Airport tunnel excavation settlement risk dynamic assessment method based on digital twinning

Abstract

The invention provides an airport tunnel excavation settlement risk dynamic assessment method based on digital twinning. The method comprises the steps that airport tunnel excavation settlement information is acquired; constructing a ground surface settlement risk assessment model, determining a mutual influence relationship between the risk event and each risk index according to domain knowledge and historical data, and establishing a dynamic Bayesian network model taking the risk indexes as network nodes; ground surface settlement risk twinning deduction: adjusting a network structure and parameters through continuously updated dynamic monitoring data, performing dynamic deduction on the ground surface settlement risk, and obtaining a risk event chain which may cause ground surface settlement by airport tunnel excavation and the probability of the risk event; and dynamically evaluating the tunnel excavation settlement risk, mapping the dynamically calculated risk event probability result to different risk levels, and obtaining a tunnel excavation settlement risk dynamic evaluation result. In this way, the problems that an existing risk assessment method is serious in actual engineering construction disjunction, insufficient in timeliness and insufficient in refinement degree can be solved.

Description

Technical Field

[0001] The present invention generally relates to the technical field of tunnel information construction, and more specifically, to a dynamic assessment method for airport tunnel excavation settlement risk based on digital twins. Background Art

[0002] Airports are not only important transportation hubs, but also key nodes of national economy and international exchanges. Any functional failure or structural damage at an airport may result in significant economic losses and impact on public trust. Therefore, when tunneling at an airport, ensuring construction safety and reducing the impact on airport operations are top priorities. Tunneling is often accompanied by a large amount of soil movement and changes in underground space, which may lead to surface settlement. In the airport area, surface settlement not only endangers the structural safety of airport buildings, but may also affect the normal use of runways, taxiways and ground transportation systems. Airport tunneling projects require a system that can monitor and predict settlement risks in real time. Such a system needs to be able to predict potential risk areas before construction, while continuously updating risk assessments during construction so that timely measures can be taken to reduce damage.

[0003] Existing tunnel excavation settlement risk assessment methods often remain at the static assessment or post-construction risk assessment stage. On the one hand, traditional static assessment models are usually based on data from the initial geological survey and design stages, which lack accuracy during the construction process and may cause the initial prediction to fail due to changes in construction activities and environmental conditions; on the other hand, post-construction risk assessment can usually only be used for post-processing, but cannot prevent or mitigate ongoing risks, which limits the timeliness and effectiveness of response measures. In summary, in order to improve the effectiveness and adaptability of risk management, a more dynamic and continuous risk assessment method is needed that can instantly update risk predictions during construction and adjust strategies in real time. Summary of the invention

[0004] In view of the shortcomings of the above-mentioned prior art, the present invention provides a method for dynamic assessment of airport tunnel excavation settlement risk based on digital twin. This scheme can be used for dynamic and immediate risk assessment of surface settlement caused by airport tunnel excavation. The method comprises the following steps:

[0005] (1) Airport tunnel excavation settlement information acquisition: including monitoring data and knowledge related to airport tunnel excavation settlement;

[0006] (2) Construction of surface subsidence risk assessment model: Based on domain knowledge and historical data, the mutual influence relationship between risk events and various risk indicators is determined, and a dynamic Bayesian network model with risk indicator network nodes is established;

[0007] (3) Twin deduction of surface subsidence risk: parameterize the risk indicator network nodes through the monitoring data, and instantly estimate the importance of each node and the degree of interaction between them; input continuously updated data to adjust the network structure and parameters, and update the parameters in combination with the maximum likelihood estimation method; dynamically deduce the surface subsidence risk to obtain the risk event chain and probability of risk events that may lead to surface subsidence due to airport tunnel excavation;

[0008] (4) Dynamic assessment of tunnel excavation settlement risk: Based on the risk levels (high, medium, and low) set in the risk grading standard, the thresholds that trigger the risk levels under different conditions are calculated through risk deduction. At the same time, the dynamically calculated risk event probability results are mapped to different risk levels to obtain the dynamic assessment results of tunnel excavation settlement risk.

[0009] Preferably, the data includes tunnel geological conditions, tunnel design, settlement monitoring data, and construction methods;

[0010] Geological conditions: including rock mass characteristics, groundwater level, rock physical parameters, etc.;

[0011] Tunnel design: including tunnel burial depth, support type, etc.;

[0012] Construction methods: including milling and excavation method, shield method, etc.;

[0013] Subsidence monitoring data: including surface monitoring sections, subsidence monitoring data at various points, etc.;

[0014] The knowledge includes airport building deformation control standards, settlement risk index system (including primary index and corresponding secondary index) and risk classification standard (describing the degree of danger of surface settlement risk events caused by airport tunnel excavation, generally high risk, medium risk and low risk). Specific airport building deformation control standards include the settlement difference between the construction area and the surrounding area, pavement inclination standards, apron or runway deformation control standards, etc.

[0015] Preferably, the use of the Bayesian method to establish a risk assessment model with risk indicators as network nodes includes:

[0016] According to the settlement risk index system, determine the key risk index I affecting surface settlement;

[0017] I={G,M,D,S,R}

[0018] Where G is the geological condition, M is the construction method, D is the building deformation, S is the monitoring data, and R is the surface settlement risk;

[0019] Abstract each indicator into a network node, and abstract the risk event "surface subsidence risk" into a network node;

[0020] Determine dependencies between nodes based on domain knowledge;

[0021] Abstracting the dependency relationship between the nodes into directed line segments, connecting the risk indicators and risk events to obtain a network topology graph;

[0022] Mathematical modeling of network structure, based on historical data and expert knowledge, to determine the dependencies between nodes in the network and their conditional probability tables;

[0023] The network topology diagram is mathematically modeled in turn to obtain a risk assessment model.

[0024] Preferably, in the twin deduction of surface subsidence risk, the monitoring data is firstly subjected to data preprocessing operations, including data cleaning, feature extraction, and data conversion, and then the probability distribution is determined:

[0025] For each node, its probability distribution is determined based on the collected data, and the prior probability of the root node and the conditional probability of the intermediate nodes are calculated using the rough set classification principle:

[0026]

[0027] In the above formula, P is the prior probability or conditional probability, which indicates the probability of a certain risk state or indicator state, and I 1 , I 2 , I 3 is a certain type of key risk indicator, i, j, k are risk status (such as high, medium and low risk), m is the frequency, and n is the total number of samples;

[0028] Inputting continuously updated data to adjust the network structure and parameters means using real-time or the latest monitoring data to adjust and optimize the parameters in the Bayesian network, and combining the maximum likelihood estimation method to update the parameters:

[0029]

[0030] θ is the Bayesian network parameter, which represents the parameter set of the conditional probability distribution of each indicator node in the network, n is the number of samples, i is a sample, P is the above probability function, x i is the observed value of the i-th node in the Bayesian network, pa(x i ) is the node x i The parent node set of

[0031] Dynamically deduce the surface subsidence risk to obtain the risk event chain and probability of risk events that may lead to surface subsidence due to airport tunnel excavation;

[0032] At the same time, the dynamically calculated risk event probability results are mapped to different risk levels to obtain the dynamic assessment results of tunnel excavation settlement risk.

[0033] In this way, the problems of existing risk assessment methods being seriously out of touch with actual engineering construction, lacking timeliness, and lacking in refinement can be solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The above and other features, advantages and aspects of the embodiments of the present invention will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, wherein:

[0035] Figure 1 A flow chart of a dynamic assessment method for airport tunnel excavation settlement risk based on digital twin is shown;

[0036] Figure 2 A schematic diagram of a dynamic risk assessment model in the method of the present invention is shown;

[0037] Figure 3 A schematic diagram of the twin deduction principle of surface subsidence risk in the method of the present invention is shown. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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.

[0039] In this embodiment, a tunnel project under an airport is selected as an example, geological condition information and tunnel design information in the design and survey stage are obtained, and the milling excavation method is used for construction, and the settlement data of the arch and waist of the tunnel section are monitored in real time;

[0040] Obtain airport building deformation control standards: settlement differences between the construction area and the surrounding area, pavement inclination standards, apron or runway deformation control standards;

[0041] Obtaining the settlement risk index system: including primary indicators and corresponding secondary indicators, geological conditions (including rock mass characteristics, groundwater level, rock physical parameters, etc.), surface building types (including runways, aprons, terminals, etc.), settlement deformation (building deformation, surface settlement rate, surface displacement, etc.), construction methods (milling and excavation parameters, TBM construction parameters, support structure, etc.);

[0042] Obtain risk classification standards: Describe the degree of danger of surface settlement risk events during airport tunnel excavation, generally high risk, medium risk, and low risk.

[0043] Constructing a surface subsidence risk assessment model, and determining key risk indicators I affecting surface subsidence according to the subsidence risk indicator system;

[0044] I={G,M,d,S,R}

[0045] Including geological conditions (G), construction methods (M), building deformation (D), monitoring data (S), and surface settlement risk (R). Determine the mutual influence relationship and use the Bayesian method to establish a risk assessment model with risk indicators as network nodes.

[0046] Each indicator is abstracted as a network node, representing a random variable, and there is a probabilistic dependency between them. For example, the risk event "surface subsidence risk" is abstracted as a network node, and the dependency between nodes is determined based on domain knowledge. For example, geological conditions and surface building types directly affect tunnel excavation and lead to surface subsidence;

[0047] like Figure 2 As shown, the dependency relationship between the nodes is abstracted into directed line segments, and the risk indicators and risk events are connected to obtain a network topology diagram;

[0048] Mathematical modeling of the network structure determines the dependencies between nodes in the network and their conditional probability tables (CPTs) based on historical data and expert knowledge. For example, the risk of surface subsidence depends on the construction method and building type:

[0049] P(R|M,D)

[0050] The network topology diagram is mathematically modeled in turn to obtain a risk assessment model.

[0051] Twin deduction of surface subsidence risk, parameterize the risk indicator nodes through the monitoring data, and instantly estimate the importance of each node and the degree of interaction between them. Then input the continuously updated data to adjust the network structure and parameters, dynamically deduce the surface subsidence risk, and obtain the risk event chain and probability of risk events that may cause surface subsidence due to airport tunnel excavation. Specifically: first, perform data preprocessing operations such as data cleaning, feature extraction, and data conversion on the monitoring data, and then determine the probability distribution:

[0052] For each node, its probability distribution is determined based on the collected data. This method uses the rough set classification principle to calculate the prior probability of the root node and the conditional probability of the intermediate nodes:

[0053]

[0054] The above formula is the prior probability when the state of a certain risk indicator is i / j, m is the frequency, and n is the total number of samples.

[0055] Inputting continuously updated data to adjust the network structure and parameters means using real-time or the latest monitoring data to adjust and optimize the parameters in the Bayesian network, and combining the maximum likelihood estimation method to update the parameters:

[0056]

[0057] Dynamic deduction of surface settlement risk revealed that the risk event chain of surface settlement caused by airport tunnel excavation is: milling excavation -> soft rock deformation -> tunnel top settlement -> 10mm -> arch support -> medium settlement risk;

[0058] At the same time, the dynamically calculated risk event probability results are mapped to different risk levels, and the dynamic assessment result of tunnel excavation settlement risk is obtained as a moderate settlement risk.

[0059] According to the embodiments of the present invention, by making full use of the principles of digital twin technology, combining real-time tunnel monitoring data with risk experience knowledge to collaboratively drive the dynamic assessment of tunnel excavation settlement risk, the existing risk assessment methods can solve the problems of serious disconnection with actual engineering construction, lack of timeliness, and insufficient refinement. Provide more accurate and timely airport tunnel excavation surface settlement risk.

Claims

1. A dynamic assessment method for airport tunnel excavation settlement risk based on digital twin, characterized in that: The steps include: (1) Airport tunnel excavation settlement information acquisition: including monitoring data and knowledge related to airport tunnel excavation settlement; (2) Construction of surface subsidence risk assessment model: Based on domain knowledge and historical data, the mutual influence relationship between risk events and various risk indicators is determined, and a dynamic Bayesian network model with risk indicator network nodes is established; (3) Twin deduction of surface subsidence risk: parameterize the risk indicator network nodes through the monitoring data, and instantly estimate the importance of each node and the degree of interaction between them; Input continuously updated data to adjust the network structure and parameters, and use the maximum likelihood estimation method to update the parameters; dynamically deduce the surface subsidence risk to obtain the risk event chain and probability of risk events that may cause surface subsidence due to airport tunnel excavation; (4) Dynamic assessment of tunnel excavation settlement risk: In combination with the risk levels set in the risk classification standard, the thresholds that trigger the risk levels under different conditions are calculated through risk deduction. At the same time, the dynamically calculated risk event probability results are mapped to different risk levels to obtain the dynamic assessment results of tunnel excavation settlement risk.

2. The method for dynamic assessment of airport tunnel excavation settlement risk based on digital twin according to claim 1 is characterized in that: The data include tunnel geological conditions, tunnel design, settlement monitoring data, and construction methods; The knowledge includes airport building deformation control standards, settlement risk indicator system and risk classification standards.

3. The method for dynamic assessment of airport tunnel excavation settlement risk based on digital twin according to claim 1 is characterized in that: The Bayesian method is used to establish a risk assessment model with risk indicators as network nodes, including: According to the settlement risk index system, determine the key risk index I affecting surface settlement; I={G,M,d,S,R} Where G is the geological condition, M is the construction method, D is the building deformation, S is the monitoring data, and R is the surface settlement risk; Abstract each indicator into a network node, and abstract the risk event "surface subsidence risk" into a network node; Determine dependencies between nodes based on domain knowledge; Abstracting the dependency relationship between the nodes into directed line segments, connecting the risk indicators and risk events to obtain a network topology graph; Mathematical modeling of network structure, based on historical data and expert knowledge, to determine the dependencies between nodes in the network and their conditional probability tables; The network topology diagram is mathematically modeled in turn to obtain a risk assessment model.

4. The method for dynamic assessment of airport tunnel excavation settlement risk based on digital twin according to claim 1 is characterized in that: In the twin simulation of surface subsidence risk, the monitoring data is firstly subjected to data preprocessing operations, including data cleaning, feature extraction, and data conversion, and then the probability distribution is determined: For each node, its probability distribution is determined based on the collected data, and the prior probability of the root node and the conditional probability of the intermediate nodes are calculated using the rough set classification principle: In the above formula, P is the prior probability or conditional probability, which indicates the probability of a certain risk state or indicator state, I1, I2, I3 are certain key risk indicators, i, j, k are risk states, m is the frequency, and n is the total number of samples; Inputting continuously updated data to adjust the network structure and parameters means using real-time or the latest monitoring data to adjust and optimize the parameters in the Bayesian network, and combining the maximum likelihood estimation method to update the parameters: θ is the Bayesian network parameter, which represents the parameter set of the conditional probability distribution of each indicator node in the network, n is the number of samples, i is a sample, P is the above probability function, x i is the observed value of the ith node in the Bayesian network, pa(x i ) is the node x i The parent node set of Dynamically deduce the surface subsidence risk to obtain the risk event chain and probability of risk events that may cause surface subsidence due to airport tunnel excavation; At the same time, the dynamically calculated risk event probability results are mapped to different risk levels to obtain the dynamic assessment results of tunnel excavation settlement risk.

Images