A concrete hydration heat self-adaptive control method and system based on digital twinning

Abstract

The application discloses the technical field of building engineering and intelligent control, and relates to a concrete hydration heat self-adaptive control method and system based on digital twinning. The method comprises the following steps: S1, a concrete finite element model is constructed to simulate the hydration heat process and output simulation data of temperature and temperature stress; S2, based on the simulation data, the arrangement scheme of cooling water pipes and sensors is optimized through a composite convolutional neural network; S3, the real-time temperature and temperature stress data are collected according to the scheme; S4, the dynamic prediction model is constructed by combining the simulation data with the real-time data, adopting a BP neural network and a long short-term memory network to predict the temperature change; S5, based on the prediction results and the real-time data, a WebGL digital twinning platform is constructed to visualize the data and correct the prediction model; and S6, the cooling water pipe parameters are dynamically adjusted according to the correction results to realize the self-adaptive control of the hydration heat. The system for realizing the method can accurately control the concrete temperature and temperature stress and improve the engineering structure safety and the quality control efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of construction engineering and intelligent control technology, in particular to a concrete hydration heat self-adaptive control method and system based on digital twinning. BACKGROUND

[0002] As an indispensable basic material in modern engineering construction, concrete plays a crucial role in quality control and temperature management during construction. With increasingly stringent construction requirements, how to improve the quality of concrete pouring, especially through precise temperature control to avoid temperature-induced structural problems, has become the core issue of current research and application. However, existing temperature control technologies face many challenges in practical application. First, the acquisition of temperature control information often lags behind and is inaccurate, which leads to phenomena such as excessive temperature difference, excessive cooling amplitude and rate, and uneven temperature gradient, ultimately causing concrete cracks. Second, the deviation of operating personnel during construction is also an important factor leading to the failure of temperature control, resulting in the limitation of temperature control.

[0003] Currently, many units have developed temperature control systems that automatically control the opening and closing of water cooling pipes, and some systems can monitor concrete hydration heat. However, existing technologies still have deficiencies, especially in experimental research on hydration heat mechanisms. In the face of the use of new types of admixtures, existing systems are difficult to accurately predict the temperature development of concrete hydration heat. Most systems can only provide passive control methods and are difficult to achieve active optimization of temperature control. In addition, the data monitoring and display methods are too simple and do not fully realize the deep integration of virtual and actual space, which cannot effectively support accurate decision-making.

[0004] With the development of the construction industry towards intelligence and digitization, how to combine the automation and intelligence of concrete temperature control has become a key problem that needs to be solved. Therefore, how to improve data processing speed, optimize data display platform, and improve data utilization efficiency, and dynamically adjust temperature control parameters according to real-time data, has become an important challenge to realize the intelligentization of concrete temperature control. SUMMARY

[0005] The present application provides a concrete hydration heat self-adaptive control method and system based on digital twinning, which can accurately control the temperature and temperature stress of concrete, and achieve the technical effect of improving the safety and quality control efficiency of engineering structures. During concrete construction, temperature control can be adjusted according to real-time monitoring data, and the simulation and prediction accuracy can be optimized by using ensemble learning and composite convolutional neural networks to realize dynamic temperature management, thereby reducing temperature-related structural defects and achieving the technical effect.

[0006] In a first aspect, the present application provides a concrete hydration heat self-adaptive control method based on digital twinning, which can include:

[0007] S1. Constructing a finite element model of concrete according to preset parameters and performing hydration heat simulation prediction of concrete, and outputting simulation data of temperature and temperature stress;

[0008] S2. According to the simulation data, presetting the arrangement optimization target of cooling water pipes and sensors, and generating the arrangement scheme of cooling water pipes and sensors based on the arrangement optimization target through a composite convolutional neural network;

[0009] S3. Arranging cooling water pipes and sensors and collecting real-time data according to the arrangement scheme; the real-time data includes temperature data and temperature stress data;

[0010] S4. According to the simulation data and the real-time data, predicting the temperature change trend of concrete by using a BP neural network and a long short-term memory network, and constructing a dynamic prediction model of concrete temperature and temperature stress, and outputting a dynamic prediction result;

[0011] S5. Based on the dynamic prediction result and the real-time data, constructing a digital twin platform based on WebGL technology; the digital twin platform is used for visualizing the temperature and temperature stress changes inside and on the surface of the concrete, and correcting the dynamic prediction model, and outputting a corrected prediction result;

[0012] S6. According to the corrected prediction result output by the digital twin platform, dynamically adjusting the water flow, water flow direction and water temperature of the cooling water pipes to realize adaptive control of the hydration heat of the concrete.

[0013] In some embodiments, the preset parameters include concrete mix proportion, environmental parameters, condition parameters of concrete construction and maintenance, physical parameters and arrangement parameters of cooling water pipes; the environmental parameters include temperature, humidity, wind speed.

[0014] In the above implementation process, by considering the preset parameters such as concrete mix proportion, environmental parameters and construction conditions, the adaptability and accuracy of simulation prediction are enhanced, and the effect of optimizing the response ability of construction environment is achieved.

[0015] In some embodiments, step S1 includes constructing a concrete model using WebGL technology, and establishing a finite element simulation based thereon to simulate and analyze the change trend of simulation temperature and temperature stress of concrete hydration heat.

[0016] In the above implementation process, by using entity modeling and finite element model establishment based on WebGL technology, the continuity, accuracy of data and reliability of simulation analysis are improved, thereby achieving the effect of enhancing the practicality of simulation results.

[0017] In some embodiments, step S2 can further include:

[0018] S21. According to the simulation data, preset a layout optimization target of the cooling water pipes and the sensors; the layout optimization target is to maximize the cooling effect and minimize the layout cost;

[0019] S22. A composite convolutional neural network is used to identify the distribution of the temperature, to maximize the cooling effect, and to generate an optimal layout scheme of the cooling water pipes;

[0020] S23. On the basis of the optimal layout of the cooling water pipes, a composite convolutional neural network is used to generate a layout scheme of the sensors.

[0021] In the above implementation manner, the cooling water pipes and the sensors are automatically optimized by the composite convolutional neural network, the maximization of the cooling effect and the minimization of the cost are realized, and the construction efficiency and the economy are improved.

[0022] In some embodiments, step S23 of generating the layout scheme of the sensors includes optimizing the number of sensors by using a master-slave composite convolutional neural network; wherein the master path of the master-slave composite convolutional neural network processes the temperature change feature, and the slave path processes the humidity and the concrete proportion change.

[0023] In the above implementation manner, the number and the layout of the sensors are optimized by the master-slave composite convolutional neural network, the accuracy and the efficiency of the temperature and humidity monitoring are realized, and the comprehensiveness and the accuracy of the data monitoring are improved.

[0024] In some embodiments, step S3 can include:

[0025] S31. According to the layout scheme, a wireless sensor network is arranged;

[0026] S32. Real-time data of the concrete is collected by the wireless sensor network, and the real-time data is backed up in a local storage device; the real-time data includes temperature data and temperature stress data;

[0027] S33. The real-time data is checked by using a cyclic redundancy check method, and the checked real-time data is transmitted to a cloud platform server.

[0028] In the above implementation manner, the real-time data collection and the cyclic redundancy check of the wireless sensor network ensure the accuracy and the real-time of the data, and the effect of improving the reliability of data transmission is achieved.

[0029] In some embodiments, step S4 can include:

[0030] S41. A BP neural network is trained according to the real-time data and the simulation data, and a short-term prediction of the temperature of the concrete is made by using the trained BP neural network to obtain a short-term prediction result;

[0031] S42. inputting the simulation data, the real-time data and the short-term prediction result into a long short-term memory network, and outputting a dynamic prediction result of a temperature change trend of the concrete hydration heat in real time.

[0032] In the implementation manners described above, the combination of the BP neural network and the long short-term memory network is adopted to achieve accurate prediction of the temperature change trend of the concrete, thereby achieving the effects of optimizing the temperature control strategy and improving the structural safety.

[0033] In some embodiments, step S41 further includes optimizing key parameters of the BP neural network by using a cross-global artificial bee colony algorithm; the key parameters include input parameters, a topological structure and initial weights.

[0034] In the implementation manners described above, the key parameters of the BP neural network are optimized by using the cross-global artificial bee colony algorithm, thereby improving the accuracy of temperature prediction, and achieving more fine temperature control management and reducing the structural risk caused by temperature.

[0035] In some embodiments, step S6 includes:

[0036] S61. optimizing, based on a non-dominated sorting genetic algorithm, a water flow rate, a water flow direction and a water temperature of the cooling water pipe according to the real-time data and the dynamic prediction result;

[0037] S62. adjusting cooling zone parameters of the concrete, including a temperature rise rate, a temperature drop rate and a maximum temperature difference.

[0038] In the implementation manners described above, the parameters of the cooling water pipe are dynamically adjusted by using the non-dominated sorting genetic algorithm, thereby achieving adaptive and fine temperature control of the concrete, and achieving the effects of minimizing the risk of temperature cracks and improving the structural stability.

[0039] In a second aspect, the embodiments of the present application provide a concrete hydration heat adaptive control system based on digital twinning, which can include:

[0040] a model construction unit configured to construct a finite element model of the concrete according to preset parameters and perform hydration heat simulation, and output simulation data of temperature and temperature stress;

[0041] a layout generation unit configured to generate a layout scheme of the cooling water pipe and the sensor according to the simulation data by using a composite convolutional neural network;

[0042] a real-time monitoring unit configured to arrange the cooling water pipe and the sensor according to the layout scheme and collect real-time data;

[0043] a prediction analysis unit configured to perform temperature trend prediction based on the simulation data and the real-time data by using a BP neural network and a long short-term memory network, and construct a dynamic prediction model.

[0044] A digital twin platform unit adopts a digital twin platform based on WebGL technology to visualize and correct the temperature and stress changes inside and on the surface of the concrete, and outputs a corrected prediction result.

[0045] A control execution unit dynamically adjusts the water flow, water flow direction and water temperature of the cooling water pipe according to the corrected prediction result, so as to realize adaptive control of the concrete hydration heat.

[0046] In the above implementation manner, through the concrete hydration heat adaptive control system scheme based on digital twin, multiple technical effects are achieved. First, the system can predict and accurately grasp the temperature and temperature stress changes of the concrete in advance through accurate finite element model and real-time data monitoring, avoiding the hysteresis and inaccuracy of the traditional temperature control system. Secondly, the composite convolutional neural network optimizes the arrangement of the cooling water pipe and the sensor, improves the temperature control efficiency and system coverage, and enhances the uniformity of temperature control. Thirdly, the dynamic prediction model combining the BP neural network and the long short-term memory network effectively supports the active adjustment of the temperature control system and reduces the concrete cracks caused by excessive temperature difference. Finally, the real-time visualization and correction feedback mechanism of the digital twin platform ensures the accuracy and adaptive adjustment capability of the temperature control process, thereby improving the construction quality and engineering safety. The system significantly improves the precision, efficiency and reliability of temperature control management, and provides an intelligent solution for large-scale concrete construction.

[0047] Compared with the prior art, the beneficial effects of the present application are:

[0048] Through the concrete hydration heat adaptive control scheme based on digital twin technology, the present application achieves multi-level innovation effects. First, by constructing an accurate finite element model and simulating the concrete hydration heat, the present application achieves high-precision prediction of temperature and stress changes during concrete construction, thereby identifying potential structural risks in advance. Secondly, using a composite convolutional neural network to optimize the arrangement of cooling water pipes and sensors based on simulation data, the present application realizes more scientific and efficient resource allocation, significantly improving the efficiency of the cooling system. Thirdly, combining the BP neural network and the long short-term memory network, the present application dynamically predicts the trend of concrete temperature changes and constructs a prediction model, which makes the temperature control management more active and accurate, effectively preventing quality problems caused by temperature abnormalities. Finally, through the real-time data-driven digital twin platform, the present application not only visualizes the internal temperature and stress changes, but also adaptively adjusts the construction parameters according to the dynamic prediction results, ensuring the accuracy of temperature control and construction safety during the construction process, and significantly reducing structural problems caused by temperature. This all-round control strategy greatly improves the construction quality and ensures the long-term stability and safety of the structure. BRIEF DESCRIPTION OF DRAWINGS

[0049] Figure 1 A step schematic diagram of a concrete hydration heat self-adaptive control method based on digital twinning of embodiment 1 of the present application;

[0050] Figure 2 A method structure schematic diagram of a concrete hydration heat self-adaptive control method based on digital twinning of embodiment 2 of the present application;

[0051] Figure 3 An implementation flowchart of a concrete hydration heat self-adaptive control method based on digital twinning of embodiment 2 of the present application;

[0052] Figure 4 A system structure diagram of a concrete hydration heat self-adaptive control system based on digital twinning of embodiment 3 of the present application;

[0053] Figure 5 A cloud platform server display interface of embodiment 3 of the present application. DETAILED DESCRIPTION

[0054] The concrete hydration heat self-adaptive control method and system based on digital twinning provided by the present application will be described in detail below in combination with the drawings and specific embodiments. It should be emphasized that the embodiments are only for explaining and illustrating the technical content of the present application and should not be understood as limiting the protection scope of the present application. Any technical solution implemented based on the technical idea of the present application belongs to the protection scope of the present application. The advantages and features of the present application will be clearer in combination with the following description. It should be noted that the drawings are simplified and not to scale, which are only for the purpose of illustrating the embodiments of the present application. In the specific embodiments, how to implement the precise concrete temperature control process through the digital twinning platform according to the actual construction conditions, cooling water pipe arrangement and sensor arrangement will be shown in detail.

[0055] Embodiment 1

[0056] The applicant found in the research process that when using the traditional temperature control system, if trying to realize the instant temperature control adjustment in the concrete construction, it needs to rely on the artificial monitoring and subsequent manual adjustment and other inconvenient steps. In solving the actual engineering problems, in order to achieve the technical purpose of efficient and accurate temperature management, the prior art cannot meet the needs of real-time data processing and instant feedback. Therefore, after the applicant studied the problem, an adaptive control method based on digital twinning was proposed. When solving the technical problem of concrete temperature control, through the integration of finite element model, neural network prediction and digital twinning technical scheme, high-precision temperature prediction and real-time adaptive adjustment were realized, so as to achieve the technical effect of improving the structural safety and construction efficiency.

[0057] Please refer to Figure 1 , Figure 1A step schematic diagram of a concrete hydration heat self-adaptive control method based on digital twinning provided by the embodiments of the present application. The concrete hydration heat self-adaptive control method based on digital twinning can include:

[0058] S1. Construct a finite element model of concrete according to preset parameters and perform concrete hydration heat simulation prediction, and output simulation data of temperature and temperature stress;

[0059] S2. According to the simulation data, preset the arrangement optimization target of cooling water pipes and sensors, and generate the arrangement scheme of cooling water pipes and sensors based on the arrangement optimization target through a composite convolutional neural network;

[0060] S3. According to the arrangement scheme, arrange the cooling water pipes and sensors and collect real-time data; the real-time data include temperature data and temperature stress data;

[0061] S4. According to the simulation data and the real-time data, predict the concrete temperature change trend by using a BP neural network and a long short-term memory network, construct a dynamic prediction model of concrete temperature and temperature stress, and output a dynamic prediction result;

[0062] S5. Based on the dynamic prediction result and the real-time data, construct a digital twinning platform based on WebGL technology; the digital twinning platform is used for visualizing the temperature and temperature stress changes inside and on the surface of the concrete, correcting the dynamic prediction model, and outputting a corrected prediction result;

[0063] S6. According to the corrected prediction result output by the digital twinning platform, dynamically adjust the water flow rate, water flow direction and water temperature of the cooling water pipes to realize self-adaptive control of the concrete hydration heat.

[0064] In some embodiments, the preset parameters include concrete mix proportion, environmental parameters, condition parameters of concrete construction and maintenance, physical parameters and arrangement parameters of cooling water pipes; the environmental parameters include temperature, humidity and wind speed.

[0065] In the above implementation process, by considering the preset parameters such as concrete mix proportion, environmental parameters and construction conditions, the adaptability and accuracy of simulation prediction are enhanced, and the effect of optimizing the construction environment coping ability is achieved.

[0066] In some embodiments, step S1 includes constructing a concrete model by using WebGL technology, establishing a finite element simulation based thereon, and simulating and analyzing the change trend of the simulation temperature and temperature stress of concrete hydration heat.

[0067] In the implementation process, the continuity, accuracy and reliability of the simulation analysis are improved by adopting the entity modeling and the establishment of the finite element model based on the WebGL technology, so that the practicability of the simulation result is enhanced.

[0068] In some embodiments, step S2 can further include:

[0069] S21. According to the simulation data, preset the arrangement optimization target of the cooling water pipe and the sensor; the arrangement optimization target is to maximize the cooling effect and minimize the arrangement cost;

[0070] S22. Adopting a composite convolutional neural network to identify the distribution of temperature, and generating an optimal arrangement scheme of the cooling water pipe with the goal of maximizing the cooling effect;

[0071] S23. On the basis of the optimal arrangement of the cooling water pipe, a composite convolutional neural network is used to generate an arrangement scheme of the sensor.

[0072] In the above implementation, the cooling water pipe and the sensor arrangement are automatically optimized by the composite convolutional neural network, which maximizes the cooling effect and minimizes the cost, improves the construction efficiency and economy.

[0073] In some embodiments, step S23 generates an arrangement scheme of the sensor, which includes optimizing the number of sensors using a master-slave composite convolutional neural network; wherein the master path of the master-slave composite convolutional neural network processes temperature change characteristics, and the slave path processes humidity and concrete proportion change.

[0074] In the above implementation, the number and arrangement of sensors are optimized by the master-slave composite convolutional neural network, which realizes the accuracy and efficiency of temperature and humidity monitoring, and improves the comprehensiveness and accuracy of data monitoring.

[0075] In some embodiments, step S3 can include:

[0076] S31. According to the arrangement scheme, arrange a wireless sensor network;

[0077] S32. Collect real-time data of the concrete through the wireless sensor network, and backup the real-time data in a local storage device; the real-time data includes temperature data and temperature stress data;

[0078] S32. Adopting a cyclic redundancy check method to check the real-time data, and transmitting the checked real-time data to a cloud platform server.

[0079] In the above implementation, the real-time data collection and cyclic redundancy check of the wireless sensor network ensure the accuracy and real-time of the data, and achieve the effect of improving the reliability of data transmission.

[0080] In some embodiments, step S4 can include:

[0081] S41. Training a BP neural network according to the real-time data and the simulation data, and using the trained BP neural network to make a short-term prediction of the temperature of the concrete to obtain a short-term prediction result;

[0082] S42. Inputting the simulation data, the real-time data and the short-term prediction result into a long short-term memory network to output a dynamic prediction result of the temperature variation trend of the hydration heat of the concrete in real time.

[0083] In the above implementation manners, the combination of the BP neural network and the long short-term memory network is used to achieve accurate prediction of the temperature variation trend of the concrete, thereby achieving the effects of optimizing the temperature control strategy and improving the structural safety.

[0084] In some embodiments, step S41 further includes optimizing key parameters of the BP neural network using a cross-global artificial bee colony algorithm; the key parameters include input parameters, topological structure and initial weight.

[0085] In the above implementation manners, the key parameters of the BP neural network are optimized using the cross-global artificial bee colony algorithm, which improves the accuracy of temperature prediction, thereby achieving more precise temperature control management and reducing the structural risk caused by temperature.

[0086] In some embodiments, step S6 includes:

[0087] S61. Optimizing the water flow, water flow direction and water temperature of the cooling water pipe based on a non-dominated sorting genetic algorithm according to the real-time data and the dynamic prediction result;

[0088] S62. Adjusting the cooling zone parameters of the concrete, including the temperature rise rate, the temperature drop rate and the maximum temperature difference.

[0089] In the above implementation manners, the parameters of the cooling water pipe are dynamically adjusted using the non-dominated sorting genetic algorithm, which realizes adaptive refinement of the concrete temperature control, and achieves the effects of minimizing the risk of temperature cracks and improving the structural stability.

[0090] The concrete hydration heat adaptive control method based on digital twinning provided by the embodiments can be widely applied to multiple technical fields, such as temperature management of building engineering, infrastructure construction and large structures, and is particularly suitable for construction of key structures such as high-rise buildings and bridges. In the above implementation manners, when the concrete temperature is controlled, the temperature and stress data collected in real time can be adjusted in real time, the control strategy can be optimized by intelligent algorithms, and the temperature can be accurately adjusted, thereby effectively preventing structural cracks caused by temperature fluctuations and improving engineering safety and construction quality.

[0091] Embodiment 2

[0092] As an optimization of the foregoing embodiment, Embodiment 2 provides a specific implementation of a concrete hydration heat adaptive control method based on digital twinning,

[0093] Please refer to Figure 2 , Figure 2 is a structural schematic diagram of the method of the present application. As shown in the figure, the method includes a data acquisition layer, a data processing layer, and a control decision layer.

[0094] The data acquisition layer includes obtaining temperature and temperature stress data by using sensors for actual measurement and finite element data simulation, and uploading the data to a cloud platform server through wireless sensor network technology. During data transmission, the cyclic redundancy check method is used to check the data.

[0095] The data processing layer includes analyzing the data obtained from the data acquisition layer, including using the cross-global artificial bee colony algorithm, the BP neural network model, and the long short-term memory network model, and constructing a digital twinning modeling block using the BIM technology, the Unity3D technology, and the WebGL technology.

[0096] The control decision layer includes transmitting data in the digital twinning modeling block to the temperature control execution module, the control signal transmission module, and the feedback adjustment module. Finally, the visualization of the whole process of concrete hydration heat adaptive control can be achieved on the digital twinning platform.

[0097] Please refer to Figure 3 , Figure 3 is an implementation flowchart of a concrete hydration heat adaptive control method based on digital twinning, which specifically includes:

[0098] S1. Construction of a finite element model of concrete and simulation and prediction of hydration heat, including the following steps:

[0099] S11. First, set the concrete mix ratio, the environmental parameters calibrated in the hydration heat test (including environmental temperature, humidity, and wind speed), and consider the specific condition parameters in the concrete construction and curing process. In addition, key parameters such as boundary conditions of the model and element division also need to be set. These preset parameters provide necessary data support and physical basis for subsequent modeling and simulation.

[0100] S12. Construct a finite element model of the concrete based on the parameters preset in S11. The finite element model includes the cooling water pipes and related physical parameters such as shape and pipe diameter, as well as the specific heat capacity of the cooling water, and the flow rate, flow direction, and temperature of the water flowing into the cooling water pipes. This model is used to simulate the evolution of hydration heat temperature and temperature stress during the concrete pouring process. Through finite element analysis, the variation of temperature and temperature stress over time is obtained, which provides reference data for temperature control management and optimization.

[0101] S13. Based on the finite element model, use an ensemble learning method to correct the simulation data. By integrating the calculation results under different parameters and considering more factors that may affect the concrete hydration heat process (such as changes in construction environment and differences in material properties), the accuracy and adaptability of the simulation are improved.

[0102] S2. Optimize the arrangement of cooling water pipes and sensors based on a composite convolutional neural network:

[0103] S21. Based on the simulation data in S1, use a composite convolutional neural network to automatically identify the distribution of temperature, and adjust the arrangement parameters to maximize the cooling effect. Specific parameters include pipe diameter, number of layers, arrangement direction, and horizontal and vertical spacing. Through network calculation optimization, the best arrangement scheme of cooling water pipes is automatically selected and generated, providing accurate guidance for subsequent construction.

[0104] S22. Based on the optimized cooling water pipe arrangement in S21, use a composite convolutional neural network to further optimize the arrangement of sensors. This network can identify key feature parameters to ensure that the arrangement of sensors can comprehensively and accurately monitor the changes in internal and surface temperature and temperature stress of concrete during the hydration heat process. The main optimization is the number and location of sensors to ensure real-time data coverage of the entire construction area with high data acquisition accuracy.

[0105] S3. Design a monitoring system based on Internet of Things technology to transmit temperature and temperature stress data to a cloud platform server in real time, including the following steps:

[0106] S31. According to the optimization scheme generated in step S2, install cooling water pipes and sensors in the actual structure to ensure that the pipe position and parameters are consistent with the optimization results. Use wireless sensor network (WSN) technology to collect and transmit real-time internal and surface temperature and temperature stress data of concrete. To ensure the efficiency and stability of data transmission, high-efficiency data compression technology is used for transmission. The sensors automatically record data and backup in local storage devices to ensure data integrity. At the same time, the sensor network is equipped with a backup power supply to ensure long-term stable operation.

[0107] S32. To ensure the accuracy and reliability of the data, a cyclic redundancy check (CRC) method is used to perform integrity check on the transmitted data. The data that has passed the check will be transmitted to the cloud platform server in real time, realizing the centralized management and real-time monitoring of large-scale data. The monitoring system provides reliable raw data support for subsequent dynamic prediction and adaptive control.

[0108] S4. Construct a dynamic prediction model for concrete hydration heat temperature and temperature stress:

[0109] S41. Based on the real-time collected temperature data, temperature stress data and simulation data of concrete in step S3, a database of temperature and temperature stress is constructed for the training of BP neural network. This database provides diversified and accurate data support for the network, which helps to train a high-precision prediction model.

[0110] S42. By introducing cross-global artificial bee colony algorithm, the key parameters of BP neural network are optimized, including input parameters, topology structure, initial weight threshold. This optimization process improves the efficiency and prediction accuracy of model training through global search and local adjustment.

[0111] S43. Construct the optimized BP neural network and perform deep learning and short-term prediction to obtain the short-term prediction result. Through this model, the temperature and stress changes during the concrete hydration heat process are predicted in real time, and the prediction result will be an important basis for the automatic control system instruction release.

[0112] S44. Collect the hydration heat temperature data of mass concrete and establish the concrete temperature change curve. Input the concrete temperature change curve, short-term prediction result, real-time data and simulation data into the long short-term memory network (LSTM) model to construct a dynamic prediction model for concrete hydration heat temperature driven by physical and data. This dynamic prediction model continuously monitors the change trend of temperature and temperature stress during the concrete hydration heat process and outputs the dynamic prediction result in real time. This model provides the core prediction ability for the digital twin platform and supports the optimization of subsequent control strategy.

[0113] S5. Construct a digital twin platform for concrete hydration heat based on WebGL technology:

[0114] S51. Use Abaqus software to construct a finite element model of concrete to simulate the temperature and temperature stress changes generated during the hydration heat process of concrete. This model provides basic data support for the subsequent digital twin platform.

[0115] S52. The finite element model information (geometry, node and element information) and simulation data (node temperature field and stress field) generated by Abaqus are imported into Unity3D to generate the corresponding three-dimensional visual model. The visualization interface of the digital twin platform is built using Unity3D, ensuring that the digital twin platform can accurately reflect the temperature and temperature stress distribution of the concrete hydration heat.

[0116] S53. Based on WebGL technology, a digital twin monitoring platform is developed to combine dynamic prediction models with real-time monitoring data, achieving the following functions: comprehensive monitoring of temperature and temperature stress changes inside and on the surface of concrete; based on real-time monitoring data, continuously correcting the dynamic prediction model to improve prediction accuracy and reliability. This platform can effectively ensure the consistency of the prediction results with the actual situation, providing a scientific basis for adaptive control. The built digital twin platform can realize real-time display of hydration heat temperature and temperature stress changes. Based on WebGL technology, the platform combines big data collection and analysis of concrete temperature control, artificial intelligence early warning, virtual reality interaction, and visualization technology to transmit real-time monitoring data to the WebGL-based digital twin platform. In the digital twin platform, users can locate the position of the sensors and view various types of sensor temperature, stress historical information, generate trend curves, inner-outer temperature difference distribution graphs, inner-outer stress distribution graphs, and other visual charts.

[0117] S54. In addition, real-time data is compared with the prediction results of the dynamic prediction model built in step S4 in real time, and the prediction model is corrected according to the real-time data, so that the virtual model in the digital twin platform is consistent with the temperature and temperature stress distribution state of the physical model of the actual concrete pouring. The dynamic correction refers to the comparison of real-time data and prediction model to adjust and optimize the digital model, making it more accurately reflect the actual situation of concrete hydration heat. Through this correction process, adaptive control of the physical model is realized, ensuring the accuracy and stability of concrete temperature control.

[0118] S6. Dynamically control the cooling water pipe parameters to accurately adjust the concrete temperature.

[0119] S61. According to the monitoring and dynamic prediction results of the digital twin platform, combined with the measured data (concrete temperature, stress, temperature rise and fall rate, inner-outer temperature difference, maximum temperature difference, environment temperature, inlet temperature, etc.), comprehensive calculation and analysis are carried out. Through multi-parameter evaluation, the temperature variation law in the hydration heat process is comprehensively mastered, providing a basis for the optimization of subsequent control strategies.

[0120] S62. The Non-dominated Sorting Genetic Algorithm II (NSGA-II) optimization control method is adopted to dynamically adjust the water flow rate, water flow direction, and water temperature of the cooling water pipes, as well as other control parameters such as the pouring process parameters based on the calculation and analysis results. This achieves precise control of the concrete hydration heat, including: dynamically adjusting the flow rate, flow speed, and water temperature of the cooling water to adapt to the temperature control needs at different stages; optimizing the thickness and arrangement of the insulation materials based on the prediction results; adjusting the pouring speed or sequence if necessary to ensure uniform temperature distribution. This optimization process maximizes the reduction of temperature rise rate, ensures appropriate cooling rate, and balances the temperature difference between the interior and surface of the concrete to improve temperature control accuracy. Through precise control, the risk of temperature cracks is minimized, ensuring the quality and long-term stability of the concrete structure.

[0121] Through the above steps, an adaptive and dynamically adjusted control system is formed, which effectively responds to the temperature control needs at different stages and achieves temperature control and stress optimization during the concrete hydration heat process through fine adjustment.

[0122] In some embodiments, the method further includes conducting on-site analysis of the thermal physical properties of concrete at typical work sites to effectively address the differences between the reference values in current specifications and actual conditions. Typical work sites such as plain areas, plateaus, high-cold regions, and coastal areas are selected as research objects to ensure that the system's application can build a good experimental and theoretical foundation, thereby improving its adaptability and practicality in different construction environments.

[0123] In some embodiments, in step S13, the embodiments of the present application use ensemble learning technology to optimize the finite element hydration heat analysis model. Most existing model correction methods are deterministic methods that do not fully consider the uncertainty of structural parameters and responses, resulting in insufficient accuracy in actual engineering applications. Given the widespread uncertainty in the concrete pouring process, especially in structural parameters and responses, the finite element model correction method based on meta-model and base model selector can select the most suitable model through multiple model comparisons for hydration heat numerical simulation analysis.

[0124] The specific process is as follows: First, organize and preprocess the data of construction environmental conditions, concrete mix proportion, physical parameters and arrangement parameters of cooling water pipes, and concrete physical parameters, etc. to ensure the accuracy of modeling. Then, input environmental parameters, material parameters, boundary conditions, etc. into the finite element software to obtain multiple base models. Optimize these base models through ensemble learning technology to form meta-models and select the best model configuration. Finally, set the simulation parameters and run the simulation to predict the temperature and temperature stress trend of hydration heat to evaluate the efficiency of the cooling system.

[0125] In some embodiments, in step S21, the embodiments of the present application use a composite convolutional neural network to predict the arrangement of cooling water pipes. An optimization function is constructed to sequentially iterate single-objective optimization to maximize cooling effect and minimize cost. First, the maximum cooling effect is taken as the optimization objective, and the temperature peak and gradient are taken as the constraints to perform single-objective optimization, and the optimized cooling water pipe arrangement is obtained; then, under the premise of satisfying the local optimal solution of the cooling effect, the minimum arrangement result is taken as the optimization objective to obtain the optimal number combination of the number of cooling water pipes or sensors, while ensuring the monitoring accuracy.

[0126] By collecting the arrangement density, length, radial distribution and other parameters of the water pipes, and combining the relevant data generated by simulation (such as the temperature distribution, cooling effect and cost corresponding to different water pipe arrangements), the water pipe arrangement parameters are taken as the input of the neural network, and the corresponding cooling effect and cost are output.

[0127] Through training the neural network, the network can learn the relationship between water pipe arrangement and cooling effect and cost. Finally, according to the specific engineering requirements, the network structure or optimization algorithm parameters are adjusted to adapt to different engineering conditions and targets, and the optimal solution between cooling effect and cost is finally obtained. The multi-objective optimization method based on composite convolutional neural network not only can realize the best balance between cost and effect, but also can flexibly cope with complex environment, and provide efficient and reliable optimization scheme.

[0128] In some embodiments, in step S21, the sensor number optimization adopts a primary and secondary convolutional neural network (PasNet) network structure. The input features of the network include environmental temperature, humidity, concrete mix ratio and other data, which are preprocessed (such as normalization processing) and divided into training set and test set for training and verification of the neural network.

[0129] The “primary” path in the network structure is specially designed to handle key features (such as temperature changes), while the “secondary” path handles auxiliary features (such as humidity or concrete mix ratio changes). Through convolution operation, the network can extract and fuse these features to enhance the learning ability of complex patterns in data. Through model analysis, the sensor positions that have the greatest impact on the prediction results are identified, and the number and position of the sensors are optimized accordingly, so as to improve the efficiency and economy of the monitoring system.

[0130] Finally, through training and verification, the network parameters (such as learning rate and convolution kernel size) are adjusted to optimize the prediction accuracy of the model, and ensure that the network shows high efficiency and reliability in the sensor number optimization process.

[0131] In some embodiments, in step S42, the application introduces a cross-global artificial bee colony algorithm to optimize the key parameters of the hydration heat model. This algorithm takes concrete temperature data, simulation data, and numerical simulation data from step S1 as input features. By cross-operation to enhance population diversity and avoid local optimization, the prediction accuracy of the model is improved.

[0132] The optimized parameters will be used to run the hydration heat analysis model to simulate the effects of different environmental conditions and concrete mix proportions on the temperature drop process and temperature stress. Based on the model analysis results, precise temperature control standards are developed, including the control of base temperature difference, interlayer temperature difference, and surface-internal temperature difference, to prevent structural problems such as concrete cracking. Through further data analysis, an ideal temperature drop curve is determined, and measures such as adjusting the pouring time, modifying the mix ratio, or using specific cooling techniques are proposed to ensure project quality and reduce environmental and economic costs.

[0133] In some embodiments, in step S43, a BP neural network is used for deep learning and short-term prediction. The BP neural network contains multiple hidden layers, each of which can learn different levels of abstraction of data, gradually building complex data representations. In particular, in concrete hydration heat prediction, the deep learning network can surpass traditional statistical or machine learning methods to accurately predict the trend of hydration temperature changes.

[0134] By collecting concrete temperature change curves and combining simulation data, numerical simulation (physical model), and real-time data, a dynamic prediction model driven by physical and data is constructed using deep learning algorithms. This model automatically learns the high-level features of thermal dynamics and chemical reactions in the concrete hydration process through a hierarchical feature extraction mechanism, providing real-time and accurate temperature change predictions as an important basis for the automatic control system to issue instructions, thereby realizing the transition from passive control to active control.

[0135] In some embodiments, in step S5, the digital twin platform is built based on big data, cloud computing, the Internet of Things, and digital twin technology to build a realistic virtual concrete digital twin platform. This platform uses simulation analysis data and composite convolutional neural networks to optimize the real-time adjustment of sensor and cooling water pipe placement, while achieving intelligent diagnosis, control, and disposal of equipment and facilities. This digital twin platform can comprehensively monitor the mass concrete construction process, providing full-life-cycle monitoring, monitoring, and intelligent decision support to ensure construction quality and project safety.

[0136] In some embodiments, in the S6 step, the temperature control scheme is automatically generated in combination with the monitoring and dynamic prediction results of the digital twin platform. Through Latin hypercube sampling (LHS), based on the real-time data of the concrete mix proportion and hydration heat under different conditions, a temperature control sample database is constructed, a proxy model is constructed, and a cooling water pipe layout scheme, a sensor layout scheme and a water flow speed control scheme are automatically generated.

[0137] In combination with application cases under different environmental conditions, the model can provide a refined temperature control design scheme and output the corresponding design diagram, further optimizing the concrete temperature control effect and ensuring the temperature control accuracy and safety during construction.

[0138] The embodiments of the present application are based on digital twin technology, which realizes the fusion and interaction between virtual space and physical space, and between various elements of concrete temperature control, promotes the landing of intelligent temperature control in advanced form, directly serves the concrete temperature control and crack prevention work, and transforms the passive management mode of temperature control relying on subjective judgment and behavior of "people" into an active management mode based on data mapping of digital twin technology to prevent temperature cracks. In the concrete design and construction process, all-weather and refined online monitoring and early warning can be realized, temperature abnormal conditions can be found and responded to in a timely manner, and the informatization level of concrete maintenance work is improved.

[0139] Embodiment 3

[0140] The embodiments of the present application provide a concrete hydration heat adaptive control system based on digital twin, as shown in Figure 4 , comprising:

[0141] A model construction unit is configured to construct a finite element model of concrete according to preset parameters and perform hydration heat simulation, and output simulation data of temperature and temperature stress;

[0142] A layout generation unit is configured to generate a cooling water pipe and sensor layout scheme according to simulation data using a composite convolutional neural network;

[0143] A real-time monitoring unit is configured to arrange cooling water pipes and sensors according to the layout scheme and collect real-time data;

[0144] A prediction analysis unit is configured to perform temperature trend prediction based on simulation data and real-time data using a BP neural network and a long short-term memory network, and construct a dynamic prediction model;

[0145] A digital twin platform unit is configured to use a digital twin platform based on WebGL technology to visualize and correct temperature and stress changes inside and on the surface of concrete, and output corrected prediction results;

[0146] The control execution unit dynamically adjusts the water flow rate, water flow direction and water temperature of the cooling water pipe according to the corrected prediction result, so as to realize adaptive control of the hydration heat of the concrete.

[0147] As shown in the system composition Figure 4 , including water pump, high-power water temperature heating pipe, wireless temperature receiving controller, wireless temperature collector, concrete temperature probe and cold water temperature probe and other components. Through these key equipment, the real-time collected temperature data, stress data and other information will be automatically transmitted to the cloud platform server for centralized management and processing. Specifically, Figure 4 The working principle of the cooperation between each component is shown - the water temperature heating pipe directly affects the hydration heat process of the concrete by adjusting the water flow and temperature, and the arrangement scheme of the wireless temperature receiving controller and the sensor ensures the comprehensiveness and accuracy of data acquisition.

[0148] After the data is processed by the cloud platform server, the display interface as shown in Figure 5 , can real-time present the temperature change trend of each area, stress distribution graph and adjustment of cooling water pipe parameters. Through this interface, users can intuitively monitor the system, and correct the prediction model according to real-time data to ensure the continuous optimization of the system. The real-time and intelligent decision-making ability of this platform not only improves the response speed of the temperature control system, but also provides comprehensive data support for subsequent analysis and optimization, greatly improving the efficiency of large-scale data management.

[0149] Through the combination of Figure 4 and Figure 5 , the system realizes the whole process management from data acquisition to real-time monitoring to intelligent optimization. The digital twin platform integrates physical data with virtual space, realizes the efficient operation of the temperature control system, and provides intelligent and automated technical support for the control of concrete hydration heat.

[0150] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects.

[0151] The technical features of the above-described embodiments can be combined arbitrarily. To make the description concise, all possible combinations of the technical features in the above-described embodiments are not described, however, as long as the combination of the technical features does not exist contradictory, it should be considered as the scope of the present disclosure.

[0152] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent. It should be noted that for ordinary skilled persons in the art, without departing from the concept of the present application, several modifications and improvements can be made, which are within the scope of protection of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. A digital-twin-based self-adaptive control method for concrete hydration heat, characterized in that, The method comprises the following steps: S1. Constructing a finite element model of concrete according to preset parameters, performing hydration heat simulation prediction of concrete, and outputting simulation data of temperature and temperature stress; S2. According to the simulation data, preset the arrangement optimization target of cooling water pipes and sensors, and generate the arrangement scheme of cooling water pipes and sensors based on the arrangement optimization target through a composite convolutional neural network, comprising: S21. According to the simulation data, preset the arrangement optimization target of cooling water pipes and sensors; the arrangement optimization target is to maximize the cooling effect and minimize the arrangement cost; S22. Using a composite convolutional neural network to identify the distribution of temperature, and generating the optimal arrangement scheme of cooling water pipes with the goal of maximizing the cooling effect; S23. On the basis of the optimal arrangement of cooling water pipes, a composite convolutional neural network is used to generate the arrangement scheme of sensors; S3. According to the arrangement scheme, arrange the cooling water pipes and sensors and collect real-time data; the real-time data includes temperature data and temperature stress data; S4. According to the simulation data and the real-time data, predict the temperature change trend of concrete by using BP neural network and long short-term memory network, and construct a dynamic prediction model of concrete temperature and temperature stress, and output the dynamic prediction result; S5. Based on the dynamic prediction result and the real-time data, a digital twin platform based on WebGL technology is constructed; the digital twin platform is used to visualize the temperature and temperature stress changes inside and on the surface of the concrete, and correct the dynamic prediction model, and output the corrected prediction result; S6. According to the corrected prediction result output by the digital twin platform, dynamically adjust the water flow, water flow direction and water temperature of the cooling water pipes to realize the adaptive control of the hydration heat of concrete, comprising: S60. According to the monitoring and dynamic prediction results of the digital twin platform, combined with the measured data, including concrete temperature, temperature stress, temperature rise rate, temperature drop rate, inner and outer temperature difference, maximum temperature difference, environment temperature and inlet temperature, comprehensive calculation and analysis are carried out, and the temperature change law in the hydration heat process is evaluated by multiple parameters, which provides basis for the optimization of control strategy; S61. According to the real-time data and the dynamic prediction result, the water flow, water flow direction and water temperature of the cooling water pipes are optimized based on the non-dominated sorting genetic algorithm; S62. Adjust the cold zone parameters of concrete, including temperature rise rate, temperature drop rate and maximum temperature difference.

2. The concrete hydration heat self-adaptive control method based on digital twinning according to claim 1, characterized in that, The preset parameters include concrete mix proportion, environmental parameters, condition parameters of concrete construction and maintenance, physical parameters and arrangement parameters of cooling water pipes; the environmental parameters include temperature, humidity, wind speed.

3. The concrete hydration heat self-adaptive control method based on digital twinning according to claim 2, characterized in that, Step S1 includes using WebGL technology to construct a concrete model, on the basis of which a finite element simulation is established to simulate and analyze the change trend of simulation temperature and temperature stress of concrete hydration heat.

4. The concrete hydration heat self-adaptive control method based on digital twinning according to claim 3, characterized in that, Step S23 generates the arrangement scheme of sensors, which includes optimizing the number of sensors by using a master-slave composite convolutional neural network; wherein the master path of the master-slave composite convolutional neural network processes temperature change characteristics, and the slave path processes humidity and concrete mix proportion change.

5. The concrete hydration heat self-adaptive control method based on digital twinning according to claim 4, characterized in that, Step S3 includes: S31. Arranging a wireless sensor network according to the arrangement scheme; S32. Collecting real-time data of the concrete by the wireless sensor network and backing up the real-time data in a local storage device; the real-time data includes temperature data and temperature stress data; S33. Checking the real-time data by using a cyclic redundancy check method, and transmitting the checked real-time data to a cloud platform server.

6. The concrete hydration heat self-adaptive control method based on digital twinning according to claim 5, characterized in that, Step S4 includes: S41. Training a BP neural network according to the real-time data and the simulation data, and using the trained BP neural network to make a short-term prediction of the temperature of the concrete to obtain a short-term prediction result; S42. Inputting the simulation data, the real-time data and the short-term prediction result into a long short-term memory network to output a dynamic prediction result of the temperature change trend of the hydration heat of the concrete in real time.

7. The concrete hydration heat self-adaptive control method based on digital twinning according to claim 6, characterized in that, Step S41 further includes optimizing key parameters of the BP neural network by using a cross-global artificial bee colony algorithm; the key parameters include input parameters, topological structure and initial weight. 8.A digital-twin-based self-adaptive control system for concrete hydration heat, characterized in that, The system can implement the method of any one of claims 1 to 7 when running, including: a model construction unit configured to construct a finite element model of the concrete according to preset parameters and perform hydration heat simulation to output simulation data of temperature and temperature stress; an arrangement generation unit configured to generate an arrangement scheme of the cooling water pipe and the sensor according to the simulation data by using a composite convolutional neural network; a real-time monitoring unit configured to arrange the cooling water pipe and the sensor according to the arrangement scheme and collect real-time data; a prediction analysis unit configured to make a temperature trend prediction based on the simulation data and the real-time data by using a BP neural network and a long short-term memory network, and construct a dynamic prediction model; a digital twin platform unit configured to use a digital twin platform based on WebGL technology to visualize and correct changes in temperature and stress inside and on the surface of the concrete, and output a corrected prediction result; a control execution unit configured to dynamically adjust the water flow rate, water flow direction and water temperature of the cooling water pipe according to the corrected prediction result to achieve adaptive control of the hydration heat of the concrete.

Images