Freezing method construction monitoring method based on physical simulation and memory network

Abstract

The invention discloses a freezing method construction monitoring method based on physical simulation and a memory network. Comprising the following steps: constructing a comsol large-scale physical finite element model under different construction conditions of the freezing method to obtain temperature distribution, freezing wall thickness, soil frost heaving displacement and soil stress distribution of a construction area in the construction process of the freezing method under different environmental factor conditions at different time points through simulation calculation; taking the temperature and the soil frost heaving displacement of the construction area as input characteristics of the machine learning model, and taking the soil stress and the frozen wall thickness of the construction area as output labels of the machine learning model; acquiring soil parameters of the freezing method construction area to be monitored in real time; based on the soil parameters, the soil stress and the frozen wall thickness of the freezing method construction area are predicted in real time through the trained machine learning model; and when the soil stress and the frozen wall thickness exceed the early warning threshold range, generating an early warning notification. The problem that an existing freezing method construction monitoring method is low in efficiency is solved.

Description

Technical Field

[0001] The invention relates to the technical field of building construction, and in particular to a freezing construction monitoring method based on physical simulation and memory network. Background Art

[0002] In the field of civil engineering, freezing construction technology has been valued for its effective support in underground projects. This technology forms a frozen soil wall in the soil through freezing action, which serves as a temporary support structure for excavation operations. Although freezing construction technology has achieved certain success in practice, it still has shortcomings in monitoring and controlling the construction process. Traditional monitoring methods mainly rely on manual measurement and regular inspections. These methods are not only inefficient, but also cannot achieve continuous real-time monitoring of the construction process, resulting in the inability to timely discover and respond to problems that may arise during the construction process. In addition, these traditional methods also have limitations in data collection and analysis, making it difficult to fully utilize the large amount of data generated during the construction process, especially time series data, which limits the ability of the model to predict construction stability.

[0003] With the development of computer simulation technology, physical simulation methods have been introduced into the monitoring of freezing construction to simulate key parameters such as surface frost heave changes, frozen wall thickness, and frozen soil expansion. However, these simulation models are often based on theoretical assumptions and lack the combination with actual construction data, which leads to deviations between the simulation results and the actual situation. In addition, although machine learning technology has shown great potential in the field of construction safety monitoring, most of the existing machine learning models are trained for specific construction conditions and lack the ability to generalize under variable geological conditions and environmental factors. These models can usually only process a single type of data and fail to fully consider the interactive effects of multiple factors during the construction process, resulting in limited effectiveness in practical applications.

[0004] The information disclosed in this background technology section is only intended to enhance the understanding of the overall background of the invention and should not be regarded as an acknowledgment or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art. Summary of the invention

[0005] In order to overcome the defects of the prior art, a freezing method construction monitoring method based on physical simulation and memory network is provided to solve the problem of low efficiency of the existing freezing method construction monitoring method.

[0006] To achieve the above purpose, a freezing method construction monitoring method based on physical simulation and memory network is provided, comprising the following steps:

[0007] Construct the comsol large-scale physical finite element model under different construction conditions of the freezing method;

[0008] The temperature distribution, frozen wall thickness, soil frost heave displacement and soil stress distribution of the construction area during the freezing method construction process under different environmental factors at different time points are obtained by simulation and calculation of the Comsol large-scale physical finite element model;

[0009] A long short-term memory network (LSTM) sequential model is used as a machine learning model, and the temperature and soil frost heave displacement of the construction area are used as input features of the machine learning model, and the soil stress and frozen wall thickness of the construction area are used as output labels of the machine learning model to train the machine learning model;

[0010] Real-time acquisition of soil parameters in the freezing method construction area to be monitored;

[0011] Based on the soil parameters, the soil stress and frozen wall thickness of the freezing method construction area are predicted in real time by the trained machine learning model;

[0012] When the real-time predicted soil stress and frozen wall thickness in the freezing method construction area exceed the warning threshold range, a warning notification is generated.

[0013] Furthermore, the method also includes displaying the soil parameters of the freezing method construction area to be monitored in real time through a Web page and predicting the soil stress and frozen wall thickness of the freezing method construction area in real time.

[0014] Furthermore, after the warning notification is generated, the warning notification is notified to the monitoring personnel via SMS, email or telephone.

[0015] The beneficial effects of the present invention are that the freezing construction monitoring method based on physical simulation and memory network of the present invention uses physical simulation methods to simulate the entire freezing construction process under specific geological conditions as training data for a machine learning model; establishes a machine learning model, and uses a deep learning method to establish a machine learning model that can predict soil stress and frozen wall thickness based on input surface frost heave displacement and temperature conditions; establishes a real-time early warning platform, and uses the machine learning model to establish a platform that can monitor and warn in real time as a tool for management and maintenance of construction stability.

[0016] The freezing method construction monitoring method based on physical simulation and memory network of the present invention realizes a rapid and accurate evaluation of the stability of freezing method construction by combining the accuracy of physical simulation and the efficiency of machine learning, and provides a scientific basis for the long-term maintenance and management of construction stability. The freezing method construction monitoring method based on physical simulation and memory network of the present invention not only improves the real-time and accuracy of construction monitoring, but also enhances the generalization ability of the model, can adapt to the changing construction environment, and ensures the safety and efficiency of the construction process. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings:

[0018] Figure 1 Schematic diagram of the arrangement of monitoring points for freezing method construction according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the relevant invention, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the invention are shown in the accompanying drawings.

[0020] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0021] The present invention provides a freezing method construction monitoring method based on physical simulation and memory network, comprising the following steps:

[0022] S1. Construct a large-scale physical finite element model of Comsol under different construction conditions of the freezing method.

[0023] The soil and soil thermal parameters under different geological conditions are collected. Through the synthesis of experimental data and literature, the parameters that can reflect the soil shear strength characteristics, hydraulic characteristics and other parameters in typical soils are obtained. The necessary parameter settings will be provided for the construction of the Comsol large-scale physical finite element model in the future to ensure the accuracy of the numerical simulation.

[0024] Construct a large-scale physical finite element model of Comsol under different construction conditions. The different construction conditions specifically include:

[0025] Geological conditions, such as soil type, density, water content, permeability, etc.;

[0026] Environmental factors, such as temperature, humidity, wind speed and other climatic conditions;

[0027] Material properties, thermophysical properties of freezing pipe materials and coolants, etc.

[0028] Construction progress and construction stages, such as the initial, middle and late stages of freezing.

[0029] Considering the situations of different construction times, the stress conditions of the frozen wall and the surrounding soil in the finite element model over time can be calculated, and the initial conditions required for setting the finite element model are provided under the specific construction background.

[0030] Construct physical models under different construction conditions, consider different construction parameters, and construct geometric models and initial conditions of the entire construction area. When constructing the Comsol large-scale physical finite element model, the entire construction area is comprehensively considered by the constructed geometric model to ensure that the model can accurately simulate the actual construction conditions and processes. Through comprehensive analysis, accurate training data is provided for subsequent machine learning models, thereby achieving effective monitoring and early warning of the freezing method construction process.

[0031] S2. The temperature distribution, frozen wall thickness, soil frost heave displacement and soil stress distribution in the construction area during the freezing method construction process under different environmental factors at different time points were obtained through simulation and calculation of the Comsol large-scale physical finite element model.

[0032] The established Comsol large-scale physical finite element model can be used to simulate the freezing construction process under different environmental factors. By calculating the freezing wall handover time, the influence of different environmental factors on the construction time can be considered. The Comsol large-scale physical finite element model can be used to simulate the soil frost heave displacement and frozen soil temperature during the construction process.

[0033] Comsol large-scale physical finite element model outputs parameters such as temperature distribution, surface frost heave displacement and soil stress distribution during the construction process. Based on the simulation results, the temperature distribution, frozen wall, surface frost heave displacement and soil stress distribution in the construction area at different time points are calculated. This process is the establishment and calculation of the physical model.

[0034] S3. Use the long short-term memory network (LSTM) sequence model as the machine learning model, use the temperature and soil frost heave displacement in the construction area as the input features of the machine learning model, and use the soil stress and frozen wall thickness in the construction area as the output labels of the machine learning model to train the machine learning model.

[0035] Specifically, step S3 includes:

[0036] Step S31: Determine the input features, and select the temperature during the construction process, the ground surface frost heave displacement, etc. as the input features.

[0037] Step S32: Definition of output labels, using the soil stress and frozen wall thickness parameters obtained in the numerical simulation as output labels, that is, the target prediction values ​​of the model.

[0038] Step S33: Model selection, using the long short-term memory network LSTM sequence model as the machine learning model to consider the impact of time series information in construction stability assessment.

[0039] The long short-term memory network LSTM is used as a machine learning model to consider the impact of time series information in stability assessment. The long short-term memory network LSTM is a recurrent neural network RNN ​​variant in deep learning. The modeling effect of time series data is better. The long short-term memory network LSTM can effectively process data at different time points in the assessment. The long short-term memory network LSTM sequence model is used as a machine learning model to consider the impact of time series information in construction stability assessment.

[0040] Step S34: Model training, using the data set generated by the numerical simulation to train the model so as to adjust the model parameters to adapt to the actual construction conditions.

[0041] Step S35: Model verification and optimization: verify the model through the verification set, adjust and optimize the structure and parameters of the model to improve its generalization ability and prediction accuracy.

[0042] S4. Real-time acquisition of soil parameters of the freezing method construction area to be monitored.

[0043] See also Figure 1 , by setting up monitoring points (ground monitoring points, stress monitoring points and pipeline monitoring points, etc.) in the freezing method construction area to be monitored for data access and integration, real-time monitoring data can be obtained from the monitoring points.

[0044] S5. Based on soil parameters, the soil stress and frozen wall thickness in the freezing method construction area are predicted in real time through the trained machine learning model.

[0045] The machine learning model performs real-time predictions, using the trained long short-term memory network (LSTM) time series model to make real-time predictions on the construction stability of data acquired in real time.

[0046] When the real-time predicted soil stress and frozen wall thickness in the freezing method construction area exceed the warning threshold range, a warning notification is generated.

[0047] According to the warning rules and threshold settings, based on the risk rules and warning thresholds of the prediction results, by comparing the real-time data with the set thresholds, it is determined whether the risk is triggered to generate a warning notification.

[0048] By continuously monitoring changes in construction parameters and updating prediction results in real time, a notification system is built. Once a potential risk to construction stability is detected, an early warning notification is immediately sent through the notification system so that timely measures can be taken.

[0049] The soil parameters of the freezing method construction area to be monitored are displayed in real time through the Web page, and the soil stress and frozen wall thickness of the freezing method construction area are predicted in real time.

[0050] In this embodiment, a visual interface is constructed to display real-time data and reports, and real-time monitoring data and model prediction results are presented intuitively, so that relevant personnel can clearly understand the construction status, generate regular reports, and record historical data and early warning conditions of construction stability for post-analysis and improvement of system performance.

[0051] The freezing method construction monitoring method based on physical simulation and memory network of the present invention can realize accurate assessment and early warning of regional stability, improve the safety and sustainability of the project, and has higher prediction accuracy and stronger generalization ability, especially maintaining good performance under freezing method construction.

[0052] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the inventive concept. For example, the above features are replaced with (but not limited to) technical features with similar functions disclosed in the present application.

Claims

1. A freezing construction monitoring method based on physical simulation and memory network, characterized in that: The following steps are involved: Construct the comsol large-scale physical finite element model under different construction conditions of the freezing method; The temperature distribution, frozen wall thickness, soil frost heave displacement and soil stress distribution of the construction area during the freezing method construction process under different environmental factors at different time points are obtained by simulation and calculation of the Comsol large-scale physical finite element model; A long short-term memory network (LSTM) sequential model is used as a machine learning model, and the temperature and soil frost heave displacement of the construction area are used as input features of the machine learning model, and the soil stress and frozen wall thickness of the construction area are used as output labels of the machine learning model to train the machine learning model; Real-time acquisition of soil parameters in the freezing method construction area to be monitored; Based on the soil parameters, the soil stress and frozen wall thickness of the freezing method construction area are predicted in real time by the trained machine learning model; When the real-time predicted soil stress and frozen wall thickness in the freezing method construction area exceed the warning threshold range, a warning notification is generated.

2. The freezing method construction monitoring method based on physical simulation and memory network according to claim 1 is characterized in that: It also includes real-time display of soil parameters of the freezing method construction area to be monitored through a Web page and real-time prediction of soil stress and frozen wall thickness in the freezing method construction area.

3. The freezing method construction monitoring method based on physical simulation and memory network according to claim 1 is characterized in that: After the warning notification is generated, the warning notification is notified to the monitoring personnel via SMS, email or telephone.

Images