Tunnel construction environment monitoring and early warning method and system
By building a geological risk database and using sensor monitoring node clusters, combining convolutional networks and RBF neural networks, the problems of transmission and utilization of tunnel construction environment monitoring data are solved, real-time monitoring and risk identification of tunnel construction environment are realized, and early warning is improved Validity.
Patent Information
- Application Number
- CN202411433859.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-15
- Publication Date
- 2024-11-15
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing technology lacks effective data transmission and utilization methods in tunnel construction environment monitoring and early warning, resulting in interruption of monitoring data and slow transmission speed, which cannot meet the real-time early warning needs. Especially under the complex geological conditions of the western mountainous areas, monitoring areas increase and data Difficulty in handling.
By building a geological risk database, arranging sensors to form a monitoring node cluster, integrating data and transmitting it to the data processing center through a wireless network, and using a convolutional network and RBF neural network for data fitting and risk identification, real-time risk warning is achieved.
It improves the transmission speed of monitoring data and the real-time warning, realizes real-time monitoring and risk identification of the tunnel construction environment, and reduces the probability of accidents.
Smart Images

Figure CN118966797A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data monitoring and analysis, and in particular to a tunnel construction environment monitoring and early warning method and system. Background Art
[0002] With the increase in national infrastructure construction, especially the continuous increase in investment in road transportation in the western region, construction teams are increasingly required to work under complex mountainous conditions in the west. Tunnel construction is often required to build roads in the west, especially in the southwestern mountainous areas. These places are often in geologically complex areas. During the tunnel construction process, there may be dangerous geological conditions such as high rock temperature, high-temperature hot water, water gushing, mud gushing or gas leakage. Therefore, it is becoming more and more important to detect and warn the tunnel construction environment.
[0003] At present, the monitoring and early warning of tunnel construction environment are usually carried out by construction personnel through visual observation and limited environmental data sampling in relevant areas to obtain relevant tunnel environmental data. This usually requires a large amount of manpower to collect data in order to realize the monitoring of tunnel construction environment. In addition, due to the requirements of construction progress, there is no way to carry out long-term continuous monitoring, which easily leads to intermittent monitoring data and cannot meet the needs of tunnel construction early warning. At the same time, with the excavation of the tunnel, the area that needs to be monitored is gradually increasing, and the monitoring of the entire construction environment will become more difficult. In addition, due to the increase in monitoring data, the preservation and use of data is also a problem. In particular, how to quickly and accurately transmit these monitoring data to the data processing center and reasonably use these data for analysis and early warning is a problem that needs to be solved urgently. Summary of the invention
[0004] One of the purposes of the present invention is to provide a tunnel construction environment monitoring and early warning method, and to solve the shortcomings of the prior art in that there is a lack of transmission of tunnel construction environment monitoring data and use of monitoring data for early warning.
[0005] The present invention is implemented through the following technical scheme. A tunnel construction environment monitoring and early warning method includes the following steps: S100, constructing a geological risk database based on exploration data, and marking the risk types and safety standard thresholds of different areas in the database, arranging corresponding sensors in different areas to form sensor monitoring nodes, integrating different sensor monitoring nodes to form an environmental monitoring node cluster, and the environmental monitoring node cluster processes the data and sends the data to a data processing center; S200, the data processing center decompresses the received data, standardizes the decompressed data, and constructs the standardized data into an environmental database; S300, fitting the sensor monitoring data in the environmental database and the data in the geological risk database to obtain a tunnel risk characterization parameter model, using the risk characterization parameter model as an input sample, and inputting the convolutional network to extract feature data; S400, combining the data in the environmental database, and identifying the tunnel construction environment in real time through the relationship between the number of hidden layer nodes and the evaluation value of different risks, and finally obtaining the probability of risk occurrence of the risk characterization parameter model, and issuing an early warning based on the probability of risk occurrence.
[0006] Furthermore, the environmental monitoring node cluster deploys and plans the wireless network nodes for wireless data transmission based on the locations of the sensor monitoring nodes, including the following sub-steps: S110, deploy corresponding wireless network nodes according to the locations of the sensor monitoring nodes, aggregate these wireless network nodes to form a wireless network node set, and the wireless network nodes send monitoring data to the aggregated wireless network nodes based on the measurement data that the sensor monitoring nodes need to transmit and their own transmission rates; S120, the wireless network node set pre-processes the received monitoring data, first cleans the data, checks whether there are duplicate records in the data according to the timestamp, deletes the duplicate data, and then classifies the data; data classification includes, first classifying according to different sensor types in different areas, then dividing the data of different sensors into different time periods according to the timestamp, and finally sorting the monitoring data according to the safety threshold of the monitoring indicators, and giving priority to transmitting important data according to the sorting; S130, compressing the data after completing the data classification, and sending the compressed data to the data processing center through the TCP / IP protocol.
[0007] Furthermore, the standardization process can be performed using a min-max algorithm, which includes the following formula: ,in, For the value of a single data, x needs to be standardized. min is the minimum value of the column where the data to be standardized is located, x max The maximum value of the column where the data to be standardized is located.
[0008] Furthermore, the convolutional network includes two convolutional network layers G1 and G2, wherein the G1 convolutional network layer is used to extract the characteristic information implied by each parameter in the tunnel risk characterization parameter model; the G2 convolutional network layer is used to extract the implicit connection between the parameters in the tunnel risk characterization parameter model.
[0009] Furthermore, extracting characteristic data includes collecting the characteristic information implied by each parameter itself and the implicit relationship between each parameter as characteristic data, inputting them into the RBF neural network for classification, and then using the hidden layer selection rules of the RBF neural network to select the relationship between the number of hidden layer nodes and evaluation values of different regional risks from the classified characteristic data.
[0010] Further, the hidden layer selection rule includes, S310, determining a basic network model, which includes a convolutional network and an RBF neural network, and determining the number of convolutional network input layers and the number of RBF neural network output layers; S320, according to the formula, To determine the initial number of hidden layer nodes of the RBF neural network, where l is the number of input nodes of the RBF network; r is the number of hidden layer nodes; k is the number of output layer nodes of the RBF network; is a constant coefficient, and its value range is 0-10; S330, then the basic network model is trained to obtain the final training time And the accuracy of the basic network model test set , according to the training time And the accuracy of the test set of this network model , by using , calculate the corresponding evaluation values of different hidden layer node numbers ,in and are constant coefficients respectively; S340, according to different risk types in tunnel construction, set different hidden layer node increments , and repeat S330 to obtain a new evaluation value When the new evaluation value continues to decrease or tends to be stable, stop increasing the number of hidden layer nodes; S350, use the min-max algorithm to standardize the evaluation values corresponding to different numbers of hidden layer nodes, and use the standardized evaluation values The number of hidden layer nodes corresponding to the point with 100% accuracy is taken as the optimal number of nodes.
[0011] Another aspect of the present invention provides a tunnel construction environment monitoring and early warning system, which includes a database construction unit, a data processing unit and an early warning unit.
[0012] Among them, the database construction unit is configured to construct a geological risk database based on the exploration data, and mark the risk types and safety standard thresholds of different areas in the database, arrange corresponding sensors in different areas to form sensor monitoring nodes, integrate different sensor monitoring nodes to form an environmental monitoring node cluster, and the environmental monitoring node cluster processes the data and sends the data to the data processing center, the data processing center decompresses the received data, standardizes the decompressed data, and constructs the standardized data into an environmental database; the data processing unit is connected to the database construction unit, and is configured to fit the sensor monitoring data in the environmental database and the data in the geological risk database to obtain a tunnel risk characterization parameter model, and use the risk characterization parameter model as an input sample to input the convolutional network to extract feature data; the early warning unit is connected to the data processing unit, and is configured to, in combination with the data in the environmental database, perform real-time risk identification of the tunnel construction environment through the relationship between the number of hidden layer nodes and the evaluation value of different risks, and finally obtain the probability of risk occurrence of the risk characterization parameter model, and issue an early warning based on the probability of risk occurrence.
[0013] Furthermore, the data processing unit also includes a feature data extraction subunit and a hidden layer selection subunit.
[0014] Among them, the feature data extraction subunit is configured to collect the feature information implied by each parameter itself and the implicit relationship between each parameter as feature data, input it into the RBF neural network for classification, and then use the hidden layer selection rule of the RBF neural network to select the relationship between the number of hidden layer nodes and the evaluation value of different regional risks from the classified feature data; the hidden layer selection subunit is configured to determine the basic network model, which includes the convolutional network and the RBF neural network, and at the same time determine the number of convolutional network input layers and the number of RBF neural network output layers; according to the formula, To determine the initial number of hidden layer nodes of the RBF neural network, where l is the number of input nodes of the RBF network; r is the number of hidden layer nodes; k is the number of output layer nodes of the RBF network; is a constant coefficient with a value range of 0-10; then the basic network model is trained to obtain the final training time And the accuracy of the basic network model test set , according to the training time And the accuracy of the test set of this network model , by using , calculate the corresponding evaluation values of different hidden layer node numbers ,in and are constant coefficients respectively; according to different risk types in tunnel construction, different increments of hidden layer nodes are set , and repeat S330 to obtain a new evaluation value When the new evaluation value continues to decrease or tends to be stable, stop increasing the number of hidden layer nodes; use the min-max algorithm to standardize the evaluation values corresponding to different numbers of hidden layer nodes, and use the standardized evaluation values The number of hidden layer nodes corresponding to the point with 100% accuracy is taken as the optimal number of nodes.
[0015] In another aspect, the present invention provides a computer device, comprising: a processor; a memory storing a computer program, and when the computer program is executed by the processor, the tunnel construction environment monitoring and early warning method as described above is implemented.
[0016] In another aspect, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the tunnel construction environment monitoring and early warning method as described above.
[0017] Compared with the prior art, the present invention has the following advantages and beneficial effects:
[0018] 1. The present invention divides the risk areas in the tunnel, deploys different sensors respectively and divides these sensors into different monitoring nodes, integrates the wireless network aggregation node into the environmental monitoring node cluster, pre-processes the data in the node cluster, and transmits the pre-processed data to the data processing center based on the TCP / IP protocol, which can effectively improve the transmission speed of the monitoring data and ensure the real-time and effectiveness of the early warning.
[0019] 2. The present invention realizes real-time monitoring of the tunnel environment by collecting monitoring data of the tunnel environment and analyzing the monitoring data. At the same time, the monitoring data is extracted and screened through a neural network, and risk control and early warning are realized through feature data combined with a risk characterization parameter model. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of this application, and do not constitute a limitation of the embodiments of the present invention. In the drawings:
[0021] Figure 1 This is a flow chart of the method provided in Example 1 of the present invention.
[0022] Figure 2 A schematic diagram of the system structure provided in Example 2 of the present invention.
[0023] Figure 3 This is a schematic diagram of the system structure provided in Example 3 of the present invention.
[0024] Description of reference numerals:
[0025] 100 - tunnel construction environment monitoring and early warning system, 110 - database construction unit, 120 - data processing unit, 130 - early warning unit, 121 - feature data extraction subunit, 122 - hidden layer selection subunit, 200 - computer equipment, 210 - memory, 220 - processor. DETAILED DESCRIPTION
[0026] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions 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. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.
[0027] Example 1
[0028] like Figure 1 As shown, in this embodiment, a tunnel construction environment monitoring and early warning method is disclosed, including the following steps:
[0029] Step 1: Organize the preliminary exploration data into a geological risk database, mark the risk types and safety standard thresholds of different areas in the database, and then divide different types and levels of monitoring locations according to the specific tunnel construction conditions. (For example, when constructing tunnels in unfavorable geological areas such as fault fracture zones, it is mainly necessary to prevent sudden situations such as water gushing or mud bursting on the tunnel excavation surface. Therefore, key monitoring areas can be divided in advance according to the preliminary address exploration conditions, and corresponding sensors can be arranged.) Select different sensors according to the divided monitoring locations, and deploy them in different areas of the tunnel according to the construction monitoring needs to form sensor monitoring nodes. Different sensor monitoring nodes are integrated to form a tunnel construction environment monitoring node cluster.
[0030] Among them, different levels can be set according to the specific construction conditions, and can generally be set to three levels, namely low-risk area, medium-risk area and high-risk area; it can also be divided into different types of dangers according to the different dangerous situations to be prevented. For example, the water gushing risk area can be divided into three levels: low, medium and high; the gas leakage risk area can be divided into three levels: low, medium and high.
[0031] Specifically, the tunnel construction environment monitoring node cluster can deploy and plan the wireless network nodes for wireless data transmission based on the locations of the sensor monitoring nodes, including the following sub-steps:
[0032] According to the location of the sensor monitoring node, the corresponding wireless network nodes are deployed, and these wireless network nodes are aggregated to form a wireless network node set. The wireless network node sends monitoring data to the aggregated wireless network node based on the measurement data that the sensor monitoring node needs to transmit and its own transmission rate. It should be noted that the environment in the tunnel is complex. The data collected by the sensor is transmitted through the wireless network nodes, which does not require a lot of wiring operations and can effectively improve the transmission efficiency.
[0033] The wireless network node set preprocesses the received monitoring data. Specifically, the preprocessing includes data cleaning and data classification. First, the data is cleaned, and the data is checked for duplicate records according to the timestamp, and the duplicate data is deleted, and then the data is classified.
[0034] Data classification, first classifies according to different sensor types in different areas, then divides the data of different sensors into different time periods according to timestamps, and finally sorts the monitoring data according to the safety threshold of the monitoring indicators, and transmits important data first according to the sorting.
[0035] After completing data classification, the data is compressed and sent to the data processing center via TCP / IP protocol.
[0036] It should be noted that in the wireless network node set, the received data is pre-processed, divided into different levels, and the data of key areas are sent first to ensure the timeliness of monitoring and early warning.
[0037] Step 2: The data processing center decompresses the received data, standardizes the decompressed data, and constructs the standardized data into an environment database;
[0038] Specifically, the standardization process can be performed using the min-max algorithm, which includes the following formula:
[0039] ,in, For the value of a single data, x needs to be standardized. min is the minimum value of the column where the data to be standardized is located, x max The maximum value of the column where the data to be standardized is located.
[0040] It should be noted that the reason why the min-max algorithm is used for standardization is that the min-max algorithm has the characteristics of simple calculation method and easy-to-understand and implement formula, and it scales the data to a specified range through linear transformation, which is convenient for subsequent processing and analysis. At the same time, this method will not change the distribution form and relative relationship of the data, and can effectively retain the relationship of the original data, especially the maximum and minimum values of the data are still the maximum and minimum values in the standardized data. This feature has an important impact on the selection of the number of hidden layer nodes in the subsequent neural network algorithm and the final early warning judgment.
[0041] Step 3: Fit the sensor monitoring data in the environmental database and the data in the geological risk database to obtain the tunnel risk characterization parameter model. The risk characterization parameter model is used as an input sample and input into the convolutional network.
[0042] The convolutional network consists of two convolutional network layers G1 and G2, where the G1 convolutional network layer is used to extract the characteristic information implied by each parameter in the tunnel risk characterization parameter model; the G2 convolutional network layer is used to extract the implicit connection between the parameters in the tunnel risk characterization parameter model.
[0043] The characteristic information implied by each parameter itself and the implicit relationship between the parameters are collected as feature data and input into the RBF neural network for classification. Then, the hidden layer selection rule of the RBF neural network is used to select the relationship between the number of hidden layer nodes and evaluation values of different regional risks from the classified feature data.
[0044] Specifically, the hidden layer selection rules include: (1) first determining the basic network model, which includes a convolutional network and an RBF neural network, and at the same time determining the number of convolutional network input layers and the number of RBF neural network output layers.
[0045] (2) According to the formula, To determine the initial number of hidden layer nodes of the RBF neural network, where l is the number of input nodes of the RBF network; r is the number of hidden layer nodes; k is the number of output layer nodes of the RBF network; It is a constant coefficient, ranging from 0 to 10.
[0046] (3) Then train the basic network model to obtain the final training time And the accuracy of the basic network model test set , according to the training time And the accuracy of the test set of this network model , by using , calculate the corresponding evaluation values of different hidden layer node numbers ,in and are constant coefficients respectively;
[0047] (4) According to different risk types in tunnel construction, different increments of hidden layer nodes are set. , and repeat step (3) to get a new evaluation value , when the new evaluation value continues to decrease or tends to be stable, stop increasing the number of hidden layer nodes;
[0048] (5) Use the min-max algorithm to standardize the evaluation values corresponding to different numbers of hidden layer nodes, and use the standardized evaluation values The number of hidden layer nodes corresponding to the point with 100% accuracy is taken as the optimal number of nodes.
[0049] It should be noted that in step 3, the risk characterization parameter model is obtained by first fitting the sensor monitoring data and geological risk data, and then the convolutional neural network (CNN) is used to automatically extract the complex features in the data, which can effectively improve the accuracy and robustness of the model. Especially when there is a large amount of data, the use of convolutional neural networks can quickly realize feature extraction. And through different convolutional layers (G1, G2), the characteristics of the sensor parameters themselves and the relationship between the parameters can be captured, thereby forming a more comprehensive risk characterization, and enhancing the depth and breadth of the risk characterization parameter model. Finally, the RBF neural network is combined for classification, so as to realize the rapid identification of different types of risks, provide timely warning information for the construction team, and reduce the probability of accidents. The entire model has the ability of self-learning and optimization. By continuously inputting new monitoring data, the model can continuously optimize itself and improve the accuracy and adaptability of risk identification.
[0050] Step 4: Combined with the data in the environmental database, the real-time risk identification of the tunnel construction environment is carried out through the relationship between the number of hidden layer nodes and the evaluation value of different risks, and finally the probability of risk occurrence of the risk characterization parameter model is obtained. When the probability of occurrence is greater than 50%, the construction personnel are reminded to pay attention to the specific situation of the corresponding area.
[0051] The above steps form a complete monitoring and early warning system by combining data integration and analysis with risk identification, which can effectively realize the monitoring and early warning of the tunnel construction environment.
[0052] Example 2
[0053] like Figure 2 As shown, the tunnel construction environment monitoring and early warning system 100 includes a database construction unit 110 , a data processing unit 120 and an early warning unit 130 .
[0054] The database construction unit 110 is configured to construct a geological risk database based on the exploration data, and mark the risk types and safety standard thresholds of different areas in the database, arrange corresponding sensors in different areas to form sensor monitoring nodes, integrate different sensor monitoring nodes to form an environmental monitoring node cluster, and the environmental monitoring node cluster processes the data and sends the data to the data processing center. The data processing center decompresses the received data, standardizes the decompressed data, and constructs the standardized data into an environmental database.
[0055] The data processing unit 120 is connected to the database construction unit 110, and is configured to fit the sensor monitoring data in the environmental database and the data in the geological risk database to obtain a tunnel risk characterization parameter model, and use the risk characterization parameter model as an input sample to input the convolutional network to extract feature data.
[0056] Specifically, the data processing unit 120 may further include a feature data extraction subunit 121 and a hidden layer selection subunit 122 .
[0057] The feature data extraction subunit 121 is configured to collect the feature information implied by each parameter itself and the implicit relationship between the parameters as feature data, input them into the RBF neural network for classification, and then use the hidden layer selection rules of the RBF neural network to select the number of hidden layer nodes and the evaluation value relationship of different regional risks from the classified feature data.
[0058] The hidden layer selection subunit 122 is configured to determine a basic network model, which includes a convolutional network and an RBF neural network, and also determine the number of convolutional network input layers and the number of RBF neural network output layers; according to the formula, To determine the initial number of hidden layer nodes of the RBF neural network, where l is the number of input nodes of the RBF network; r is the number of hidden layer nodes; k is the number of output layer nodes of the RBF network; is a constant coefficient with a value range of 0-10; then the basic network model is trained to obtain the final training time And the accuracy of the basic network model test set , according to the training time And the accuracy of the test set of this network model , by using , calculate the corresponding evaluation values of different hidden layer node numbers ,in and are constant coefficients respectively; according to different risk types in tunnel construction, different increments of hidden layer nodes are set , and repeat S330 to obtain a new evaluation value When the new evaluation value continues to decrease or tends to be stable, stop increasing the number of hidden layer nodes; use the min-max algorithm to standardize the evaluation values corresponding to different numbers of hidden layer nodes, and use the standardized evaluation values The number of hidden layer nodes corresponding to the point with 100% accuracy is taken as the optimal number of nodes.
[0059] The early warning unit 130 is connected to the data processing unit 120, and is configured to combine the data in the environmental database, perform real-time risk identification on the tunnel construction environment through the relationship between the number of hidden layer nodes of different risks and the evaluation value, and finally obtain the probability of risk occurrence of the risk characterization parameter model, and issue an early warning based on the probability of risk occurrence.
[0060] Example 3
[0061] like Figure 3 As shown, a computer device 200 includes a memory 210 and a processor 220. The memory 210 stores a computer program. When the computer program is executed by the processor, a tunnel construction environment monitoring and early warning method as described in Example 1 is implemented.
[0062] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.
Claims
1. A tunnel construction environment monitoring and early warning method, It is characterized in that The monitoring and early warning method comprises: S100, constructing a geological risk database based on the exploration data, and marking the risk types and safety standard thresholds of different areas in the database, arranging corresponding sensors in different areas to form sensor monitoring nodes, integrating different sensor monitoring nodes to form an environmental monitoring node cluster, and the environmental monitoring node cluster processes the data and sends the data to the data processing center; S200, the data processing center decompresses the received data, performs standardization processing on the decompressed data, and constructs the standardized data into an environment database; S300, fitting the sensor monitoring data in the environmental database and the data in the geological risk database to obtain a tunnel risk characterization parameter model, and using the risk characterization parameter model as an input sample to input into a convolutional network to extract feature data; S400. Combined with the data in the environmental database, the real-time risk identification of the tunnel construction environment is carried out through the relationship between the number of hidden layer nodes of different risks and the evaluation value, and finally the probability of risk occurrence of the risk characterization parameter model is obtained, and an early warning is issued according to the probability of risk occurrence.
2. The tunnel construction environment monitoring and early warning method according to claim 1, It is characterized in that The environment monitoring node cluster performs deployment planning for wireless network nodes for wireless data transmission based on the locations of the sensor monitoring nodes, including the following sub-steps: S110, deploying corresponding wireless network nodes according to the positions of the sensor monitoring nodes, aggregating these wireless network nodes to form a wireless network node set, wherein the wireless network nodes send monitoring data to the aggregating wireless network nodes based on the measurement data required to be transmitted by the sensor monitoring nodes and their own transmission rates; S120, the wireless network node set pre-processes the received monitoring data, first cleans the data, checks whether there are duplicate records in the data according to the timestamp, deletes the duplicate data, and then classifies the data; the data classification includes first classifying according to different sensor types in different areas, then dividing the data of different sensors into different time periods according to the timestamp, and finally sorting the monitoring data according to the safety threshold of the monitoring indicator, and transmitting important data first according to the sorting; S130, after completing the data classification, compress the data and send the compressed data to the data processing center through the TCP / IP protocol.
3. The tunnel construction environment monitoring and early warning method according to claim 1, It is characterized in that The standardization process uses the min-max algorithm for standardization, and the min-max algorithm includes the following formula: ,in, For the value of a single data, x needs to be standardized. min is the minimum value of the column where the data needs to be standardized, x max The maximum value of the column where the data to be standardized is located.
4. The tunnel construction environment monitoring and early warning method according to claim 1, It is characterized in that The convolutional network includes two convolutional network layers G1 and G2. Among them, the G1 convolutional network layer is used to extract the characteristic information implied by each parameter in the tunnel risk characterization parameter model; the G2 convolutional network layer is used to extract the implicit connection between the parameters in the tunnel risk characterization parameter model.
5. The tunnel construction environment monitoring and early warning method according to claim 1, It is characterized in that The extracting feature data comprises: The characteristic information implied by each parameter itself and the implicit relationship between the parameters are collected as feature data and input into the RBF neural network for classification. Then, the hidden layer selection rule of the RBF neural network is used to select the relationship between the number of hidden layer nodes and evaluation values of different regional risks from the classified feature data.
6. The tunnel construction environment monitoring and early warning method according to claim 5, It is characterized in that The hidden layer selection rules include: S310, determining a basic network model, which includes a convolutional network and an RBF neural network, and determining the number of convolutional network input layers and the number of RBF neural network output layers; S320, according to the formula, To determine the initial number of hidden layer nodes of the RBF neural network, where l is the number of input nodes of the RBF network; r is the number of hidden layer nodes; k is the number of output layer nodes of the RBF network; is a constant coefficient, ranging from 0 to 10; S330, then train the basic network model to obtain the final training time And the accuracy of the basic network model test set , according to the training time And the accuracy of the test set of this network model , by using , calculate the corresponding evaluation values of different hidden layer node numbers ,in and are constant coefficients respectively; S340, according to different risk types in tunnel construction, set different hidden layer node number increments , and repeat S330 to obtain a new evaluation value , when the new evaluation value continues to decrease or tends to be stable, stop increasing the number of hidden layer nodes; S350, use the min-max algorithm to standardize the evaluation values corresponding to different numbers of hidden layer nodes, and the standardized evaluation values The number of hidden layer nodes corresponding to the point with 100% accuracy is taken as the optimal number of nodes.
7. A tunnel construction environment monitoring and early warning system, It is characterized in that The monitoring and early warning system includes: Database construction unit, data processing unit and early warning unit, wherein: The database construction unit is configured to construct a geological risk database based on the exploration data, and mark the risk types and safety standard thresholds of different areas in the database, arrange corresponding sensors in different areas to form sensor monitoring nodes, integrate different sensor monitoring nodes to form an environmental monitoring node cluster, and the environmental monitoring node cluster processes the data and sends the data to the data processing center, the data processing center decompresses the received data, standardizes the decompressed data, and constructs the standardized data into an environmental database; The data processing unit is connected to the database construction unit and is configured to fit the sensor monitoring data in the environmental database and the data in the geological risk database to obtain a tunnel risk characterization parameter model, and use the risk characterization parameter model as an input sample to input the convolutional network to extract feature data; The early warning unit is connected to the data processing unit and is configured to, in combination with the data in the environmental database, identify the risks of the tunnel construction environment in real time through the relationship between the number of hidden layer nodes of different risks and the evaluation value, and finally obtain the probability of risk occurrence of the risk characterization parameter model, and issue an early warning based on the probability of risk occurrence.
8. The tunnel construction environment monitoring and early warning system according to claim 7, It is characterized in that The data processing unit also includes a feature data extraction subunit and a hidden layer selection subunit, wherein: The feature data extraction subunit is configured to collect the feature information implied by each parameter itself and the relationship implied between the parameters as feature data, input them into the RBF neural network for classification, and then use the hidden layer selection rule of the RBF neural network to select the relationship between the number of hidden layer nodes and evaluation values of different regional risks from the classified feature data; The hidden layer selection subunit is configured to determine a basic network model, which includes a convolutional network and an RBF neural network, and also determine the number of convolutional network input layers and the number of RBF neural network output layers; according to the formula, To determine the initial number of hidden layer nodes of the RBF neural network, where l is the number of input nodes of the RBF network; r is the number of hidden layer nodes; k is the number of output layer nodes of the RBF network; is a constant coefficient with a value range of 0-10; then the basic network model is trained to obtain the final training time And the accuracy of the basic network model test set , according to the training time And the accuracy of the test set of this network model , by using , calculate the corresponding evaluation values of different hidden layer node numbers ,in and are constant coefficients respectively; according to different risk types in tunnel construction, different increments of hidden layer nodes are set , and repeat S330 to obtain a new evaluation value When the new evaluation value continues to decrease or tends to be stable, stop increasing the number of hidden layer nodes; use the min-max algorithm to standardize the evaluation values corresponding to different numbers of hidden layer nodes, and use the standardized evaluation values The number of hidden layer nodes corresponding to the point with 100% accuracy is taken as the optimal number of nodes.
Citation Information
Patent Citations
Bridge health monitoring method and system based on wireless sensor network
CN114894411A
Health portrait data acquisition method and cloud computing system
CN115512798A
Well drilling risk identification model, establishment method, identification method and computer equipment
CN115640526A
Construction engineering safety monitoring and early warning analysis method
CN116433024A
Tunnel safety risk identification method and device and safety management platform
CN117522149A
Cited By
Intelligent monitoring and early warning method and system for cofferdam construction process
CN119250547A
Tunnel construction environment detection system and method
CN120141567A
Construction safety monitoring method and equipment based on Internet of Things, and medium
CN121520543A
A construction safety monitoring method, device and medium based on the Internet of Things
CN121520543B
