A BIM-based tunnel engineering safety risk monitoring system and method

Abstract

The present invention belongs to the field of building information model information technology, and in particular, relates to a BIM-based tunnel engineering safety risk monitoring system and method. The on-site monitoring module of the system is arranged in the monitoring area of ​​the tunnel construction site, and is used to monitor the mechanical performance data of the support system and the surrounding rock stability state data of the monitoring points in the current monitoring area; and transmit the data to the data server; through the risk assessment model, the risk assessment results of the monitoring points in the monitoring area are determined and transmitted to the database; a three-dimensional BIM monitoring model is created and converted into an IFC monitoring model, and the IFC monitoring information of the IFC monitoring model is expanded, and the expanded IFC monitoring model is obtained and loaded using a visual early warning platform; the IFC monitoring model of the present invention is used as an interactive standard for BIM monitoring model data, and the IFC attribute set of the monitoring model is written to improve the use efficiency of the monitoring information, and the amount of information carried by the model is expanded to meet the requirements of visual monitoring, so as to realize real-time dynamic safety early warning of tunnel engineering.

Description

Technical Field

[0001] The present invention belongs to the field of building information model information technology, and in particular relates to a tunnel engineering safety risk monitoring system and method based on BIM. Background Art

[0002] With the rapid economic development of my country and the significant improvement of its comprehensive national strength, the infrastructure field has also ushered in a leap-forward development. However, the construction characteristics of tunnel projects themselves have also caused frequent safety accidents. The low efficiency of information transmission during the construction process, the lack of construction experience of relevant personnel, and the incomplete risk management methods and technical means will directly lead to problems such as untimely alarms and inadequate crisis handling, which will cause major engineering accidents. Therefore, the rapid development of tunnel projects has also made the application of tunnel monitoring increasingly massive and complex, which has brought new challenges to the collection, analysis, and processing of tunnel monitoring information.

[0003] At present, most tunnel projects use traditional manual monitoring methods. Although this method is technically mature, it has great subjective uncertainty and requires high quality of monitoring personnel. It also has defects such as low accuracy, low efficiency and safety issues of monitoring personnel. This is not suitable for the technical characteristics of current tunnel construction. With the development of new technologies, relevant technicians have made a series of explorations on information monitoring, and three-dimensional laser scanning technology, automatic sensor technology and intelligent stress gauge technology have emerged. However, new monitoring technologies still have problems such as high labor cost, low visualization and native data processing process. In addition, the current information monitoring technology of tunnels also has the following problems: (1) The three-dimensional model and monitoring and early warning are not effectively combined, and the three-dimensional model and monitoring data are relatively independent; (2) The processing of monitoring data is relatively lagging, and the method of predicting the safety risk of tunnel projects is not perfect; (3) The information transmission is lagging in the process of defect disposal management of tunnel construction, and it is impossible to track, process and close the quality hazards; (4) The monitoring data early warning technology is not yet mature. Therefore, how to use the rapidly developing digital equipment to realize remote visualization monitoring to improve the management efficiency and safety level of tunnel projects is a problem that needs to be solved.

[0004] At present, BIM technology is developing steadily in my country, and its application in many fields in my country is becoming more and more extensive. BIM has the characteristics of physical visualization and information integration throughout the life cycle of the project. It brings the unique three-dimensional visualization advantages of BIM technology into the field of tunnel monitoring, which will greatly improve the informatization of the monitoring system. Moreover, combining BIM technology with risk assessment models can greatly improve the risk management level of the construction site, reduce economic losses such as personnel and property, and provide technical support for the digital management of tunnel projects.

[0005] At present, the visualization and risk prediction methods of BIM-based tunnel engineering safety monitoring are not in-depth enough at home and abroad. Most databases can only store and read and write monitoring information. At the same time, the expansion of monitoring information in the IFC standard still needs to be improved, and it is impossible to conduct dynamic risk assessment of tunnel engineering in a scientific and reasonable manner. The application of BIM technology remains in local applications. Summary of the invention

[0006] The present invention provides a BIM-based tunnel engineering safety risk monitoring system and method.

[0007] The present invention adopts the following technical solution:

[0008] The BIM-based tunnel engineering safety risk monitoring system of the present invention comprises a field monitoring module, a data server, a risk assessment module, a database, a BIM monitoring model creation module, an IFC monitoring model extension module, a visual early warning platform, and a mobile terminal quality inspection module;

[0009] The field monitoring module is arranged in the monitoring area of ​​the tunnel construction site, and is used to monitor the mechanical performance data of the support system and the surrounding rock stability state data of the monitoring points in the current monitoring area; and transmit the data to the data server;

[0010] The support system mechanical property data and surrounding rock stability state data acquired by the data server are sent to the risk assessment module in real time; the risk assessment module is used to calculate and analyze the support system mechanical property data and surrounding rock stability state data, determine the risk assessment results of the monitoring points in the monitoring area and transmit them to the database; the BIM monitoring model creation module is used to create a three-dimensional BIM monitoring model according to the design parameters of the CAD drawings of the monitoring points, convert the three-dimensional BIM monitoring model into an IFC monitoring model through the IFC monitoring model extension module, and extend the IFC monitoring information of the IFC monitoring model to obtain the extended IFC monitoring model;

[0011] The visual early warning platform is used to load the IFC monitoring model to visualize it, generate risk assessment results, and issue quality inspection tasks to be transmitted to the mobile terminal quality inspection module. The mobile terminal quality inspection module is used for supervisors to receive and process quality inspection tasks.

[0012] In the BIM-based tunnel engineering safety risk monitoring system of the present invention, the support system mechanical performance data monitored by the field monitoring module can include: anchor axial force, steel frame internal force, support internal stress, lining concrete stress, lining concrete strain, contact pressure between initial support and secondary lining, and concrete external load;

[0013] The surrounding rock stability data include: displacement in the surrounding rock mass, vault settlement, geological and support observations, surface settlement, settlement of surrounding buildings, tunnel water pressure, pore water pressure, surrounding rock pressure and tunnel uplift.

[0014] The monitoring method of the tunnel engineering safety risk monitoring system based on BIM of the present invention comprises the following steps:

[0015] Step 1: Collect the mechanical property data of the support system and the stability state data of the surrounding rock at the monitoring points in the monitoring area through the on-site monitoring module;

[0016] Step 2, computing and analyzing the monitoring data: using a data server to collect the support system mechanical performance data and surrounding rock stability data of the monitoring point in the monitoring area and send them to the risk assessment module, determine the risk assessment results of the monitoring area where the monitoring point is located through the risk assessment module, and transmit the risk assessment results and monitoring data to the database in real time; the risk assessment results include safety risk level and risk probability;

[0017] Step 3: Create a three-dimensional BIM monitoring model according to the design parameters of the monitoring point CAD drawings, and encode the components of the monitoring model; convert the BIM monitoring model into an IFC monitoring model, and extend the IFC monitoring information of the IFC monitoring model to obtain the extended IFC monitoring model;

[0018] Step 4: Load the extended IFC monitoring model according to the visual early warning platform, associate the monitoring data in the database and the risk assessment results of the monitoring area with the extended IFC monitoring model; issue early warnings and quality verification tasks according to the risk level;

[0019] Step 5: Supervisors receive and process quality inspection tasks.

[0020] In the monitoring method of the BIM-based tunnel engineering safety risk monitoring system of the present invention, in step 2, the monitoring data is calculated and analyzed to determine the risk assessment result of the monitoring area where the monitoring point is located, including:

[0021] Step 2.1, count the safety accidents during the tunnel construction process, and classify and summarize the safety accidents according to the tunnel accident type, i.e. historical monitoring data;

[0022] The process of tunnel construction is decomposed structurally, and a list of tunnel engineering safety risk factors is constructed based on the above classified tunnel accident types;

[0023] Based on the above-mentioned list of tunnel engineering safety risk factors, a risk factor network topology structure is constructed using the Bayesian network topology structure;

[0024] The fuzzy theory is used to calculate the parameters of each node of the risk factor network topology structure, and the Bayesian network topology structure is used to quantitatively analyze the safety risk factors of tunnel engineering to obtain the risk factor weight value;

[0025] Step 2.2, based on the network topology of risk factors, establish the corresponding relationship between monitoring data and risk factors, use the GS risk classification model to calculate the risk classification results and risk probability corresponding to the monitoring data, and thus determine the risk level;

[0026] Step 2.2.1, randomly extract data samples from the above historical monitoring data, and normalize the sample data. The formula is as follows:

[0027]

[0028] Among them, x max is the maximum value in the sample, x min is the minimum value in the sample, x i is a sample data, x′ is the normalized data;

[0029] Step 2.2.2, use the radial basis kernel function as the kernel function of the support vector machine (SVM) to classify the historical monitoring data, and use the genetic algorithm (GA) to calculate the optimal parameters gamma and C of the radial basis kernel function, where the parameter gamma is the inverse of the influence radius of the support vector sample and the parameter C is the penalty relaxation variable; construct a GS risk classification model;

[0030] Step 2.2.3, use the Sigmoid function to obtain the classification results and classification probabilities of the posterior probability calculation. The posterior probability formula is as follows:

[0031]

[0032] Where f(x) is the output of the decision function of the GS risk classification model; A and B are the Sigmoid parameters to be fitted; P(y=1|f(x)) is the degree to which the sample belongs to category 1 under the condition of the output value f(x);

[0033] The parameters A and B are calculated using the following formula:

[0034]

[0035] in, Indicates The values ​​of parameters A and B when they are at their minimum values; t i is the target probability, i=1,2,…,n,N + and N - are the number of positive and negative examples in the model respectively; y i represents the sample, and n represents the total number of samples;

[0036] Step 2.2.4, randomly extracting predicted monitoring data from the above historical monitoring data excluding sample data; substituting the predicted monitoring data into the GS risk classification model to verify the model evaluation results;

[0037] Step 2.2.5, substitute the monitored support system mechanical property data and surrounding rock stability state data into the GS risk classification model to calculate the risk classification results and risk probability, so as to determine the risk level of the monitoring area;

[0038] The risk probability is 0-0.25, the risk level is level 1, and the risk is negligible;

[0039] The risk probability is 0.25-0.5, the risk level is level 2, and the risk is acceptable;

[0040] The risk probability is 0.5-0.8, the risk level is level 3, and the risk needs to be handled;

[0041] The risk probability is 0.8-1, the risk level is level four, the risk is unacceptable and needs to be dealt with immediately.

[0042] In the monitoring method of the BIM-based tunnel engineering safety risk monitoring system of the present invention, the database in step 2 includes a monitoring point information data table, a monitoring data data table, and a risk information data table.

[0043] The monitoring point information data table is used to store monitoring point codes, monitoring types, monitoring areas and monitoring point warning values;

[0044] The monitoring data table is used to store monitoring point codes, monitoring data and monitoring time;

[0045] The risk information data table is used to store monitoring area codes, risk probabilities, risk levels and colors corresponding to the risk levels.

[0046] The monitoring method of the BIM-based tunnel engineering safety risk monitoring system of the present invention, the step 3 creates a three-dimensional BIM monitoring model according to the design parameters of the monitoring point CAD drawing, the BIM monitoring model is obtained according to the construction joint division during the tunnel construction process, the monitoring model between two construction joints is classified as a monitoring area, and a unique digital code is given to the monitoring area;

[0047] Export the BIM monitoring model to a BIM model in IFC format, namely, an IFC monitoring model; the IFC monitoring model is uniquely identified using a globally unique identifier GUID;

[0048] Use ArchiCAD software to expand the IFC monitoring model with IFC monitoring information;

[0049] The extended IFC monitoring information includes monitoring point code, installation time, collection time, monitoring data type, warning value upper limit, warning value lower limit, monitoring area code, transmission protocol name, transmission protocol version number characteristic attributes;

[0050] In step 4, the visual early warning platform loads the IFC monitoring model after information expansion; through the correspondence between the GUID identifier of the IFC monitoring model after information expansion and the monitoring point code, the data in the monitoring point information data table, the monitoring data data table, and the risk information data table in the database are read into the expanded IFC monitoring model;

[0051] The visual early warning platform displays monitoring data curves, monitoring point information, monitoring data over-limit reminders, IFC monitoring model real-time colors, and risk level information;

[0052] When the risk level is level one, the BIM monitoring model is displayed in the initial color;

[0053] When the risk level is level 2, the BIM monitoring model is displayed in yellow;

[0054] When the risk level is level three, the BIM monitoring model is displayed in orange;

[0055] When the risk level is level 4, the BIM monitoring model is displayed in red;

[0056] When the risk level is level 3 or 4, or the monitoring data is greater than the set warning threshold, an early warning will be issued. The warning information includes the monitoring area, risk level, problem description, and quality verification task information reminder;

[0057] When the risk level is level three or level four, or the monitoring data is greater than the set warning threshold, a quality verification instruction will be issued and supervision personnel will be arranged to conduct inspections at the construction site.

[0058] Beneficial Effects

[0059] The present invention provides a tunnel engineering safety risk monitoring system and method based on BIM. The IFC monitoring model is used as the interaction standard of the BIM monitoring model data, and the IFC attribute set of the monitoring model is written to improve the use efficiency of the monitoring information, expand the amount of information carried by the model to meet the requirements of visual monitoring, and enrich the information display content of the visualization platform;

[0060] The present invention provides a tunnel engineering safety risk monitoring system and method based on BIM, which combines BIM technology with risk assessment to conduct risk evaluation, scientifically predict risk levels, effectively prevent the occurrence of safety accidents through dynamic risk assessment and visual early warning, and can help managers make more scientific and reasonable decisions, thereby strengthening active control of risks, effectively avoiding risks, and reducing the probability of risk occurrence;

[0061] The present invention provides a tunnel engineering safety risk monitoring system and method based on BIM, which combines a visual early warning platform with a mobile terminal construction quality inspection process, improves collaborative work efficiency, reduces the risk of information loss, and facilitates historical information query. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 The structure diagram of a BIM-based tunnel engineering safety risk monitoring system and method.

[0063] Figure 2 It is the BIM model and component coding diagram for tunnel secondary lining stress monitoring.

[0064] Figure 3 This is the Bayesian network structure topology diagram of tunnel safety risk.

[0065] Figure 4 It is a curve chart of surrounding rock displacement monitoring data of the visual early warning platform.

[0066] Figure 5 This is a display diagram of the monitoring area when the risk level of the visual early warning platform is level three.

[0067] Figure 6 Release interface image for mobile tasks.

[0068] Figure 7 Interface diagram for uploading construction site photos and treatment plans to the mobile terminal.

[0069] Figure 8 The photos and solutions for quality defect treatment are displayed on the visual early warning platform. DETAILED DESCRIPTION

[0070] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation methods, and the above-mentioned and / or other advantages of the present invention will become more clear, but this should not be understood as the protection scope of the present invention being limited to the following embodiments. All technical solutions implemented based on the contents of the present invention fall within the protection scope of the present invention.

[0071] like Figure 1 As shown: The BIM-based tunnel engineering safety risk monitoring system includes a field monitoring module, a data server, a risk assessment module, a database, a BIM monitoring model creation module, an IFC monitoring model extension module, a visual early warning platform, and a mobile terminal quality inspection module;

[0072] The on-site monitoring module is used to collect two types of monitoring data, namely, the mechanical properties of the support system and the stability of the surrounding rock at the monitoring point during the on-site construction of the tunnel project;

[0073] The monitoring data of the monitoring point are the mechanical properties of the support system and the stability of the surrounding rock. The mechanical properties of the support system include: anchor axial force, steel frame internal force, support internal stress, lining concrete stress, lining concrete strain, contact pressure between initial support and secondary lining, and concrete external load;

[0074] The stability state of surrounding rock includes: displacement in surrounding rock mass, vault subsidence, geological and support observation, surface settlement, settlement of surrounding buildings, tunnel water pressure, pore water pressure, surrounding rock pressure, and tunnel uplift.

[0075] The data server periodically sends monitoring commands to the field monitoring system to obtain monitoring data. In short, medium and long tunnel projects, the field monitoring module sends the collected monitoring data to the data server via the GPRS (General packet radio service) network. In long and extra-long tunnel projects, the field monitoring module uses the GPRS network and the wireless transmission relay module to transmit the collected monitoring data to the data server; the data server transmits the monitoring data to the risk assessment module in real time. In this embodiment, the wireless transmission relay module uses an RS-485 industrial-grade repeater - Penghe Electronics Source Manufacturer.

[0076] The risk assessment module is used to calculate and analyze the monitoring data, determine the risk assessment results of the monitoring area where the monitoring point is located, and transmit the risk assessment results and monitoring data to the database in real time; the risk assessment results include safety risk level and risk probability; the field monitoring module includes monitoring equipment including various sensors and monitors, such as AR-SS-SZY02 static level, HC-5100 fixed inclinometer, HC-2200 multi-point displacement meter, 9430 electronic convergence meter, 9500 vibrating wire piezometer, 9980 standard thermometer, HC-2400 crack meter, 9000 / 9050 vibrating wire support strain gauge, 9011 vibrating wire steel bar stress gauge, 9900 vibrating wire anchor meter, etc.

[0077] Database for storing monitoring data and risk assessment results of the monitoring areas;

[0078] The BIM monitoring model creation module is used to create a three-dimensional BIM monitoring model according to the design parameters of the monitoring point CAD drawing. The attribute information of the monitoring model includes the monitoring point, monitoring type, monitoring area, monitoring equipment installation time, monitoring equipment model and monitoring model coding information; the monitoring model is component-encoded to obtain the monitoring point code, and the coding format is: construction section-monitoring area-monitoring object-monitoring type-serial number;

[0079] For example: 1-DKD15-ECCQ-JZC-064, Standard 1-Long Span 15-Secondary Lining-Stress-064;

[0080] The IFC monitoring model extension module is used to convert the BIM monitoring model into the IFC monitoring model and extend the IFC monitoring information of the IFC monitoring model to obtain the extended IFC monitoring model so that the monitoring data and the risk assessment results of the monitoring area can be stored in the IFC model;

[0081] Visual early warning platform, used to load the extended IFC monitoring model, associate the monitoring data in the database and the risk assessment results of the monitoring area with the extended IFC monitoring model, realize visual query of the monitoring data and the risk assessment results of the monitoring area, timely risk early warning, and issue quality verification tasks;

[0082] The visual early warning platform is built using Unreal Engine 4. The IFC monitoring model is loaded using Unreal Engine 4, and the model is recompiled and its functions are applied, ultimately completing the development of a visual early warning platform based on BIM technology.

[0083] The mobile terminal quality inspection module is used for supervisors to receive and process quality inspection tasks. The processing of quality inspection tasks includes supervisors using mobile terminals to take photos of quality defect areas at the construction site and upload them, and initiate remediation tasks. After the construction unit receives the remediation task, it processes the defective areas. The supervisors take photos of the tunnel site after treatment, upload the photos and treatment plans after treatment, and copy the photos and treatment plans after treatment to the supervisory leaders and business reviewers; the photos and treatment plans after treatment that have passed the review can be displayed on the visual early warning platform for easy viewing of historical records.

[0084] The BIM-based tunnel engineering safety risk monitoring method uses the BIM-based tunnel engineering safety risk monitoring system, and the steps are:

[0085] Step 1, collecting monitoring data;

[0086] Step 2: Calculate and analyze the monitoring data to determine the risk assessment results of the monitoring area where the monitoring point is located, and transmit the risk assessment results and monitoring data to the database in real time; the risk assessment results include the safety risk level and risk probability;

[0087] Step 2.1: Use the Bayesian network topology structure to quantitatively analyze the safety risk factors of tunnel engineering and obtain the risk factor network topology structure. The specific steps are as follows:

[0088] Firstly, the safety accidents in the process of tunnel construction are counted and classified and summarized according to the types of tunnel accidents; the safety accidents are obtained through questionnaires, reference materials, accident cases and news bulletins;

[0089] The following table shows some examples;

[0090]

[0091] Secondly, the process of tunnel construction is structurally decomposed, and a list of tunnel accident risk factors is constructed in combination with the types of tunnel accidents, as shown in the following table; then, a Bayesian network topology is constructed based on the causal relationship between tunnel safety risk factors, such as Figure 3 As shown; finally, fuzzy theory is used to calculate the parameters of each node of the risk factor network topology structure, for example: over-excavation parameter is: 0.056, anchor failure parameter is: 0.863, concrete strength deficiency parameter is: 0.162, water seepage parameter is: 0.098, etc., to achieve quantitative analysis of tunnel construction safety risk factors.

[0092] Checklist of risk factors in tunnel construction

[0093]

[0094]

[0095] Then, according to the list of tunnel engineering safety risk factors, the Bayesian network topology structure is used to construct the risk factor network topology structure;

[0096] Finally, fuzzy theory is used to calculate the parameters of each node of the risk factor network topology structure to achieve quantitative analysis of tunnel engineering safety risk factors.

[0097] In this step, the risk factors of the construction process are summarized, the topological structure of the risk factors is constructed using the Bayesian network topology, and the node parameters are calculated using fuzzy theory to obtain the weight values ​​of the risk factors. On this basis, the key risk factors are identified and the key risk factors are monitored first during the tunnel construction process, which can effectively reduce the possibility of risk occurrence. In addition, the greater the weight of the risk factor, the greater its impact on the risk. The risk contribution rate of each monitoring indicator is obtained and brought into the risk assessment model as a parameter for quantitative risk analysis.

[0098] Step 2.2, based on the network topology of risk factors, establish the corresponding relationship between monitoring data and risk factors, and use the GS risk model formed by combining the genetic algorithm GA (Genetic Algorithm) and the support vector machine SVM (Support Vector Machine) to calculate the risk classification results and risk probabilities corresponding to the monitoring data, so as to determine the risk level; the steps are as follows:

[0099] Step 2.2.1, establish the corresponding relationship between the monitoring data indicators and risk factors of the monitoring project. The following table shows some examples.

[0100]

[0101] Step 2.2.2, randomly extract data samples from historical monitoring data and normalize the sample data. The formula is as follows:

[0102]

[0103] Among them, x max is the maximum value in the sample, x min is the minimum value in the sample, x i is a sample data, x′ is the normalized data;

[0104] Step 2.2.3, use the radial basis kernel function as the kernel function of the support vector machine to classify the historical monitoring data, and use the GA algorithm to calculate the optimal parameters gamma and C of the radial basis kernel function. The parameter gamma is the inverse of the influence radius of the support vector sample, and the parameter C is the penalty relaxation variable. The core code is as follows:

[0105] load monitoring_SVM; #monitoring SVM monitoring data support vector machine

[0106] Training data ①model=svmtrain(train_label,train_data,'options'); #train_label training set label, train_data training set attribute matrix, options parameter options include:

[0107] s - svm type;

[0108] t-kernel function type;

[0109] The degree setting in the d-degree kernel function (for polynomial kernel functions);

[0110] g-gama: gamma function setting in kernel function (for polynomial / rbf / sigmoid kernel function);

[0111] r-coef0: coef0 setting in the kernel function (for polynomial / sigmoid kernel function);

[0112] c-cost: sets the parameters (loss function) of C-SVC, epsilon-SVR and nu-SVC;

[0113] n-nu: Set the parameters of nu-SVC, one-class SVM and nu-SVR;

[0114] p-epsilon: Set the value of the loss function epsilon in epsilon-SVR;

[0115] m-cachesize: Set the cache memory size in MB;

[0116] e-eps: Set the allowed termination criteria;

[0117] h-shrinking: whether to use heuristics, 0 or 1;

[0118] wi-weight: Set the parameter C of the class to weight*C (C in C-SVC);

[0119] Prediction data②[predict_label,accuracy,dec_value]=svmpredict(test_label,test_data,model); #predict_label prediction set label, accuracy represents classification accuracy, dec_value is decision value, test_label represents test set label, test_data represents test set attribute matrix, model=svmtrain(train_label,train_data,'options');

[0120] ③[bestacc,bestc,bestg]=ga SVMcg For Class[train_monitoring_labels,train_monitoring,ga_parameter]

[0121] Among them, bestacc represents the best classification accuracy, bestc represents the best parameter c, bestg represents the best parameter g, ga SVMcg For Class represents genetic algorithm parameter optimization, train_monitoring represents the training sample set; train_monitoring_labels represents the label of the training sample set; ga_parameter is used to store the parameters obtained during the GA calculation optimization process.

[0122] Use Matlab software for programming.

[0123] Step 2.2.4, use the Sigmoid function to obtain the classification results and classification probabilities of the posterior probability calculation. The posterior probability formula is as follows:

[0124]

[0125] Where f(x) is the output of the decision function of the GS risk classification model; A and B are the Sigmoid parameters to be fitted; P(y=1|f(x)) is the degree to which the sample belongs to category 1 under the condition of the output value f(x);

[0126] The parameters A and B are calculated using the following formula:

[0127]

[0128] in, Indicates The values ​​of parameters A and B when they are at their minimum values; t i is the target probability, i=1,2,…,n,N + and N - are the number of positive and negative examples in the model respectively; y i represents the sample, and n represents the total number of samples;

[0129] Step 2.2.5, randomly extracting predicted monitoring data from the above historical monitoring data excluding sample data; substituting the predicted monitoring data into the GS risk classification model to verify the model evaluation results;

[0130] Step 2.2.6, substitute the monitored support system mechanical property data and surrounding rock stability data into the GS risk classification model to calculate the risk classification results and risk probability, thereby determining the risk level of the monitored area;

[0131] The risk level of the monitored area is determined accordingly. The classification of safety risk levels is shown in the table below.

[0132]

[0133] The risk probability is 0-0.25, the risk level is level 1, and the risk is negligible;

[0134] The risk probability is 0.25-0.5, the risk level is level 2, and the risk is acceptable;

[0135] The risk probability is 0.5-0.8, the risk level is level 3, and the risk needs to be handled;

[0136] The risk probability is 0.8-1, the risk level is level four, the risk is unacceptable and needs to be dealt with immediately.

[0137] The database uses SQL Server as the backend data storage management system to store monitoring data and the risk level and probability of the monitoring area, and designs the table structure in the database to ensure the smoothness of monitoring information reading and storage and the completeness of the database. The data table structure design is shown in the following table:

[0138]

[0139] The traditional support vector machine algorithm has a good effect on linearly separable classification calculations, but the operation is more complicated for nonlinear separable problems. The GS risk classification model in this step introduces the idea of ​​posterior probability, which not only solves the multi-classification problem, but also obtains the probability of the classification result by combining the Bayesian decision rule. Comprehensively judging the urgency of the accident through the classification results and classification probabilities can help managers improve the scientific nature of decision-making and is more valuable for practical engineering.

[0140] The database includes monitoring point information data table, monitoring data data table and risk information data table;

[0141] The monitoring point information data table is used to store the monitoring point code, monitoring type, monitoring area and monitoring point warning value;

[0142] The monitoring data table is used to store monitoring point codes, monitoring data and monitoring time;

[0143] The risk information data table is used to store monitoring area codes, risk probabilities, risk levels and colors corresponding to the risk levels.

[0144] Step 3: Create a 3D BIM monitoring model based on the design parameters of the monitoring point CAD drawings;

[0145] like Figure 2As shown in the figure: Bentley's MicroStation software is used to create a three-dimensional BIM model according to the design parameters of the monitoring point CAD drawings, and the monitoring model is coded. The BIM monitoring model is obtained according to the construction joints during the tunnel construction process. The monitoring model between two construction joints is classified as a monitoring area, and a unique digital code is given to the monitoring area. The monitoring model is coded in the following format: construction section-monitoring area-monitoring object-monitoring type-serial number.

[0146] For example: 1-DKD15-ECCQ-JZC-064, DKD15 indicates the monitoring area.

[0147] The BIM monitoring model is exported to the BIM model in IFC format, i.e., the IFC monitoring model. The IFC monitoring model is uniquely identified by the globally unique identifier GUID. The original BIM monitoring model is the monitoring model divided according to the construction joints during the tunnel construction process. The monitoring model between the two construction joints is classified as a monitoring area, and a unique digital code is assigned to the monitoring area. Figure 2 As shown, the monitoring area is: DKD15, large span section 15.

[0148] Use ArchiCAD software to expand the IFC monitoring information of the IFC monitoring model to facilitate data storage. The expanded IFC monitoring information includes characteristic attributes such as monitoring point code, installation time, collection time, monitoring data type, warning value upper limit, warning value lower limit, monitoring area code, transmission protocol name and transmission protocol version number. For the original BIM monitoring model, use the MicroStation software export function to export it to an IFC format BIM model, that is, an IFC monitoring model. The IFC monitoring model uses a unique identifier GUID (Globally Unique Identifier) ​​for unique identification; use ArchiCAD software to expand the IFC monitoring information of the IFC monitoring model so that the monitoring data and information can be stored in the IFC model. Other BIM software such as Revit, 3dsMax, Rhino, CATIA, etc. can also be used to convert the BIM model into an IFC model.

[0149] The IFC monitoring model monitoring information extension refers to supplementing the characteristic attributes of each monitoring model in the native framework of IFC to facilitate data storage. The IFC extended monitoring information format is shown in the following table.

[0150]

[0151]

[0152] Step 4, the visual early warning platform loads the extended IFC monitoring model; through the correspondence between the GUID identifier of the IFC monitoring model and the monitoring point code, the data in the monitoring point information data table, the monitoring data data table and the risk information data table in the database are read into the extended IFC monitoring model; the extended IFC monitoring model is loaded on the visual early warning platform, the monitoring point data information and risk level and probability in the SQL Server database are called, and the data information is attached to the IFC monitoring model through the correspondence between the GUID identifier of the IFC monitoring model and the monitoring point code. The GUID identifier is a specific unique identifier in the IFC monitoring model.

[0153] like Figure 4 As shown: In the visual early warning platform, you can click to view the monitoring data curve, monitoring point related information, monitoring data over-limit reminder, 3D model real-time color and risk level information, photos and plans after construction defect treatment. Click the monitoring point in the visual early warning platform to display the monitoring curve of the monitoring point and the surrounding rock displacement monitoring data curve.

[0154] The visual early warning platform retrieves the monitoring risk level and color type calculated by the risk assessment module from the SQL Server database in real time.

[0155] When the risk level is level one, the BIM monitoring model is displayed in the initial color;

[0156] When the risk level is level 2, the BIM monitoring model is displayed in yellow;

[0157] When the risk level is level three, the BIM monitoring model is displayed in orange;

[0158] When the risk level is level 4, the BIM monitoring model is displayed in red;

[0159] When the risk level is level 3 or 4, or the monitoring data is greater than the set warning threshold, an early warning will be issued. The warning information includes the monitoring area, risk level, problem description, and quality verification task information reminder;

[0160] The visual early warning platform retrieves monitoring data from the SQL Server database in real time, and associates the monitoring data with the BIM model by monitoring the correspondence between the model component code and the GUID identifier; and retrieves the design thresholds of various monitoring indicators in the SQL Server database and compares them with the monitoring data. When the risk level is level three or level four or the monitoring data is greater than the set design early warning threshold, the visual early warning platform issues warning reminders and quality verification tasks.

[0161] When the risk level is level three or level four, or the monitoring data is greater than the set warning threshold, a quality verification instruction will be issued and supervision personnel will be arranged to conduct inspections at the construction site.

[0162] Step 5. After the supervisor receives the verification instruction issued by the visual early warning platform, the supervisor goes to the construction site and uses the mobile terminal to take photos of the quality defect areas and initiates a remediation task. After the construction unit receives the remediation task, it handles the defective area. The supervisor takes photos of the tunnel site after treatment and uploads the treatment method. The supervisor copies the photos and plans to the supervisor leader and the owner for review. The approved photos and plans are displayed on the visual early warning platform. The data dictionary for supervisors to perform tasks using the mobile terminal is shown in the following table. Supervisors issue tasks on the mobile terminal as follows: Figure 6 As shown, the construction site information is uploaded as Figure 7 shown.

[0163]

[0164]

[0165] When the supervisor uploads the processed photos and plans on the mobile terminal, he / she checks the monitoring area code and monitoring point code where the construction defect is located, and uses the relationship established between the code and the visualization platform monitoring model. The quality defect treatment photos and plans can be displayed on the visualization early warning platform, such as Figure 8 shown.

[0166] According to the characteristics of tunnel engineering projects, the BIM-based tunnel engineering safety monitoring visualization and risk prediction method of the present invention can realize the visual linkage between monitoring data and BIM models, and automatically calculate, dynamically update and visualize the safety risks of tunnel engineering construction, thereby improving the risk warning effect and reducing economic losses.

[0167] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A BIM-based tunnel engineering safety risk monitoring system, characterized by: It includes on-site monitoring module, data server, risk assessment module, database, BIM monitoring model creation module, IFC monitoring model extension module, visual early warning platform, and mobile terminal quality inspection module; The field monitoring module is arranged in the monitoring area of ​​the tunnel construction site, and is used to monitor the mechanical performance data of the support system and the surrounding rock stability state data of the monitoring points in the current monitoring area; and transmit the data to the data server; The support system mechanical property data and surrounding rock stability state data acquired by the data server are sent to the risk assessment module in real time; the risk assessment module is used to calculate and analyze the support system mechanical property data and surrounding rock stability state data, determine the risk assessment results of the monitoring points in the monitoring area, and transmit them to the database; The BIM monitoring model creation module is used to design parameters according to the monitoring point CAD drawings; Create a three-dimensional BIM monitoring model, convert the three-dimensional BIM monitoring model into an IFC monitoring model through the IFC monitoring model extension module, and extend the IFC monitoring information of the IFC monitoring model to obtain the extended IFC monitoring model; The visual early warning platform is used to load the IFC monitoring model to visualize it, generate risk assessment results, and issue quality inspection tasks to be transmitted to the mobile terminal quality inspection module. The mobile terminal quality inspection module is used for supervisors to receive and process quality inspection tasks.

2. The BIM-based tunnel engineering safety risk monitoring system according to claim 1 is characterized in that: The mechanical performance data of the support system monitored by the on-site monitoring module can include: anchor axial force, steel frame internal force, support internal stress, lining concrete stress, lining concrete strain, contact pressure between initial support and secondary lining, and concrete external load; The surrounding rock stability data include: displacement in the surrounding rock mass, vault settlement, geological and support observations, surface settlement, settlement of surrounding buildings, tunnel water pressure, pore water pressure, surrounding rock pressure and tunnel uplift.

3. The monitoring method of the BIM-based tunnel engineering safety risk monitoring system according to any one of claims 1 to 2, characterized in that: Here are the steps: Step 1: Collect the mechanical property data of the support system and the stability state data of the surrounding rock at the monitoring points in the monitoring area through the on-site monitoring module; Step 2, computing and analyzing the monitoring data: using a data server to collect the support system mechanical performance data and surrounding rock stability data of the monitoring point in the monitoring area and send them to the risk assessment module, determine the risk assessment results of the monitoring area where the monitoring point is located through the risk assessment module, and transmit the risk assessment results and monitoring data to the database in real time; the risk assessment results include safety risk level and risk probability; Step 3: Create a three-dimensional BIM monitoring model according to the design parameters of the monitoring point CAD drawings, and encode the components of the monitoring model; convert the BIM monitoring model into an IFC monitoring model, and extend the IFC monitoring information of the IFC monitoring model to obtain the extended IFC monitoring model; Step 4: Load the extended IFC monitoring model according to the visual early warning platform, associate the monitoring data in the database and the risk assessment results of the monitoring area with the extended IFC monitoring model; issue early warnings and quality verification tasks according to the risk level; Step 5: Supervisors receive and process quality inspection tasks.

4. The monitoring method of the BIM-based tunnel engineering safety risk monitoring system according to claim 3 is characterized in that: In step 2, the monitoring data is calculated and analyzed to determine the risk assessment results of the monitoring area where the monitoring point is located, including: Step 2.1, count the safety accidents during the tunnel construction process, and classify and summarize the safety accidents according to the tunnel accident type, i.e. historical monitoring data; The process of tunnel construction is decomposed structurally, and a list of tunnel engineering safety risk factors is constructed based on the above classified tunnel accident types; Based on the above-mentioned list of tunnel engineering safety risk factors, a risk factor network topology structure is constructed using the Bayesian network topology structure; The fuzzy theory is used to calculate the parameters of each node of the risk factor network topology structure, and the Bayesian network topology structure is used to quantitatively analyze the safety risk factors of tunnel engineering to obtain the risk factor weight value; Step 2.2, based on the network topology of risk factors, establish the corresponding relationship between monitoring data and risk factors, use the GS risk classification model to calculate the risk classification results and risk probability corresponding to the monitoring data, and thus determine the risk level; Step 2.2.1, randomly extract data samples from the above historical monitoring data, and normalize the sample data. The formula is as follows: Among them, x max is the maximum value in the sample, x min is the minimum value in the sample, x i is a sample data, x′ is the normalized data; Step 2.2.2, use the radial basis kernel function as the kernel function of the support vector machine (SVM) to classify the historical monitoring data, and use the genetic algorithm (GA) to calculate the optimal parameters gamma and C of the radial basis kernel function, where the parameter gamma is the inverse of the influence radius of the support vector sample and the parameter C is the penalty relaxation variable; construct a GS risk classification model; Step 2.2.3, use the Sigmoid function to obtain the classification results and classification probabilities of the posterior probability calculation. The posterior probability formula is as follows: Where f(x) is the output of the decision function of the GS risk classification model; A and B are the Sigmoid parameters to be fitted; P(y=1|f(x)) is the degree to which the sample belongs to category 1 under the condition of the output value f(x); The parameters A and B are calculated using the following formula: in, Indicates The values ​​of parameters A and B when they are at their minimum values; t i is the target probability, i=1,2,…,n,N + and N - are the number of positive and negative examples in the model respectively; y i represents the sample, and n represents the total number of samples; Step 2.2.4, randomly extracting predicted monitoring data from the historical monitoring data excluding the sample data; substituting the predicted monitoring data into the GS risk classification model to verify the model evaluation results; Step 2.2.5, substitute the monitored support system mechanical property data and surrounding rock stability state data into the GS risk classification model to calculate the risk classification results and risk probability, so as to determine the risk level of the monitoring area; The risk probability is 0-0.25, the risk level is level 1, and the risk is negligible; The risk probability is 0.25-0.5, the risk level is level 2, and the risk is acceptable; The risk probability is 0.5-0.8, the risk level is level 3, and the risk needs to be handled; The risk probability is 0.8-1, the risk level is level four, the risk is unacceptable and needs to be dealt with immediately.

5. The monitoring method of the BIM-based tunnel engineering safety risk monitoring system according to claim 3 is characterized in that: In step 2, the database includes a monitoring point information table, a monitoring data table, and a risk information table. The monitoring point information data table is used to store monitoring point codes, monitoring types, monitoring areas and monitoring point warning values; The monitoring data table is used to store monitoring point codes, monitoring data and monitoring time; The risk information data table is used to store monitoring area codes, risk probabilities, risk levels and colors corresponding to the risk levels.

6. The monitoring method of the BIM-based tunnel engineering safety risk monitoring system according to claim 5 is characterized in that: The step 3 creates a three-dimensional BIM monitoring model according to the design parameters of the monitoring point CAD drawings. The BIM monitoring model is obtained according to the construction joint division during the tunnel construction process, and the monitoring model between two construction joints is classified into one monitoring area, and a unique digital code is given to the monitoring area; Export the BIM monitoring model to a BIM model in IFC format, namely, an IFC monitoring model; the IFC monitoring model is uniquely identified using a globally unique identifier GUID; Use ArchiCAD software to expand the IFC monitoring model with IFC monitoring information; The extended IFC monitoring information includes monitoring point code, installation time, collection time, monitoring data type, warning value upper limit, warning value lower limit, monitoring area code, transmission protocol name, transmission protocol version number characteristic attributes; In step 4, the visual early warning platform loads the IFC monitoring model after information expansion; through the correspondence between the GUID identifier of the IFC monitoring model after information expansion and the monitoring point code, the data in the monitoring point information data table, the monitoring data data table, and the risk information data table in the database are read into the expanded IFC monitoring model; The visual early warning platform displays monitoring data curves, monitoring point information, monitoring data over-limit reminders, IFC monitoring model real-time colors, and risk level information; When the risk level is level one, the BIM monitoring model is displayed in the initial color; When the risk level is level 2, the BIM monitoring model is displayed in yellow; When the risk level is level three, the BIM monitoring model is displayed in orange; When the risk level is level 4, the BIM monitoring model is displayed in red; When the risk level is level 3 or 4, or the monitoring data is greater than the set warning threshold, an early warning will be issued. The warning information includes the monitoring area, risk level, problem description, and quality verification task information reminder; When the risk level is level three or level four, or the monitoring data is greater than the set warning threshold, a quality verification instruction will be issued and supervision personnel will be arranged to conduct inspections at the construction site.

Images