Deep rock mass crustal stress stochastic analysis method and device and processing equipment

Abstract

The invention provides a deep rock mass crustal stress stochastic analysis method, a deep rock mass crustal stress stochastic analysis device and processing equipment, which are used for configuring a stochastic geologic model and maximum posterior probability Markov iteration from the angle of spatial correlation and the like of lithologic distribution of an engineering area aiming at the stochastic characteristics of lithologic distribution of a deep complex geologic body, so that the geologic stress stochastic analysis method and the deep rock mass crustal stress stochastic analysis device are obtained. According to the method, the three-dimensional geological random field of the deep rock mass engineering is constructed, and the distribution characteristics of the crustal stress field in each key aspect in the inversion crustal stress field are analyzed, so that the problem that the precision of the crustal stress analysis result of the deep rock mass obtained in the prior art is relatively poor is effectively solved, and an effective reference basis is provided for the design and construction process of the rock mass engineering.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of geology, in particular to a deep rock mass geostress random analysis method, device and processing equipment. BACKGROUND

[0002] With the continuous increase of the buried depth, the rock mass environment of the geotechnical engineering structure is increasingly complex, and the information collection means that can be applied is increasingly limited.

[0003] Under this feature, in the actual situation, only the main stratum structure and geological structure are usually considered in the deep underground engineering, and for the area where the stratum structure information is missing, the extension interpolation method is often used, and the significant change of the lithology in a small range is ignored to realize the geostress field inversion process, which greatly increases the subjectivity of the inversion model and the error of the inverted geostress field. The analysis result of the deep rock mass geostress often cannot reasonably evaluate the true stress environment of the surrounding rock of the engineering environment, that is, there is a problem of low analysis accuracy. SUMMARY

[0004] The present application provides a deep rock mass geostress random analysis method, device and processing equipment, which is used for the random characteristics of the lithology distribution of the deep complex geological body. The present application starts from the spatial correlation of the lithology distribution of the engineering area and configures a random geological model and a maximum posterior probability Markov iteration to construct a three-dimensional geological random field of the deep rock mass engineering and analyze the distribution characteristics of the geostress field in each key aspect in the inverted geostress field, so as to effectively solve the problem of poor accuracy of the deep rock mass geostress analysis result obtained by the prior art and provide an effective reference basis for the design and construction process of the rock mass engineering.

[0005] In a first aspect, the present application provides a deep rock mass geostress random analysis method, which comprises: For a target geotechnical engineering, on the basis of the engineering area geological background, the engineering area structure condition, the geological drilling information in the engineering area and the lithology information exposed by the hole wall in the engineering area, the corresponding deep rock mass geological information is configured; On the basis of the deep rock mass geological information, a three-dimensional cubic grid model including a non-drilling unit and a drilling unit corresponding to the drilling unit in the drilling area is constructed; The three-dimensional cubic grid model is subjected to a maximum posterior probability Markov iteration whose iteration number is controlled by the minimum posterior energy, and the iteration is completed when the variation coefficient value of the average information entropy of the calculated non-drilling unit meets the convergence requirement, so as to obtain a random geological model; The geostress field is inverted through the random geological model, and on the basis of the inverted geostress field result, the distribution characteristics of the geostress field in the measurement point layer, the measurement section layer and the regional layer are analyzed.

[0006] In a second aspect, the application provides a deep rock mass in-situ stress random analysis device, which comprises: A configuration unit is configured to, for a target geotechnical engineering, configure corresponding deep rock mass geological information based on engineering area geological background, engineering area tectonic conditions, geological drilling information in the engineering area, and hole wall exposed lithology information in the engineering area. A construction unit is configured to construct a three-dimensional cubic lattice model comprising non-drilling units and drilling area corresponding drilling units based on the deep rock mass geological information. An iteration unit is configured to perform maximum a posteriori Markov iteration of an iteration number of the three-dimensional cubic lattice model controlled by a minimum posterior energy, and complete the iteration when a variation coefficient value of the average information entropy of the non-drilling units calculated meets convergence requirements, to obtain a random geological model. An analysis unit is configured to perform in-situ stress field inversion through the random geological model, and analyze in-situ stress field distribution characteristics of a measurement point layer, a measurement section layer, and a regional layer based on the in-situ stress field inversion result.

[0007] In a third aspect, the application provides a processing device comprising a processor and a memory, the memory storing a computer program, and the processor executing the computer program in the memory to perform the method provided in the first aspect or any possible implementation manner of the first aspect of the application.

[0008] In a fourth aspect, the application provides a computer readable storage medium storing a plurality of instructions, the instructions being adapted to be loaded by a processor to execute the method provided in the first aspect or any possible implementation manner of the first aspect of the application.

[0009] From the above, the application has the following beneficial effects: Under the target of deep rock mass in-situ stress analysis, in view of the random characteristics of the lithology distribution of the deep complex geological body, the application starts from the spatial correlation of the lithology distribution of the engineering area, configures a random geological model and maximum a posteriori Markov iteration, constructs a three-dimensional geological random field of the deep rock mass engineering, and analyzes in-situ stress field distribution characteristics of each key aspect in the in-situ stress field inversion, to effectively solve the problem of poor accuracy of the deep rock mass in-situ stress analysis result obtained by the prior art, and provide an effective reference for the design and construction process of the rock mass engineering. BRIEF DESCRIPTION OF DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0011] Figure 1 A flowchart of a deep rock mass in-situ stress random analysis method; Figure 2 A scene diagram of a stratum spatial correlation parameter in a Gaussian autocorrelation function calculation of the application; Figure 3 A scene diagram of a neighborhood division condition in a maximum a posteriori probability Markov stratum optimization method in the application; Figure 4 A scene diagram of a unit update sequence in a maximum a posteriori probability Markov stratum optimization method of a deep rock mass in the application; Figure 5 A scene diagram of stress and displacement boundary conditions in a deep rock mass in-situ stress field inversion process of the application; Figure 6 A structural diagram of a deep rock mass in-situ stress random analysis device of the application; Figure 7 A structural diagram of a processing device of the application. DETAILED DESCRIPTION

[0012] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only some of the embodiments of the application, but not all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative work are within the protection scope of the application.

[0013] The terms "first", "second", and the like in the specification and claims of the application and the above-described drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or device that includes a series of steps or modules does not necessarily have to be limited to those steps or modules clearly listed, but can include other steps or modules that are not clearly listed or inherent to the process, method, product, or device. The naming or numbering of the steps appearing in the application does not mean that the steps in the method flow must be executed in the order / time sequence indicated by the naming or numbering, and the named or numbered flow steps can be executed in a different order according to the technical purpose to be achieved, as long as the same or similar technical effects can be achieved.

[0014] The division of the modules appearing in the present application is a logical division. In actual application, another division manner can be adopted, for example, a plurality of modules can be combined or integrated in another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the modules shown or discussed can be through an interface, and the indirect coupling or communication connection between the modules can be electrical or other similar forms, which are not limited in the present application. In addition, the modules or sub-modules described as separate components can or can not be physically separated, can or can not be physical modules, or can be distributed in a plurality of circuit modules. According to actual needs, part or all of the modules can be selected to achieve the purpose of the present application.

[0015] Before introducing the deep rock mass ground stress random analysis method provided in the present application, the background content involved in the present application is first introduced.

[0016] The deep rock mass ground stress random analysis method, device and computer readable storage medium provided in the present application can be applied to a processing device, and is used for the random characteristics of the lithology distribution of a deep complex geological body. The present application starts from the spatial correlation of the lithology distribution of an engineering area, configures a random geological model and a maximum posterior probability Markov iteration, constructs a three-dimensional geological random field of a deep rock mass engineering, and analyzes and inverts the ground stress field distribution characteristics of each key aspect in the ground stress field, so as to effectively solve the problem of poor precision of the deep rock mass ground stress analysis result obtained by the prior art, and provide an effective reference basis for the design and construction process of a rock mass engineering.

[0017] The deep rock mass ground stress random analysis method mentioned in the present application can be executed by a deep rock mass ground stress random analysis device, or a server, a physical host or a user equipment (UE) integrated with the deep rock mass ground stress random analysis device, or different types of processing devices. The deep rock mass ground stress random analysis device can be realized by hardware or software. The UE can be a terminal device such as a smart phone, a tablet computer, a notebook computer, a desktop computer or a personal digital assistant (PDA). The processing device can be set by a device cluster.

[0018] In specific application, considering that the present application is mainly based on existing data for data processing, the specific device type and device deployment form of the processing device executing the deep rock mass ground stress random analysis method of the present application or carrying the corresponding application service of the deep rock mass ground stress random analysis method of the present application are flexible, and can be adjusted flexibly according to actual conditions.

[0019] For example, the processing device can be a background device, and provide remote reference data support for the engineering site through a corresponding web service.

[0020] For example, the processing device can be a background device, and provide remote reference data support for the engineering site through a corresponding web service.

[0021] In addition, if the application further involves application requirements such as content display, the specific device type and device deployment form of the processing device can also be adjusted adaptively.

[0022] For example, the processing device can include a first device for performing main data processing in the laboratory background and a second device for entering data and outputting data processing results on site, and the first device and the second device can communicate and interact based on a web service type communication connection.

[0023] Next, the deep rock mass ground stress analysis method provided by the application will be introduced.

[0024] First, refer to Figure 1 , Figure 1 A flowchart of the deep rock mass ground stress random analysis method is shown. The deep rock mass ground stress random analysis method provided by the application can include the following steps S101 to S104: Step S101, for a target geotechnical engineering, on the basis of engineering regional geological background, engineering regional tectonic conditions, engineering regional geological drilling information and engineering regional borehole wall exposed lithology information, configure corresponding deep rock mass geological information; It can be understood that, corresponding to the deep rock mass ground stress analysis target faced by the application, the application can start from the configuration and processing of deep rock mass geological information in the early stage.

[0025] Correspondingly, for the current analysis target, i.e., the target geotechnical engineering, in the process of configuring its deep rock mass geological information, the specific configuration and processing can be based on information of four aspects: engineering regional geological background, engineering regional tectonic conditions, engineering regional geological drilling information and engineering regional borehole wall exposed lithology information.

[0026] Among them, the engineering regional geological background and the engineering regional tectonic conditions can reflect the main trend of the stratum distribution in the engineering region, and the engineering regional geological drilling information and the engineering regional borehole wall exposed lithology information can determine the types and distribution positions of the involved lithology, and obtain the rock mechanics parameters of the lithology in the engineering region according to the rock cores obtained by drilling.

[0027] As an exemplary embodiment here, the specific imported or configured deep rock mass geological information can specifically include engineering geological conditions, geological drilling hole mouth positions, geological drilling hole axis directions, drilling core lithology distribution information, engineering area rock types and engineering area mechanical parameters.

[0028] It can be understood that the embodiment here aims to give a specific implementation scheme for the deep rock mass geological information concerned in the early stage of the scheme of the application from a very detailed and very practical operation level.

[0029] Step S102, on the basis of the deep rock mass geological information, a three-dimensional cubic grid model including non-drilling units and drilling area corresponding drilling units is constructed; It can be understood that the application is aimed at the deep rock mass stress analysis target, and the corresponding scheme is set up on the basis of the configured random field theory, and specifically can involve the three-dimensional cubic grid model composed of a large number of units here, wherein in the model construction process, the units planned in the model can be specifically divided into two categories of non-drilling units and drilling units, which lay a good foundation for subsequent configuration processing of the deep engineering three-dimensional geological random field as a random geological model.

[0030] It can be easily seen that the construction processing of the three-dimensional cubic grid model here is specifically based on the previously configured deep rock mass geological information, which contains the effective information required for constructing different units.

[0031] Further, as an exemplary embodiment here, on the basis of the deep rock mass geological information, the three-dimensional cubic grid model including non-drilling units and drilling area corresponding drilling units is constructed, which can specifically include the following processing contents: (1) Research area planning link On the basis of the deep rock mass geological information, the research range containing each exploration hole, in-situ stress test point and inversion target area is constructed in combination with the spatial positional relationship between the engineering actual exploration hole and the inversion target, and a three-dimensional cubic grid model is preliminarily constructed, wherein the initial three-dimensional cubic grid model contains different units; (2) Non-drilling / drilling unit generation link On the basis of the initial three-dimensional cubic grid model, by analyzing the positions of the exploration hole mouths, the inclination angles of the exploration hole mouths and the lithology distribution revealed by the drilling hole walls, the correlation between each unit and the hole axis is judged, and the corresponding relationship between each lithology, drilling number and burial depth is established, so as to configure drilling units and non-drilling units in the different units contained in the model; (3) Gaussian autocorrelation function calculation link The Gaussian autocorrelation function value between the drilling units and the non-drilling units is calculated, and the calculation formula is as follows: , wherein, ρ(i,j) is the Gaussian autocorrelation function value between the unit i and the unit j, i=1,2,...n e , j=1,2,...n e , n e is the total number of all units in the calculation range, ρ(i,j) is used to measure the spatial correlation between the unit i and the unit j, d m , d n and d p are the distance between the unit i and the unit j, the projection component length of the rock formation strike, the rock formation layer normal and the rock formation layer inclination, I m , I n and I p are the influence range of the corresponding direction under different lithology conditions. Specifically, for the processing herein, it can also be more intuitively understood in combination with the scene diagram of the spatial correlation parameter in the Gaussian autocorrelation function calculation of the application shown in Figure 2 , wherein, 1 is the rock formation layer, 2 is the rock formation inclination, 3 is the rock formation strike, 4 is the drilling unit, 5 is the rock formation strike axis direction, 6 is the rock formation normal axis direction, 7 is the rock formation inclination axis direction, 8 is the non-drilling unit, 9 is the distance d m (the projection component length) between the units in the rock formation strike, 10 is the distance d n (the projection component length) between the units in the rock formation normal, and 11 is the distance d p (the projection component length) between the units in the rock formation inclination.

[0032] Next, based on the basic idea of the random field theory configured by the application, the current calculation of the lithology type / class of the non-drilling unit depends on the lithology type / class of the drilling unit which has a higher spatial correlation with the current calculation of the non-drilling unit.

[0033] (4) Lithology type determination link Based on the Gaussian autocorrelation function value between the drilling units and the non-drilling units, the lithology type in which the current calculation of the non-drilling unit is located is determined, and the calculation formula is as follows: , wherein, P k (i) is the probability that the current calculation of the non-drilling unit i is the lithology type k, ρ k(i, q) is a current calculation of the non-drilling unit i and the lithology type k unit q Gaussian type autocorrelation function value, X(k, q) is an indicator function, which is 1 when and only when k = q, and 0 in other cases, n h is the number of drilling units, n q is the number of lithology types in the calculation area. It can be understood that by P k (i) reveals the probability of the current calculation of the non-drilling unit for different lithology types, and then the final lithology type can be assigned by traversing all non-drilling units.

[0034] (5) According to a generated random number s1 on the interval (0, 1] that meets the uniform distribution, the probability interval of the lithology type of the current calculation of the non-drilling unit is determined, and the configuration of the random initial model is completed, wherein the calculation formula involved is: When a = 1, , Thus, the current calculation of the non-drilling unit is assigned to the lithology type k, and the random initial model of the study area is obtained.

[0035] It can be understood that in the embodiment, the construction process of the random initial model of the study area is based on very specific parameter characteristics, and the same very specific quantitative formula is used to provide a very application valuable implementation scheme.

[0036] At the same time, for the random initial model constructed in the foregoing, that is, a three-dimensional cubic grid model of a stage, other configuration work can also be involved to further improve the application value of the model, so that the subsequent processing based on the model can be better or better quality.

[0037] Correspondingly, as an exemplary embodiment, corresponding to the model unit size design link that can continue to be added, the method of the present application can also include: Calculate the index value and determine the unit size in the subsequent simulation process when the index value is stable, wherein the calculation formula involved is: , , Wherein, LUI iLUI is a quantitative index for quantitatively representing the lithology distribution of the geological model, and the LUI of a specific number of non-drilling units is the cumulative value of the individual LUI values of the specific number of non-drilling units, and n is the number of units of the geological model, and the LUI is achieved by statistically calculating the average expected value of the non-drilling units for each lithology label, so as to reflect the differences in the geological model under different unit sizes, and as the unit division is gradually refined and the lithology distribution is gradually stable, the proportion of the units of each lithology category to the non-drilling units gradually reaches a stable state, and the LUI gradually tends to be stable as the unit size decreases.

[0038] It can be seen that, through the arrangement of the embodiment, the model constructed by the application determines the appropriate unit size for the subsequent simulation process according to the unit size when the LUI is stable, which helps to further improve the processing efficiency and processing accuracy in detail.

[0039] In step S103, the three-dimensional cubic grid model is subjected to a maximum a posteriori Markov iteration controlled by the number of iterations of the lowest posterior energy, and the iteration is completed when the variation coefficient value of the average information entropy of the calculated non-drilling units meets the convergence requirement, and a random geological model is obtained. It can be understood that, on the basis of the three-dimensional cubic grid model configured in the foregoing, the maximum a posteriori Markov iteration processing configured by the application can be continued to iteratively optimize the model, and finally obtain a random geological model that can be put into the subsequent geostress field inversion.

[0040] It can be understood that the maximum a posteriori Markov iteration processing is developed on the basis of the existing concept, and the application also gives a further supporting scheme adapted to the specific application scenario in which the application is used.

[0041] Specifically, as an exemplary embodiment, the maximum a posteriori Markov iteration has the following three aspects of processing content: (1) In the neighborhood system specification, when the lithology category of the currently calculated non-drilling unit i in all units S (in the calculation area) is determined, it is considered that the lithology category is affected by the lithology distribution of the surrounding units, and the set composed of the surrounding units is regarded as the neighborhood N of the currently calculated non-drilling unit i. i | ∀i∈S}; Specifically, it can also be combined with Figure 3 The three-dimensional space maximum a posteriori Markov stratum optimization method of the application is shown in a scene schematic diagram of neighborhood division, in which 12 is the neighborhood radius, 13 is the neighborhood unit, and 14 is the non-neighborhood unit.

[0042] (2) In the aspect of class division, according to the distance between the stratigraphic unit and the adjacent borehole, the units at different distances from the borehole are divided into different classes, so that the units contained in different classes are calculated in sequence according to different classes, and the sequence is used as the sequence of stratigraphic updating; More specifically, first, the first class closest to the borehole is calculated, and then the classes farther from the borehole are gradually transitioned to. In the calculation process of each class, all units in the class are randomly traversed, and the neighborhood discrimination and posterior probability calculation process are sequentially performed, until the calculation of all units in the class is completed, and the next class is calculated, until the calculation of all classes and non-borehole units is completed.

[0043] Specifically, the method can also be combined with Figure 4 The application can be more intuitively understood by referring to a scene diagram of the unit updating sequence in the deep rock mass maximum posterior probability Markov stratigraphic optimization method shown in the application, wherein 4 is a borehole unit, 15 is a first-class unit, 16 is a second-class unit, and 17 is a third-class unit.

[0044] (3) In the aspect of maximum posterior probability calculation, when the current calculation non-borehole unit i is traversed, the posterior probability formula of the lithology type k of the current calculation non-borehole unit i under the current neighborhood condition is calculated, wherein the calculation formula is: , wherein Z is a partition function, Z is the lithology type k of the current calculation non-borehole unit i, and Z is the cumulative sum of the conditional probability from 1 to n f , so that the cumulative sum of the existence probability of different lithologies of the current calculation non-borehole unit i is 1, r i (f i ) is the influence of the original lithology type of the current calculation non-borehole unit i on the current calculation non-borehole unit i, r i,i′ (f i ,f i′ ) is the influence of the lithology type of the neighborhood unit on the current calculation non-borehole unit i, wherein the calculation formula is: , , wherein T is a temperature, T is used to indicate the clarity of the distribution, and the larger T is, the more uniform the distribution is, and the distribution tends to be uniform, and vice versa. V1(f i ) is a likelihood function of the initial lithology type of the current calculation non-borehole unit i and the current calculation lithology type, V2(f i ,f i′ ) is a likelihood function of the lithology type of the neighborhood unit and the current calculation non-borehole unit i, both of which are constants, and the specific value depends on whether there is a difference between the lithology types, wherein the calculation formula is: , , wherein, p(i, i') is the value of the Gaussian-type autocorrelation function between the current computing non-drilling unit i and the neighborhood unit i', and the specific quantification formula corresponds to the Gaussian-type autocorrelation function calculation link mentioned above, According to the aforementioned calculation, the posterior probability of the current unit under different lithology types is obtained, then a random number s2 conforming to the uniform distribution on the interval (0, 1] is generated, and the probability interval of the lithology type in which the current computing non-drilling unit i is located is judged, wherein the calculation formula involved is: When a = 1, , Thus, the current unit is assigned with the lithology type k, and the lithology distribution of all units of this type is finally obtained, and after assigning the corresponding lithology type, the calculation of the next type is performed.

[0045] At the same time, as an exemplary embodiment, corresponding to the maximum posterior probability Markov iteration, there are the following processing contents: (1) In terms of posterior energy calculation, the energy function value after the current round of iteration is calculated, wherein the calculation formula involved is as follows: , Based on the minimum energy criterion, the energy function value of each unit in the study area after the current round of iteration is calculated, multiple iterations are completed until the energy function value gradually converges and is lower than the expected threshold, the iteration is terminated, and the establishment of the maximum posterior Markov random field is completed; (2) In terms of the calculation of the coefficient of variation of average information entropy, in order to determine the number of implementations to be established, the present application specifically adopts the coefficient of variation of the average information entropy of the non-drilling units in the study area for analysis, for this, the average information entropy of the non-drilling units (in the study area) is calculated, wherein the calculation formula involved is: , , wherein, E E (i) is the information entropy of the current computing non-drilling unit i, i = 1, 2, 3,... n-n h , E E is the average information entropy of the current constructed random model.

[0046] On the basis of the above processing content, the average value and the standard deviation of the average information entropy of the random geological model can be statistically obtained, and the coefficient of variation of the corresponding average information entropy is calculated. If the convergence condition is not met, multiple iterations are required until the value of the coefficient of variation meets the convergence requirement, thereby obtaining a numerical model containing various random geological results, and applying it to the subsequent inversion of the in-situ stress field and statistical analysis of the random characteristics of the in-situ stress field.

[0047] It can be understood that the above two embodiments, similar to the previous ones, are based on very specific parameter characteristics and combined with very specific quantitative formulas to provide very application-valuable implementation schemes.

[0048] Step S104, in-situ stress field inversion is performed through the random geological model, and on the basis of the in-situ stress field inversion result, the distribution characteristics of the in-situ stress field at the measurement point level, the measurement section level and the regional level are analyzed.

[0049] After the iterative optimization is completed and the convergent random geological model is obtained, the corresponding in-situ stress inversion processing can be carried out based on the model, and on the basis of the in-situ stress inversion result, the distribution characteristics of the in-situ stress field at the measurement point level, the measurement section level and the regional level are analyzed.

[0050] Specifically, in terms of details, to provide more detailed and practical implementation schemes, as an exemplary embodiment here, there are: (1) In the inversion boundary condition, the stress environment of the deep engineering rock mass is fully considered. Specifically, the in-situ stress field inversion stress condition can include X-direction compression, Y-direction compression, Z-direction compression, XY plane shear, XZ plane shear, YZ plane shear and self-weight action. In addition, the stress and displacement boundary conditions can be combined Figure 5 The stress and displacement boundary conditions in the deep rock mass in-situ stress field inversion process of the present application can be more intuitively understood by the scene schematic diagram shown in the figure, in which Figure 5 In the figure, 18 is the X-direction compression condition, 19 is the Y-direction compression condition, 20 is the Z-direction compression condition, 21 is the rock mass self-weight condition, 22 is the XY plane shear condition, 23 is the XZ plane shear condition, 24 is the YZ plane shear condition, 25 is the stress boundary condition, and 26 is the displacement boundary condition.

[0051] (2) The in-situ stress inversion method can specifically use multivariate linear regression method, neural network method, genetic algorithm or machine learning algorithm, etc. (3) Corresponding to the stress field distribution characteristics of the measuring point level, the average values of the stress component values and the principal stress orientation of the target position are statistically inverted, and the variances of the stress component values and the principal stress orientation of the target position are statistically inverted to consider the influence of geological randomness on the inversion results of the in-situ stress field; (4) Corresponding to the stress field distribution characteristics of the measuring section level, for the maximum horizontal stress, the minimum horizontal stress and the vertical stress at each position of the measuring section, the average values and variances are respectively statistically inverted, and the orientation distribution of the maximum horizontal stress and the minimum horizontal stress at each position of the measuring section is statistically inverted to judge the random characteristics of the in-situ stress of the research area; (5) Corresponding to the stress field distribution characteristics of the regional level, the characteristics of each stress component in the inversion results of the in-situ stress field are respectively statistically inverted, the average values and variance distributions of the principal stress values, the principal stress orientation angle and the principal stress inclination angle on each section of the model are respectively calculated, and the regions with high uncertainty of the principal stress value and the principal stress orientation and the regions with significant concentration of the principal stress (two aspects of the regions) are determined to further provide effective reference for the design and construction process of the project.

[0052] After the deep rock mass in-situ stress analysis and processing are completed, further related data application processing can be involved.

[0053] For example, the in-situ stress field distribution characteristics of the measuring point level, the measuring section level and the regional level can be locally stored, stored in other places, results can be displayed, forwarded, outputted to complete the analysis prompt or further data analysis, etc. to meet the diversified application goals under actual conditions, which is easy to understand. The data processing involved can be flexibly configured according to the pre-configuration and real-time configuration of the data application strategy / rules.

[0054] Finally, for the above scheme content, in general, under the deep rock mass in-situ stress analysis target, aiming at the random characteristics of the lithology distribution of the deep complex geological body, the present application starts from the spatial correlation of the lithology distribution of the engineering area, configures the random geological model and the maximum posterior probability Markov iteration to construct the three-dimensional geological random field of the deep rock mass engineering and analyze the in-situ stress field distribution characteristics of each key aspect in the inversion of the in-situ stress field, so as to effectively solve the problem of poor precision of the deep rock mass in-situ stress analysis results obtained by the prior art and provide effective reference for the design and construction process of the rock mass engineering.

[0055] The above is the introduction of the deep rock mass in-situ stress random analysis method provided by the present application. In order to better implement the deep rock mass in-situ stress random analysis method provided by the present application, the present application also provides a deep rock mass in-situ stress random analysis device from the functional module angle.

[0056] Reference Figure 6 , Figure 6It is a structural schematic diagram of the deep rock mass geostress random analysis device. In the present application, the deep rock mass geostress random analysis device 600 can specifically include the following structures: The configuration unit 601 is configured to, for a target rock-soil engineering, on the basis of an engineering area geological background, an engineering area tectonic condition, engineering area geological drilling information and engineering area borehole wall exposed lithology information, configure corresponding deep rock mass geological information; The construction unit 602 is configured to, on the basis of the deep rock mass geological information, construct a three-dimensional cubic grid model including non-drilling units and drilling area corresponding drilling units; The iteration unit 603 is configured to perform a maximum a posteriori Markov iteration of an iteration number controlled by a minimum posterior energy on the three-dimensional cubic grid model, and complete the iteration when a variation coefficient value of an average information entropy of the non-drilling units calculated satisfies a convergence requirement, to obtain a random geological model; The analysis unit 604 is configured to perform geostress field inversion through the random geological model, and analyze geostress field distribution characteristics of a measurement point layer, a measurement section layer and a region layer on the basis of a geostress field inversion result.

[0057] In an exemplary embodiment, the deep rock mass geological information specifically includes an engineering geological condition, a geological drilling hole position, a geological drilling hole axis direction, drilling core lithology distribution information, an engineering area rock type and an engineering area mechanical parameter.

[0058] In another exemplary embodiment, the construction unit 602 is specifically configured to: On the basis of the deep rock mass geological information, a three-dimensional cubic grid model containing a research range of each exploration hole, a geostress test point and an inversion target region is preliminarily constructed in combination with a spatial position relationship between the exploration hole and the inversion target, wherein the initial three-dimensional cubic grid model contains different units; On the basis of the initial three-dimensional cubic grid model, a correlation between each unit and a hole axis is judged by analyzing a hole position of each exploration hole, a hole axis inclination angle of each exploration hole and a lithology distribution exposed by a borehole wall, and a corresponding relationship among each lithology, a drilling number and a buried depth is established, so as to configure drilling units and non-drilling units in different units contained in the model; A Gaussian autocorrelation function value between the drilling units and the non-drilling units is calculated, wherein a calculation formula involved is as follows: , Wherein, p(i,j) is a Gaussian autocorrelation function value between unit i and unit j, i=1,2,...n e , j=1,2,...n e , n eρ(i, j) is used to measure the strata space correlation between unit i and unit j, d m n p d m n p are the influence ranges of different lithology conditions in corresponding directions; Based on the Gaussian autocorrelation function value between the borehole unit and the non-borehole unit, the lithology category in which the current calculation non-borehole unit is located is determined, and the calculation formula involved is: P k (i) is the probability that the current calculation non-borehole unit i is of lithology category k, ρ k (i, q) is the Gaussian autocorrelation function value of the current calculation non-borehole unit i and the unit q of lithology category k, X(k, q) is an indicator function, which is 1 when and only when k=q, and 0 otherwise, n h is the number of borehole units, and n q is the number of lithology categories in the calculation area; According to a random number s1 generated on the interval (0, 1] that conforms to a uniform distribution, the probability interval of the lithology category in which the current calculation non-borehole unit is located is determined, and the configuration of the random initial model is completed, and the calculation formula involved is: When a=1, .

[0059] In another exemplary embodiment, the unit construction 602 is further configured to: Calculate the index value and determine the unit size in the subsequent simulation process when the index value is stable, and the calculation formula involved is: LUI i is the individual LUI value of the current calculation non-borehole unit i, Y(k) is the weight value when the current lithology category is k, LUI is the cumulative value of the individual LUI values of a specific number of non-borehole units, n is the number of geological model units, and LUI is achieved by statistically calculating the average expected value of the non-borehole units for each lithology label to reflect the differences of the geological model under different unit sizes. As the unit division gradually becomes fine and the lithology distribution gradually becomes stable, the proportion of units of each lithology category to non-borehole units gradually reaches a stable state, and LUI gradually tends to be stable as the unit size decreases.​​​​​​​

[0060] In another exemplary embodiment, corresponding to the maximum a posteriori probability Markov iteration, there are: In the neighborhood system definition aspect, when determining the lithology type of the current non-drilling unit i in all units S, it is considered that the lithology type is affected by the lithology distribution of the surrounding units, and the set of surrounding units is taken as the neighborhood N={N i | ∀i∈S} of the current non-drilling unit i. In the class division aspect, according to the distance between the stratigraphic unit and the adjacent drilling, the units at different distances from the drilling are divided into different classes, so that the units in different classes are calculated in turn according to different classes, and the order of stratigraphic update is taken as the order of stratigraphic update. In the maximum a posteriori probability calculation aspect, when the current non-drilling unit i is traversed, the posterior probability formula of the lithology type k of the current non-drilling unit i under the current neighborhood condition is calculated, wherein the calculation formula is: , wherein Z is the partition function, Z is the lithology type of the current non-drilling unit i, and Z is the lithology type of the current non-drilling unit i, respectively, and the cumulative sum of the conditional probability from 1 to n f is 1, r i (f i ) is the influence of the original lithology type of the current non-drilling unit i on the current non-drilling unit i, r i,i′ (f i ,f i′ ) is the influence of the lithology type of the units in the neighborhood on the current non-drilling unit i, wherein the calculation formula is: , , wherein T is the temperature, V1(f i ) is the likelihood function of the initial lithology type of the current non-drilling unit i and the current calculated lithology type, and V2(f i ,f i′ ) is the likelihood function of the lithology type of the units in the neighborhood and the current non-drilling unit i, both of which are constants, wherein the calculation formula is: , , wherein p(i,i') is the Gaussian-type autocorrelation function value between the current non-drilling unit i and the neighborhood unit i', and A random number s2 conforming to a uniform distribution on the interval (0, 1] is generated, and a probability interval of the lithology type of the current non-drilling unit i is determined, wherein the calculation formula is: When a = 1, .

[0061] In another exemplary embodiment, corresponding to the maximum a posteriori probability Markov iteration, there are: In terms of posterior energy calculation, the energy function value after the current round of iteration is calculated, and the calculation formula is as follows: , Based on the minimum energy criterion, the energy function value of each unit in the study area after the current round of iteration is calculated, and multiple iterations are completed until the energy function value gradually converges and is lower than the expected threshold, the iteration is terminated, and the establishment of the maximum a posteriori Markov random field is completed; In terms of the coefficient of variation of average information entropy, the average information entropy of the non-drilling unit is calculated, and the calculation formula is: , , Wherein, E E (i) is the information entropy of the current non-drilling unit i, i = 1, 2, 3,... n-n h , E E is the average information entropy of the current constructed random model.

[0062] In another exemplary embodiment, the corresponding geostress field inversion stress condition includes X-direction compression, Y-direction compression, Z-direction compression, XY-plane shear, XZ-plane shear, YZ-plane shear and gravity action; The geostress inversion method adopts a multiple linear regression method, a neural network method, a genetic algorithm or a machine learning algorithm; Corresponding to the geostress field distribution characteristics of the measuring point level, the average value of each stress component value and the principal stress direction of the inversion target position is calculated, and the variance of each stress component value and the principal stress direction of the inversion target position is calculated. Corresponding to the geostress field distribution characteristics of the measuring section level, for the maximum horizontal stress, the minimum horizontal stress and the vertical stress at each position of the measuring section, the average value and the variance are calculated, and the orientation distribution of the maximum horizontal stress and the minimum horizontal stress at each position of the measuring section is calculated. Based on the regional distribution characteristics of the geostress field, characteristic statistics were performed on each stress component in the geostress field inversion results. The average value and variance distribution of the principal stress force, principal stress azimuth angle and principal stress pitch angle on each cross section of the model were calculated to determine the regions where the height of the principal stress force and principal stress azimuth is uncertain and the regions where the principal stress is significantly concentrated.

[0063] This application also provides a processing device from a hardware architecture perspective, see [link / reference]. Figure 7 , Figure 7 This diagram illustrates a structural schematic of the processing device of this application. Specifically, the processing device may include a processor 701, a memory 702, and an input / output device 703. The processor 701 executes the computer program stored in the memory 702 to implement, for example... Figure 1 The corresponding steps of the stochastic analysis method for deep rock mass in the embodiment; or, when the processor 701 executes the computer program stored in the memory 702, it implements as follows: Figure 6 Corresponding to the functions of each unit in the embodiment, the memory 702 is used to store the functions executed by the processor 701 as described above. Figure 1 The computer program required for the random analysis method of deep rock mass in the corresponding embodiment.

[0064] For example, a computer program may be divided into one or more modules / units, one or more of which are stored in memory 702 and executed by processor 701 to complete this application. One or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in a computer device.

[0065] The processing device may include, but is not limited to, processor 701, memory 702, and input / output device 703. Those skilled in the art will understand that the illustrations are merely examples of the processing device and do not constitute a limitation on the processing device. It may include more or fewer components than illustrated, or combine certain components, or different components. For example, the processing device may also include network access devices, buses, etc., and processor 701, memory 702, input / output device 703, etc., are connected via a bus.

[0066] The processor 701 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 gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor and the like. The processor is the control center of the processing device, and connects various parts of the entire device through various interfaces and lines.

[0067] The memory 702 can be used to store computer programs and / or modules, and the processor 701 realizes various functions of the computer device by running or executing the computer programs and / or modules stored in the memory 702, and calling the data stored in the memory 702. The memory 702 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required by a function, etc.; and the data storage area can store data created according to the use of the processing device, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state memory devices.

[0068] When the processor 701 is used to execute the computer programs stored in the memory 702, the following functions can be realized: For the target geotechnical engineering, on the basis of the engineering area geological background, the engineering area tectonic condition, the engineering area geological drilling information and the engineering area hole wall exposed lithology information, the corresponding deep rock mass geological information is configured; On the basis of the deep rock mass geological information, a three-dimensional cubic grid model including non-drilling units and drilling area corresponding drilling units is constructed; The three-dimensional cubic grid model is subjected to a maximum a posteriori Markov iteration with the iteration number controlled by the lowest posterior energy, and the iteration is completed when the variation coefficient value of the average information entropy of the calculated non-drilling unit meets the convergence requirement, so as to obtain a random geological model; The geostress field inversion is carried out through the random geological model, and on the basis of the geostress field inversion result, the geostress field distribution characteristics of the measurement point layer, the measurement section layer and the regional layer are analyzed.

[0069] Those skilled in the art can clearly understand that, for the convenience and brevity of the description, the specific working process of the deep rock mass ground stress random analysis device, the processing device and the corresponding units thereof described above can be referred to as Figure 1 The description of the deep rock mass ground stress random analysis method in the corresponding embodiments will not be repeated here.

[0070] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructions or controlled by related hardware through instructions, which can be stored in a computer readable storage medium and loaded and executed by a processor.

[0071] Therefore, the present application provides a computer readable storage medium, which stores a plurality of instructions capable of being loaded by a processor to execute the present application as Figure 1 The steps of the deep rock mass ground stress random analysis method in the corresponding embodiments can be referred to as Figure 1 The description of the deep rock mass ground stress random analysis method in the corresponding embodiments will not be repeated here.

[0072] The computer readable storage medium can include a read only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0073] Due to the instructions stored in the computer readable storage medium, the present application as Figure 1 The steps of the deep rock mass ground stress random analysis method in the corresponding embodiments, therefore, the present application as Figure 1 The beneficial effects of the deep rock mass ground stress random analysis method in the corresponding embodiments can be achieved, as described above, and will not be repeated here.

[0074] The deep rock mass ground stress random analysis method, device, processing device and computer readable storage medium provided by the present application are described in detail above, and the principles and implementation modes of the present application are described by applying specific examples; the above embodiment is only used to help understand the core idea of the present application; at the same time, for those skilled in the art, according to the idea of the present application, the specific implementation mode and application range will be changed; in view of the above, the content of the specification should not be understood as a limitation of the present application.

Claims

1. A method for random analysis of deep rock mass ground stress, characterized in that, The method comprises: For a target geotechnical engineering, on the basis of engineering regional geological background, engineering regional tectonic conditions, engineering regional geological drilling information and engineering regional borehole wall exposed lithology information, corresponding deep rock mass geological information is configured; On the basis of the deep rock mass geological information, a three-dimensional cubic grid model including non-drilling units and drilling units corresponding to drilling units is constructed; The three-dimensional cubic grid model is subjected to maximum a posteriori Markov iteration with the number of iterations controlled by minimum posterior energy, and iteration is completed when the variation coefficient value of the average information entropy of the non-drilling units calculated meets the convergence requirement, to obtain a random geological model; Through the random geological model, a geostress field is inverted, and on the basis of the geostress field inversion result, the distribution characteristics of the geostress field at the level of a measuring point, a measuring section and a region are analyzed.

2. The method of claim 1, wherein, The deep rock mass geological information specifically comprises engineering geological conditions, geological drilling hole position, geological drilling hole axis direction, drilling core lithology distribution information, engineering regional rock types and engineering regional mechanical parameters.

3. The method of claim 1, wherein, On the basis of the deep rock mass geological information, a three-dimensional cubic grid model including non-drilling units and drilling units corresponding to drilling units is constructed, comprising: On the basis of the deep rock mass geological information, a three-dimensional cubic grid model is preliminarily constructed by combining the spatial position relationship between the engineering actual exploration holes and the inversion target, wherein the initial three-dimensional cubic grid model contains different units. On the basis of the initial three-dimensional cubic grid model, by analyzing the hole position of each exploration hole, the hole axis inclination angle of each exploration hole and the lithology distribution exposed by the borehole wall, the correlation between each unit and the hole axis is determined, and the corresponding relationship among each lithology, drilling number and depth is established, so as to configure the drilling units and the non-drilling units in the different units contained in the model. The Gaussian autocorrelation function value between the drilling units and the non-drilling units is calculated, wherein the calculation formula involved is as follows: ， Wherein, ρ(i,j) is the value of the Gaussian autocorrelation function between the unit i and the unit j, i=1, 2, …n e , j=1, 2, …n e , n e is the total number of all units in the calculation range, ρ(i,j) is used to measure the spatial correlation between the unit i and the unit j, d m , d n and d p are the distance between the unit i and the unit j, the projection component length of the rock formation trend, the rock formation layer normal and the rock formation layer inclination, I m , I n and I p are the influence range of the corresponding direction under different lithology conditions; Based on the Gaussian autocorrelation function value between the drilling units and the non-drilling units, the lithology category in which the current non-drilling unit is located is determined, wherein the calculation formula involved is as follows: ， where P k (i) is the current computation probability of non-drilling cell i being of lithology class k, p k (i,q) is the current computation value of the Gaussian-type autocorrelation function of non-drilling cell i and cell q of lithology class k, X(k,q) is an indicator function, which is 1 if and only if k=q, and 0 otherwise, n h is the number of drilling cells, n q is the number of lithology classes in the computation region; According to a random number s1 generated on the interval (0, 1] in accordance with a uniform distribution, the probability interval of the lithology category in which the current non-drilling unit is located is determined, and the configuration of the random initial model is completed, wherein the calculation formula involved is as follows: when a = 1, .

4. The method of claim 3, wherein, The method further comprises: The index value is calculated, and the unit size in the subsequent simulation process when the index value is stable is determined, wherein the calculation formula involved is as follows: ， ， wherein LUI i is the individual LUI value of the current non-drilling cell i, Y(k) is the weight value when the current lithology type is k, LUI is the cumulative value of the individual LUI values of a specific number of the non-drilling cells, n is the number of the geological model cells, LUI is achieved by statistically calculating the average expectation value of the non-drilling cells for each lithology label to reflect the difference of the geological model under different cell sizes, and as the cell division gradually becomes fine and the lithology distribution gradually becomes stable, the proportion of the cells of each lithology type in the non-drilling cells gradually reaches a stable state, and LUI gradually tends to be stable as the cell size decreases.

5. The method of claim 4, wherein, Corresponding to the maximum a posteriori Markov iteration, in terms of class division, units at different distances from the drilling hole are divided into different classes according to the distance from the drilling hole, and the units contained in the different classes are sequentially calculated according to the different classes, and the order of stratum update is determined accordingly; In the neighborhood system definition aspect, when all the unit S currently calculates the lithology type of the non-drilling unit i, it is considered that the lithology type is affected by the lithology distribution of the surrounding units, and the set composed of the surrounding units is taken as the neighborhood N of the currently calculated non-drilling unit i i | ∀i∈S} ​ In the maximum a posteriori probability calculation, when the current calculation non-drilling unit i is traversed, the posterior probability formula of the lithology type k of the current calculation non-drilling unit i under the current neighborhood condition is calculated, wherein the calculation formula involved is: ， wherein Z is a partition function, Z is the current computing non-drilling unit i lithology category, respectively, 1 to n f the cumulative conditions probability, so that the current computing non-drilling unit i different lithology existence probability calculation results are accumulated to 1, r i (f i ) is the current computing non-drilling unit i original lithology category influence on the current computing non-drilling unit i, r i,i′ (f i ,f i′ ) is the lithology category of the unit in the neighborhood influence on the current computing non-drilling unit i, wherein the formula involved is: ， ， where T is temperature, V1(f i ) is the likelihood function of the initial lithology class of the current computing non-drilling cell i and the current computing lithology class, V2(f i ,f i′ ) is the likelihood function of the neighborhood cell lithology class and the current computing non-drilling cell i, both of which are constants, where the calculation formula involved is: ， ， Wherein, ρ(i,i') is the value of the Gaussian autocorrelation function between the current calculation non-drilling unit i and the neighborhood unit i', A random number s2 conforming to the uniform distribution on the interval (0, 1] is generated, and the probability interval of the lithology type in which the current calculation non-drilling unit i is located is judged, wherein the calculation formula involved is: when a = 1, .

6. The method of claim 5, wherein, Corresponding to the maximum a posteriori probability Markov iteration, there are: In the posterior energy calculation, the energy function value after the current round of iteration is calculated, and the calculation formula involved is as follows: ， Based on the minimum energy criterion, the energy function values of each unit in the research area after the current round of iteration are calculated, multiple iterations are completed until the energy function values gradually converge and are lower than the expected threshold, the iteration is terminated, and the establishment of the maximum a posteriori Markov random field is completed; In the coefficient of variation of the average information entropy calculation, the average information entropy of the non-drilling unit is calculated, and the calculation formula involved is: ， ， where E E (i) is the information entropy of the current calculation of non-drilling unit i, i = 1, 2, 3,... n-n h , E E is the average information entropy of the current constructed random model.

7. The method of claim 1, wherein, The corresponding ground stress field inversion ground stress inversion stress conditions include X-direction compression, Y-direction compression, Z-direction compression, XY plane shear, XZ plane shear, YZ plane shear and gravity action; The ground stress inversion method adopts a multivariate linear regression method, a neural network method, a genetic algorithm or a machine learning algorithm; Corresponding to the ground stress field distribution characteristics of the measuring point level, the average values of each stress component value and the principal stress direction of the inversion target position are counted, and the variances of each stress component value and the principal stress direction of the inversion target position are counted; Corresponding to the ground stress field distribution characteristics of the measuring section level, for the maximum horizontal stress, the minimum horizontal stress and the vertical stress at each position of the measuring section, the average values and the variances are counted respectively, and the direction distribution of the maximum horizontal stress and the minimum horizontal stress at each position of the measuring section is counted; Corresponding to the ground stress field distribution characteristics of the region level, the feature statistics of each stress component in the ground stress field inversion result are performed, the average values and the variance distributions of the principal stress value, the principal stress direction angle and the principal stress inclination angle on each section of the model are calculated respectively, and the region with high uncertainty of the principal stress value and the principal stress direction and the region with significant concentration of the principal stress are determined.

8. A device for random analysis of deep rock mass ground stress, characterized in that, The device comprises: The configuration unit is configured to, for a target geotechnical engineering, on the basis of an engineering region geological background, an engineering region tectonic condition, engineering region geological drilling information and engineering region borehole wall exposed lithology information, configure corresponding deep rock mass geological information; The construction unit is configured to, on the basis of the deep rock mass geological information, construct a three-dimensional cubic grid model comprising non-drilling units and drilling region corresponding drilling units; An iteration unit is configured to perform maximum a posteriori Markov iteration on the three-dimensional cubic grid model for a number of iterations controlled by a minimum posterior energy, and complete the iteration when a variation coefficient value of the average information entropy of the non-drilling unit calculated satisfies a convergence requirement, to obtain a random geological model; An analysis unit is configured to perform geostress field inversion based on the random geological model, and analyze distribution characteristics of the geostress field at a measurement point level, a measurement section level and a region level based on the geostress field inversion result.

9. A processing device, characterized by The computer readable storage medium stores a plurality of instructions adapted to be loaded by the processor to execute the method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a plurality of instructions adapted to be loaded by the processor to execute the method of any one of claims 1 to 7.

Images