the construction environment changes significantly, the weights of temperature, humidity, stress, strain, and position deviation need to be readjusted using the following formula: ； ； In the formula, are the weights of the adjusted temperature, humidity, stress, strain, and position deviation, is the adjustment factor, is the current characteristic value of each construction data, is the historical mean of each construction data.

2. The construction dynamic monitoring method based on BIM and multi-source data according to claim 1 is characterized in that: An automated script was used to identify outliers in the collected construction data and then normalize all the construction data as input data including: ； ； ； ； ； In the formula, , , , , They are temperature data, humidity data, stress data, strain data, and position deviation data. At 20°C, At 35°C, is 40%, is 90%, 10MPa, and Select according to the specific environment.

3. The construction dynamic monitoring method based on BIM and multi-source data according to claim 2 is characterized in that: The neural network model includes: A feature extraction layer receiving input data, extracting features of multi-source input data; The fully connected layer receives the results of the feature extraction layer and performs weighted fusion on the extracted features; The output layer connected to the fully connected layer predicts the probability of various results based on the fusion results of the fully connected layer.

4. The construction dynamic monitoring method based on BIM and multi-source data according to claim 3 is characterized in that: The feature extraction layer includes using an LSTM layer to extract long-term dependency and periodicity features of temperature and humidity data and stress and strain data.

5. The construction dynamic monitoring method based on BIM and multi-source data according to claim 4 is characterized in that: The fully connected layer includes receiving the time series features of the temperature and humidity data and stress and strain data output by the LSTM layer and the directly input position deviation data for weighted fusion, as follows: ； In the formula, is the fusion function, , , , They are the temperature, humidity, stress and strain features after LSTM layer processing. , , , , The weights correspond to each data respectively, and the weights of each data are allocated according to the construction focus of the current stage.

6. The construction dynamic monitoring method based on BIM and multi-source data according to claim 5 is characterized in that: The output layer includes output categories such as temperature and humidity exceeding the standard, stress exceeding the limit, position deviation being too large, and everything being normal. The sensitivity and threshold of the output category are dynamically adjusted according to the real-time data and the construction stage.

7. The construction dynamic monitoring method based on BIM and multi-source data according to claim 1 is characterized in that: According to the historical data of each construction data, the change threshold of its characteristics is set to determine whether the construction environment has changed significantly; The adjusted weights of temperature, humidity, stress, strain, and position deviation need to be normalized so that the sum of all weights is 1.

8. The construction dynamic monitoring method based on BIM and multi-source data according to claim 1 is characterized in that: Dynamic weight adjustment uses a smooth adjustment mechanism to make a gradual transition to weight adjustment. The following formula is used for smoothing: ； In the formula, represent The smoothed weights, is the weight adjustment caused by changes in the construction environment, is the smoothing factor.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the construction dynamic monitoring method based on BIM and multi-source data described in any one of claims 1 to 8 is implemented.

Images