Method and system for predicting unfavorable geology of laneway based on spectrum and geophysical prospecting joint inversion

Abstract

The invention discloses a laneway unfavorable geology prediction method and system based on spectrum and geophysical prospecting joint inversion, and belongs to the technical field of laneway unfavorable geology prediction.The method comprises the steps that a rock sample library is constructed, a geophysical prospecting means and a spectrum technology are adopted for testing, and a rock sample multi-source database is constructed; in combination with a machine learning algorithm, constructing a geophysical prospecting information-spectral information-rock information complex equivalent model; aiming at an unexcavated section in front of a laneway tunnel face, geophysical prospecting data of a front geologic body are obtained through testing by a geophysical prospecting means, and the position, the shape and the scale of a front unfavorable geologic body are obtained through interpretation; geophysical prospecting data are input into the geophysical prospecting information-spectral information-rock information complex equivalent model, spectral data and rock information of the unfavorable geologic body are obtained through inversion, the type and the property of the unfavorable geologic body are obtained through qualitative and quantitative interpretation, the position, the shape and the scale of the unfavorable geologic body are corrected according to rock information constraint, and the unfavorable geologic body is obtained. And precise, intelligent and quantitative comprehensive prediction of the unfavorable geology of the laneway is completed.

Description

Technical Field

[0001] The present invention belongs to the technical field of prediction of bad geological conditions in shafts and tunnels, and in particular, relates to a method and system for predicting bad geological conditions in shafts and tunnels based on combined inversion of spectroscopy and geophysical exploration. Background Art

[0002] In the process of coal mining, tunnel construction often encounters adverse geological zones such as fault fracture zones, karst, and alteration zones, which will lead to major geological disasters, equipment damage, casualties, and other accidents. Therefore, accurate prediction and forecasting of adverse geology during the construction of mine tunnels is of great significance to ensure the construction safety of mine tunnels and coal mining.

[0003] At present, the prediction of poor geology in shafts and tunnels is mostly carried out by geophysical means. Geophysical exploration can be used to identify the location and shape of poor geological bodies, such as using seismic wave reflection method to identify fault zones and broken areas, and using electrical or electromagnetic methods to detect water bodies in front of the face. However, on the one hand, only using geophysical means can only identify the "shape" of poor geological bodies (i.e. the location, shape, scale, etc. of poor geological bodies), but cannot identify the "nature" of poor geological bodies (i.e. the nature, type, etc. of poor geological bodies), and cannot achieve comprehensive prediction and forecast of poor geological bodies. Although spectral testing can be combined to more accurately quantitatively describe elemental mineral information, thereby identifying the "nature" of poor geology, spectral testing is limited to the face and cannot predict and forecast the poor geology of rocks or rock masses in the unexcavated geological bodies in front. That is, in actual engineering, it is still impossible to effectively obtain the "nature" of the unexcavated geological body in front of the face.

[0004] On the other hand, geophysical exploration has the problem of multiple solutions, that is, based on the same set of geophysical data, there may be multiple different geological structure interpretations. For example, the uncertainty of the spatial distribution of geophysical parameters, the insufficient coverage of geophysical data, the sensitivity and resolution of instruments, the complexity of geology, etc. will all lead to multiple solutions. Therefore, geological analysis or drilling methods are needed to provide geophysical exploration with prior knowledge constraints to solve this multi-solution problem. However, although drilling can accurately obtain the geological data of the geological body ahead, drilling has the disadvantage of "one-hole view". Its prediction is not comprehensive and its accuracy is poor. In addition, drilling is time-consuming and labor-intensive, which greatly affects the efficiency of tunnel excavation. In actual engineering, geological analysis is mostly done by geological experts entering the tunnel and combining the geological sketch of the face, the information revealed by on-site construction and the survey data for empirical judgment. It cannot provide quantitative identification results and often results in misjudgment and missed judgments. Summary of the invention

[0005] In order to address the deficiencies of the above-mentioned prior art, the present invention provides a method and system for predicting poor geological conditions in shafts and tunnels based on joint inversion of spectroscopy and geophysical exploration. For the unexcavated section ahead of the shaft and tunnel heading face, the method and system combine geophysical exploration methods and spectral technology with machine learning algorithms to conduct a comprehensive "shape" and "property" comprehensive prediction of the poor geological conditions in the geological body ahead, thereby realizing the precise, intelligent and quantitative prediction of poor geological conditions in shafts and tunnels, improving the prediction accuracy and efficiency, and effectively ensuring the rapid and safe construction of shafts and tunnels.

[0006] In a first aspect, the present invention provides a method for predicting poor geology in shafts and tunnels based on joint inversion of spectroscopy and geophysical exploration.

[0007] A method for predicting poor geological conditions in shafts and tunnels based on combined spectral and geophysical inversion, including: Through preliminary geological survey, collect rock samples of normal surrounding rock and adverse geological impact areas along the mine and tunnel, and build a rock sample library; Use geophysical exploration methods and spectral technology to test rock samples in the rock sample library, obtain multi-source data of rock samples, and build a multi-source database of rock samples; Based on the multi-source database of rock samples and combined with machine learning algorithms, a complex equivalent model of geophysical information-spectral information-rock information is constructed; For the unexcavated section in front of the tunnel face, geophysical data of the geological body in front is tested and obtained through geophysical exploration, and the location, shape and scale of the unfavorable geological body in front are obtained through interpretation; The geophysical data of the poor geological body ahead is input into the complex equivalent model of geophysical information-spectral information-rock information for inversion, and the spectral data and rock information of the poor geological body ahead are obtained. Qualitative and quantitative interpretation is performed based on the spectral data to obtain the type and nature of the poor geological body. The position, shape and scale of the poor geological body ahead obtained by geophysical inversion are corrected according to the rock information constraints to complete the comprehensive prediction of poor geology.

[0008] In a further technical solution, the rock sample library includes normal rocks and bad geological rocks of different properties; The normal rocks include sandstone, shale, limestone, basalt, granite, gneiss and slate; The unfavorable geological body rocks include rocks in the fault influence zone, rocks in the fault core, rocks with different degrees of fragmentation, weak interlayer rocks, rocks with different degrees of alteration, rocks in the karst area and rocks in water-rich sections.

[0009] In a further technical solution, the multi-source data includes geophysical data, spectral data and the types and properties of unfavorable geological bodies; the constructed multi-source database of rock samples includes geophysical data and spectral data of rocks of different properties; Use geophysical methods to test rocks and obtain geophysical data, including: Use the seismic wave reflection method to obtain the wave velocity data of the rock, use the resistivity method to obtain the resistivity data of the rock, and use the electromagnetic method to obtain the conductivity data of the rock; The rock is tested by using spectral technology, that is, the rock is tested by using spectra of different wavelengths to obtain spectral data of the rock; the spectra of different wavelengths include visible light-near infrared spectrum, mid-infrared spectrum, and Raman spectrum, and the spectral data include visible light-near infrared spectrum data, mid-infrared spectrum data, and Raman spectrum data.

[0010] According to a further technical solution, the complex equivalent model of geophysical information-spectral information-rock information includes a complex equivalent model of geophysical information-spectral information and a complex equivalent model of spectral information-rock information.

[0011] A further technical solution is that the construction of the complex equivalent model of geophysical information-spectral information-rock information includes: Based on the rock sample multi-source database, the geophysical characteristics and spectral characteristics of each rock sample, as well as the type and property characteristics of the adverse geological body, are extracted; the geophysical characteristics include wave velocity characteristics, resistivity characteristics and conductivity characteristics, and the spectral characteristics include absorption peak characteristics at different wavelengths; Taking the geophysical characteristics of rocks of different properties as input features and the absorption peak characteristics at different wavelengths as output features, a complex equivalent model of geophysical information-spectral information is built and trained; Taking the absorption peak characteristics at different wavelengths as input features and the types and properties of adverse geological bodies as output features, a complex equivalent model of spectral information and rock information is built and trained.

[0012] A further technical solution is to input the geophysical data of the bad geological body ahead into a complex equivalent model of geophysical information-spectral information-rock information for inversion to obtain the spectral data of the bad geological body ahead, including: The wave velocity characteristics, resistivity characteristics, and conductivity characteristics of the bad geological body in front of the tunnel face are input into the complex equivalent model of geophysical information-spectral information, and the absorption peak characteristics of the bad geological body in front at different wavelengths are inverted; The absorption peak characteristics of the bad geological body at different wavelengths are input into the complex equivalent model of spectral information and rock information, and the type and property characteristics of the bad geological body are inverted.

[0013] A further technical solution is to collect the type, properties, geophysical data and spectral data of the unfavorable geological body at the same position in the subsequent excavation section of the shaft and tunnel during the excavation process, verify the prediction results based on the collected data, and dynamically optimize and continuously update the complex equivalent model of geophysical information-spectral information-rock information, forming a data-driven adaptive modeling cycle, including: Obtain geophysical data of the geological body at the current excavation location, spectral information of the geological body at the corresponding location, and update the rock sample library and the rock sample multi-source database; Based on the updated rock sample library and rock sample multi-source database, the complex equivalent model of geophysical information-spectral information-rock information is corrected in real time.

[0014] In a second aspect, the present invention provides a system for predicting poor geological conditions in shafts and tunnels based on joint inversion of spectroscopy and geophysical exploration.

[0015] A system for predicting bad geological conditions in shafts and tunnels based on combined spectral and geophysical inversion, including: The database construction module is used to collect rock samples of normal surrounding rocks and unfavorable geological influence areas along the wells and tunnels through preliminary geological surveys to build a rock sample library; use geophysical exploration methods and spectral technology to test the rock samples in the rock sample library, obtain multi-source data of rock samples, and build a multi-source database of rock samples; Equivalent model building module, which is used to build a complex equivalent model of geophysical information-spectral information-rock information based on a multi-source database of rock samples combined with a machine learning algorithm; The unfavorable geological prediction module is used to test and obtain the geophysical data of the geological body ahead through geophysical means for the unexcavated section ahead of the shaft and tunnel face, and interpret the position, shape and scale of the unfavorable geological body ahead; the geophysical data of the unfavorable geological body ahead is input into the complex equivalent model of geophysical information-spectral information-rock information for inversion, and the spectral data and rock information of the unfavorable geological body ahead are obtained. Based on the spectral data, qualitative and quantitative interpretation is performed to obtain the type and properties of the unfavorable geological body, and the position, shape and scale of the unfavorable geological body ahead obtained by geophysical inversion is corrected according to the rock information constraints to complete the comprehensive prediction of unfavorable geology.

[0016] Further technical solutions include: The model dynamic correction module is used to collect the type, properties, geophysical data and spectral data of the unfavorable geological body at the same position in the subsequent excavation section of the shaft and tunnel during the shaft and tunnel excavation process, verify the prediction results based on the collected data, and dynamically optimize and continuously update the complex equivalent model of geophysical information-spectral information-rock information, forming a data-driven adaptive modeling cycle.

[0017] In a third aspect, the present invention further provides an electronic device comprising a memory and a processor, and computer instructions stored in the memory and executed on the processor, wherein when the computer instructions are executed by the processor, the steps of the method described in the first aspect are completed.

[0018] In a fourth aspect, the present invention further provides a computer-readable storage medium for storing computer instructions, which, when executed by a processor, complete the steps of the method described in the first aspect.

[0019] One or more of the above technical solutions have the following beneficial effects: 1. The present invention provides a method and system for predicting poor geology in shafts and tunnels based on joint inversion of spectroscopy and geophysical prospecting. For the unexcavated section in front of the shaft and tunnel face, the present invention combines geophysical prospecting means and spectral technology, deeply integrates quantitative geological data and geophysical information of spectral analysis, and combines machine learning algorithms to accurately and efficiently predict the "shape" and "properties" of poor geology in the geological body ahead. Compared with the traditional method of combining geophysical prospecting with drilling or geological experience analysis, the present invention can achieve more accurate, intelligent and quantitative prediction of poor geology in the unexcavated section in front of the shaft and tunnel face, avoid the shortcoming of "one-sided view" in geophysical prospecting combined with drilling, improve the comprehensiveness of poor geological body prediction, effectively improve prediction accuracy and efficiency, and ensure rapid and safe construction of shafts and tunnels.

[0020] 2. The present invention constructs a complex equivalent model of geophysical information-spectral information-rock information based on a complex equivalent model of geophysical information-spectral information and a complex equivalent model of spectral information-rock information by building multiple neural network models. First, the spectral information of the geological body behind the face is indirectly obtained based on the geophysical data through the constructed geophysical-spectral model, and the information on the "nature" of the unfavorable geology is accurately predicted according to the spectral information. Then, the detailed geological data such as the type and property of the predicted unfavorable geological body is obtained through the constructed spectral-unfavorable geology model, and the information on the "shape" of the unfavorable geology obtained by geophysical inversion is further constrained and corrected, so as to solve the problem of multiple solutions when only geophysical information is used for inversion, and ensure the accuracy of the predicted information. The present invention not only realizes the prediction of the type and property characteristics of unfavorable geology, but also provides quantitative geological data for the geophysical inversion of the position, shape and scale characteristics of unfavorable geology, and can perform comprehensive prediction of unfavorable geology efficiently and in real time. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0022] Figure 1 It is a flow chart of the method for predicting bad geological conditions of wells and tunnels based on the combined inversion of spectroscopy and geophysical exploration according to an embodiment of the present invention; Figure 2 It is a schematic diagram of a complex equivalent model of geophysical information-spectral information constructed in an embodiment of the present invention; Figure 3 It is a schematic diagram of the complex equivalent model of spectral information-rock information constructed in an embodiment of the present invention. DETAILED DESCRIPTION

[0023] It should be noted that the following detailed descriptions are exemplary only, are intended to describe specific embodiments, are intended to provide further explanation of the present invention, and are not intended to limit exemplary embodiments according to the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those of ordinary skill in the art to which the present invention belongs. In addition, it should also be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0024] Embodiment 1 This embodiment provides a method for predicting bad geological conditions in wells and tunnels based on spectral and geophysical joint inversion. Figure 1 As shown, the specific steps include: Step S1: Through preliminary geological survey, collect rock samples of normal surrounding rocks and rock samples of adverse geological impact areas along the mine and tunnel areas to build a rock sample library.

[0025] Specifically, preliminary geological surveys are carried out before shaft and tunnel excavation, and rock samples of normal surrounding rocks and adverse geological impact areas along the shaft and tunnel are collected to construct a rock sample library, which includes normal rocks and adverse geological rocks of different properties. Normal rocks include sandstone, shale, limestone, basalt, granite, gneiss and slate; adverse geological rocks include fault influence zone rocks, fault core rocks, rocks with different degrees of fragmentation, weak interlayer rocks, rocks with different degrees of alteration, karst zone rocks and water-rich section rocks.

[0026] Step S2: Use geophysical exploration methods and spectral technology to test rock samples in the rock sample library, obtain multi-source data of the rock samples, and build a multi-source database of rock samples.

[0027] In the above step S2, the multi-source data includes geophysical data, spectral data, and information such as the type and properties of unfavorable geological bodies; the constructed rock sample multi-source database includes geophysical data, spectral data, and information such as the type and properties of unfavorable geological bodies of rocks of different properties.

[0028] Specifically, geophysical methods are used to test rocks and obtain geophysical data, including: using seismic wave reflection method to obtain rock wave velocity data, using resistivity method to obtain rock resistivity data, and using electromagnetic method to obtain rock conductivity data.

[0029] The rock is tested by spectral technology, wherein before the spectral test, the sample is ground and the spectrometer parameters are adjusted, including wavelength range, light source intensity, spectral resolution, etc.; then, the rock is tested using spectra of different wavelengths, including visible light-near infrared spectrum, mid-infrared spectrum, Raman spectrum, etc.; after the spectral test, the acquired spectral data is preprocessed, which includes smoothing the acquired data using a smoothing algorithm to remove noise, and using polynomial fitting to perform background subtraction, restore the actual signal, and select representative spectral bands through methods such as principal component analysis, and discard the remaining bands to reduce the workload. Through the above test, spectral data of the rock at different wavelengths are obtained, such as visible light-near infrared spectrum data, mid-infrared spectrum data, Raman spectrum data, etc.

[0030] In addition, through field investigation and recording of poor geological bodies and collecting rock samples at key locations of the poor geological bodies, the type, properties and other information of the poor geological bodies (referred to as poor geological bodies) are determined through subsequent laboratory rock thin section identification, physical and mechanical property testing and chemical analysis. This information can also be called rock information.

[0031] Step S3: Based on the rock sample multi-source database and combined with the machine learning algorithm, a complex equivalent model of geophysical information-spectral information-rock information is constructed.

[0032] Since the geophysical data reflect the differences in the physical properties of the geological structures ahead, such as wave velocity, resistivity, conductivity and other parameters, these physical property parameters can only roughly "describe" the location and size of rocks of different properties ahead. For example, the physical property parameters on the left and right sides of the displayed geophysical data are inconsistent. However, on the basis of clarifying the physical property parameters of normal rocks, the area of ​​poor geological bodies can be determined; further, the geophysical data can only reflect the information of the "shape" of poor geology, but cannot reflect the "property" information such as the intrinsic composition of poor geology, and the geophysical response information has the problem of multiple solutions and cannot directly and accurately reflect the geological information. Considering that the "property" of poor geology is the physical composition of rocks, which is an intrinsic property, the mineral properties can be obtained through the absorption peak characteristics of the spectral curve. The composition and content of the material can be used to obtain the information on the "nature" of the adverse geology, that is, to determine whether the specific composition of the adverse geological body in front is a fault alteration material or a water-containing body, etc. However, the spectrum can only be used at the tunnel face and cannot predict the geological body behind the tunnel face. For this reason, in this embodiment, a geophysical-spectral model is first constructed to indirectly obtain the spectral information of the geological body behind the tunnel face based on the geophysical data, and the information on the "nature" of the adverse geology is accurately predicted based on the spectral information. Then, a spectral-adverse geology model is constructed to predict detailed geological data such as the type and properties of the adverse geological body. Based on the predicted rock information, the information on the "shape" of the adverse geology obtained by the geophysical inversion is further constrained and corrected, so as to solve the problem of multiple solutions in inversion using only geophysical information and ensure the accuracy of the predicted information.

[0033] Therefore, in the above step S3, the complex equivalent model of geophysical information-spectral information-rock information constructed includes a complex equivalent model of geophysical information-spectral information and a complex equivalent model of spectral information-rock information.

[0034] The construction of the complex equivalent model of geophysical information-spectral information-rock information is as follows: Step S3.1: Based on the rock sample multi-source database, extract the geophysical characteristics and spectral characteristics of each rock sample, as well as the type and property characteristics of the bad geological body. The geophysical characteristics extracted from the geophysical data include wave velocity characteristics, resistivity characteristics, and conductivity characteristics, etc. The spectral characteristics extracted from the spectral data include absorption peak characteristics at different wavelengths, etc. In addition, in combination with the above step S2, extract the type and property characteristics of the bad geological body in the rock sample multi-source database.

[0035] Step S3.2: Figure 2 As shown, the geophysical characteristics of rocks of different properties are used as input features, including wave velocity characteristics, resistivity characteristics, conductivity characteristics, etc., and the absorption peak characteristics at different wavelengths are used as output features, including visible light-near infrared spectrum absorption peak, mid-infrared spectrum absorption peak, Raman spectrum absorption peak, etc., to build and train a complex equivalent model of geophysical information-spectral information.

[0036] Step S3.3: Figure 3 As shown, the absorption peak characteristics at different wavelengths are used as input features, including visible light-near infrared spectrum absorption peak, mid-infrared spectrum absorption peak, Raman spectrum absorption peak, etc., and the type and property characteristics of unfavorable geological bodies are used as output features, including unfavorable geological types, unfavorable geological properties, etc., to build and train a complex equivalent model of spectral information-rock information.

[0037] Furthermore, the machine learning algorithm adopted above can be implemented by a neural network. Preferably, any neural network based on an input layer, a hidden layer and an output layer structure can be adopted. This embodiment is implemented by a BP neural network, which includes an input layer, an output layer and a hidden layer (also called a hidden layer), and the hidden layer is located between the input layer and the output layer; the input feature feature After passing through the input layer, the input feature data is transformed through the hidden layer to extract the intermediate features , so that the output layer can better complete the prediction task, and the output features generated are Output by the output layer. Among them, the number of nodes in the input layer and output layer of the constructed neural network are determined, respectively related to the input features and output features Correspondingly, represents the number of input features (i.e. the number of input layer nodes), represents the number of output features (i.e. the number of output layer nodes), while the number of hidden layer nodes in the network (which is the same as the number of intermediate features) Corresponding) is uncertain and can be determined according to the empirical formula: ,in, Indicates the number of hidden layer nodes, parameter is a regulation constant between 1 and 10. By building the above network, the relationship between the input layer nodes and the hidden layer nodes is established and learned. , the relationship between hidden layer nodes and output layer nodes , so as to achieve the mapping between various features.

[0038] Step S4: For the unexcavated section in front of the tunnel face, geophysical data of the geological body in front are tested and obtained by geophysical means, and the position, shape and scale of the unfavorable geological body in front are obtained by interpretation.

[0039] Specifically, geophysical testing is performed on the unexcavated section in front of the tunnel face, including: using seismographs, geophones and seismic sources to obtain the wave velocity characteristics of the unexcavated section in front; using resistivity meters to obtain the resistivity characteristics of the unexcavated section in front; using electromagnetic induction meters to obtain the conductivity characteristics of the unexcavated section in front, etc. Based on the acquired geophysical data / geophysical characteristics, the location, shape, scale and other "shape" information of the unfavorable geology such as water-bearing bodies and fault structures in the front geological body is obtained through interpretation. It should be clear that the "shape" information of the front geological body obtained here is the preliminary interpretation.

[0040] Step S5, input the geophysical data of the bad geological body ahead into the complex equivalent model of geophysical information-spectral information-rock information for inversion, obtain the spectral data and rock information of the bad geological body ahead, perform qualitative and quantitative interpretation based on the spectral data, obtain the type and property of the bad geological body, and correct the position, shape and scale of the bad geological body ahead obtained by geophysical inversion according to the rock information constraints to complete the comprehensive prediction of bad geology.

[0041] By inverting the spectral data / spectral information in step S5, the spectral information is qualitatively and quantitatively interpreted to obtain information on the type and properties of the adverse geological body, thereby achieving a comprehensive prediction of the adverse geological "shape" and "properties".

[0042] Specifically, the above inversion process is: The wave velocity characteristics, resistivity characteristics, and conductivity characteristics of the bad geological body in front of the tunnel face are input into the complex equivalent model of geophysical information-spectral information, and the absorption peak characteristics of the bad geological body in front at different wavelengths are inverted; The absorption peak characteristics of the bad geological body at different wavelengths are input into the complex equivalent model of spectral information and rock information, and the type and property characteristics of the bad geological body are inverted.

[0043] Furthermore, the prediction of the "properties" of the bad geological bodies includes: using the spectral absorption peak characteristics obtained by inversion of geophysical information and the complex equivalent model of spectral information-rock information to obtain the type and properties of the bad geological bodies. Among them, the qualitative and quantitative interpretation of spectral information means: matching and identifying the characteristic peaks of the absorption peak characteristics obtained by inversion of geophysical information with the spectral data in the multi-source database of collected rock samples to obtain the main components of the rock, and calculating the quantitative information of the main components based on the absorption peak intensity.

[0044] Furthermore, the above-mentioned comprehensive prediction of the "shape" and "properties" of adverse geology also includes: using detailed geological data such as the type and properties of the predicted adverse geological body to constrain and correct the "shape" information such as the location, shape, and scale of the adverse geological body obtained by geophysical inversion, so as to solve the problem of multiple solutions when only using geophysical information inversion. At this time, the acquired adverse geological "shape" information is the final accurate prediction information.

[0045] Specifically, in this embodiment, taking the fault as an example of an unfavorable geological body, there are a large number of cracks and broken rocks inside it and it contains a lot of groundwater, which is manifested in geophysical exploration as a decrease in wave velocity, a decrease in resistivity and an increase in conductivity; and the groundwater aquifer, as an unfavorable geological body, contains a lot of dissolved salts and water-containing bodies, which makes it similar to the fault in geophysical inversion; therefore, only the geophysical means are used to invert and obtain the polysemicarity of the existence of faults and groundwater aquifers in the unfavorable geological body. On this basis, considering that common minerals such as montmorillonite, illite, and chlorite in the fault have significant absorption peak characteristics in the near-infrared and mid-infrared bands, they can be used to identify faults; and the water molecules in the aquifer are only at 1.4 and 1.9 The nearby OH (group) stretching vibration absorption peak has obvious absorption peak characteristics, and the spectral characteristics are relatively simple. Therefore, this embodiment uses spectral characteristics to obtain characteristics such as the type and properties of the adverse geological body, and can also perform geological constraints and corrections on the geophysical inversion to further improve the accuracy of the prediction.

[0046] Step S6: During the excavation of shafts and tunnels, the type, properties, geophysical information, and spectral information of the adverse geological body at the same position in the subsequent excavation section of the shafts and tunnels are collected, the prediction results are verified based on the collected information, and the complex equivalent model of geophysical information-spectral information-rock information is dynamically optimized and continuously updated to form a data-driven adaptive modeling cycle.

[0047] Specifically, during the excavation of the shaft and tunnel, the geophysical data of the geological body at the current excavation position and the spectral information of the geological body at the corresponding position are obtained, and the rock sample library and the rock sample multi-source database are updated; based on the updated rock sample library and the rock sample multi-source database, the complex equivalent model of geophysical information-spectral information-rock information is corrected in real time, that is, according to the collected wave velocity characteristics, resistivity characteristics, conductivity characteristics and the absorption peak characteristics at different wavelengths of the geological body at the corresponding position, the complex equivalent model of geophysical information-spectral information is updated; then, according to the collected absorption peak characteristics at different wavelengths and the type and property characteristics of the unfavorable geological body at the corresponding position, the complex equivalent model of spectral information-rock information is updated, so as to form a data-driven adaptive modeling cycle. In this way, the accuracy of the model inversion is guaranteed and the accuracy of the comprehensive prediction of unfavorable geology is improved.

[0048] Embodiment 2 This embodiment provides a system for predicting poor geological conditions in shafts and tunnels based on combined spectral and geophysical inversion, including: The database construction module is used to collect rock samples of normal surrounding rocks and unfavorable geological influence areas along the wells and tunnels through preliminary geological surveys to build a rock sample library; use geophysical exploration methods and spectral technology to test the rock samples in the rock sample library, obtain multi-source data of rock samples, and build a multi-source database of rock samples; Equivalent model building module, which is used to build a complex equivalent model of geophysical information-spectral information-rock information based on a multi-source database of rock samples combined with a machine learning algorithm; The unfavorable geological prediction module is used to test and obtain the geophysical data of the geological body ahead through geophysical means for the unexcavated section ahead of the shaft and tunnel face, and interpret the position, shape and scale of the unfavorable geological body ahead; the geophysical data of the unfavorable geological body ahead is input into the complex equivalent model of geophysical information-spectral information-rock information for inversion, and the spectral data and rock information of the unfavorable geological body ahead are obtained. Based on the spectral data, qualitative and quantitative interpretation is performed to obtain the type and properties of the unfavorable geological body, and the position, shape and scale of the unfavorable geological body ahead obtained by geophysical inversion is corrected according to the rock information constraints to complete the comprehensive prediction of unfavorable geology.

[0049] Furthermore, the above system also includes: The model dynamic correction module is used to collect the type, properties, geophysical data and spectral data of the unfavorable geological body at the same position in the subsequent excavation section of the shaft and tunnel during the shaft and tunnel excavation process, verify the prediction results based on the collected data, and dynamically optimize and continuously update the complex equivalent model of geophysical information-spectral information-rock information, forming a data-driven adaptive modeling cycle.

[0050] Embodiment 3 This embodiment provides an electronic device, including a memory and a processor, and computer instructions stored in the memory and running on the processor. When the computer instructions are run by the processor, the steps in the method for predicting poor geological conditions in shafts and tunnels based on joint inversion of spectroscopy and geophysical exploration as described above are completed.

[0051] Embodiment 4 This embodiment also provides a computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the steps in the method for predicting poor geological conditions in shafts and tunnels based on joint inversion of spectroscopy and geophysical exploration as described above are completed.

[0052] The steps involved in the above embodiments 2 to 4 correspond to the method embodiment 1, and the specific implementation methods can refer to the relevant description part of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method in the present invention.

[0053] Those skilled in the art should understand that the modules or steps of the present invention described above can be implemented by a general-purpose computer device, or alternatively, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0054] The above description is only a preferred embodiment of the present invention. Although the specific implementation mode of the present invention is described in conjunction with the accompanying drawings, it is not a limitation of the protection scope of the present invention. Those skilled in the art should understand that on the basis of the technical solution of the present invention, various modifications or variations that can be made by those skilled in the art without creative work are still within the protection scope of the present invention.

Claims

1. A method for predicting poor geological conditions in shafts and tunnels based on combined spectral and geophysical inversion, characterized in that: include: Through preliminary geological survey, collect rock samples of normal surrounding rock and adverse geological impact areas along the mine and tunnel, and build a rock sample library; Use geophysical exploration methods and spectral technology to test rock samples in the rock sample library, obtain multi-source data of rock samples, and build a multi-source database of rock samples; Based on the multi-source database of rock samples and combined with machine learning algorithms, a complex equivalent model of geophysical information-spectral information-rock information is constructed; For the unexcavated section in front of the tunnel face, geophysical data of the geological body in front is tested and obtained through geophysical exploration, and the location, shape and scale of the unfavorable geological body in front are obtained through interpretation; The geophysical data of the poor geological body ahead is input into the complex equivalent model of geophysical information-spectral information-rock information for inversion, and the spectral data and rock information of the poor geological body ahead are obtained. Qualitative and quantitative interpretation is performed based on the spectral data to obtain the type and nature of the poor geological body. The position, shape and scale of the poor geological body ahead obtained by geophysical inversion are corrected according to the rock information constraints to complete the comprehensive prediction of poor geology.

2. A method for predicting bad geological conditions in shafts and tunnels based on combined spectral and geophysical inversion according to claim 1, characterized in that: The rock sample library includes normal rocks and unfavorable geological rocks of different properties; The normal rocks include sandstone, shale, limestone, basalt, granite, gneiss and slate; The unfavorable geological body rocks include rocks in the fault influence zone, rocks in the fault core, rocks with different degrees of fragmentation, weak interlayer rocks, rocks with different degrees of alteration, rocks in the karst area and rocks in water-rich sections.

3. The method for predicting bad geological conditions in shafts and tunnels based on combined spectral and geophysical inversion according to claim 1, characterized in that: The multi-source data includes geophysical data, spectral data and the types and properties of unfavorable geological bodies; the constructed rock sample multi-source database includes geophysical data and spectral data of rocks of different properties; Use geophysical methods to test rocks and obtain geophysical data, including: using seismic wave reflection method to obtain rock wave velocity data, using resistivity method to obtain rock resistivity data, and using electromagnetic method to obtain rock conductivity data; The rock is tested by using spectral technology, that is, the rock is tested by using spectra of different wavelengths to obtain spectral data of the rock; the spectra of different wavelengths include visible light-near infrared spectrum, mid-infrared spectrum, and Raman spectrum, and the spectral data include visible light-near infrared spectrum data, mid-infrared spectrum data, and Raman spectrum data.

4. The method for predicting bad geological conditions in shafts and tunnels based on combined spectral and geophysical inversion according to claim 1, characterized in that: The complex equivalent model of geophysical information-spectral information-rock information includes a complex equivalent model of geophysical information-spectral information and a complex equivalent model of spectral information-rock information; The construction of the complex equivalent model of geophysical information-spectral information-rock information includes: Based on the rock sample multi-source database, the geophysical characteristics and spectral characteristics of each rock sample, as well as the type and property characteristics of the adverse geological body, are extracted; the geophysical characteristics include wave velocity characteristics, resistivity characteristics and conductivity characteristics, and the spectral characteristics include absorption peak characteristics at different wavelengths; Taking the geophysical characteristics of rocks of different properties as input features and the absorption peak characteristics at different wavelengths as output features, a complex equivalent model of geophysical information-spectral information is built and trained; Taking the absorption peak characteristics at different wavelengths as input features and the types and properties of adverse geological bodies as output features, a complex equivalent model of spectral information and rock information is built and trained.

5. The method for predicting bad geological conditions in shafts and tunnels based on combined spectral and geophysical inversion according to claim 1, characterized in that: The step of inputting the geophysical data of the bad geological body ahead into the complex equivalent model of geophysical information-spectral information-rock information for inversion to obtain the spectral data of the bad geological body ahead includes: The wave velocity characteristics, resistivity characteristics, and conductivity characteristics of the bad geological body in front of the tunnel face are input into the complex equivalent model of geophysical information-spectral information, and the absorption peak characteristics of the bad geological body in front at different wavelengths are inverted; The absorption peak characteristics of the bad geological body at different wavelengths are input into the complex equivalent model of spectral information and rock information, and the type and property characteristics of the bad geological body are inverted.

6. The method for predicting bad geological conditions in shafts and tunnels based on combined spectral and geophysical inversion according to claim 1, characterized in that: During the excavation of the shaft and tunnel, the type, properties, geophysical data, and spectral data of the adverse geological body at the same location in the subsequent excavation section of the shaft and tunnel are collected. The prediction results are verified based on the collected data, and the complex equivalent model of geophysical information-spectral information-rock information is dynamically optimized and continuously updated to form a data-driven adaptive modeling cycle, including: Obtain geophysical data of the geological body at the current excavation location, spectral information of the geological body at the corresponding location, and update the rock sample library and the rock sample multi-source database; Based on the updated rock sample library and rock sample multi-source database, the complex equivalent model of geophysical information-spectral information-rock information is corrected in real time.

7. A system for predicting bad geological conditions in shafts and tunnels based on combined spectral and geophysical inversion, characterized in that: include: The database construction module is used to collect rock samples of normal surrounding rocks and unfavorable geological influence areas along the wells and tunnels through preliminary geological surveys to build a rock sample library; use geophysical exploration methods and spectral technology to test the rock samples in the rock sample library, obtain multi-source data of rock samples, and build a multi-source database of rock samples; Equivalent model building module, which is used to build a complex equivalent model of geophysical information-spectral information-rock information based on a multi-source database of rock samples combined with a machine learning algorithm; The unfavorable geological prediction module is used to test and obtain the geophysical data of the geological body ahead through geophysical means for the unexcavated section ahead of the shaft and tunnel face, and interpret the position, shape and scale of the unfavorable geological body ahead; the geophysical data of the unfavorable geological body ahead is input into the complex equivalent model of geophysical information-spectral information-rock information for inversion, and the spectral data and rock information of the unfavorable geological body ahead are obtained. Based on the spectral data, qualitative and quantitative interpretation is performed to obtain the type and properties of the unfavorable geological body, and the position, shape and scale of the unfavorable geological body ahead obtained by geophysical inversion is corrected according to the rock information constraints to complete the comprehensive prediction of unfavorable geology.

8. A system for predicting bad geological conditions in shafts and tunnels based on combined spectral and geophysical inversion as claimed in claim 7, characterized in that: Also includes: The model dynamic correction module is used to collect the type, properties, geophysical data and spectral data of the unfavorable geological body at the same position in the subsequent excavation section of the shaft and tunnel during the shaft and tunnel excavation process, verify the prediction results based on the collected data, and dynamically optimize and continuously update the complex equivalent model of geophysical information-spectral information-rock information, forming a data-driven adaptive modeling cycle.

9. An electronic device, characterized in that: The method comprises a memory and a processor and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps of a method for predicting poor geology of shafts and tunnels based on joint inversion of spectroscopy and geophysical exploration as described in any one of claims 1 to 6 are completed.

10. A computer-readable storage medium, characterized in that: Used to store computer instructions, which, when executed by a processor, complete the steps of a method for predicting poor geological conditions in shafts and tunnels based on combined spectral and geophysical inversion as described in any one of claims 1-6.

Images