Tunnel construction safety risk management intelligent decision support platform

Abstract

The invention discloses an intelligent decision support platform for tunnel construction safety risk management, which belongs to the field of tunnel construction risk management and mainly comprises a data collection module, a risk analysis module, a selection decision module, a risk feedback module and the like. Firstly, a data collection module collects relevant data including geological data, construction parameters, engineering environment and the like by monitoring parameters of a shield tunneling machine in real time; and then, the risk analysis module processes and analyzes the data based on an artificial intelligence algorithm, and generates risk early warning identification information at the same time. And in the selection decision module, decision analysis is carried out according to the early warning identification information, and a most reasonable risk disposal scheme is given. And the risk feedback module feeds back a processing result to the system to realize loop optimization. The system collects and analyzes construction data in real time, provides an effective decision-making basis, facilitates safe and effective management of tunnel construction, and improves the safety and efficiency of constructional engineering.

Description

Technical Field

[0001] The present invention belongs to the field of tunnel construction risk management, and more particularly relates to an intelligent decision support platform for tunnel construction safety risk management. Background Art

[0002] Safety risk management is a very important link in the process of tunnel construction. Because tunnel engineering involves many factors such as geology, structure, and construction, there are many uncertainties and risks. Traditional risk management methods mainly rely on manual judgment and experience, which cannot comprehensively and accurately identify and evaluate risks, are prone to loopholes, and are relatively complicated to operate.

[0003] With the development of science and technology, technologies such as artificial intelligence and data analysis are gradually applied to various fields. For the safety risk management of tunnel construction, the application of artificial intelligence and data analysis technology can achieve quantitative measurement and accurate analysis of various risk factors, improving the accuracy and efficiency of risk management. However, most of the existing risk management systems on the market have only achieved data collection and monitoring, and are not intelligent enough for data analysis and interpretation, risk prediction and early warning, and decision-making recommendations for risks.

[0004] In addition, most of the existing tunnel construction risk management systems are based on known risk factors for risk prevention, and are often insufficient in dealing with unknown and sudden risk factors. In addition, the existing systems are active in risk warning and risk feedback, but there is little research on how to choose a reasonable risk disposal plan for decision-making guidance.

[0005] Therefore, how to use advanced technology to develop a product that can not only accurately warn of tunnel construction risks, but also provide risk disposal decision support, and realize full process automation and intelligence of risk management is a major challenge and demand currently faced by the field of tunnel construction. Based on the above background technology and demand, the present invention proposes an intelligent decision support platform for tunnel construction safety risk management, in order to solve the problems existing in the prior art. Summary of the invention

[0006] The present invention aims to solve the problems existing in the traditional tunnel construction risk management methods, such as the difficulty in comprehensively and accurately identifying and assessing risks, the insufficient response to unknown and sudden risks, and the difficulty in providing real-time and effective risk disposal decision guidance. To this end, the present invention proposes an intelligent decision support platform for tunnel construction safety risk management, which uses modern data analysis and artificial intelligence technology to achieve comprehensive monitoring, accurate prediction, real-time warning of tunnel construction risks, and intelligent risk disposal decision suggestions, thereby improving the efficiency and accuracy of tunnel construction risk management.

[0007] In order to achieve the above object, the present invention is implemented by adopting the following technical solutions: the platform comprises: The data collection module collects geological data, construction parameters and engineering environment data by real-time monitoring of shield machine parameters; The risk analysis module processes and analyzes data based on artificial intelligence algorithms and generates risk warning identification information; Select the decision-making module, conduct decision analysis based on the early warning identification information, and provide the most reasonable risk disposal plan; The risk feedback module feeds back the processing results to the system to achieve cyclic optimization.

[0008] In one embodiment, the data collection module includes: Select specialized monitoring equipment appropriate to the characteristics of tunnel construction, including: sensors for measuring shield machines, geological survey equipment, and environmental monitoring equipment; The monitoring equipment monitors the tunnel construction process in real time and automatically collects measurement data, including: shield machine operating parameters: cutter head torque, thrust pressure, thrust speed; geological data: stratum type, soil properties; construction parameters: construction sequence, construction speed, construction quality; and engineering environment data: temperature, humidity, noise, vibration; After the data collection module collects the data, it performs preliminary processing on the original data to remove noise, eliminate outliers, and perform data calibration.

[0009] In one solution, the risk analysis module uses the LSTM algorithm for risk warning; the details are as follows: The risk analysis module reads the pre-processed construction data uploaded by the data collection module.

[0010] Extract features from the read data and use principal component analysis (PCA) to reduce the dimensionality of the original data; Input the feature vector obtained in the previous step into the pre-trained long short-term memory network LSTM for risk prediction; The warning threshold is set according to the prediction results of the model. Once the predicted risk indicator exceeds this threshold, the system automatically generates a warning message.

[0011] In one embodiment, the selection decision module includes: Reading of early warning information: The decision-making module first reads the early warning identification information generated by the risk analysis module, including the type of risk, risk level and impact range; Risk assessment: Based on the early warning identification information, the system uses the risk assessment model to conduct qualitative and quantitative evaluation of the risk level; Decision-making plan generation: Based on the risk assessment results, the system ranks all response measures; Decision result feedback: Select the decision module to feed back the final decision result to the executor to guide the actual risk response activities.

[0012] In one embodiment, the risk feedback module includes: The risk feedback module first collects the decision results generated by the selection decision module. This information includes the selected risk treatment plan, the expected implementation effect, and the problems and risks that arise; Implementation monitoring: When the selected risk treatment plan is implemented, its implementation status is monitored in real time; Feedback information processing: The risk feedback module analyzes and processes existing problems based on the collected feedback information; Model optimization: Based on the results of processing feedback information, the original risk prediction model and decision-making model are optimized so that these two modules can make more accurate predictions and decisions when processing subsequent data.

[0013] In one embodiment, the decision solution is generated using a linear programming algorithm: Assume that there are n risk response measures, the cost of each measure is c[i], the benefit is b[i], and the decision result is x[i], which can be expressed as the following linear programming problem: The goal is to minimize the total cost: , The constraint for all i is that the benefit must be greater than a threshold T: , For all i, the value of each decision outcome is between 0 and 1: , The specific values ​​of the above items are determined by the specific nature of the risk assessment and decision-making plan.

[0014] Beneficial effects of the present invention: Comprehensive data collection and processing: Through real-time monitoring equipment, comprehensive geological data, construction parameters and engineering environment data are collected and initially processed to improve the accuracy and effectiveness of the data.

[0015] Artificial intelligence algorithms are used to process and analyze data to achieve accurate prediction of risks. In addition, risk warning identification information can be automatically generated to achieve timely warning of risks.

[0016] According to the early warning identification information, the decision-making module is selected to conduct decision analysis on risk handling and give the most reasonable risk disposal plan, so as to efficiently deal with possible problems.

[0017] Risk feedback information will be used to optimize risk prediction models and risk handling strategies, allowing the system to continuously optimize itself and improve the efficiency and effectiveness of risk management.

[0018] Improve project efficiency and safety: With the help of this platform, you can more effectively monitor and manage project risks, avoid potential problems, improve project efficiency, and greatly improve project safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a system block diagram of the platform of the present invention; Figure 2 A flow chart for implementing the risk analysis of the present invention; Figure 3 A decision-making implementation flow chart for selecting the present invention; Figure 4 This is a risk feedback flow chart of the present invention. DETAILED DESCRIPTION

[0020] In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. Typical embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described in the present invention. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0021] Unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as commonly understood by technicians in the technical field to which the present invention belongs. The terms used in the present invention and in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. In order to facilitate the understanding of the present invention, the present invention will be described in more detail below with reference to the relevant drawings. Typical embodiments of the present invention are shown in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described in the present invention. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Figure 1 As shown, an intelligent decision support platform for tunnel construction safety risk management, the platform includes: Data collection module,First, the data collection module collects relevant data including geological data,,construction parameters and engineering environment by real-time monitoring of,the shield machine parameters.

[0022] Selection and installation of monitoring equipment: Select specialized monitoring equipment suitable for the characteristics of tunnel construction, such as sensors for measuring shield machines, geological exploration equipment, and environmental monitoring equipment. Install the equipment and ensure its normal operation.

[0023] Data collection: The data collection module monitors the tunnel construction process in real time through monitoring equipment and automatically collects measurement data. These data include but are not limited to shield machine operating parameters (such as cutter head torque, thrust pressure, thrust speed, etc.), geological data (such as stratum type, soil properties, etc.), construction parameters (such as construction sequence, construction speed, construction quality, etc.) and engineering environment data (such as temperature, humidity, noise, vibration, etc.).

[0024] Data processing: After the data collection module collects data, it will perform preliminary processing on the original data to remove noise, eliminate outliers, and perform data calibration to improve the accuracy of the data.

[0025] The collected and processed data needs to be uploaded to the backend storage system according to certain regulations. Data upload should be carried out through a safe and reliable communication protocol to avoid data loss or tampering. In order to ensure data availability, the data collection module will back up the collected data in real time to prevent the loss of original data.

[0026] The risk analysis module processes and analyzes data based on artificial intelligence algorithms and generates risk warning identification information.

[0027] like Figure 2 As shown in the figure, in the intelligent decision support platform for tunnel construction safety risk management, the implementation process of the risk analysis module is as follows: S101, data reading: the risk analysis module reads the pre-processed construction data uploaded by the data collection module.

[0028] S102, feature extraction: extract features from the read data, and use some dimensionality reduction techniques such as principal component analysis (PCA) to reduce the dimensionality of the original data.

[0029] S103, build a risk prediction model: input the feature vector obtained in the previous step into the pre-trained long short-term memory network LSTM to perform risk prediction. The specific model selection can be based on the actual data characteristics and problem background. Here, the specific mathematical model form depends on the specific algorithm we use.

[0030] The long short-term memory network (LSTM) is used for risk prediction. As a special recurrent neural network (RNN), LSTM uses forget gate, input gate and output gate to avoid the problem of gradient disappearance or explosion, so it performs well in dealing with lag effects. The mathematical model is as follows: The hidden state ht and memory ct at time t are updated from the state at time t-1 as:

[0031]

[0032]

[0033]

[0034]

[0035]

[0036] ft: The value of the forget gate at time t, which determines how much past information the unit will forget based on its new information.

[0037] it: The value of the input gate at time t, which determines how much new information the unit will store.

[0038] ot: The value of the output gate at time t, which outputs a new version of its current state based on its new input.

[0039] c~t: The value to be saved to the candidate collection at time t.

[0040] ct: internal state value at time t, the "memory" in LSTM, that is, the sum of all short-term memories, stores long-term information.

[0041] ht: The hidden state at time t, that is, the memory that ot selectively forgets after ct passes through tanh.

[0042] σ: sigmoid function, used to adjust the weight so that the output value is between 0 and 1.

[0043] *[ht-1,xt]: represents the concatenation of the hidden state at time t-1 and the input state at time t.

[0044] Wf, Wi, Wo, Wc: are the weight matrices of the forget gate, input gate, output gate and candidate collections respectively.

[0045] bf, bi, bo, bc: are the bias of the forget gate, input gate, output gate and candidate collection respectively, and the output value of the neuron when the input is zero.

[0046] S104, risk warning: set a warning threshold based on the model's prediction results. Once the predicted risk indicator exceeds this threshold, the system automatically generates a warning message.

[0047] Select the decision-making module, conduct decision analysis based on the early warning identification information, and provide the most reasonable risk disposal plan.

[0048] like Figure 3 As shown in the figure, the selection decision module is the part that performs decision analysis based on the warning identification information. The detailed implementation process is as follows: S201, early warning information reading: The selection decision module first needs to read the early warning identification information generated by the risk analysis module, which includes the possible risk type, risk level and impact range.

[0049] S202, Risk Assessment: Based on the early warning identification information, the system conducts qualitative and quantitative evaluation of the risk level through the risk assessment model, using the fault tree analysis method.

[0050] S203, decision-making solution generation: Based on the risk assessment results, the system ranks all possible response measures and uses linear programming to generate decision-making solutions.

[0051] Assume that there are n risk response measures, the cost of each measure is c[i], the benefit is b[i], and the decision result is x[i], which can be expressed as the following linear programming problem: The goal is to minimize the total cost: , The constraint for all i is that the benefit must be greater than a threshold T: , For all i, the value of each decision outcome is between 0 and 1: , The specific values ​​of the above items are determined by the specific nature of the risk assessment and decision-making plan.

[0052] S204. Feedback on decision results: Finally, the decision module is selected to feed back the final decision results to the executors to guide the actual risk response activities.

[0053] like Figure 4 As shown, the risk feedback module feeds back the processing results to the system to achieve cyclic optimization.

[0054] S301. Feedback information collection: The risk feedback module first needs to collect the decision results generated by the decision-making module. This information includes the selected risk disposal plan, the expected implementation effect, and possible problems and risks.

[0055] S302, implementation status monitoring: When the selected risk disposal plan is implemented, its implementation status needs to be monitored in real time. Based on the monitoring data, the risk feedback module needs to evaluate the actual effect of the risk disposal plan, and the evaluation results will be input into the system as feedback information.

[0056] S303, feedback information processing: The risk feedback module analyzes and processes existing problems based on the collected feedback information. The analysis results will be used to optimize the risk prediction model and decision-making model to improve the accuracy of its prediction and decision-making.

[0057] S304, model optimization: based on the results of processing feedback information, optimize the original risk prediction model and decision-making model so that these two modules can make more accurate predictions and decisions when processing subsequent data.

[0058] S305, system update: Update the optimized model to the system so that more accurate predictions and decisions can be provided when dealing with new risk situations, thereby achieving cyclic optimization of the entire system.

[0059] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM).

[0060] It should be understood that the detailed description of the technical solutions of the present invention by means of the preferred embodiments is illustrative rather than restrictive. A person skilled in the art may modify the technical solutions described in the embodiments, or replace some of the technical features by equivalents, based on reading the specification of the present invention; and these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent decision support platform for tunnel construction safety risk management, characterized by: The platform includes: The data collection module collects geological data, construction parameters and engineering environment data by real-time monitoring of shield machine parameters; The risk analysis module processes and analyzes data based on artificial intelligence algorithms and generates risk warning identification information; Select the decision-making module, conduct decision analysis based on the early warning identification information, and provide the most reasonable risk disposal plan; The risk feedback module feeds back the processing results to the system to achieve cyclic optimization.

2. The intelligent decision support platform for tunnel construction safety risk management according to claim 1, characterized in that: The data collection module comprises: Select specialized monitoring equipment appropriate to the characteristics of tunnel construction, including: sensors for measuring shield machines, geological survey equipment, and environmental monitoring equipment; The monitoring equipment monitors the tunnel construction process in real time and automatically collects measurement data, including: shield machine operating parameters: cutter head torque, thrust pressure, thrust speed; geological data: stratum type, soil properties; construction parameters: construction sequence, construction speed, construction quality; and engineering environment data: temperature, humidity, noise, vibration; After the data collection module collects the data, it will perform preliminary processing on the original data to remove noise, eliminate outliers, and perform data calibration.

3. The intelligent decision support platform for tunnel construction safety risk management according to claim 1, characterized in that: The risk analysis module uses the LSTM algorithm for risk warning; the details are as follows: The risk analysis module reads the pre-processed construction data uploaded by the data collection module; Extract features from the read data and use principal component analysis to reduce the dimensionality of the original data; Input the feature vector obtained in the previous step into the pre-trained long short-term memory network LSTM for risk prediction; The warning threshold is set according to the prediction results of the model. Once the predicted risk indicator exceeds this threshold, the system automatically generates a warning message.

4. The intelligent decision support platform for tunnel construction safety risk management according to claim 1 is characterized by: The selection decision module comprises: Reading of early warning information: The decision-making module first reads the early warning identification information generated by the risk analysis module, including the type of risk, risk level and impact range; Risk assessment: Based on the early warning identification information, the system uses the risk assessment model to conduct qualitative and quantitative evaluation of the risk level; Decision-making plan generation: Based on the risk assessment results, the system ranks all response measures; Decision result feedback: Select the decision module to feed back the final decision result to the executor to guide the actual risk response activities.

5. The intelligent decision support platform for tunnel construction safety risk management according to claim 1, characterized in that: The risk feedback module includes: The risk feedback module first collects the decision results generated by the selection decision module. This information includes the selected risk treatment plan, the expected implementation effect, and the problems and risks that arise; Implementation monitoring: When the selected risk treatment plan is implemented, its implementation status is monitored in real time; Feedback information processing: The risk feedback module analyzes and processes existing problems based on the collected feedback information; Model optimization: Based on the results of processing feedback information, the original risk prediction model and decision-making model are optimized so that these two modules can make more accurate predictions and decisions when processing subsequent data.

6. The intelligent decision support platform for tunnel construction safety risk management according to claim 4, characterized in that: The decision-making scheme is generated by using a linear programming algorithm: Assume that there are n risk response measures, the cost of each measure is c[i], the benefit is b[i], and the decision result is x[i], which can be expressed as the following linear programming problem: The goal is to minimize the total cost: ， The constraint for all i is that the benefit must be greater than a threshold T: ， For all i, the value of each decision outcome is between 0 and 1: 。

Images