Tunnel working face geological risk prediction method and device based on multi-source information fusion

Abstract

The application provides a tunnel working face geological risk prediction method and device based on multi-source information fusion, the method comprising: acquiring a tunnel excavation working face image, preprocessing the tunnel excavation working face image to obtain a tunnel excavation working face target image; inputting the tunnel excavation working face target image into a trained underground water flow state recognition model to obtain an underground water flow state classification result corresponding to the tunnel excavation working face image; inputting the tunnel excavation working face target image into a trained fracture and fragmentation degree recognition model to obtain a rock mass fragmentation degree classification result corresponding to the tunnel excavation working face image; obtaining a geological condition comprehensive score of the tunnel excavation working face image based on the underground water flow state classification result and the rock mass fragmentation degree classification result, and determining a geological risk grade corresponding to the tunnel excavation working face image based on the geological condition comprehensive score. The application can improve the accuracy and efficiency of tunnel working face geological risk prediction.

Description

Technical Field

[0001] This invention relates to the field of geological risk prediction technology, and in particular to a method and apparatus for predicting geological risks at tunnel working faces based on multi-source information fusion. Background Technology

[0002] Tunnel engineering belongs to underground structures, which are diverse, and the construction methods and technologies for constructing them are also diverse. The construction methods for tunnels are related to the characteristics of the underground structures. The characteristics of tunnel construction include: high concealment and many unknown factors; limited working space, narrow working faces, and significant interference between construction procedures; strong cyclical nature of the construction process, as tunnels are longitudinal, and construction must strictly follow a certain sequence of cyclical operations, such as excavation which must follow the sequence of "drilling—charging—blasting—ventilation—muck removal"; strong comprehensiveness of construction operations, with multiple procedures (tunneling, support, lining, etc.) carried out in the same working environment; the changing geomechanical state and physical and mechanical properties of the surrounding rock during construction, making construction dynamic; harsh working environment, narrow working space, high construction noise, dust, smoke, humidity, low light, poor geological conditions, and safety issues, all of which create unfavorable working conditions for construction personnel; and high operational risk, which is related to the concealment and dynamic nature of the work, requiring construction personnel to constantly monitor the risks of tunnel construction during the process.

[0003] Groundwater is a prevalent risk factor during tunnel construction, especially when traversing aquifers or fractured rock strata, where water inrush can pose serious safety hazards. Currently, the risks posed by water inrush and fractures in tunnels are primarily assessed through manual observation combined with experience. This method is highly subjective, has a high time lag, and limited coverage, resulting in low accuracy and untimely risk prediction at the tunnel working face. Therefore, how to predict geological risks at the tunnel working face in a timely and accurate manner is an urgent technical problem to be solved. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide a method and apparatus for predicting geological risks at tunnel working faces based on multi-source information fusion, so as to eliminate or improve one or more defects existing in the prior art.

[0005] One aspect of the present invention provides a method for predicting geological risks at tunnel working faces based on multi-source information fusion, the method comprising the following steps:

[0006] Acquire images of the tunnel excavation face, and preprocess the images to obtain a target image of the tunnel excavation face;

[0007] The target image of the tunnel excavation face is input into the trained groundwater flow pattern recognition model to obtain the groundwater flow pattern classification result corresponding to the tunnel excavation face image;

[0008] The target image of the tunnel excavation face is input into the trained fracture degree recognition model to obtain the rock mass fracture degree classification result corresponding to the tunnel excavation face image;

[0009] Based on the groundwater flow regime classification results and the rock mass fracture degree classification results, a comprehensive geological condition score is obtained for the tunnel excavation face image, and the geological risk level corresponding to the tunnel excavation face image is determined based on the comprehensive geological condition score.

[0010] In some embodiments of the present invention, preprocessing the tunnel excavation face image to obtain a target image of the tunnel excavation face includes:

[0011] The tunnel excavation face image is denoised using a nonlocal mean denoising method.

[0012] The denoised tunnel excavation face image is subjected to adaptive histogram equalization to obtain the tunnel excavation face image with enhanced image contrast.

[0013] The target image of the tunnel excavation face is obtained by performing geometric correction on the image of the tunnel excavation face after the image contrast has been enhanced.

[0014] In some embodiments of the present invention, the groundwater flow pattern identification model includes a U2Net network and a classifier.

[0015] In some embodiments of the present invention, the target image of the tunnel excavation face is input into a trained fracture fragmentation degree recognition model to obtain the rock mass fragmentation degree classification result corresponding to the tunnel excavation face image, including:

[0016] The cracks in the target image of the tunnel excavation face are determined based on the edge detection algorithm;

[0017] The target image of the tunnel excavation face is divided into multiple grids, and the length, width and number of cracks in each grid are calculated;

[0018] Cluster analysis is performed on multiple grids based on the crack length, width, and number within each grid using a clustering algorithm;

[0019] Based on the cluster analysis results, the rock mass fracture degree classification result corresponding to the tunnel excavation face image was determined.

[0020] In some embodiments of the present invention, a comprehensive geological condition score for the tunnel excavation face image is obtained based on the groundwater flow regime classification results and the rock mass fracturing degree classification results, including:

[0021] The water flow regime score is determined based on the groundwater flow regime classification results.

[0022] The degree of fracturing is determined based on the classification results of the rock mass fracturing degree.

[0023] A comprehensive geological condition score for the tunnel excavation face image is calculated based on the water flow regime score and the degree of fragmentation score.

[0024] In some embodiments of the present invention, the formula for calculating the comprehensive geological condition score is as follows:

[0025] S total =W water ×S water +W fracture ×S fracture；

[0026] Among them, S total S is a comprehensive score for geological conditions. water For water flow regime scoring, S fracture To score the degree of breakage, W water W represents the weight of the water flow state. fracture Weights for the degree of fragmentation.

[0027] In some embodiments of the present invention, preprocessing the tunnel excavation face image to obtain a target image of the tunnel excavation face further includes:

[0028] Acquire images of the tunnel excavation face at different resolutions and scales;

[0029] Multiple images of the tunnel excavation face with different resolutions and scales are fused together.

[0030] In some embodiments of the present invention, the groundwater flow regime classification result is dry, water-bearing, or flowing; and / or,

[0031] The rock mass is classified as broken, relatively broken, or intact.

[0032] Another aspect of the present invention provides a geological risk prediction system for tunnel working faces based on multi-source information fusion, including a processor, a memory, and a computer program stored in the memory. The processor is used to execute the computer program, and when the computer program is executed, the system implements the steps of the method as described in any of the above embodiments.

[0033] Another aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method as described in any of the above embodiments.

[0034] This invention relates to a method and apparatus for predicting geological risks at tunnel working faces based on multi-source information fusion. It identifies groundwater flow regime classification results using a groundwater flow regime identification model and rock mass fracture degree classification results using a fracture degree identification model. Finally, it combines the groundwater flow regime classification results and the rock mass fracture degree classification results to determine the final geological risk level. This method and apparatus utilize advanced computer vision technology to comprehensively monitor tunnel working faces, enabling more timely and accurate prediction of potential geological risks. This allows for timely and effective preventative and response measures, reducing the accident rate and improving tunnel construction efficiency.

[0035] Additional advantages, objects, and features of the invention will be set forth in part in the description which follows, and will also become apparent in part to those skilled in the art upon studying the description, or may be learned by practice of the invention. The objects and other advantages of the invention can be realized and obtained by means of the structures specifically pointed out in the description and drawings.

[0036] Those skilled in the art will understand that the objectives and advantages achievable with the present invention are not limited to those specifically described above, and that the above and other objectives achievable with the present invention will become clearer from the following detailed description. Attached Figure Description

[0037] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, are not intended to limit the scope of the invention. The components in the drawings are not drawn to scale but are merely illustrative of the principles of the invention. For ease of illustration and description of certain parts of the invention, corresponding portions in the drawings may be enlarged, i.e., may appear larger relative to other components in an exemplary device actually manufactured according to the invention. In the drawings:

[0038] Figure 1 This is a flowchart illustrating a method for predicting geological risks at tunnel working faces based on multi-source information fusion, according to one embodiment of the present invention.

[0039] Figure 2 This is a schematic diagram of the architecture of a tunnel working face geological risk prediction system based on multi-source information fusion in one embodiment of the present invention. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the embodiments and accompanying drawings. Here, the illustrative embodiments and descriptions of this invention are used to explain the invention, but are not intended to limit the invention.

[0041] It should also be noted that, in order to avoid obscuring the invention with unnecessary details, only the structures and / or processing steps closely related to the solution according to the invention are shown in the accompanying drawings, while other details that are not closely related to the invention are omitted.

[0042] It should be emphasized that the term "including / comprises" as used herein refers to the presence of a feature, element, step, or component, but does not exclude the presence or addition of one or more other features, elements, steps, or components.

[0043] It should also be noted that, unless otherwise specified, the term "connection" in this article can refer not only to a direct connection, but also to an indirect connection involving an intermediary.

[0044] Currently, traditional tunnel face geological risk prediction mainly relies on manual observation and experience-based judgment. This leads to inconsistencies in judgments regarding the same phenomenon due to differences in the experience and understanding of different observers, resulting in inconsistent early warning results. Furthermore, existing manual observation methods cannot achieve real-time monitoring, easily missing the optimal early warning window. Moreover, manual observation typically only targets local areas, making it difficult to comprehensively monitor the entire tunnel face. Based on these factors, it is clear that existing traditional tunnel face geological risk prediction methods suffer from low accuracy and untimely warnings, resulting in low safety and efficiency in tunnel construction. To address these technical problems in existing tunnel face geological risk prediction methods, this application provides a tunnel face geological risk prediction method and device based on multi-source information fusion. During tunnel construction, this method and device utilize advanced computer vision technology to comprehensively monitor the tunnel face, specifically including the identification of groundwater flow patterns and the degree of rock mass fracture. By combining information from these two aspects, a comprehensive scoring mechanism is established, forming a comprehensive early warning system, ultimately achieving real-time and accurate early warning of geological risks at the tunnel face.

[0045] In the following description, embodiments of the invention will be illustrated with reference to the accompanying drawings. In the drawings, the same reference numerals represent the same or similar parts, or the same or similar steps.

[0046] Figure 1 This is a flowchart illustrating a method for predicting geological risks at tunnel working faces based on multi-source information fusion, according to an embodiment of this application. Figure 1As shown, the method for predicting geological risks at tunnel working faces based on multi-source information fusion includes at least steps S10 to S40.

[0047] Step S10: Obtain an image of the tunnel excavation face and preprocess the image to obtain a target image of the tunnel excavation face.

[0048] In this step, high-resolution camera equipment suitable for the tunnel environment can be used to acquire images of the tunnel excavation face to ensure clear images can be obtained in low-light conditions. Additionally, considering the environment inside the tunnel, industrial-grade cameras with waterproof and dustproof properties can be selected to acquire images of the tunnel excavation face.

[0049] In addition, preprocessing of tunnel excavation face images may include image denoising and image enhancement. In one embodiment, preprocessing the tunnel excavation face image to obtain a target image of the tunnel excavation face may include: denoising the tunnel excavation face image based on a non-local means denoising method; performing adaptive histogram equalization on the denoised tunnel excavation face image to obtain a tunnel excavation face image with enhanced image contrast; and performing geometric correction on the tunnel excavation face image with enhanced image contrast to obtain the target image of the tunnel excavation face. In this embodiment, an advanced denoising algorithm—the non-local means algorithm—is used to reduce noise in the image and enhance image quality; adaptive histogram equalization (CLAHE) and other techniques are used to improve image contrast and highlight the characteristics of groundwater flow; furthermore, to correct image distortion caused by camera angle or equipment installation position and ensure the accuracy of image data, geometric correction is further performed on the tunnel excavation face image acquired by the image acquisition device. It is understood that the image preprocessing methods and steps listed in this embodiment are only examples. In other embodiments, specific tunnel excavation face images can be preprocessed based on actual needs, as long as the target image of the tunnel excavation face obtained after preprocessing can meet the requirements of the groundwater flow pattern recognition model and the fracture degree recognition model for the input image.

[0050] In another embodiment, the preprocessing of the tunnel excavation face image, in addition to denoising and geometric correction, may also include the following steps: acquiring tunnel excavation face images of different resolutions and scales; and fusing multiple tunnel excavation face images of different resolutions and scales. This embodiment further fuses images of different resolutions and scales, improving the ability to express image details, thereby enabling better extraction of feature information from the image.

[0051] Step S20: Input the target image of the tunnel excavation face into the trained groundwater flow pattern recognition model to obtain the groundwater flow pattern classification result corresponding to the tunnel excavation face image.

[0052] In this step, the groundwater flow regime in the target image of the tunnel excavation face is identified based on the pre-trained groundwater flow regime recognition model. The groundwater flow regime can be divided into three types: dry, water-bearing, or flowing. Therefore, the groundwater flow regime classification result corresponding to the target image of the tunnel excavation face obtained in this step is dry, water-bearing, or flowing.

[0053] In one embodiment, the groundwater flow pattern identification model includes a U2Net network and a classifier. The U2Net network is a powerful image segmentation network that can accurately segment regions of interest in complex backgrounds. In addition, the output of the U2Net network is connected to the classifier, which further sends the region of interest information output by the U2Net network to the classifier for classification, thereby obtaining the groundwater flow pattern classification result of the target image of the tunnel excavation face.

[0054] Furthermore, the classifier can specifically be a softmax classifier, meaning that the groundwater flow pattern identification model includes a U2Net network and a softmax classifier. In this embodiment, computer vision technology based on a U2Net network is used to analyze the target image of the tunnel excavation face, identify the groundwater flow pattern, and obtain the category of the groundwater flow pattern based on the softmax classifier. The category of the groundwater flow pattern can be dry, water-bearing, or flowing, etc.

[0055] In addition, in order to obtain a well-trained groundwater flow pattern recognition model, one embodiment of the tunnel working face geological risk prediction method may further include the following steps: constructing an initial groundwater flow pattern recognition model and a first sample dataset, wherein each sample data in the first sample dataset includes target image samples of the tunnel excavation working face and groundwater flow pattern result samples; further, based on the constructed first sample dataset, the initial groundwater flow pattern recognition model is pre-trained to update the model parameters of the initial groundwater flow pattern recognition model to obtain a well-trained groundwater flow pattern recognition model.

[0056] Step S30: Input the target image of the tunnel excavation face into the trained fracture degree recognition model to obtain the rock mass fracture degree classification result corresponding to the tunnel excavation face image.

[0057] In this step, the rock mass fracture degree in the target image of the tunnel excavation face is identified based on the pre-trained fracture degree recognition model. The rock mass fracture degree can be divided into three levels: fractured, relatively fractured, and intact. Therefore, the rock mass fracture degree classification result corresponding to the target image of the tunnel excavation face obtained in this step is fractured, relatively fractured, or intact.

[0058] In one embodiment, the target image of the tunnel excavation face is input into a trained fracture fragmentation degree recognition model to obtain the rock mass fragmentation degree classification result corresponding to the tunnel excavation face image. This includes: determining the fractures in the target image of the tunnel excavation face based on an edge detection algorithm; dividing the target image of the tunnel excavation face into multiple grids and calculating the length, width, and number of fractures in each grid; performing cluster analysis on the multiple grids based on the fracture length, width, and number of fractures in each grid using a clustering algorithm; and determining the rock mass fragmentation degree classification result corresponding to the tunnel excavation face image based on the cluster analysis results. In this embodiment, the target image of the tunnel excavation face is first processed based on an edge detection algorithm to identify fractures on the structural surface, and then the fracture length is statistically analyzed using a gridding method to assess the rock mass fragmentation degree.

[0059] For example, the edge detection algorithm used can be the Laplace edge detection algorithm, a classic edge detection method that can effectively identify edge features in an image. Further, the target image of the tunnel excavation face processed by the Laplace edge detection algorithm is divided into structured statistical units according to a certain grid size. Within each grid unit, the length, width, and number of cracks are statistically analyzed to quantify the degree of rock mass fracture. In this example, the Laplace edge detection algorithm is used to process the target image of the tunnel excavation face. The Laplace edge detection algorithm identifies edges by calculating the image grayscale change rate, effectively detecting cracks. It is understood that in other embodiments, edge detection algorithms other than Laplace can also be used to identify cracks in the image.

[0060] In addition, image processing algorithms can be used for statistical calculations when calculating the length and width of cracks within each grid cell. After obtaining the length, width, and number of cracks in each grid cell, clustering algorithms can be used to classify the degree of rock fragmentation based on the statistical results, in order to identify whether the rock fragmentation degree of the target image of the tunnel excavation face is broken, relatively broken, or intact. Specifically, for multiple grid cells, they can first be clustered according to the length, width, and number of cracks within each grid cell. Grid cells with similar crack lengths, widths, and numbers are grouped into one category. Finally, the category containing the most grid cells is taken as the category to which the rock fragmentation degree of the target image of the tunnel excavation face belongs. For example, when the number of grid cells classified into the broken category is the largest, the rock fragmentation degree of the target image of the tunnel excavation face is determined to be broken. For example, the clustering algorithm used in this embodiment can be the K-means algorithm, that is, the rock mass fracture degree of the target image of the tunnel excavation face is classified based on the K-means algorithm; it is understood that the clustering algorithm listed in this example is only an example, and other clustering algorithms can be used to classify the rock mass fracture degree in other embodiments.

[0061] Similarly, in order to obtain a trained fracture degree identification model, a tunnel working face geological risk prediction method of one embodiment may further include the following steps: constructing an initial fracture degree identification model and a second sample dataset, wherein each sample data in the second sample dataset includes target image samples of the tunnel excavation working face and rock mass fracture degree classification result samples; further, based on the constructed second sample dataset, the initial fracture degree identification model is pre-trained to update the model parameters of the initial fracture degree identification model to obtain a trained fracture degree identification model.

[0062] Step S40: Based on the groundwater flow pattern classification results and the rock mass fracture degree classification results, obtain the comprehensive geological condition score of the tunnel excavation face image, and determine the geological risk level corresponding to the tunnel excavation face image based on the comprehensive geological condition score.

[0063] In this step, a comprehensive risk assessment system is established, combining groundwater flow regime classification results and rock mass fracturing degree classification results to comprehensively score the tunnel excavation face images, obtaining a geological risk level matching the comprehensive score. Furthermore, before determining the geological risk level, a correspondence between the comprehensive score and the geological risk level can be pre-established.

[0064] In one embodiment, obtaining a comprehensive geological condition score for the tunnel excavation face image based on the groundwater flow regime classification result and the rock mass fracturing degree classification result includes: determining a flow regime score based on the groundwater flow regime classification result; determining a fracturing degree score based on the rock mass fracturing degree classification result; and calculating a comprehensive geological condition score for the tunnel excavation face image based on the flow regime score and the fracturing degree score. Similarly, to determine the flow regime score and fracturing degree score based on the groundwater flow regime classification result and the rock mass fracturing degree classification result, respectively, a correspondence between the groundwater flow regime classification result and the flow regime score, and a correspondence between the rock mass fracturing degree classification result and the fracturing degree score, can be pre-constructed.

[0065] For example, the formula for calculating the comprehensive geological condition score is as follows:

[0066] S total =W water ×S water +W fracture ×S fracture；

[0067] Among them, S total S is a comprehensive score for geological conditions. water For water flow regime scoring, S fracture To score the degree of breakage, W water W represents the weight of the water flow state. fracture Weights for the degree of fragmentation.

[0068] In the above embodiments, different weights are assigned to the groundwater flow regime and the degree of structural surface fragmentation according to actual engineering needs. Based on a comprehensive scoring formula, the water flow regime score and the degree of fragmentation score are combined to calculate the comprehensive geological condition score of the rock mass. When determining the weights of the water flow regime and the degree of fragmentation, the optimal weight allocation can be determined based on expert experience or historical data analysis. Alternatively, the comprehensive scoring model can be dynamically adjusted based on real-time data using machine learning algorithms to optimize the accuracy of the comprehensive score.

[0069] Furthermore, based on the correspondence between the comprehensive score and geological risk standards, multiple early warning levels (which can also be understood as risk levels) can be set, such as low, medium, and high early warning levels. Each early warning level corresponds to a different comprehensive score. After determining the comprehensive geological condition score of the tunnel excavation face image based on the above embodiments, the calculated comprehensive geological condition score can be mapped to the early warning level using a fuzzy logic algorithm. Additionally, an early warning value can be further set. When the early warning level corresponding to the comprehensive geological condition score reaches the warning value, an early warning of geological risk at the tunnel working face is achieved. Furthermore, to further improve the reliability of the early warning module, in one embodiment, the early warning value can be optimized in real time based on historical data and big data analysis methods.

[0070] As can be seen from the above embodiments, the tunnel working face geological risk prediction method based on multi-source information fusion of this application uses the advanced U2Net network to identify groundwater flow patterns and the Laplace edge detection algorithm to identify fractures, thereby improving the accuracy and reliability of geological risk identification and reducing human error. Through comprehensive scoring and early warning indicators, it can provide scientific and objective risk assessment and early warning, helping construction managers to take timely measures to prevent accidents. Furthermore, this method also achieves automated monitoring and early warning, reducing manual intervention and improving work efficiency and safety.

[0071] Accordingly, the present invention also provides a geological risk prediction system for tunnel working faces based on multi-source information fusion, including a processor, a memory, and a computer program stored in the memory. The processor is used to execute the computer program, and when the computer program is executed, the system implements the steps of the method as described in any of the above embodiments.

[0072] For example, such as Figure 2 As shown, the system also includes at least an image acquisition module 100, a water flow pattern recognition module 200, a fracture degree recognition module 300, and a geological risk assessment module 400. The image acquisition module 100 acquires images of the tunnel excavation face and preprocesses these images to obtain a target image of the tunnel excavation face. The water flow pattern recognition module 200 obtains a groundwater flow pattern classification result corresponding to the tunnel excavation face image based on a trained groundwater flow pattern recognition model. The fracture degree recognition module 300 obtains a rock mass fracture degree classification result corresponding to the tunnel excavation face image based on a trained fracture fracture degree recognition model. The geological risk assessment module 400 obtains a comprehensive geological condition score for the tunnel excavation face image based on the groundwater flow pattern classification result and the rock mass fracture degree classification result, and determines the geological risk level corresponding to the tunnel excavation face image based on the comprehensive geological condition score.

[0073] The geological risk prediction system described in the above embodiment can monitor the groundwater flow pattern and structural surface fragmentation degree of the tunnel working face in real time, achieving all-weather, full-coverage risk monitoring. Through a comprehensive scoring of the groundwater flow pattern and structural surface fragmentation degree, it can more comprehensively and accurately assess the geological risks of the tunnel working face.

[0074] This invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the aforementioned edge computing server deployment method. The computer-readable storage medium can be a tangible storage medium, such as random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, floppy disks, hard disks, removable storage disks, CD-ROMs, or any other form of storage medium known in the art.

[0075] Those skilled in the art will understand that the exemplary components, systems, and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software, or a combination of both. Whether implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention. When implemented in hardware, it can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this invention are programs or code segments used to perform the desired tasks. The programs or code segments can be stored in a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried in a carrier wave.

[0076] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.

[0077] In this invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with or in place of features of other embodiments.

[0078] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, various modifications and variations of the embodiments of the present invention are possible. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for predicting geological risks at tunnel working faces based on multi-source information fusion, characterized in that, The method includes the following steps: A tunnel excavation face image is acquired, and the image is denoised using a nonlocal mean denoising method. The denoised image is then subjected to adaptive histogram equalization to obtain a contrast-enhanced tunnel excavation face image. Geometric correction is performed on the contrast-enhanced image to obtain a corrected image. Multiple corrected images of different resolutions and scales are then fused to obtain the target image of the tunnel excavation face. The target image of the tunnel excavation face is input into a trained groundwater flow pattern recognition model to obtain the groundwater flow pattern classification result corresponding to the tunnel excavation face image. The groundwater flow pattern recognition model includes a U2Net network and a softmax classifier. The U2Net network is used to perform groundwater flow pattern recognition on the target image of the tunnel excavation face to obtain regional information and send the regional information to the softmax classifier. The softmax classifier is used to classify the regional information to obtain the groundwater flow pattern classification result to which the target image of the tunnel excavation face belongs. The cracks in the target image of the tunnel excavation face are determined based on an edge detection algorithm; the target image of the tunnel excavation face is divided into multiple grids, and the length, width, and number of cracks in each grid are calculated; based on the length, width, and number of cracks in each grid, a clustering algorithm is used to perform cluster analysis on the multiple grids; based on the cluster analysis results, the rock mass fracture degree classification result corresponding to the tunnel excavation face image is determined; Based on the groundwater flow regime classification results and the rock mass fracturing degree classification results, a comprehensive geological condition score is obtained for the tunnel excavation face image. Based on this comprehensive geological condition score, the geological risk level corresponding to the tunnel excavation face image is determined. The formula for calculating the comprehensive geological condition score is as follows: S total = W water × S water + W fracture × S fracture； in, S total A comprehensive score for geological conditions. S water To score the water flow pattern, S fracture To score the degree of breakage, W water For the water flow state weight, W fracture The weights are determined by the degree of breakage. When determining the weights of the water flow state and the degree of breakage, the optimal weight allocation can be determined based on expert experience or historical data analysis. In addition, the comprehensive scoring model can be dynamically adjusted based on real-time data using machine learning algorithms to optimize the accuracy of the comprehensive score.

2. The method for predicting geological risks at tunnel working faces based on multi-source information fusion according to claim 1, characterized in that, Based on the groundwater flow regime classification results and the rock mass fracturing degree classification results, a comprehensive geological condition score for the tunnel excavation face image is obtained, including: The water flow regime score is determined based on the groundwater flow regime classification results. The degree of fracturing is determined based on the classification results of the rock mass fracturing degree. A comprehensive geological condition score for the tunnel excavation face image is calculated based on the water flow regime score and the degree of fragmentation score.

3. The method for predicting geological risks at tunnel working faces based on multi-source information fusion according to any one of claims 1 to 2, characterized in that, The groundwater flow regime classification result is dry, water-bearing, or flowing; and / or, The rock mass is classified as broken, relatively broken, or intact.

4. A geological risk prediction system for tunnel working faces based on multi-source information fusion, comprising a processor, a memory, and a computer program stored in the memory, characterized in that, The processor is configured to execute the computer program, and when the computer program is executed, the system implements the steps of the method as described in any one of claims 1 to 3.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 3.

Images