Foundation pit tunnel engineering safety risk early warning detection and judgment method

Abstract

The invention discloses a foundation pit tunnel engineering safety risk early-warning measurement and judgment method, and relates to the technical field of tunnel risk early warning, and the method comprises the following steps: S1, the foundation pit tunnel engineering safety risk early-warning measurement and judgment method is applied to the safety risk measurement and judgment of a subway tunnel engineering underneath passing existing line project, and the foundation pit tunnel engineering is divided into sections; the method comprises the following steps: constructing a tunnel section, collecting ground surface settlement data of the constructed tunnel section, preprocessing the collected ground surface settlement data of the tunnel, and constructing a ground surface settlement database based on the preprocessed ground surface settlement data; the invention provides a foundation pit tunnel engineering safety risk early-warning measurement and judgment method which is applied to measurement and judgment of the safety risk of a subway tunnel engineering underneath passing existing line project, and comprises the following steps: performing section division on a tunnel, and performing ground surface settlement data acquisition on a constructed tunnel section in the tunnel; and the collected ground surface settlement data is preprocessed and a ground surface settlement database is constructed.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel risk early warning, and in particular to a method for early warning and determining safety risks of foundation pit tunnel engineering. Background Art

[0002] In recent years, my country's railway construction has tended to be located in complex and dangerous areas such as plateaus and high-altitude areas. In particular, tunnel projects have the characteristics of deep burial, super-long, and complex geology. In the Chinese patent application number 202310453471.7, "A tunnel construction safety risk intelligent monitoring and early warning system, in the field of tunnel safety monitoring, includes: a signal sending unit for generating a first monitoring signal; a monitoring unit for monitoring the deformation of the inner wall of the tunnel based on the first monitoring signal to obtain a second monitoring signal; an analysis unit for analyzing the second monitoring signal to determine whether there is a safety risk in the tunnel, and if there is a safety risk, obtaining the analysis result and recording the time T when the second monitoring signal corresponding to the safety risk is generated; a collection unit for collecting video information of the monitoring unit; a communication unit for transmitting the analysis result and video information to the background processing center; the background processing center for analyzing the video information corresponding to before and after the time T, determining whether there is an accidental touch on the monitoring unit and issuing an early warning based on the judgment result. The present invention can solve the problem of false alarms in existing tunnel construction safety risk monitoring."; This existing technology only solves the problem that the current fiber optic sensor monitoring is prone to false alarms, resulting in slow progress in tunnel construction. It does not take into account that when issuing safety risk warnings for constructed tunnel sections, the current risk level of the constructed tunnel section should be evaluated from multiple aspects based on the relevant data of the constructed tunnel. During subsequent tunnel construction, in order to facilitate the prediction of changes in the risk level of the constructed tunnel section, the tunnel surface settlement is selected as the prediction indicator, and the tunnel surface settlement is predicted by constructing a tunnel construction risk prediction model, and the risk level of the tunnel section is obtained in combination with cloud model technology. Summary of the invention

[0003] The purpose of the present invention is to provide a method for early warning and determining safety risks of foundation pit tunnel engineering to solve the problems raised in the above-mentioned background technology.

[0004] To achieve the above object, the present invention provides the following technical solution: a method for early warning and determination of safety risks in foundation pit tunnel engineering, comprising the following steps: S1: A method for early warning and determination of safety risk of foundation pit tunnel engineering is applied to the safety risk determination of subway tunnel engineering under the existing line project. The foundation pit tunnel engineering is divided into sections, and the surface settlement data of the constructed tunnel sections are collected. The collected tunnel surface settlement data are preprocessed, and a surface settlement database is constructed based on the preprocessed surface settlement data. S2: Construct a tunnel construction risk assessment framework, and determine the weights of the tunnel construction risk assessment indicators within the tunnel construction risk assessment framework through the combined weighting method, and then determine the comprehensive risk cloud map of the constructed tunnel section in combination with the cloud model technology; S3: Based on the tunnel surface settlement, with the purpose of predicting the tunnel surface settlement, the tunnel construction risk prediction model is constructed and the performance of the tunnel construction risk prediction model is optimized to achieve the prediction of the tunnel surface settlement; S4: Based on the prediction results of the tunnel construction risk prediction model for the tunnel surface settlement, the cumulative settlement and average daily settlement deformation rate data are obtained, and then the risk level of the tunnel surface settlement change is determined to achieve risk warning of the tunnel surface settlement.

[0005] Preferably, step S1 also includes dividing the foundation pit tunnel project into sections based on the foundation pit tunnel project construction plan, and naming each connected section tunnel in sequence without duplication starting from the tunnel entrance, and laying out a surface settlement monitoring network for the constructed tunnel sections through the Internet of Things technology, and setting up surface settlement monitoring section measuring points and longitudinal surface settlement monitoring points above the tunnel axis. Each monitoring section is provided with 9 measuring points, which are evenly distributed perpendicular to the tunnel excavation direction. The interval between two adjacent monitoring sections is 30m, and the distance between two adjacent measuring points in each section is 5m. The entire monitoring duration lasts until the completion of the tunnel. The surface settlement data of the constructed tunnel section is collected through the set surface settlement monitoring network, and is associated with the corresponding tunnel section.

[0006] Preferably, step S1 also includes preprocessing the collected tunnel surface settlement data, specifically eliminating abnormal data in the collected tunnel surface settlement data, filling in the eliminated gaps and the originally collected missing tunnel surface settlement data through interpolation, and in order to eliminate the influence of different dimensions and units in the surface settlement data that make it impossible to directly compare the data, the surface settlement data is normalized to the range of 0-1 through the maximum and minimum normalization algorithm, and a surface settlement database is constructed based on the preprocessed surface settlement data.

[0007] Preferably, step S2 also includes constructing a tunnel construction risk assessment framework, and the specific operation is to set geological conditions, tunnel design conditions, underpass building conditions and tunnel construction conditions as first-level tunnel construction risk assessment indicators, and select surrounding rock grade, groundwater level and water inflow from geological conditions as second-level tunnel construction risk assessment indicators, select tunnel burial depth, tunnel span and tunnel maximum slope from tunnel design conditions as second-level tunnel construction risk assessment indicators, select minimum vertical clearance and crossing space angle from underpass building conditions as second-level tunnel construction risk assessment indicators, select grouting pressure, excavation speed, excavation pressure and soil bin pressure from tunnel construction conditions as second-level tunnel construction risk assessment indicators, so as to realize the construction of tunnel construction risk assessment framework, and tunnel construction management personnel divide each second-level tunnel construction risk assessment indicator into 1-5 risk levels based on expert experience, and determine the maximum boundary value and minimum boundary value of the second-level tunnel construction risk assessment indicator at 1-5 risk levels, and determine the weight of each second-level tunnel construction risk assessment indicator in the tunnel construction risk assessment framework by combined weighting method.

[0008] Preferably, step S2 also includes determining a comprehensive risk cloud map for the constructed tunnel sections in combination with the cloud model technology, and the specific operation is to obtain the corresponding data of each secondary tunnel construction risk assessment index in each constructed tunnel section through the Internet of Things technology, and further convert the dimensionless results obtained by converting the corresponding data of each secondary tunnel construction risk assessment index into the expectation, entropy and super entropy digital characteristics of the cloud model, and the specific calculation formula is as follows:

[0009] In the formula, Express expectations, represents entropy, represents super entropy, It indicates the minimum boundary value of the risk level corresponding to the corresponding data of the secondary tunnel construction risk assessment index. Indicates the maximum boundary value of the risk level corresponding to the corresponding data of the secondary tunnel construction risk assessment index, It represents the number of cloud droplets, and integrates the digital features of the cloud model obtained by transforming the corresponding data of the risk assessment indicators of each secondary tunnel construction. At the same time, a comprehensive risk cloud map is generated through a forward normal cloud generator to obtain a comprehensive risk cloud map of each constructed tunnel section. The risk level of the constructed tunnel section is determined by observing the position of the risk cloud in the standard cloud in the comprehensive risk cloud map, which facilitates risk monitoring of the constructed tunnel section.

[0010] Preferably, step S3 also includes, for the purpose of predicting the surface settlement of the tunnel, selecting a convolutional neural network and a long short-term memory model as base models, and constructing a tunnel construction risk prediction model by a model combination method. The specific operation is that the CNN layer is composed of alternatingly superimposed convolutional layers and pooling layers, and a relu activation function is used between each convolutional layer and pooling layer to accelerate the convergence of the tunnel construction risk prediction model. In the tunnel construction risk prediction model, after the tunnel surface settlement data is processed by the CNN layer, all features are fused to form a feature description of the CNN layer, and the feature description is passed to the LSTM layer for further processing. After the LSTM layer obtains the input features, it uses the sigmoid activation function to determine the input features that need to be kept and discarded, and transfers the new state and hidden state to the next layer. A fully connected layer is added after the LSTM layer to output the final prediction result.

[0011] Preferably, step S3 also includes optimizing the performance of the tunnel construction risk prediction model, specifically selecting the sparrow optimization algorithm to be combined with the tunnel construction risk prediction model to avoid tedious manual adjustments, quickly and accurately improving the prediction accuracy of the tunnel construction risk prediction model, and improving the performance of the tunnel construction risk prediction model. The tunnel surface settlement data samples are crawled by a web crawler, and a model training data set is constructed based on the crawled tunnel surface settlement data samples. The tunnel construction risk prediction model is trained by the model training data set, and the trained tunnel construction risk prediction model is subjected to accuracy verification. Based on the tunnel surface settlement data in the surface settlement database, the tunnel surface settlement data of any number of days in the future are predicted by the trained tunnel construction risk prediction model, thereby obtaining the tunnel surface settlement data of the constructed section tunnel of any number of days in the future.

[0012] Preferably, step S4 also includes obtaining tunnel surface settlement data of the constructed section tunnel in any number of days in the future, adding the data to the original tunnel surface settlement data to obtain tunnel surface cumulative settlement data, dividing the data by the predicted number of days to obtain tunnel average daily settlement deformation rate data, and dividing the tunnel surface cumulative settlement data and the tunnel average daily settlement deformation rate data into tunnel surface settlement risk levels, wherein the specific division operation is as follows: when the tunnel surface cumulative settlement is ≤13.3mm, it is judged as a first-level cumulative settlement risk; when 13.3mm<tunnel surface cumulative settlement is ≤16.5mm, it is judged as a second-level cumulative settlement risk; when 16.5mm<tunnel surface cumulative settlement is ≤19.8mm, it is judged as a third-level cumulative settlement risk; when 19.8mm<tunnel surface cumulative settlement is ≤25mm, it is judged as a fourth-level cumulative settlement risk; and when 25mm<tunnel surface cumulative settlement is ≤30mm, it is judged as a fifth-level cumulative settlement risk. Settlement risk: when the average daily settlement deformation rate data of the tunnel is ≤0.4mm / d, it is judged as the first-level average daily settlement deformation risk; when 0.4mm / d<the average daily settlement deformation rate data of the tunnel is ≤0.8mm / d, it is judged as the second-level average daily settlement deformation risk; when 0.8mm / d<the average daily settlement deformation rate data of the tunnel is ≤1.2mm / d, it is judged as the third-level average daily settlement deformation risk; when 1.2mm / d<the average daily settlement deformation rate data of the tunnel is ≤1.6mm / d, it is judged as the fourth-level average daily settlement deformation risk; when 1.6mm / d<the average daily settlement deformation rate data of the tunnel is ≤2.0mm / d, it is judged as the fifth-level average daily settlement deformation risk. Based on the cumulative settlement risk level and the average daily settlement deformation risk level corresponding to the tunnel surface cumulative settlement data and the tunnel average daily settlement deformation rate data respectively, the risk prediction level of the corresponding tunnel section is determined by cloud model technology, thereby realizing the risk warning of tunnel surface settlement.

[0013] Compared with the prior art, the beneficial effects of the present invention at least include: the present invention proposes a method for early warning and determination of safety risks of foundation pit tunnel engineering, which is applied to the safety risk determination of subway tunnel engineering underpass existing line projects, divides the tunnel into sections and collects surface settlement data of the constructed tunnel sections therein, pre-processes the collected surface settlement data and constructs a surface settlement database, constructs a tunnel construction risk assessment framework and determines the weights of each secondary tunnel construction risk assessment index within the tunnel construction risk assessment framework, and obtains the corresponding data of each secondary tunnel construction risk assessment index in each constructed tunnel section, thereby determining the risk level of the currently constructed tunnel section, and then, for the purpose of predicting the surface settlement of the tunnel, by constructing a tunnel construction risk prediction model, the surface settlement data of the constructed tunnel section at any number of days in the future are obtained, thereby determining the risk level of the constructed tunnel section at any number of days in the future, which is convenient for risk early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 A schematic flow chart of a method for early warning and determining safety risks in foundation pit tunnel engineering provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0015] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0016] See also Figure 1 The present invention provides a technical solution: a method for early warning and determination of safety risks in foundation pit tunnel engineering, comprising the following steps: S1: A method for early warning and determination of safety risk of foundation pit tunnel engineering is applied to the safety risk determination of subway tunnel engineering under the existing line project. The foundation pit tunnel engineering is divided into sections, and the surface settlement data of the constructed tunnel sections are collected. The collected tunnel surface settlement data are preprocessed, and a surface settlement database is constructed based on the preprocessed surface settlement data. S2: Construct a tunnel construction risk assessment framework, and determine the weights of the tunnel construction risk assessment indicators within the tunnel construction risk assessment framework through the combined weighting method, and then determine the comprehensive risk cloud map of the constructed tunnel section in combination with the cloud model technology; S3: Based on the tunnel surface settlement, with the purpose of predicting the tunnel surface settlement, the tunnel construction risk prediction model is constructed and the performance of the tunnel construction risk prediction model is optimized to achieve the prediction of the tunnel surface settlement; S4: Based on the prediction results of the tunnel construction risk prediction model for the tunnel surface settlement, the cumulative settlement and average daily settlement deformation rate data are obtained, and then the risk level of the tunnel surface settlement change is determined to achieve risk warning of the tunnel surface settlement.

[0017] Step S1 also includes dividing the foundation pit tunnel project into sections based on the foundation pit tunnel project construction plan, and naming each connected section tunnel in sequence without duplication starting from the tunnel entrance, and laying out a surface settlement monitoring network for the constructed tunnel section through the Internet of Things technology, setting up surface settlement monitoring section measuring points and longitudinal surface settlement monitoring points above the tunnel axis, setting up 9 measuring points for each monitoring section, and evenly distributing them perpendicular to the tunnel excavation direction, with an interval of 30m between two adjacent monitoring sections, and a distance of 5m between two adjacent measuring points in each section. The entire monitoring duration lasts until the tunnel is completed, and the surface settlement data of the constructed tunnel section is collected through the set surface settlement monitoring network, and associated with the corresponding tunnel section; Step S1 also includes preprocessing the collected tunnel surface settlement data, specifically removing abnormal data from the collected tunnel surface settlement data, filling in the removed gaps and the originally collected missing tunnel surface settlement data by interpolation, and in order to eliminate the influence of different dimensions and units in the surface settlement data that make it impossible to directly compare the data, the surface settlement data is normalized to the range of 0-1 by a maximum and minimum normalization algorithm, and a surface settlement database is constructed based on the preprocessed surface settlement data; Step S2 also includes constructing a tunnel construction risk assessment framework, which specifically sets geological conditions, tunnel design conditions, underpass building conditions and tunnel construction conditions as first-level tunnel construction risk assessment indicators, and selects surrounding rock grade, groundwater level and water inflow from geological conditions as second-level tunnel construction risk assessment indicators, selects tunnel burial depth, tunnel span and tunnel maximum slope from tunnel design conditions as second-level tunnel construction risk assessment indicators, selects minimum vertical clearance and crossing space angle from underpass building conditions as second-level tunnel construction risk assessment indicators, selects grouting pressure, excavation speed, excavation pressure and soil bin pressure from tunnel construction conditions as second-level tunnel construction risk assessment indicators, thereby realizing the construction of the tunnel construction risk assessment framework, and the tunnel construction management personnel divide each second-level tunnel construction risk assessment indicator into 1-5 risk levels based on expert experience, and determines the maximum boundary value and minimum boundary value of the second-level tunnel construction risk assessment indicator under the 1-5 risk levels, and determines the weight of each second-level tunnel construction risk assessment indicator in the tunnel construction risk assessment framework by a combined weighting method; Step S2 also includes determining a comprehensive risk cloud map for the constructed tunnel sections in combination with the cloud model technology. The specific operation is to obtain the corresponding data of the construction risk assessment indicators of each secondary tunnel in each constructed tunnel section through the Internet of Things technology, and further convert the dimensionless results obtained by converting the corresponding data of the construction risk assessment indicators of each secondary tunnel into the expectation, entropy and super entropy digital characteristics of the cloud model. The specific calculation formula is as follows:

[0018] In the formula, Express expectations, represents entropy, represents super entropy, It indicates the minimum boundary value of the risk level corresponding to the corresponding data of the secondary tunnel construction risk assessment index. Indicates the maximum boundary value of the risk level corresponding to the corresponding data of the secondary tunnel construction risk assessment index, Indicates the number of cloud droplets, synthesizes the digital features of the cloud model obtained by transforming the corresponding data of each secondary tunnel construction risk assessment index, and generates a comprehensive risk cloud map through a forward normal cloud generator to obtain a comprehensive risk cloud map of each constructed tunnel section. By observing the position of the risk cloud in the standard cloud in the comprehensive risk cloud map, the risk level of the constructed tunnel section is determined, thereby facilitating risk monitoring of the constructed tunnel section; Step S3 also includes, for the purpose of predicting the surface settlement of the tunnel, selecting a convolutional neural network and a long short-term memory model as base models, and constructing a tunnel construction risk prediction model by a model combination method. The specific operation is that the CNN layer is composed of alternating convolutional layers and pooling layers, and a relu activation function is used between each convolutional layer and pooling layer to accelerate the convergence of the tunnel construction risk prediction model. In the tunnel construction risk prediction model, after the tunnel surface settlement data is processed by the CNN layer, all features are fused to form a feature description of the CNN layer, and the feature description is passed to the LSTM layer for further processing. After the LSTM layer obtains the input features, the sigmoid activation function is used to determine the input features that need to be kept and discarded, and the new state and hidden state are transferred to the next layer. A fully connected layer is added after the LSTM layer to output the final prediction result; Step S3 also includes optimizing the performance of the tunnel construction risk prediction model, specifically selecting the sparrow optimization algorithm and combining it with the tunnel construction risk prediction model to avoid tedious manual adjustments, quickly and accurately improving the prediction accuracy of the tunnel construction risk prediction model, and improving the performance of the tunnel construction risk prediction model. The tunnel surface settlement data samples are crawled by a web crawler, and a model training data set is constructed based on the crawled tunnel surface settlement data samples. The tunnel construction risk prediction model is trained by the model training data set, and the trained tunnel construction risk prediction model is verified for accuracy. Based on the tunnel surface settlement data in the surface settlement database, the tunnel surface settlement data of any number of days in the future are predicted by the trained tunnel construction risk prediction model, so as to obtain the tunnel surface settlement data of the constructed section tunnel of any number of days in the future. Step S4 also includes adding the acquired tunnel surface settlement data of the constructed section tunnel in any number of days in the future to the original tunnel surface settlement data to obtain the tunnel surface cumulative settlement data, dividing it by the predicted number of days to obtain the tunnel average daily settlement deformation rate data, and dividing the tunnel surface cumulative settlement data and the tunnel average daily settlement deformation rate data into tunnel surface settlement risk levels, and the specific division operation is as follows: when the tunnel surface cumulative settlement is ≤13.3mm, it is judged as a first-level cumulative settlement risk; when 13.3mm<tunnel surface cumulative settlement is ≤16.5mm, it is judged as a second-level cumulative settlement risk; when 16.5mm<tunnel surface cumulative settlement is ≤19.8mm, it is judged as a third-level cumulative settlement risk; when 19.8mm<tunnel surface cumulative settlement is ≤25mm, it is judged as a fourth-level cumulative settlement risk; when 25mm<tunnel surface cumulative settlement is ≤30mm, it is judged as a fifth-level cumulative settlement risk. When the average daily settlement deformation rate data of the tunnel is ≤0.4mm / d, it is judged as the first-level average daily settlement deformation risk; when 0.4mm / d<the average daily settlement deformation rate data of the tunnel is ≤0.8mm / d, it is judged as the second-level average daily settlement deformation risk; when 0.8mm / d<the average daily settlement deformation rate data of the tunnel is ≤1.2mm / d, it is judged as the third-level average daily settlement deformation risk; when 1.2mm / d<the average daily settlement deformation rate data of the tunnel is ≤1.6mm / d, it is judged as the fourth-level average daily settlement deformation risk; when 1.6mm / d<the average daily settlement deformation rate data of the tunnel is ≤2.0mm / d, it is judged as the fifth-level average daily settlement deformation risk. Based on the cumulative settlement risk level and the average daily settlement deformation risk level corresponding to the tunnel surface cumulative settlement data and the tunnel average daily settlement deformation rate data respectively, the risk prediction level of the corresponding tunnel section is determined by the cloud model technology, thereby realizing the risk warning of tunnel surface settlement.

[0019] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0020] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for early warning and determination of safety risks in foundation pit tunnel engineering, characterized in that The following steps are involved: S1: A method for early warning and determination of safety risk of foundation pit tunnel engineering is applied to the safety risk determination of subway tunnel engineering under the existing line project. The foundation pit tunnel engineering is divided into sections, and the surface settlement data of the constructed tunnel sections are collected. The collected tunnel surface settlement data are preprocessed, and a surface settlement database is constructed based on the preprocessed surface settlement data. S2: Construct a tunnel construction risk assessment framework, and determine the weights of the tunnel construction risk assessment indicators within the tunnel construction risk assessment framework through the combined weighting method, and then determine the comprehensive risk cloud map of the constructed tunnel section in combination with the cloud model technology; S3: Based on the tunnel surface settlement, with the purpose of predicting the tunnel surface settlement, the tunnel construction risk prediction model is constructed and the performance of the tunnel construction risk prediction model is optimized to achieve the prediction of the tunnel surface settlement; S4: Based on the prediction results of the tunnel construction risk prediction model for the tunnel surface settlement, the cumulative settlement and average daily settlement deformation rate data are obtained, and then the risk level of the tunnel surface settlement change is determined to achieve risk warning of the tunnel surface settlement.

2. A method for early warning and determination of safety risks in foundation pit tunnel engineering according to claim 1, characterized in that: The step S1 also includes dividing the foundation pit tunnel project into sections based on the foundation pit tunnel project construction plan, and naming each connected section tunnel in sequence without duplication starting from the tunnel entrance, and laying out a surface settlement monitoring network for the constructed tunnel sections through the Internet of Things technology, setting up surface settlement monitoring section measuring points and longitudinal surface settlement monitoring points above the tunnel axis, and setting up 9 measuring points for each monitoring section, which are evenly distributed perpendicular to the tunnel excavation direction, with an interval of 30m between two adjacent monitoring sections, and a distance of 5m between two adjacent measuring points in each section. The entire monitoring duration lasts until the completion of the tunnel, and the surface settlement data of the constructed tunnel sections are collected through the set surface settlement monitoring network, and are associated with the corresponding tunnel sections.

3. A method for early warning and determination of safety risks in foundation pit tunnel engineering according to claim 1, characterized in that: The step S1 also includes preprocessing the collected tunnel surface settlement data, specifically removing abnormal data from the collected tunnel surface settlement data, filling in the missing tunnel surface settlement data through interpolation, and in order to eliminate the influence of different dimensions and units in the surface settlement data that make it impossible to directly compare the data, the surface settlement data is normalized to the range of 0-1 through the maximum and minimum normalization algorithm, and a surface settlement database is constructed based on the preprocessed surface settlement data.

4. A method for early warning and determination of safety risks in foundation pit tunnel engineering according to claim 1, characterized in that: The step S2 also includes constructing a tunnel construction risk assessment framework, which specifically sets geological conditions, tunnel design conditions, underpass building conditions and tunnel construction conditions as first-level tunnel construction risk assessment indicators, and selects surrounding rock grade, groundwater level and water inflow from geological conditions as second-level tunnel construction risk assessment indicators, selects tunnel burial depth, tunnel span and tunnel maximum slope from tunnel design conditions as second-level tunnel construction risk assessment indicators, selects minimum vertical clearance and crossing space angle from underpass building conditions as second-level tunnel construction risk assessment indicators, and selects grouting pressure, excavation speed, excavation pressure and soil bin pressure from tunnel construction conditions as second-level tunnel construction risk assessment indicators, thereby realizing the construction of the tunnel construction risk assessment framework, and the tunnel construction management personnel divide each second-level tunnel construction risk assessment indicator into 1-5 risk levels based on expert experience, and determines the maximum boundary value and minimum boundary value of the second-level tunnel construction risk assessment indicator under the 1-5 risk levels, and determines the weight of each second-level tunnel construction risk assessment indicator in the tunnel construction risk assessment framework by a combined weighting method.

5. The method for early warning and determination of safety risks in foundation pit tunnel engineering according to claim 1 is characterized by: The step S2 also includes determining a comprehensive risk cloud map for the constructed tunnel sections in combination with the cloud model technology, and the specific operation is to obtain the corresponding data of each secondary tunnel construction risk assessment indicator in each constructed tunnel section through the Internet of Things technology, and further convert the dimensionless results obtained by converting the corresponding data of each secondary tunnel construction risk assessment indicator into the expectation, entropy and super entropy digital characteristics of the cloud model, and synthesize the cloud model digital characteristics obtained by converting the corresponding data of each secondary tunnel construction risk assessment indicator, and generate a comprehensive risk cloud map through a forward normal cloud generator to obtain a comprehensive risk cloud map for each constructed tunnel section, and determine the risk level of the constructed tunnel section by observing the position of the risk cloud in the standard cloud in the comprehensive risk cloud map, so as to facilitate risk monitoring of the constructed tunnel section.

6. A method for early warning and determination of safety risks in foundation pit tunnel engineering according to claim 1, characterized in that: The step S3 also includes, for the purpose of predicting the surface settlement of the tunnel, selecting a convolutional neural network and a long short-term memory model as base models, and constructing a tunnel construction risk prediction model by a model combination method. The specific operation is that the CNN layer is composed of alternatingly superimposed convolutional layers and pooling layers, and a relu activation function is used between each convolutional layer and pooling layer to accelerate the convergence of the tunnel construction risk prediction model. In the tunnel construction risk prediction model, after the tunnel surface settlement data is processed by the CNN layer, all features are fused to form a feature description of the CNN layer, and the feature description is passed to the LSTM layer for further processing. After the LSTM layer obtains the input features, it uses the sigmoid activation function to determine the input features that need to be kept and discarded, and transfers the new state and hidden state to the next layer. A fully connected layer is added after the LSTM layer to output the final prediction result.

7. A method for early warning and determination of safety risks in foundation pit tunnel engineering according to claim 1, characterized in that: The step S3 also includes optimizing the performance of the tunnel construction risk prediction model, specifically selecting the sparrow optimization algorithm and combining it with the tunnel construction risk prediction model to avoid tedious manual adjustments, quickly and accurately improve the prediction accuracy of the tunnel construction risk prediction model, and improve the performance of the tunnel construction risk prediction model. The tunnel surface settlement data samples are crawled by a web crawler, and a model training data set is constructed based on the crawled tunnel surface settlement data samples. The tunnel construction risk prediction model is trained by the model training data set, and the trained tunnel construction risk prediction model is verified for accuracy. Based on the tunnel surface settlement data in the surface settlement database, the tunnel surface settlement data of any number of days in the future are predicted by the trained tunnel construction risk prediction model, so as to obtain the tunnel surface settlement data of the constructed section tunnel of any number of days in the future.

8. The method for early warning and determination of safety risks in foundation pit tunnel engineering according to claim 1 is characterized by: The step S4 also includes obtaining tunnel surface settlement data of the constructed section tunnel in any number of days in the future, adding the data to the original tunnel surface settlement data to obtain tunnel surface cumulative settlement data, dividing the data by the predicted number of days to obtain tunnel average daily settlement and deformation rate data, and classifying the tunnel surface cumulative settlement data and the tunnel average daily settlement and deformation rate data into tunnel surface settlement risk levels, and determining the risk prediction level of the corresponding tunnel section through cloud model technology based on the cumulative settlement risk level and the average daily settlement and deformation risk level corresponding to the tunnel surface cumulative settlement data and the tunnel average daily settlement and deformation rate data, thereby realizing risk warning of tunnel surface settlement.

Images