Tunnel construction system based on tunnel advanced geological forecast

By combining forecasting methods in the tunnel construction system to verify and build an optimization model, the problems of low accuracy and efficiency in tunnel advance geological forecasting in the prior art are solved, and more efficient and reliable forecast results are achieved.

CN119045080APending Publication Date: 2024-11-29GUIZHOU SHUNKANG TESTING CO LTD
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411118861.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-15
Publication Date
2024-11-29

Smart Images

  • Figure CN119045080A_ABST
    Figure CN119045080A_ABST
Patent Text Reader

Abstract

The invention discloses a tunnel construction system based on tunnel advanced geological forecast, which relates to the technical field of tunnel geological forecast, and is characterized in that forecast results of different combination methods are verified, the accuracy rates of the different combination methods are analyzed, and a data set is constructed based on the accuracy rates of the different combination methods; meanwhile, a tunnel advanced geological forecast optimization model is constructed, a data set is divided into a training set and a test set to train and evaluate the tunnel advanced geological forecast optimization model, and finally the actual application of the tunnel advanced geological forecast optimization model is evaluated, so that the accuracy of an output result of the tunnel advanced geological forecast optimization model is improved; the problem that an automatic selection combination forecasting method is lacked in the prior art is solved, the selection accuracy of the combination forecasting method is improved, guarantee is provided for the accuracy and reliability of subsequent forecasting, the forecasting time cost is reduced, and the forecasting effect and efficiency are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of tunnel geological prediction, and in particular to a tunnel construction system based on tunnel advanced geological prediction. Background Art

[0002] The rapid development of the transportation industry has accelerated the construction of transportation infrastructure, and at the same time, great attention has been paid to the safety of traffic road construction. In particular, in the process of tunnel construction, adverse geological phenomena such as large karst caves, underground rivers and faults are often encountered. These situations are very likely to cause geological disasters such as landslides, water inrush and mud inrush, which seriously threaten the safety of life and property of construction workers. The advanced geological forecast of tunnels can timely discover the location, type and possible geological risks of adverse geological bodies in front of the face, and effectively prevent the occurrence of geological disasters.

[0003] Prior art, such as an efficient and comprehensive advance geological prediction method disclosed in the invention patent application with publication number CN112485823A, includes: S100 before tunnel construction, surface survey of the target area is conducted according to existing data, multiple single prediction methods are selected based on the surface survey results to form a combined prediction scheme, and geological prediction before construction is conducted; S200 tunnel construction is conducted according to the tunnel construction scheme of geological prediction before construction, and during the tunnel construction process, geological information is fed back according to tunnel excavation, and a long-distance prediction is conducted in front of the excavation using the TSP advance detection method to verify whether it is consistent with the geological prediction before construction; S300 during the tunnel construction process, if there is any inconsistency with the geological prediction before construction, the construction is suspended, and multiple single prediction methods are re-selected according to the newly discovered geological conditions to form a new combined prediction scheme, geological prediction is re-conducted with reference to the geological prediction method before construction, and then the tunnel construction scheme is adjusted according to the new geological prediction to continue the tunnel construction. The present invention improves the accuracy and efficiency of prediction.

[0004] Existing technologies such as the paper "Study on the Selection and Interpretation Threshold of Geophysical Exploration Methods for Tunnel Advanced Geological Forecasting" discloses a comprehensive geophysical exploration method for advanced prediction of poor geology in tunnels based on the analytic hierarchy process, including: the necessity of quantifying the process of selecting geophysical exploration methods, establishing a hierarchical analysis method model, calculating the importance weights of the TSP method, the GPR method and the TEM method in the Tianchi Tunnel advanced geological prediction scheme through the analytic hierarchy process, selecting the TSP method with the largest weight for the first prediction considering all influencing factors, and selecting a geophysical exploration method for re-surveying the poor geology through the analytic hierarchy process when poor geology is found in the first prediction; establishing a hierarchical structure model by comprehensively considering all factors affecting the selection of geophysical exploration methods for tunnel advanced geological prediction, such as prediction accuracy, prediction distance, technical reliability and economic rationality, outputting the important weights of multiple prediction geophysical exploration methods, and then selecting one or more geophysical exploration prediction methods for re-surveying in descending order according to the weights.

[0005] The above scheme has at least the following shortcomings: 1. There are multiple combination forecasting methods formed by multiple single forecasting methods. Selecting the combination forecasting method with the highest accuracy is conducive to improving the accuracy of the forecast. However, the above efficient comprehensive advanced geological forecasting method does not directly monitor and analyze the accuracy of the forecasts of multiple combination methods, and cannot quickly select the combination method with the highest accuracy for forecasting, thereby failing to improve the accuracy and efficiency of the forecast.

[0006] 2. In the above-mentioned paper, it is necessary to first use a single prediction method with the largest weight for prediction. After the occurrence of unfavorable geology, one or more geophysical prediction methods are selected for re-testing, which greatly increases the time cost of prediction and reduces the efficiency of prediction. In addition, the above-mentioned paper establishes a hierarchical model by using all factors that affect the selection of geophysical methods for tunnel advance geological prediction, which cannot reflect the importance of prediction accuracy in prediction. At the same time, when evaluating the accuracy of the prediction, the above-mentioned paper only evaluates the accuracy of the prediction based on the type of predicted unfavorable geology. Unfavorable geology includes karst and faults. The location, size and filling properties of unfavorable geology all reflect the accuracy of the prediction method. However, the above-mentioned paper did not analyze the accuracy of the prediction based on basic information such as the location of the unfavorable geology, thereby failing to improve the reliability of the accuracy analysis results of the prediction method, failing to provide an effective reference for the selection of the combination method, and failing to guarantee the reliability of subsequent prediction results.

[0007] 3. The above scheme does not construct a tunnel advanced geological forecast optimization model based on the forecast conditions of different combined forecast methods, and cannot dynamically extract the optimal forecast combination method for different tunnels and different geological conditions, thereby increasing the time cost of the forecast, and cannot guarantee the accuracy of the forecast, reducing the efficiency and effectiveness of the forecast. Summary of the invention

[0008] In view of the above-mentioned technical deficiencies, the object of the present invention is to provide a tunnel construction system based on advanced tunnel geological prediction.

[0009] In order to solve the above technical problems, the present invention adopts the following technical solution: The present invention provides a tunnel construction system based on tunnel advanced geological prediction, including: a prediction analysis unit, a model building unit and an execution terminal.

[0010] The forecast tracking unit includes a forecast module and a forecast verification module.

[0011] The forecast module is used to combine various forecast methods to obtain various combined forecast methods, use various combined forecast methods to collect and analyze the geophysical data of the test tunnel, and obtain forecast conclusions drawn by various combined forecast methods.

[0012] The prediction verification module is used to carry out actual excavation of the test tunnel and record a description of the actual verification situation.

[0013] The model building unit includes a model building and training module and a model evaluation and verification module.

[0014] The model building and training module is used to build a data set, then build a tunnel advanced geological prediction model, and train the tunnel advanced geological prediction model.

[0015] The model evaluation and verification module is used to test the tunnel advanced geological prediction model using a test set and evaluate the application effect of the tunnel advanced geological prediction model.

[0016] The execution terminal is used to provide feedback on the application effect of the tunnel advanced geological prediction model.

[0017] The beneficial effects of the present invention are: 1. The present invention provides a tunnel construction system based on tunnel advance geological prediction, verifies the prediction results of different combination methods, analyzes the accuracy of different combination methods, and constructs a data set based on the accuracy of different combination methods. At the same time, it constructs a tunnel advance geological prediction optimization model, and divides the data set into a training set and a test set to train and evaluate the tunnel advance geological prediction optimization model. Finally, the actual application of the tunnel advance geological prediction optimization model is evaluated, and the accuracy of the output results of the tunnel advance geological prediction optimization model is improved, which solves the problem of lack of automatic selection of combined prediction methods in the prior art, improves the accuracy of the selection of combined prediction methods, provides guarantee for the accuracy and reliability of subsequent predictions, reduces the time cost of predictions, and improves the effect and efficiency of predictions.

[0018] 2. When analyzing the accuracy of each combined prediction method, the present invention analyzes the accuracy of the prediction of each combined prediction method according to the prediction situation of karst and fault corresponding to each combined prediction method, and analyzes the accuracy of different combined methods from multiple dimensions, thereby providing an effective reference for selecting the combined method, increasing the authenticity and objectivity of the data set, and improving the training and testing effects of the tunnel advanced geological prediction optimization model. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 It is a schematic diagram of the system structure connection of the present invention. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] See also Figure 1 As shown, a tunnel construction system based on advanced tunnel geological prediction includes: a prediction analysis unit, a model building unit and an execution terminal.

[0023] The forecast tracking unit includes a forecast module and a forecast verification module.

[0024] The forecast module is used to combine various forecast methods to obtain various combined forecast methods, use various combined forecast methods to collect and analyze the geophysical data of the test tunnel, and obtain forecast conclusions drawn by various combined forecast methods.

[0025] In the above, the forecasting methods include seismic wave reflection method, geological radar method and transient electromagnetic method. Various combined forecasting methods: seismic wave reflection method + geological radar method, seismic wave reflection method + transient electromagnetic method, geological radar method + transient electromagnetic method and seismic wave reflection method + geological radar method + transient electromagnetic method.

[0026] It should be noted that geophysical data include but are not limited to the properties, location and thickness of the detected strata of the geological body. The properties of the geological body include weak zones, broken zones, faults and water content.

[0027] The forecast conclusions obtained by various combined forecasting methods include the forecasted karst positions, forecasted starting pile numbers, characteristic data of karst formation, the forecasted fault positions of fault formation, the forecasted starting pile numbers and characteristic data of the forecasted fault positions of fault formation, etc.

[0028] Among the above, the characteristic data of karst include scale, spatial position and filling properties; the scale includes predicted length, predicted width and predicted height; the filling properties include mud filling, groundwater filling and no filling; the characteristic data of fault include scale, spatial position and groundwater development; the scale includes predicted length, predicted width and predicted height; the groundwater development includes dry, wet, rainy and flowing.

[0029] Among them, seismic wave reflection method is the abbreviation of tunnel seismic wave reflection tomography technology. The basic principle of this technology is that when seismic waves encounter an interface with acoustic impedance difference, part of the signal is reflected back and part of the signal is transmitted into the medium in front. Changes in acoustic impedance usually occur at the interface of geological rock layers or discontinuous interfaces in the rock mass. The reflected seismic signal is received by a high-sensitivity seismic signal sensor and is used to understand the properties, location and scale of the geological body in front of the tunnel working face, including weak zones, broken zones, faults and water-bearing zones. When seismic waves propagate from soft rock to hard surrounding rock, the deflection polarity of the echo is consistent with the wave source. When there is a fracture zone inside the rock mass, the polarity of the echo will be reversed. The larger the size of the reflector, the greater the difference in acoustic impedance, the more obvious the echo, and the easier it is to detect.

[0030] The acquisition process of the seismic wave reflection method is as follows: hammering on the epicenter point, while hammering the rock mass to generate seismic waves, the trigger generates a trigger signal to the base station, and then the base station issues a command to collect seismic waves to the wireless remote module, and transmits the data sent back by the remote module to the laptop to complete the geophysical data acquisition.

[0031] The geological radar method is based on Maxwell's classical electromagnetic theory. Its core content is that the static charge distribution or current distribution excites a stable electric field, which does not change with time and does not radiate energy outward. Because there are electrical differences in underground media, electromagnetic waves will be reflected during propagation, and the reflected signal will be received by the radar. As an electromagnetic wave detection technology, geological radar follows the principles of geometric optics. It determines the target distance based on the one-way travel time of the echo and the propagation speed of the electromagnetic wave in the corresponding medium, and judges the nature of the target through comprehensive analysis.

[0032] The working principle of geological radar is that the control unit first emits a set of electromagnetic waves of fixed frequency to the detection stratum. When the electromagnetic waves propagate into the underground medium and encounter the interface between different media, part of the electromagnetic wave energy will be converted into reflected waves and return to the ground, and the remaining electromagnetic waves will continue to propagate forward. When encountering the interface, part of the electromagnetic waves will be reflected back to the ground. When electromagnetic waves propagate in different media, their attenuation coefficient, reflection coefficient, and reflected wave frequency are different. Therefore, when the radar antenna receives these signals and analyzes them, it can obtain geophysical data such as the thickness of the detected stratum, the integrity of the rock mass, and the water content of the rock stratum.

[0033] Transient electromagnetic method is a time domain electromagnetic method. It uses step waveform electromagnetic pulse excitation and uses an ungrounded loop to transmit a primary field underground. After the primary field is powered off, the underground medium will generate an induced secondary field. Due to the heat loss of the induced current in the good conductive medium, the secondary field decays with time roughly according to the exponential law, forming a transient magnetic field. The secondary field mainly comes from the induced current in the good conductive medium, so it contains geological information related to the underground medium. The secondary field is observed by receiving the loop, and the observed data is analyzed and processed to solve the time domain electromagnetic method related to geological problems.

[0034] The transient electromagnetic field can be divided into three stages according to the attenuation process: early, middle and late. In the early stage of the transient electromagnetic field attenuation process, the high-frequency components are dominant, so the eddy current is mainly distributed near the tunnel face. Therefore, in the early stage of the transient process, the electromagnetic field mainly reflects the geophysical data at a close distance. In the late stage of the transient process, the high-frequency components are absorbed by the conductive medium, and the low-frequency components dominate. In this stage, the geophysical data at a long distance are mainly reflected.

[0035] The prediction verification module is used to carry out actual excavation of the test tunnel and record a description of the actual verification situation.

[0036] When the staff conducts actual excavation, they record the actual verification situation description in real time in the tunnel construction system. The actual verification situation description includes the karst formation verification results of each predicted karst position, the verification starting pile number, the verification characteristic data of each karst, the fault formation verification results of each predicted fault position, the verification starting pile number of each predicted fault position, and the verification characteristic data of each fault.

[0037] The verification characteristic data of the above-mentioned karst include verification scale size, verification spatial position and verification filling properties; the verification characteristic data of the fault include verification length, verification width and verification height; among which the verification scale size includes verification length, verification width and verification height.

[0038] The model building unit includes a model building and training module and a model evaluation and verification module.

[0039] The model building and training module is used to build a data set, then build a tunnel advanced geological prediction model, and train the tunnel advanced geological prediction model.

[0040] In a specific embodiment, the data set is constructed, and the specific process is as follows: S1, obtaining each predicted karst position and predicted starting pile number of karst formation from the forecast conclusions obtained by various combined forecasting methods, obtaining the karst formation verification result and the verification starting pile number of each predicted karst position from the actual verification situation description, and comparing each predicted karst position of karst formation in various combined forecasting methods with the karst formation verification result of each predicted karst position, obtaining the qualitative judgment accuracy of each predicted karst position in various combined forecasting methods, the qualitative judgment accuracy includes the values ​​of 1 and 0, and each predicted karst position with a qualitative judgment accuracy of 1 is recorded as each karst, and S2 is executed at the same time.

[0041] In the above, the specific process of obtaining the qualitative judgment accuracy of each predicted karst location in the various combined forecasting methods is as follows: in, It represents the qualitative judgment accuracy of the i′th predicted karst position in the f′th combined forecasting method. Qualitative verification is correct: the karst formation verification result corresponding to the i′th predicted karst position in the f′th combined forecasting method is karst formation; qualitative verification is wrong: the karst formation verification result corresponding to the i′th predicted karst position in the f′th combined forecasting method is that karst is not formed. f represents the number of each type of combined forecasting method, i′ represents the number of the predicted karst position, and both f and i′ are positive integers.

[0042] S2. Obtain the characteristic data of each karst from the forecast conclusions obtained from various combined forecasting methods, where the characteristic data of karst include scale, spatial position and filling properties; obtain the verification characteristic data of each karst from the description of actual verification situation, where the verification characteristic data of karst include verification scale, verification spatial position and verification filling properties; analyze the forecast accuracy, spatial position accuracy, scale accuracy and filling property accuracy of each karst in various combined forecasting methods.

[0043] In the above, the analysis process of the prediction accuracy, spatial position accuracy, scale accuracy and filling property accuracy of each karst in the various combined prediction methods is as follows: The analysis formula of prediction accuracy is: In the formula, K fi , α1 fi , α2 fi , α3 fi They respectively represent the prediction accuracy, qualitative judgment accuracy, spatial position accuracy, scale accuracy, and filling property accuracy of the i-th karst in the f-th type combined prediction method. i represents the number of each karst and is a positive integer.

[0044] Analysis formula for spatial position accuracy: In the formula, Z′ fi , Z fi, W f ' i , W fi They respectively represent the forecast starting pile number, verification starting pile number, forecast spatial position, and verification spatial position of the i-th karst in the f-th type combined forecasting method.

[0045] It should be noted that |Z′ fi -Z fi The values ​​of | are shown in the following table:

[0046] |Z′ fi -Z fi |Value example table

[0047] <![CDATA[Z′ fi -Z fi Value of]]> Value result 0≤x<1 100% 1≤x<2 90% 2≤x<3 80% 3≤x<4 70% 4≤x<5 60% >5 0

[0048] Analysis formula for scale accuracy:

[0049] Where L1 fi 、L2 fi 、L3 fi , L1′ fi , L2′ fi , L3′ fi They respectively represent the verification length, verification width, verification height, prediction length, prediction width and prediction height of the i-th karst in the f-th type combined prediction method.

[0050] Expression for accuracy of filling properties: In the formula, the qualitative verification is correct: the filling properties and verification filling properties of the i-th karst in the f-th combined forecasting method are the same, and the qualitative verification is wrong: the filling properties and verification filling properties of the i-th karst in the f-th combined forecasting method are not the same.

[0051] S3. Obtain the predicted fault positions and predicted starting pile numbers of the formed faults from the forecast conclusions obtained by various combined forecasting methods, obtain the fault formation verification results and verification starting pile numbers of each predicted fault position from the description of the actual verification situation, and compare the predicted fault positions of the formed faults in various combined forecasting methods with the fault formation verification results of each predicted fault position to obtain the qualitative judgment accuracy of each predicted fault position in various combined forecasting methods. The qualitative judgment accuracy includes the values ​​of 1 and 0. Each predicted fault position with a qualitative judgment accuracy of 1 is recorded as each fault, and S4 is executed at the same time.

[0052] In the above, the process of obtaining the qualitative judgment accuracy of each predicted fault position in the various combined prediction methods is as follows: Among them, φ fj′It represents the qualitative judgment accuracy of the j′ predicted fault position in the f-th combined forecast method. The qualitative verification is correct: the fault formation verification result corresponding to the j′ predicted fault position in the f-th combined forecast method is that a fault is formed; the qualitative verification is wrong: the fault formation verification result corresponding to the j′ predicted fault position in the f-th combined forecast method is that no fault is formed. f represents the number of each type of combined forecast method, j′ represents the number of each predicted fault position, and both f and j′ are positive integers.

[0053] S4. Obtain the characteristic data of each fault from the forecast conclusions obtained from various combined forecasting methods, and obtain the verification characteristic data of each fault from the description of actual verification conditions; the characteristic data of the fault includes scale, spatial position and groundwater development conditions, and the verification characteristic data of the fault includes verification scale, verification spatial position and verification groundwater development conditions. Analyze the forecast accuracy, spatial position accuracy, scale accuracy and groundwater development accuracy of each fault in various combined forecasting methods.

[0054] In the above, the analysis process of the forecast accuracy, spatial position accuracy, scale accuracy and groundwater development accuracy of each fault in the various combined forecasting methods is as follows: The analysis formula of forecast accuracy is: In the formula, F fj ,φ fj ,δ1 fj , δ2 fj , δ3 fj They respectively represent the prediction accuracy, qualitative judgment accuracy, spatial position accuracy, scale accuracy, and groundwater development accuracy of the j-th fault in the f-th type combined forecasting method. j represents the number of each fault and is a positive integer.

[0055] Analysis formula for spatial position accuracy: Where Z′ fj , Z fj , W f ' j , W fj They respectively represent the prediction starting pile number, verification starting pile number, prediction spatial position, and verification spatial position of the jth fault in the fth type combined prediction method.

[0056] It should be noted that: |Z′ fj -Z fj The values ​​of | are shown in the following table:

[0057] |Z′ fj -Z fj |Value example table

[0058] <![CDATA[Z′ fj -Z fj Value of]]> Value result 0≤x<1 100% 1≤x<2 90% 2≤x<3 80% 3≤x<4 70% 4≤x<5 60% >5 0

[0059] Analysis formula for scale accuracy:

[0060] Where L1 fj 、L2 fj 、L3 fj , L1′ fj , L2′ fj , L3′ fj They respectively represent the verification length, verification width, verification height, prediction length, prediction width and prediction height of the j-th fault in the f-th combined prediction method.

[0061] The expression of groundwater development accuracy is: The qualitative verification is correct: the groundwater development situation of the jth fault in the fth type of combined forecasting method is the same as the verified groundwater development situation. The qualitative verification is correct or incorrect: the groundwater development situation of the jth fault in the fth type of combined forecasting method is not the same as the verified groundwater development situation.

[0062] S5. Using the prediction accuracy of each karst and the prediction accuracy of each fault in various combined prediction methods, the prediction accuracy of karst geological bodies, the prediction accuracy of fault geological bodies and the comprehensive prediction accuracy in various combined prediction methods are analyzed. The data set is composed of the qualitative judgment accuracy, prediction accuracy, spatial position accuracy, scale accuracy, groundwater development accuracy of each fault in various combined prediction methods, the qualitative judgment accuracy of each fault, the prediction accuracy of each fault, the spatial position accuracy of each fault, the scale accuracy of each fault, the groundwater development accuracy of each fault, the prediction accuracy of karst geological bodies, the prediction accuracy of fault geological bodies and the comprehensive prediction accuracy.

[0063] In the above, the analysis process of the karst geological body prediction accuracy, fault geological body prediction accuracy and comprehensive prediction accuracy in the various combined prediction methods is as follows: The karst geological body prediction accuracy analysis formula is: In the formula, K f It represents the prediction accuracy of karst geological body by the fth type of combined prediction method, and n represents the number of karst.

[0064] The analysis formula for the prediction accuracy of fault geological bodies is: In the formula, F f represents the prediction accuracy of the fault geological body of the fth type of combined prediction method, and m represents the number of faults.

[0065] The comprehensive forecast accuracy analysis formula is: In the formula, CFA f Represents the comprehensive forecast accuracy of the f-th type of combined forecasting method.

[0066] When analyzing the accuracy of each combined prediction method, the present invention analyzes the accuracy of the prediction of each combined prediction method according to the prediction situation of karst and fault corresponding to each combined prediction method, and analyzes the accuracy of different combined methods from multiple dimensions, thereby providing an effective reference for selecting the combined method, increasing the authenticity and objectivity of the data set, and improving the training and testing effects of the tunnel advanced geological prediction optimization model.

[0067] The model evaluation and verification module is used to test the tunnel advanced geological prediction model using a test set and evaluate the application effect of the tunnel advanced geological prediction model.

[0068] In a specific embodiment, the specific process of the model construction and training module is as follows: based on the machine learning algorithm, a tunnel advanced geological prediction optimization model is constructed. The model calculates and outputs the prediction accuracy of each combined prediction method based on the geophysical data of the tunnel, and then selects the optimal combined prediction method for the tunnel.

[0069] Based on the random forest model, a training set and a test set are divided from the data set. The training set is used to train the tunnel advanced geological prediction optimization model, and the test set is used to evaluate the tunnel advanced geological prediction optimization model, obtain evaluation indicators, and provide feedback.

[0070] The ratio of training set to test set is 8:2.

[0071] It should be noted that if there is non-numeric data in the data set, the label encoding function in the programming software is used to convert the non-numeric data.

[0072] For example: convert various combined forecasting methods: seismic wave reflection method + geological radar method, seismic wave reflection method + transient electromagnetic method, geological radar method + transient electromagnetic method and seismic wave reflection method + geological radar method + transient electromagnetic method into label codes 0, 1, 2, 3.

[0073] It should also be noted that when the test set evaluates the tunnel advanced geological prediction optimization model, the tunnel construction system obtains evaluation indicators and provides feedback to the staff. The staff can make adjustments based on the feedback. The evaluation indicators include accuracy, recall rate, and F1 score.

[0074] In another specific embodiment, the application effect of the tunnel advance geological prediction model is evaluated, and the specific process is as follows: the geophysical data of the test tunnel is input into the tunnel advance geological prediction model, and the tunnel advance geological prediction model outputs the optimal combination prediction method and the prediction conclusion of the optimal combination prediction method, and analyzes the comprehensive prediction accuracy of the tunnel advance geological prediction model according to the actual verification situation description, recorded as CFA, and substitutes it into the application effect analysis formula Wherein, η represents the application effect of the tunnel advance geological prediction model, CFA′ represents the preset comprehensive prediction accuracy threshold, and the application effect of the tunnel advance geological prediction model includes the values ​​of 1 and 0. When the application effect of the tunnel advance geological prediction model is 1, it indicates that the application effect of the tunnel advance geological prediction model is good. When the application effect of the tunnel advance geological prediction model is 0, it indicates that the application effect of the tunnel advance geological prediction model is poor.

[0075] It should be noted that the analysis method of the comprehensive prediction accuracy of the tunnel advanced geological prediction model is the same as steps S1-S5, which will not be repeated here. The comprehensive prediction accuracy threshold is the maximum comprehensive prediction accuracy among the comprehensive prediction accuracy of various combined prediction methods.

[0076] The execution terminal is used to provide feedback on the application effect of the tunnel advanced geological prediction model.

[0077] The embodiment of the present invention verifies the prediction results of different combination methods, analyzes the accuracy of different combination methods, and constructs a data set based on the accuracy of different combination methods. At the same time, it constructs a tunnel advanced geological prediction optimization model, and divides the data set into a training set and a test set to train and evaluate the tunnel advanced geological prediction optimization model. Finally, the practical application of the tunnel advanced geological prediction optimization model is evaluated, and the accuracy of the output results of the tunnel advanced geological prediction optimization model is improved. The problem of lack of automatic selection of combined prediction methods in the prior art is solved, the accuracy of the selection of combined prediction methods is improved, the accuracy of the selection of combined prediction methods is improved, the accuracy and reliability of subsequent forecasts are guaranteed, the time cost of forecasts is reduced, and the effect and efficiency of forecasts are improved.

[0078] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they shall all fall within the protection scope of the present invention.

Claims

1. A tunnel construction system based on advanced tunnel geological prediction, characterized in that: include: Forecast analysis unit, model building unit and execution terminal; The forecast tracking unit includes a forecast module and a forecast verification module; The prediction module is used to combine various prediction methods to obtain various combined prediction methods, collect and analyze the geophysical data of the test tunnel using the various combined prediction methods, and obtain the prediction conclusions obtained by the various combined prediction methods; The prediction verification module is used to perform actual excavation of the test tunnel and record a description of the actual verification situation; The model building unit includes a model building and training module and a model evaluation and verification module; The model building and training module is used to build a data set, then build a tunnel advanced geological prediction model, and train the tunnel advanced geological prediction model; The model evaluation and verification module is used to test the tunnel advanced geological prediction model using the test set to evaluate the application effect of the tunnel advanced geological prediction model; The execution terminal is used to provide feedback on the application effect of the tunnel advanced geological prediction model.

2. A tunnel construction system based on advanced tunnel geological prediction according to claim 1, characterized in that: The prediction methods include seismic wave reflection method, geological radar method and transient electromagnetic method.

3. A tunnel construction system based on advanced tunnel geological prediction according to claim 1, characterized in that: The specific process of constructing the data set is as follows: S1. Obtain each predicted karst position and predicted starting pile number of karst formation from the prediction conclusions obtained by various combined prediction methods, obtain the karst formation verification result and verification starting pile number of each predicted karst position from the description of the actual verification situation, and compare each predicted karst position of karst formation in various combined prediction methods with the karst formation verification result of each predicted karst position, obtain the qualitative judgment accuracy of each predicted karst position in various combined prediction methods, the qualitative judgment accuracy includes the values ​​of 1 and 0, and record each predicted karst position with the qualitative judgment accuracy of 1 as each karst, and execute S2 at the same time; S2. Obtain the characteristic data of each karst from the forecast conclusions obtained by various combined forecasting methods, where the characteristic data of karst include scale, spatial position and filling properties; obtain the verification characteristic data of each karst from the description of actual verification situation, where the verification characteristic data of karst include verification scale, verification spatial position and verification filling properties; analyze the forecast accuracy, spatial position accuracy, scale accuracy and filling property accuracy of each karst in various combined forecasting methods; S3, obtaining the predicted fault positions and predicted starting pile numbers of the formed faults from the prediction conclusions obtained by various combined prediction methods, obtaining the fault formation verification results and verification starting pile numbers of each predicted fault position from the actual verification situation description, and comparing the predicted fault positions of the formed faults in various combined prediction methods with the fault formation verification results of each predicted fault position, obtaining the qualitative judgment accuracy of each predicted fault position in various combined prediction methods, the qualitative judgment accuracy includes the values ​​of 1 and 0, recording each predicted fault position with a qualitative judgment accuracy of 1 as each fault, and executing S4 at the same time; S4. Obtain the characteristic data of each fault from the forecast conclusions obtained by various combined forecasting methods, and obtain the verification characteristic data of each fault from the description of actual verification conditions; wherein the characteristic data of the fault includes scale, spatial position and groundwater development conditions, and the verification characteristic data of the fault includes verification scale, verification spatial position and verification groundwater development conditions, and analyze the forecast accuracy, spatial position accuracy, scale accuracy and groundwater development accuracy of each fault in various combined forecasting methods; S5. Using the prediction accuracy of each karst and the prediction accuracy of each fault in various combined prediction methods, the prediction accuracy of karst geological bodies, the prediction accuracy of fault geological bodies and the comprehensive prediction accuracy in various combined prediction methods are analyzed. The data set is composed of the qualitative judgment accuracy, prediction accuracy, spatial position accuracy, scale accuracy, groundwater development accuracy of each fault in various combined prediction methods, the qualitative judgment accuracy of each fault, the prediction accuracy of each fault, the spatial position accuracy of each fault, the scale accuracy of each fault, the groundwater development accuracy of each fault, the prediction accuracy of karst geological bodies, the prediction accuracy of fault geological bodies and the comprehensive prediction accuracy.

4. A tunnel construction system based on advanced tunnel geological prediction according to claim 3, characterized in that: The specific process of obtaining the qualitative judgment accuracy of each predicted karst location in the various combined forecasting methods is as follows: in, It represents the qualitative judgment accuracy of the i′th predicted karst position in the f′th combined forecasting method. Qualitative verification is correct: the karst formation verification result corresponding to the i′th predicted karst position in the f′th combined forecasting method is karst formation; qualitative verification is wrong: the karst formation verification result corresponding to the i′th predicted karst position in the f′th combined forecasting method is that karst is not formed. f represents the number of each type of combined forecasting method, i′ represents the number of the predicted karst position, and both f and i′ are positive integers.

5. A tunnel construction system based on advanced tunnel geological prediction according to claim 4, characterized in that: The analysis process of the prediction accuracy, spatial position accuracy, scale accuracy and filling property accuracy of each karst in the various combined prediction methods is as follows: Analytical formula for forecast accuracy: In the formula, K fi , α1 fi , α2 fi , α3 fi They represent the prediction accuracy, qualitative judgment accuracy, spatial position accuracy, scale accuracy, and filling property accuracy of the i-th karst in the f-th type combined prediction method, respectively. i represents the number of each karst, and i is a positive integer. Analysis formula for spatial position accuracy: In the formula, Z′ fi , Z fi , W f ' i , W fi They represent the forecast starting pile number, verification starting pile number, forecast spatial position, and verification spatial position of the i-th karst in the f-th type combined forecast method respectively; Analysis formula for scale accuracy: Where L1 fi 、L2 fi 、L3 fi , L1′ fi , L2′ fi , L3′ fi They represent the verification length, verification width, verification height, prediction length, prediction width, and prediction height of the i-th karst in the f-th type combined prediction method respectively; Expression for accuracy of filling properties: In the formula, the qualitative verification is correct: the filling properties and verification filling properties of the i-th karst in the f-th combined forecasting method are the same, and the qualitative verification is wrong: the filling properties and verification filling properties of the i-th karst in the f-th combined forecasting method are not the same.

6. A tunnel construction system based on advanced tunnel geological prediction according to claim 3, characterized in that: The process of obtaining the qualitative judgment accuracy of each predicted fault position in the various combined prediction methods is as follows: Among them, φ fj′ It represents the qualitative judgment accuracy of the j′ predicted fault position in the f-th combined forecast method. The qualitative verification is correct: the fault formation verification result corresponding to the j′ predicted fault position in the f-th combined forecast method is that a fault is formed; the qualitative verification is wrong: the fault formation verification result corresponding to the j′ predicted fault position in the f-th combined forecast method is that no fault is formed. f represents the number of each type of combined forecast method, j′ represents the number of each predicted fault position, and both f and j′ are positive integers.

7. A tunnel construction system based on advanced tunnel geological prediction according to claim 6, characterized in that: The analysis process of the forecast accuracy, spatial position accuracy, scale accuracy and groundwater development accuracy of each fault in the various combined forecasting methods is as follows: Analytical formula for forecast accuracy: In the formula, F fj ,φ fj ,δ1 fj , δ2 fj , δ3 fj They represent the prediction accuracy, qualitative judgment accuracy, spatial position accuracy, scale accuracy, and groundwater development accuracy of the jth fault in the fth type of combined prediction method, respectively. j represents the number of each fault and is a positive integer. Analysis formula for spatial position accuracy: Where Z′ fj , Z fj , W f ' j , W fj They represent the prediction starting pile number, verification starting pile number, prediction spatial position, and verification spatial position of the jth fault in the fth type combined prediction method respectively; Analysis formula for scale accuracy: Where L1 fj 、L2 fj 、L3 fj , L1′ fj , L2′ fj , L3′ fj They represent the verification length, verification width, verification height, prediction length, prediction width, and prediction height of the jth fault in the fth type combined prediction method respectively; The expression of groundwater development accuracy is: The qualitative verification is correct: the groundwater development situation of the jth fault in the fth type of combined forecasting method is the same as the verified groundwater development situation. The qualitative verification is correct or incorrect: the groundwater development situation of the jth fault in the fth type of combined forecasting method is not the same as the verified groundwater development situation.

8. A tunnel construction system based on advanced tunnel geological prediction according to claim 7, characterized in that: The analysis process of the karst geological body prediction accuracy, fault geological body prediction accuracy and comprehensive prediction accuracy in the various combined prediction methods is as follows: The analysis formula for karst geological body prediction accuracy is: In the formula, K f represents the prediction accuracy of karst geological bodies by the f-th type of combined prediction method, and n represents the number of karsts; The analysis formula for the prediction accuracy of fault geological bodies is: In the formula, F f represents the prediction accuracy of the fault geological body of the fth type of combined prediction method, and m represents the number of faults; The comprehensive forecast accuracy analysis formula is: In the formula, CFA f Represents the comprehensive forecast accuracy of the f-th type of combined forecasting method.

9. The tunnel construction system based on advanced tunnel geological prediction according to claim 1, characterized in that: The specific process of the model construction and training module is as follows: Based on the machine learning algorithm, a tunnel advanced geological prediction optimization model is constructed. The model calculates and outputs the prediction accuracy of each combined prediction method based on the geophysical data of the tunnel, and then selects the optimal combined prediction method for the tunnel; Based on the random forest model, a training set and a test set are divided from the data set. The training set is used to train the tunnel advanced geological prediction optimization model, and the test set is used to evaluate the tunnel advanced geological prediction optimization model, obtain evaluation indicators, and provide feedback.

10. The tunnel construction system based on advanced tunnel geological prediction according to claim 1, characterized in that: The specific process of evaluating the application effect of the tunnel advanced geological prediction model is as follows: The geophysical data of the test tunnel is input into the tunnel advance geological prediction model. The tunnel advance geological prediction model outputs the optimal combination prediction method and the prediction conclusion of the optimal combination prediction method. According to the actual verification situation description, the comprehensive prediction accuracy of the tunnel advance geological prediction model is analyzed, recorded as CFA, and substituted into the application effect analysis formula Wherein, η represents the application effect of the tunnel advance geological prediction model, CFA′ represents the preset comprehensive prediction accuracy threshold, and the application effect of the tunnel advance geological prediction model includes the values ​​of 1 and 0. When the application effect of the tunnel advance geological prediction model is 1, it indicates that the application effect of the tunnel advance geological prediction model is good. When the application effect of the tunnel advance geological prediction model is 0, it indicates that the application effect of the tunnel advance geological prediction model is poor.

Citation Information

Cited By

  • Tunnel advanced geology forecasting method based on analytic hierarchy process

    CN119291803A

  • TSP and TEM fusion tunnel water judgment method based on random forest

    CN120577892A

  • Complex terrain karst tunnel disaster identification method and system based on advanced drilling

    CN120670800A