Advanced geological forecasting method and device for highway tunnel construction site

Abstract

The invention discloses an advanced geological forecasting method and device for a highway tunnel construction site, and relates to the technical field of highway tunnel intelligent construction. The method comprises the following steps: constructing an advanced geological forecasting system, and deploying the advanced geological forecasting system to head-mounted mixed reality equipment; the method comprises the following steps: classifying and collecting initial multi-source geological information in highway tunnel construction through an advanced geological forecast system, and carrying out data noise reduction and feature extraction to obtain multi-source data features; based on a multi-source data fusion algorithm, carrying out data fusion processing on the multi-source data features, carrying out weighted calculation, and obtaining and outputting security level information; the on-site environment of the highway tunnel is collected through the head-mounted mixed reality equipment, and the safety level information is anchored to the on-site environment of the highway tunnel through the space anchor points and displayed to construction technicians wearing the head-mounted mixed reality equipment. By the adoption of the method, effective space mapping of digital information and physical information can be achieved, and the highway tunnel construction quality can be improved in an assisted mode.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent construction of highway tunnels, in particular to a geological prediction method and device for a highway tunnel construction site. BACKGROUND

[0002] In the process of highway tunnel construction, advanced geological prediction is an important means to ensure construction safety, reduce sudden geological risks, guide support design and optimize construction organization. The current advanced prediction techniques include geological radar, seismic wave reflection method, electromagnetic detection, infrared thermal imaging, drilling sampling and borehole television, etc. These methods identify abnormal geological bodies such as soft interlayer, karst cave, fault fracture zone, and aquifer in the front rock layer by obtaining the electromagnetic wave response, elastic wave propagation characteristics, temperature change or imaging results of rock mass medium. In order to improve the detection accuracy, multiple physical detection means are often used in engineering, and numerical simulation means are used for comprehensive judgment.

[0003] However, the existing advanced prediction system still faces a series of technical bottlenecks in actual engineering. On the one hand, the operation of traditional equipment is complex, and the information expression method is limited. Most equipment relies on professional technicians to operate, and the input parameters involve a large number of geological characteristics, wave velocity calibration and algorithm adjustment, which is difficult for on-site construction personnel to master independently, and is seriously dependent on data processing and technical interpretation in the rear. At the same time, the system lacks intuitive three-dimensional interface or visual prediction result presentation method, the information interaction efficiency is low, and the rapid response and emergency capability of the construction site are affected. On the other hand, there are significant limitations in data sensing and processing. In complex surrounding rock environment, sensors are easily affected by metal support components, cables, underground water, etc., resulting in signal distortion and imaging blur. For example, the attenuation of geological radar in wet or conductive medium is serious, the seismic wave method has insufficient identification ability for micro-cracks and weak reflection interface, and drilling sampling has strong hysteresis and obvious local problems. Current multi-source fusion analysis still mainly relies on static rules and manual interpretation, lacks a unified data processing framework and compatible algorithm, and is prone to signal conflict, information redundancy or judgment error, affecting the prediction accuracy.

[0004] In addition, the lack of system intelligence and self-adaptive ability is also a major difficulty in engineering application. Most of the existing advanced prediction equipment has a lag in data acquisition frequency, automatic analysis, model updating, etc. The data refresh cycle of some equipment is as long as several hours, which cannot meet the demand for "quasi-real-time" warning in high-risk area construction. At the same time, there is a lack of feedback mechanism and intelligent calibration function for the dynamic evolution of rock mass parameters, which leads to a lag of the prediction results behind the actual geological changes, and reduces the sensing and disposal ability of the system to sudden disasters such as sudden water gushing, mud gushing or surrounding rock instability.

[0005] In summary, the current tunnel advanced geological prediction technology has deficiencies in equipment interactivity, environmental adaptability, and data intelligent analysis, and is difficult to fully meet the safe, fast, and visual prediction requirements under complex geological conditions, resulting in low accuracy, practicability, and response efficiency of on-site prediction. SUMMARY

[0006] To solve the technical problems of low accuracy, practicability, and response efficiency of on-site prediction in the prior art, the embodiments of the present application provide an advanced geological prediction method and device for a highway tunnel construction site. The technical solution is as follows:

[0007] On the one hand, an advanced geological prediction method for a highway tunnel construction site is provided, which is realized by an advanced geological prediction device for a highway tunnel construction site, and the method comprises the following steps:

[0008] S1, an advanced geological prediction system based on interaction between a person, an environment, and a computer is constructed, and the advanced geological prediction system is deployed to a head-mounted mixed reality device;

[0009] S2, initial multi-source geological information in highway tunnel construction is classified and collected through the advanced geological prediction system, data noise reduction and feature extraction are performed on the collected initial multi-source geological information, and multi-source data features are obtained;

[0010] S3, data fusion processing is performed on the multi-source data features based on a multi-source data fusion algorithm, weighted calculation is performed according to the data after fusion processing, safety level information is obtained and output;

[0011] S4, the on-site environment of the highway tunnel is collected through the head-mounted mixed reality device, the safety level information is anchored in the on-site environment of the highway tunnel based on the mixed reality technology through a spatial anchor point, and is displayed to a construction technical personnel wearing the head-mounted mixed reality device.

[0012] On the other hand, an advanced geological prediction device for a highway tunnel construction site is provided, which is applied to an advanced geological prediction method for a highway tunnel construction site, and the device comprises the following:

[0013] A construction unit is configured to construct an advanced geological prediction system based on interaction between a person, an environment, and a computer, and deploy the advanced geological prediction system to a head-mounted mixed reality device;

[0014] An acquisition unit is configured to classify and collect initial multi-source geological information in highway tunnel construction through the advanced geological prediction system, perform data noise reduction and feature extraction on the collected initial multi-source geological information, and obtain multi-source data features;

[0015] The processing unit is configured to perform data fusion processing on the multi-source data features based on a multi-source data fusion algorithm, and perform weighted calculation based on the data after the fusion processing to obtain and output safety level information.

[0016] The display unit is configured to collect a live environment of the highway tunnel through a head-mounted mixed reality device, anchor the safety level information in the live environment of the highway tunnel based on a mixed reality technology through a spatial anchor point, and display the safety level information to a construction technician wearing the head-mounted mixed reality device.

[0017] In another aspect, a device for advanced geological prediction of a highway tunnel construction site is provided, and the device includes a processor and a memory having computer readable instructions stored thereon, which, when executed by the processor, implement any one of the above methods for advanced geological prediction of a highway tunnel construction site.

[0018] In another aspect, a computer readable storage medium is provided, and the storage medium stores at least one instruction, which is loaded and executed by a processor to implement any one of the above methods for advanced geological prediction of a highway tunnel construction site.

[0019] The technical solutions provided by the embodiments of the present application have at least the following beneficial effects:

[0020] The present application provides a comprehensive prediction system integrating multi-source perception, three-dimensional space modeling, augmented reality interaction and intelligent early warning mechanism, which realizes perception enhancement and man-machine collaborative auxiliary judgment of complex geological environment, promotes the transformation of tunnel construction from "experience interpretation" to "data-driven and model-supported" intelligent prediction, and improves the accuracy, practicality and response efficiency of on-site prediction. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0022] Figure 1 is a flow chart of a method for advanced geological prediction of a highway tunnel construction site provided by the embodiments of the present application;

[0023] Figure 2 is a block diagram of a device for advanced geological prediction of a highway tunnel construction site provided by the embodiments of the present application;

[0024] Figure 3A structure schematic diagram of an advanced geological prediction device for a highway tunnel construction site is provided in the embodiments of the present application. DETAILED DESCRIPTION

[0025] The technical solutions in the present application will be described below with reference to the drawings.

[0026] In the embodiments of the present application, the words such as "example", "for example" are used to represent an example, illustration or description. Any embodiment or design scheme described as "example" in the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the word "example" is intended to present the concept in a specific manner. In addition, in the embodiments of the present application, the meaning expressed by "and / or" can be both, or can be one of the two.

[0027] In the embodiments of the present application, "image" and "picture" can be used interchangeably at times, and it should be pointed out that the meanings expressed are consistent when the distinction is not emphasized. "Of", "corresponding" and "corresponding" can be used interchangeably at times, and it should be pointed out that the meanings expressed are consistent when the distinction is not emphasized.

[0028] In the embodiments of the present application, sometimes the subscript such as W1 can be written in the form of non-subscript such as W1, and the meanings expressed are consistent when the distinction is not emphasized.

[0029] To make the technical problems, technical solutions and advantages to be solved by the present application clearer, specific embodiments will be described in detail below with reference to the drawings.

[0030] The embodiments of the present application provide an advanced geological prediction method for a highway tunnel construction site, which can be implemented by an advanced geological prediction system for a highway tunnel construction site, which can include a field sensor, a local device, a head-mounted mixed reality device and the like. As shown in the flow chart of the advanced geological prediction method for a highway tunnel construction site, the processing flow of the method can include the following steps: Figure 1

[0031] S1, construct an advanced geological prediction system based on the interaction among human, environment and computer, and deploy the advanced geological prediction system to a head-mounted mixed reality device.

[0032] Optionally, S1 of constructing an advanced geological prediction system based on the interaction among human, environment and computer, and deploying the advanced geological prediction system to a head-mounted mixed reality device, comprises:

[0033] ​Build the UI interface of the advanced geological prediction system for human-environment-computer interaction, and build the corresponding response event of the indication button.

[0034] In a feasible implementation, the UI interface of the advanced geological prediction system for human-environment-computer interaction is built based on Figma_Unity, and the interface content includes safety level indication, camera interface indication, audio interface indication, and information feedback indication.

[0035] Based on the four interface indication buttons, the corresponding response events are built to realize the gesture operation on the interface through the head-mounted mixed reality device based on the spatial interaction of the tunnel construction personnel.

[0036] In the subsequent steps, the obtained multi-dimensional geological information is transmitted to the head-mounted mixed reality device through cloud service and edge computing based on the head-mounted mixed reality device; the algorithm model and calculation realize the calculation and visualization output of the tunnel construction geological safety level; the construction personnel will feed back the information to the project management department through the information feedback indication on the interface according to the real information on the scene, so as to realize the collaborative construction mechanism.

[0037] Deploy the system to the head-mounted mixed reality device to realize the algorithm-driven interaction function, wherein the resources developed by Unity are deployed to the head-mounted mixed reality device based on Visual Stuodio through the same local area network IP address.

[0038] S2, through the advanced geological prediction system, the initial multi-source geological information in the highway tunnel construction is classified and collected, the collected initial multi-source geological information is subjected to data noise reduction and feature extraction, and multi-source data features are obtained.

[0039] In a feasible implementation, after the initial multi-source geological information is obtained through the field sensor, the initial multi-source geological information is transmitted to the advanced geological prediction system of the head-mounted mixed reality device, and the data noise reduction and feature extraction are performed by the advanced geological prediction system.

[0040] Optionally, the data noise reduction and feature extraction of the collected initial multi-source geological information in S2 to obtain multi-source data features is processed through a multi-scale residual attention network, which can specifically include the following steps S21-S24:

[0041] S21, the collected initial multi-source geological information is represented as an input matrix.

[0042] Optionally, the collected initial multi-source geological information is multi-dimensional and multi-attribute, and can include relative elevation, increment, cumulative amount and subsidence rate of vault subsidence, design surrounding rock grade of a working face, groundwater pressure and change rate, acoustic emission energy, uniaxial compressive strength of rock, fracture density, basic quality index of rock mass, peripheral displacement increment and cumulative amount, and convergence rate.

[0043] In a feasible implementation, the input matrix X can be represented as the following formula (1):

[0044] (1)

[0045] wherein the parameters of each row and each column of the input matrix represent different attributes, respectively representing each data of the collected initial multi-source geological information, and T represents matrix transposition, which is a structured expression of the initial multi-source geological information.

[0046] S22, performing multi-scale parallel convolution on the input matrix to extract local space-time features to obtain convolution feature mappings under fusion of multiple receptive fields.

[0047] In a feasible implementation, for the input matrix X, the embodiment of the application constructs N different size convolution branches in parallel to extract local space-time features through multi-scale parallel convolution, as the following formula (2) and (3):

[0048] (2)

[0049] (3)

[0050] wherein each convolution branch corresponds to a different receptive field with a convolution kernel size extracts local space-time and physical quantity features through a weight matrix and a bias vector . Then, the convolution feature mappings under fusion of multiple receptive fields are obtained to realize full-scale coverage of the prediction information.

[0051] S23, obtaining attention weighted features according to the channel attention and the convolution feature mappings.

[0052] In a feasible implementation, for channel attention is introduced to adaptively strengthen key features:

[0053] (4-1)

[0054] (4-2)

[0055] wherein the global average pooling ) and max-pooling respectively aggregated channel statistics, multi-layer perceptron (MLP) to map channel attention in a dimension-reduce-activation-dimension-increase structure, sigmoid function to ensure the output channel attention weight vector , dynamically emphasizing the characteristic channels sensitive to vault settlement, surrounding rock deformation and physical and mechanical parameters. The attention weighted features are The channel product operation is

[0056] S24, residual fusion is performed on the convolutional feature map and the attention weighted feature to obtain multi-source data features.

[0057] In a feasible implementation, the attention weighted feature and the original multi-scale feature are residual fused to preserve global information and strengthen key responses, as shown in the following formula (5):

[0058] (5)

[0059] Wherein, Y has global information and attention enhanced key responses, and can be directly used as input of a subsequent model. The network structure not only covers settlement, deformation, geology and mechanics multi-dimensional parameters, but also facilitates subsequent expansion of more environmental or rock mass indicators, significantly improving the reliability and accuracy of the advanced geological prediction system.

[0060] S3, based on a multi-source data fusion algorithm, data fusion processing is performed on the multi-source data features, and weighted calculation is performed according to the data after fusion processing to obtain and output safety level information.

[0061] In a feasible implementation, the comprehensive geological safety score A obtained by calculation is used to divide the safety level In addition, the vault settlement trend prediction and the underground water pressure rise risk evolution trend model are selectively constructed for multi-source data.

[0062] The multi-source data features are subjected to data fusion processing and weighted calculation, which can be realized by the head-mounted mixed reality device. Specifically, after data denoising and feature extraction by the advanced geological prediction system, the obtained multi-source data features are subjected to fusion processing. In addition, in order to reduce the calculation amount of the head-mounted mixed reality device and reduce the calculation burden, the head-mounted mixed reality device can transmit the multi-source data features to the local device, and the local device can perform data fusion and weighted calculation to obtain the safety level information. After obtaining the safety level information, the local device transmits the safety level information to the head-mounted mixed reality device.

[0063] Optionally, the multi-source data fusion algorithm of S3 performs data fusion processing on the multi-source data features, including:

[0064] S31, splicing the feature components of the multi-source data features, obtaining a node feature matrix according to the spliced vector and spatial graph convolution.

[0065] Optionally, the feature components of the multi-source data features can include: relative elevation of vault subsidence , relative elevation of vault , settlement increment , cumulative amount , and settlement rate ; surrounding rock grade of the working face design ; groundwater pressure and its change rate ; acoustic emission energy ; uniaxial compressive strength of rock ; fracture density RQD; basic quality index BQ of rock mass; peripheral displacement increment , cumulative amount , and convergence rate . In addition, in order to further improve the prediction accuracy, the principal stress components , and (stress in three directions), acoustic wave velocity and (two dimensions of transverse and longitudinal), change rate of pore water pressure and other multi-dimensional parameters are selected.

[0066] On this basis, optionally, further trend modeling and risk prediction processing are performed on the multi-source index data to generate more rich structural state information output. Specifically, based on the sequence characteristics of the settlement increment, cumulative settlement and settlement rate, a vault settlement trend prediction model is constructed to realize the judgment of the future deformation trend; the risk evolution trend of the rise of groundwater pressure is evaluated by using the parameters such as groundwater pressure.

[0067] The above multi-dimensional output helps the system to expand from static grade division to dynamic trend prediction, providing more targeted and forward-looking risk prompt support for tunnel construction. In addition, in order to realize the future state prediction of the tunnel vault settlement, a long short-term memory neural network (LSTM) is selected to model the settlement time series data, which has good nonlinear fitting ability and long-term dependence modeling ability. The input data is the constructed settlement sequence , wherein is the combined features of the relative elevation of vault , settlement increment , settlement rate at time t, and the model structure is a double-layer LSTM network combined with a fully connected layer to predict the settlement at future time points (such as t+1 to t+K). The prediction expression is , where FC denotes a fully connected layer, is the predicted settlement value at time t+k, where k represents the kth step of prediction after the current time T, and K represents the total number of steps or the total prediction length. The output includes the predicted settlement curve trend, the maximum settlement rate point, and the risk time of exceeding the limit, forming a dynamic vault safety assessment reference.

[0068] To achieve short-term trend prediction of groundwater pressure, a first-order exponential smoothing model is used to model the observation sequence. Let the observation sequence of groundwater pressure at the current time t be then the predicted value at the next time t+1 is which can be expressed as

[0069] , where, denotes the predicted groundwater pressure at time t+1, is the current observed pressure value, is the predicted value at the previous time, is the smoothing coefficient, which controls the degree of response of the prediction to the current observation. A larger emphasizes the current observation value and is suitable for situations where the trend fluctuates quickly; a smaller is smoother and is suitable for slow-changing scenarios. This model structure is simple and has strong real-time performance, making it suitable for monitoring data that is relatively stable or for sampling periods that are relatively dense in groundwater dynamic prediction analysis.

[0070] The above multi-dimensional output helps the system to expand from static level division to dynamic trend prediction, providing more targeted and forward-looking risk warning support for tunnel construction.

[0071] In a feasible implementation, in the multi-modal spatio-temporal graph convolutional network, the input matrix is actually a three-order tensor containing three dimensions of time sequence, space (node), and mode (feature group) at the same time. The feature tensor can be represented by the following formula (6):

[0072] (6)

[0073] where T is the time sequence length (time step), N is the number of spatial nodes (observation position or number of measuring points), F is the feature dimension of each node at each time (i.e., the concatenation length of all modal features), is the input feature tensor of the multi-modal spatio-temporal graph convolutional network (representing the initial multi-modal input matrix), and R represents the set of real numbers (used to indicate the type of elements in the tensor).

[0074] For each time t and each node i, the geological-structure-environment original parameters are concatenated as follows (7): is the input feature vector of node i (e.g. tunnel measuring point, structural unit, sensor location, etc.) at time t, which integrates the multi-modal data perceived or calculated by the node at the time t:

[0075] (7)

[0076] The node feature matrix at time t is shown in the following formula (8), and all t = 1, …, T are stacked to obtain (i.e. the matrix of N node features at each time Stacked by time dimension, a three-order tensor is obtained , as shown in formula (8).

[0077] (8)

[0078] At each time t, the corresponding spatial graph convolution is completed by . In the formula, is the input feature matrix of the l-th layer, represents the spatial adjacency relationship, is the degree matrix, is the trainable weight, takes the ReLU activation function, is the feature dimension number extracted by the l-th layer, and T in formulas (7), (8) and (10) is the transpose symbol. This step effectively integrates the geological connectivity and mechanical coupling information between tunnel measuring points.

[0079] S32, apply one-dimensional time convolution to each node time sequence in the node feature matrix.

[0080] In a feasible implementation, one-dimensional time convolution is applied to each node time sequence, as shown in the following formula (9):

[0081] (9)

[0082] wherein is a time convolution kernel with a length of K’, which slides to cover the historical information of the previous K’ time, thereby capturing the time evolution trend and historical dependence.

[0083] S33, obtain a high-order feature tensor according to the feature matrix after time convolution and the cross-modal attention mechanism.

[0084] In a feasible implementation, in order to further eliminate the redundancy between multi-source modalities and adaptively focus on key information, cross-modal dot attention is introduced, as shown in the following formula (10):

[0085] (10)

[0086] wherein Q, K, V are generated by linear mapping from different modal (such as settlement, ground stress, etc.) characteristic matrices respectively, d is a scaling factor, and M is a fusion feature representation matrix (which is an intermediate calculation result constituting a high-order tensor, not the final high-order tensor) calculated by a cross-modal attention mechanism. In the whole multi-modal spatio-temporal graph convolution network, from the original input tensor , the graph convolution, time convolution, cross-modal attention (i.e. formula (10)) are sequentially performed for modal-level fusion, and M is output and used as the input of the next layer , and after multiple iterations, the final high-order feature tensor is formed . This mechanism can dynamically weight the correlation strength between each modal and realize deep fusion of multi-source information.

[0087] After the above "spatial graph convolution → time convolution → cross-modal attention" is alternately stacked for L layers, a high-order feature tensor is obtained: wherein, represents the feature dimension of the final output layer of the graph neural network, and finally it is mapped to the output result matrix by global pooling or node-by-node linear mapping. This algorithm design not only maintains the spatio-temporal topology structure, but also takes into account the multi-modal collaboration, and has a significant improvement effect on the accuracy and robustness of the advanced geological prediction.

[0088] Optionally, S3 performs weighted calculation on the data after fusion processing to obtain and output safety level information, including:

[0089] S34, an initial judgment matrix is obtained, and according to the initial judgment matrix, the data after fusion processing and a consistency calculation method, an index to which a weight needs to be assigned is determined. In this embodiment, various geological structure parameters extracted from the multi-modal fusion data are used as the basic input variables of the risk assessment model, and the weights of the indexes are determined according to the initial judgment matrix constructed by the industry experts' experience, and the weights are adjusted by the consistency ratio test method, so that the importance of each index to the safety level evaluation is reasonably reflected. The fusion data provides an objective basis for the evaluation, and the expert judgment guides the weight direction of the model, and the two together constitute a geological risk quantitative evaluation system driven by multi-source information.

[0090] In a feasible implementation, the initial judgment matrix is constructed based on the scoring of the multi-dimensional information of the advanced geological prediction by the industry experts in the field on the tunnel construction, and after the consistency ratio is calculated, the weight T of each index is determined i , and the sum of the weights of all indexes is 1.

[0091] S35, the entropy weight method is used to assign a weight value to each index based on the data dispersion degree.

[0092] In a feasible implementation, the entropy weight method is used to automatically assign weight sizes based on the data dispersion degree, and the smaller the entropy value, the higher the weight of the index, as shown in the following formulas (11) and (12):

[0093] (11)

[0094] (12)

[0095] where i represents the i-th index currently being calculated, j represents the sum of all indexes in the denominator, and k represents the k-th data sample or classification interval, is the entropy value of the index i, m represents the number of samples, and n represents the number of evaluation indexes, is the assigned weight of each index.

[0096] S36, according to the weight value, a threshold segmentation function is constructed as an index function.

[0097] In a feasible implementation, then a threshold segmentation function is used as an index function, as shown in the following formula (13):

[0098] (13)

[0099] where, is the value of the i-th geological survey information (i.e., the current observation value of the i-th index), is the safety critical value in the specification, is the critical value of the dangerous value in the specification.

[0100] S37, according to the index function, a linear weighted model is used to divide the geological safety level during tunnel construction.

[0101] In a feasible implementation, on the basis of the above, a linear weighted model is used to divide the geological safety level during tunnel construction, as shown in the following formulas (14) and (15):

[0102] (14)

[0103] (15)

[0104] where, is the weight T i limit value of the index.

[0105] The multi-scale residual attention network (S2) performs deep noise reduction and high-order feature extraction on the original multi-source geological prediction data X, and outputs a feature tensor Y with rich channel information and attention enhancement. The multi-modal spatio-temporal graph convolution network (S3) takes the output Y of S2 as the initial input matrix ) is regarded as a high-dimensional feature representation of each moment and each node, and then according to the spatial adjacency relationship between the tunnel measuring points and the correlation between the multi-source modalities, spatial graph convolution, one-dimensional time convolution and cross-modality attention are sequentially applied, so as to mine the spatial and temporal dependence and modality coupling between nodes at a higher level. The channel attention residual in S2 helps the network to retain the original multi-scale features and highlight the key signals; the cross-modality attention in S3 and the graph convolution residual also ensure smooth information flow and stable gradient in the multi-source information fusion process.

[0106] In addition, in order to realize the prediction of the future state of the tunnel vault settlement, a long short-term memory neural network (LSTM) is selected to model the settlement time series data, which has good nonlinear fitting ability and long-term dependence modeling ability.

[0107] S4, through the head-mounted mixed reality device, the on-site environment of the highway tunnel is collected, the safety level information is anchored in the on-site environment of the highway tunnel based on the mixed reality technology through the spatial anchor point, and is displayed to the construction technical personnel wearing the head-mounted mixed reality device.

[0108] The safety level information is used for prompting or visual information for the construction technical personnel, so that the construction technical personnel make decisions.

[0109] The embodiment of the application realizes the perception enhancement and man-machine collaborative auxiliary judgment of the complex geological environment by a comprehensive prediction system integrating multi-source perception, three-dimensional space modeling, augmented reality interaction and intelligent early warning mechanism, promotes the intelligent prediction transformation of the tunnel construction from "experience interpretation" to "data driving and model support", and improves the accuracy, practicality and response efficiency of the on-site prediction.

[0110] Figure 2 is a kind of advanced geological prediction device block diagram for highway tunnel construction site provided by the embodiment of the application, the device is used for a kind of advanced geological prediction method for highway tunnel construction site. Refer to Figure 2 The device includes a construction unit 210, an acquisition unit 220, a processing unit 230 and a display unit 240. Wherein:

[0111] The construction unit 210 is used to construct an advanced geological prediction system based on the interaction between man, environment and computer, and deploy the advanced geological prediction system to a head-mounted mixed reality device;

[0112] The acquisition unit 220 is used to classify and collect the initial multi-source geological information in the highway tunnel construction through the advanced geological prediction system, perform data noise reduction and feature extraction on the collected initial multi-source geological information, and obtain multi-source data features;

[0113] The processing unit 230 is configured to perform data fusion processing on the multi-source data features based on a multi-source data fusion algorithm, and perform weighted calculation according to the data after fusion processing to obtain and output safety level information.

[0114] The display unit 240 is configured to collect a field environment of the highway tunnel through a head-mounted mixed reality device, anchor the safety level information in the field environment of the highway tunnel based on a mixed reality technology through a spatial anchor point, and display the safety level information to a construction technician wearing the head-mounted mixed reality device.

[0115] The embodiment of the present application realizes the perception enhancement and man-machine collaborative auxiliary judgment of the complex geological environment by a comprehensive prediction system integrating multi-source perception, three-dimensional space modeling, augmented reality interaction and intelligent early warning mechanism, promotes the intelligent prediction transformation of the tunnel construction from 'experience interpretation' to 'data driving and model support', and improves the accuracy, practicality and response efficiency of the field prediction.

[0116] Figure 3 is a structural schematic diagram of an advanced geological prediction equipment for a highway tunnel construction site provided by the embodiment of the present application, as Figure 3 indicated, the advanced geological prediction equipment for the highway tunnel construction site can include the advanced geological prediction device for the highway tunnel construction site shown in the above Figure 2 Optionally, the advanced geological prediction equipment 310 for the highway tunnel construction site can include the first processor 2001.

[0117] Optionally, the advanced geological prediction equipment 310 for the highway tunnel construction site can further include the memory 2002 and the transceiver 2003.

[0118] The first processor 2001, the memory 2002 and the transceiver 2003 can be connected through a communication bus.

[0119] The following will be specifically introduced to each constituent part of the advanced geological prediction equipment 310 for the highway tunnel construction site: Figure 3

[0120] ​The first processor 2001 is a control center of the advanced geological prediction device 310 for a highway tunnel construction site, and can be one processor or a collective term of multiple processing elements. For example, the first processor 2001 is one or more central processing units (CPUs), and can also be an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present application, such as one or more digital signal processors (DSPs), or one or more field programmable gate arrays (FPGAs).

[0121] Optionally, the first processor 2001 can execute various functions of the advanced geological prediction device 310 for a highway tunnel construction site by running or executing software programs stored in the memory 2002 and calling data stored in the memory 2002.

[0122] In a specific implementation, as an embodiment, the first processor 2001 can include one or more CPUs, such as the CPU0 and the CPU1 shown in FIG. 2. Figure 3

[0123] In a specific implementation, as an embodiment, the advanced geological prediction device 310 for a highway tunnel construction site can also include multiple processors, such as the first processor 2001 and the second processor 2004 shown in FIG. 2. Each of the processors can be a single-CPU or a multi-CPU. The processor herein can refer to one or more devices, circuits, and / or processing cores for processing data (such as computer program instructions). Figure 3

[0124] The memory 2002 is configured to store software programs for implementing the schemes of the present application, and the first processor 2001 is configured to control execution. For specific implementation manners, reference can be made to the above method embodiments, and details are not described herein again.

[0125] ​​Optionally, the memory 2002 can be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (RAM) or other type of dynamic storage device that can store information and instructions, an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disk storage, a magnetic disk storage or other magnetic storage devices, or any other medium capable of storing desired program code in the form of instructions or data structures and that can be accessed by a computer, but is not limited to this. The memory 2002 can be integrated with the first processor 2001 or exist independently and be coupled with the first processor 2001 through an interface circuit (not shown in the figure) of the advanced geological forecast device 310 for a highway tunnel construction site, and the embodiments of the present application do not make specific limitations hereon. Figure 3

[0126] The transceiver 2003 is configured to communicate with a network device or a terminal device.

[0127] Optionally, the transceiver 2003 can include a receiver and a transmitter (not shown in the figure separately). The receiver is configured to implement a receiving function, and the transmitter is configured to implement a transmitting function. Figure 3

[0128] Optionally, the transceiver 2003 can be integrated with the first processor 2001 or exist independently and be coupled with the first processor 2001 through an interface circuit (not shown in the figure) of the advanced geological forecast device 310 for a highway tunnel construction site, and the embodiments of the present application do not make specific limitations hereon. Figure 3

[0129] It should be noted that the structure of the advanced geological forecast device 310 for a highway tunnel construction site shown in the figure does not constitute a limitation on the router, and the actual knowledge structure identification device can include more or fewer components than those shown in the figure, or combine certain components, or different component arrangements. Figure 3

[0130] In addition, the technical effects of the advanced geological forecast device 310 for a highway tunnel construction site can refer to the technical effects of the advanced geological forecast method for a highway tunnel construction site described in the above method embodiments, which will not be repeated here. ​​​​

[0131] It should be appreciated that the first processor 2001 in the embodiments of the present application can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0132] It should also be understood that the memory in the embodiments of the present application can be a volatile memory or a nonvolatile memory, or can include both volatile and nonvolatile memory. Among them, the nonvolatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically EPROM (EEPROM) or a flash memory. The volatile memory can be a random access memory (RAM) used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM) and direct memory bus random access memory (direct rambus RAM, DR RAM).

[0133] The above-described embodiments can be implemented in whole or in part by software, hardware (such as a circuit), firmware, or any combination thereof. When implemented in software, the above-described embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are wholly or partially generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transferred from one website, computer, server, or data center to another website, computer, server, or data center through a wired (such as infrared, wireless, microwave, etc.) manner. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. containing one or more available medium collections. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state disk.

[0134] It should be understood that the term "and / or" herein merely describes the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B can represent the following three cases: A exists alone, A and B exist together, and B exists alone, where A and B can be singular or plural. In addition, the character " / " herein generally represents that the associated objects before and after it are in an "or" relationship, but it can also represent an "and / or" relationship, which can be understood according to the context before and after it.

[0135] In the present application, "at least one" means one or more, and "multiple" means two or more. "At least one of the following" or the like means any combination of these items, including any combination of single or multiple items. For example, at least one of a, b, or c can mean a, b, c, a-b, a-c, b-c, or a-b-c, where a, b, and c can be single or multiple.

[0136] It should be understood that in various embodiments of the present application, the size of the sequence number of the above-described processes does not mean the order of execution, and the execution order of the processes should be determined according to their functions and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0137] Those skilled in the art can clearly understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are performed 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 the present application.

[0138] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working processes of the devices, apparatuses and units described above can refer to the corresponding processes in the foregoing method embodiments, which will not be repeated here.

[0139] In several embodiments provided by the present application, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, for example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0140] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0141] In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can exist physically independently, or two or more units can be integrated into one unit.

[0142] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application or the parts of the present application that essentially contribute to the prior art or the parts of the technical solutions can be embodied in the form of software products. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0143] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for advanced geological prediction for a highway tunnel construction site, characterized in that, The method comprises: S1, constructing an advanced geological prediction system based on the interaction among human, environment and computer, and deploying the advanced geological prediction system to a head-mounted mixed reality device; S2, classifying and collecting initial multi-source geological information in highway tunnel construction through the advanced geological prediction system, performing data noise reduction and feature extraction on the collected initial multi-source geological information to obtain multi-source data features; S3, performing data fusion processing on the multi-source data features based on a multi-source data fusion algorithm, performing weighted calculation according to the data after fusion processing, and obtaining and outputting safety level information; S4, collecting the on-site environment of the highway tunnel through the head-mounted mixed reality device, anchoring the safety level information in the on-site environment of the highway tunnel based on the mixed reality technology through a spatial anchor point, and displaying the safety level information to a construction technician wearing the head-mounted mixed reality device.

2. The advanced geological prediction method for a highway tunnel construction site according to claim 1, characterized in that, The S2 of performing data noise reduction and feature extraction on the collected initial multi-source geological information to obtain multi-source data features comprises: S21, representing the collected initial multi-source geological information as an input matrix; S22, performing multi-scale parallel convolution on the input matrix to extract local space-time features and obtain convolution feature mapping under fusion of multiple receptive fields; S23, obtaining attention weighted features according to channel attention and convolution feature mapping; S24, performing residual fusion on the convolution feature mapping and the attention weighted features to obtain multi-source data features.

3. The advanced geological prediction method for a highway tunnel construction site according to claim 2, characterized in that, The collected initial multi-source geological information comprises relative elevation, increment, cumulative amount and settlement rate of vault subsidence, design surrounding rock grade of the working face, underground water pressure and change rate, acoustic emission energy, uniaxial compressive strength of rock, fracture density, basic quality index of rock mass, peripheral displacement increment and cumulative amount, and convergence rate.

4. The advanced geological prediction method for a highway tunnel construction site according to claim 3, characterized by, The S3 of performing data fusion processing on the multi-source data features based on a multi-source data fusion algorithm comprises: S31, splicing feature components of the multi-source data features, and obtaining a node feature matrix according to the spliced vector and spatial graph convolution; S32, applying one-dimensional time convolution to each node time series in the node feature matrix; S33, obtaining a high-order feature tensor according to the feature matrix after time convolution and a cross-modal attention mechanism.

5. The advanced geological prediction method for a highway tunnel construction site according to claim 4, characterized in that, The S3 of performing weighted calculation according to the data after fusion processing to obtain and output safety level information comprises: S34, obtaining an initial judgment matrix constructed in advance, determining indexes that need to be assigned weights according to the initial judgment matrix, the data after fusion processing and a consistency calculation method; S35, assigning weight values to each index based on the data dispersion degree by using an entropy weight method; S36, constructing a threshold segmentation function as an index function according to the weight values; S37, dividing the geological safety level during tunnel construction according to the index function by using a linear weighting model.

6. The advanced geological prediction method for a highway tunnel construction site according to claim 1, characterized by, The S1 of constructing an advanced geological prediction system based on the interaction among human, environment and computer, and deploying the advanced geological prediction system to a head-mounted mixed reality device comprises: Building a UI interface of the advanced geological prediction system for the interaction among human, environment and computer, and constructing a response event corresponding to an indication button.

7. A device for advanced geological prediction for a highway tunnel construction site for implementing the method for advanced geological prediction for a highway tunnel construction site according to any one of claims 1 to 6, characterized in that, The device comprises: a construction unit configured to construct an advanced geological prediction system based on interactions among a human, an environment, and a computer, and deploy the advanced geological prediction system to a head-mounted mixed reality device; a collection unit configured to collect initial multi-source geological information in highway tunnel construction through the advanced geological prediction system, classify the initial multi-source geological information, and perform data denoising and feature extraction on the collected initial multi-source geological information to obtain multi-source data features; a processing unit configured to perform data fusion processing on the multi-source data features based on a multi-source data fusion algorithm, perform weighted calculation based on the data after the fusion processing, and obtain and output safety level information; a display unit configured to collect a live environment of the highway tunnel through the head-mounted mixed reality device, anchor the safety level information in the live environment of the highway tunnel based on a mixed reality technology through a spatial anchor point, and display the safety level information to a construction technician wearing the head-mounted mixed reality device.

8. The advanced geological prediction device for a highway tunnel construction site according to claim 7, characterized by, The collection unit is configured to: S21, represent the collected initial multi-source geological information as an input matrix; S22, perform multi-scale parallel convolution on the input matrix to extract local spatiotemporal features and obtain convolution feature mapping under multiple receptive fields; S23, obtain attention weighted features based on channel attention and the convolution feature mapping; S24, perform residual fusion on the convolution feature mapping and the attention weighted features to obtain multi-source data features.

9. An advanced geological prediction device for a highway tunnel construction site, characterized by, The advanced geological prediction device for a highway tunnel construction site comprises: a processor; a memory having computer readable instructions stored thereon, wherein the computer readable instructions, when executed by the processor, implement the method of any one of claims 1 to 6.

10. A computer readable storage medium, characterized in that, The computer readable storage medium stores program code, and the program code can be called and executed by the processor to implement the method of any one of claims 1 to 6.

Images