A rock fracture prediction method and system based on a resistivity data assimilation algorithm

Abstract

The application discloses a rock fracture prediction method and system based on a resistivity data assimilation algorithm, and comprises the following steps: establishing a numerical model for simulating the gradual fracture process of a rock mass under load action according to the boundary conditions and initial conditions of the rock mass; simulating the stress failure process of the rock mass by using a numerical calculation method to obtain prediction data of the stress field and the fracture damage field distribution of the rock mass; monitoring the failure process of the rock mass pattern in real time by using a high-density electrical method to obtain observation data; obtaining the fracture damage distribution of the rock mass based on the observation data; combining the information of the observation data and the prediction data by using an assimilation algorithm to correct the initial conditions or parameters of the numerical model; and further obtaining the corrected rock fracture prediction data. The application fuses different data and model simulation results in the dynamic framework of rock fracture by using the data assimilation algorithm, constantly adjusts the model trajectory automatically by relying on observation, obtains higher-precision and more consistent analysis results, and improves the prediction accuracy and predictability of the model.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of geotechnical engineering, and in particular to a rock fracture prediction method and system based on a resistivity data assimilation algorithm. BACKGROUND

[0002] Deep underground engineering is in a "three high and one disturbance" environment, and the disaster incubation process is the result of the interaction of the rock mass itself and environmental quantities. Due to the increase in mining depth and complexity of mining conditions, the stability of the rock mass is poorer, and the rock mass fracture-induced engineering disasters such as landslides, subsidence, collapse, water inrush, rock burst, roof fall, rib spalling, and the like are increasingly prominent, and have become a major bottleneck restricting the mining of mineral resources. Such rock mass engineering disasters are closely related to the evolution process of the internal cracks of the rock, and are the result of the progressive damage and fracture of the rock.

[0003] In the study of the process of underground engineering activities, obtaining rock fracture information plays an important role in effectively predicting the state and evolution of underground behavior. There are two methods to obtain rock fracture information. One is through numerical simulation means, based on different assumptions, setting different boundary conditions, discretizing the medium into mesoscopic units, introducing the non-uniformity of the medium in the model, and introducing a reasonable constitutive relationship and progressive failure criterion, thereby realizing the simulation of the whole process of material progressive fracture. Numerical calculation methods of rock mechanics are mainly divided into two categories. One is based on continuum mechanics, such as the finite element method and the finite difference method. The other is based on non-continuous medium mechanics, including the discrete element method and the discrete fracture network method. The second is to obtain deep rock fracture information through observation data collected by various sensors, such as acoustic emission / microseismic data and resistivity data during rock fracture. Rock acoustic emission / microseismic monitoring technology receives the acoustic waves generated by micro-fractures in the rock medium, understands the time, location and intensity of micro-fracture occurrence and the physical properties of the rock (wave velocity, damping, etc.), and can realize non-destructive, rapid and accurate real-time positioning of the seismic source location, dynamic monitoring and evaluation of the rock; electrical monitoring technology receives the resistivity changes during rock failure to understand the rock fracture damage area and characteristics.

[0004] The existing methods have the following advantages and disadvantages: the two methods have their own advantages and have their own application scenarios, for example, the numerical simulation method can flexibly design the medium type and the unit size, and can obtain the simulation of the progressive whole process of the material, and the inversion relying on the sensor observation data can truly obtain the rock damage and fracture information in the scene. However, the above two methods have errors for the rock damage and fracture process, the numerical simulation method assumes the initial boundary condition, but the control variable is different from the actual deep geological engineering, the reliability of the model and the parameter lacks the verification of the actual physical feedback information, which often leads to inconsistency with the actual rock damage process, and the rock damage and fracture obtained by relying on the sensor observation data inversion often has errors due to the system error of the sensor itself and the non-uniqueness problem of the inversion algorithm itself, and the rock damage and fracture mechanism also lacks rigorous mechanical support, which also leads to inconsistency with the actual rock damage and fracture process. SUMMARY

[0005] The technical problem to be solved by the present application is to provide a rock fracture prediction method and system based on a resistivity data assimilation algorithm to solve the defects in the prior art.

[0006] The technical scheme adopted by the present application to solve the technical problem is:

[0007] The present application provides a rock fracture prediction method based on a resistivity data assimilation algorithm, which comprises the following steps:

[0008] A numerical model simulating the progressive fracture process of the rock mass under load is established according to the boundary conditions and initial conditions of the rock mass; the stress failure process of the rock mass is simulated by using a numerical calculation method to obtain the prediction data of the stress field and the fracture damage field distribution of the rock mass;

[0009] The destruction process of the rock mass is monitored in real time by using a high-density electrical method to obtain observation data; and the rock fracture damage distribution is obtained based on the observation data;

[0010] An assimilation algorithm is used to combine the information of the observation data and the prediction data, and to correct the initial conditions or parameters of the numerical model; and then the corrected rock fracture prediction data is obtained.

[0011] Further, the method for obtaining prediction data in the present application specifically comprises:

[0012] A continuous / discontinuous method is used to establish a numerical model, the geometric model of the rock mass is divided into finite element grids to form solid elements, and joint elements with no thickness are inserted between adjacent solid elements to simulate the initiation and fracture of the crack and the discontinuous deformation behavior;

[0013] The joint unit is initially without thickness, the stress-displacement constitutive of the fracture plastic zone theory is used to express the mechanical behavior of the joint unit, and the internal force of the joint unit is obtained;

[0014] When the joint unit is damaged and fails, the solid element connected with the joint unit forms a potential contact pair, the potential contact pairs are identified through a contact search algorithm, contact detection is carried out, and finally, all the contact pairs detected are subjected to contact force calculation, that is, the prediction data of the stress field and the rupture damage field distribution of the rock mass are obtained.

[0015] Further, the continuous / discontinuous method of the present application specifically comprises:

[0016] The motion control equation of the continuous / discontinuous method is represented as follows:

[0017]

[0018] Wherein, and are the mass matrix and the damping matrix of the finite element grid; is the node displacement vector; , , and are the internal force of the solid element, the internal force of the joint element, the contact force and the combined external force respectively;

[0019] Based on the finite element calculation method of the continuous method and the elastoplastic constitutive theory, the internal force of the solid element is represented as:

[0020]

[0021] Wherein, is a solid, is the shape function gradient matrix of the solid element node, is the Cauchy stress tensor of the solid element.

[0022] Further, the method for obtaining the internal force of the joint element of the present application specifically comprises:

[0023] The stress of the joint element between two adjacent solid elements can be calculated by the following formula:

[0024]

[0025]

[0026] The internal force of the joint element is represented as:

[0027]

[0028] Wherein, and is the normal force and shear force of the joint element; and are the relative normal displacement and tangential displacement between two adjacent entity elements, respectively; and is the elastic limit of and ; and is the plastic fracture limit of the normal displacement and tangential displacement; and is the tensile strength and shear strength of the joint element, is the material softening coefficient.

[0029] Further, the method for calculating the contact force of the present application specifically comprises:

[0030] The potential-based contact force calculation method is represented as:

[0031]

[0032]

[0033] wherein, is the contact force between the contact pair, is the normal contact stiffness, is the contact potential gradient inside the contact block, is the contact potential gradient inside the target block.

[0034] Further, the method for obtaining the observation data of the present application specifically comprises:

[0035] The electrical property change of the rock mass is represented by the conductivity or resistivity, so as to reflect the internal fracture development and expansion change and quantify the internal damage of the rock;

[0036] The acoustic-electric sensor is arranged on the surface of the rock sample, and the rock is subjected to external force by using the loading device, so as to simulate the real underground stress condition; the resistivity of the rock mass is calculated by collecting the resistivity signal of the rock mass by the acoustic-electric sensor;

[0037] The observation potential difference and the potential difference obtained by the forward calculation are minimized by inversion calculation; the resistivity calculation result is updated by continuous iteration, the objective function value is gradually reduced, and the rock mass damage resistivity cloud picture is obtained as the observation data.

[0038] Further, the formula for calculating the resistivity of the present application is:

[0039]

[0040] wherein: is the apparent resistivity, is a geometry factor, is a measured potential difference, is an injected current;

[0041] The objective function used in the inversion calculation is expressed as:

[0042]

[0043] where, is the potential difference of the th measurement, is the potential difference calculated from the resistivity, is a regularization term for smoothing the resistivity calculation, is a regularization parameter controlling the weight of the regularization term.

[0044] Further, the assimilation algorithm method of the present invention specifically comprises:

[0045] a prediction stage, calculating predicted data from the current state variable or parameter estimate using the numerical model;

[0046] an analysis / update stage, comparing the predicted data with the measured data, calculating the difference between them, using a gain matrix to adjust the weight of the predicted data and the observed data, generating an updated state variable or parameter estimate;

[0047] an iteration process, repeatedly performing the steps of prediction and analysis / update stages until a convergence condition is reached or the number of iterations is reached.

[0048] Further, the assimilation algorithm of the present invention is expressed as:

[0049]

[0050] where, is the state variable or parameter variable of the th iteration, i.e. the optimal estimate value of the model state variable or parameter, is the updated state variable or parameter vector, i.e. the optimal estimate value of the th iteration, is the gain matrix of the th iteration, used to adjust the weight of the observed data and the model predicted data, is the measured observed data vector, obtained by the electrode sensor, is a model operator, mapping the state variable or parameter variable to the predicted value in the observation space, i.e. the observed data.

[0051] The application provides a rock fracture prediction system based on a resistivity data assimilation algorithm, comprising: a temperature and pressure control device, a loading device, an acoustic electrode sensor, an electrode switch, a high-density electrical method instrument and a data processing device; wherein:

[0052] The loading device is used for applying external force to the rock sample.

[0053] The temperature and pressure control device is connected with the loading device and is used for controlling the temperature and pressure of the loading device to simulate the real underground stress condition.

[0054] The acoustic electrode sensor is arranged on the surface of the rock sample and is used for collecting the conductivity or resistivity of the rock sample.

[0055] The electrode switch is connected with the acoustic electrode sensor at one end and is connected with the high-density electrical method instrument at the other end.

[0056] The high-density electrical method instrument is used for receiving the resistivity signal.

[0057] The data processing device is connected with the high-density electrical method instrument, wherein a computer program is arranged in the data processing device and is used for realizing the rock fracture prediction method based on the resistivity data assimilation algorithm.

[0058] The application has the following beneficial effects:

[0059] 1. The application integrates different sources and different resolution resistivity observation data and model simulation results in the dynamic framework of rock fracture through the data assimilation algorithm, so that the rock fracture process model and various observation operators are constantly adjusted according to the observation to automatically adjust the model trajectory, obtain more accurate and consistent analysis results, and improve the prediction accuracy and predictability of the model.

[0060] 2. The method of the application adjusts the boundary condition by comparing with the measured data, truly feeds back the rock failure process caused by damage and fracture, provides sufficient experimental basis for the rock damage and fracture mechanism, and can effectively realize the prediction of rock failure.

[0061] 3. In the establishment of the numerical model, the continuous / discontinuous method is adopted, the finite element grid is divided on the rock mass geometric model to form a solid element model, then a joint element with no thickness is inserted between adjacent solid elements, and the initiation and fracture of the crack and other non-continuous deformation behaviors can be effectively simulated. In addition, the joint element has no thickness at the beginning and the internal force cannot be directly analyzed by using the existing constitutive theory, therefore, the stress-displacement constitutive of the fracture plastic zone theory is used to express the mechanical behavior of the joint element.

[0062] 4. The present application combines the actual observation data (resistivity data) and the information of the driving numerical model (prediction value) by assimilation algorithm to correct the initial conditions or parameters of the numerical model, so that it better meets the actual observation data, thereby improving the accuracy and reliability of the model. BRIEF DESCRIPTION OF DRAWINGS

[0063] The present application will be further described below in conjunction with the accompanying drawings and embodiments, wherein:

[0064] Figure 1 is the assimilation process flowchart of the embodiment of the present application;

[0065] Figure 2 is the observation data acquisition and processing system of the embodiment of the present application;

[0066] Figure 3 is the potential-based contact force calculation principle diagram of the embodiment of the present application. DETAILED DESCRIPTION

[0067] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0068] Embodiment 1

[0069] As shown in Figure 1 , the rock fracture prediction method based on resistivity data assimilation algorithm of the embodiment of the present application, the method comprises the following steps:

[0070] Step 1, according to the boundary conditions and initial conditions of the rock mass, a numerical model simulating the gradual fracture process of the rock mass under the action of load is established; the stress failure process of the rock mass is simulated by using numerical calculation method, and the prediction data of the stress field and the fracture damage field distribution of the rock mass are obtained;

[0071] Step 2, the destruction process of the rock mass sample is monitored in real time by using high-density electrical method, and observation data is obtained; the rock mass fracture damage distribution is obtained based on the observation data;

[0072] Step 3, the assimilation algorithm is used to combine the information of the observation data and the prediction data, and the initial conditions or parameters of the numerical model are corrected; and then the corrected rock fracture prediction data is obtained.

[0073] In the preferred embodiment of the present application, the method for obtaining prediction data in step 1 specifically comprises:

[0074] The numerical model is established by using continuous / discontinuous method, the geometric model of the rock mass is divided into finite element grids to form solid elements, and the joint elements with no thickness are inserted between adjacent solid elements to simulate the non-continuous deformation behavior of crack initiation and fracture;

[0075] The joint element is initially without thickness, the stress-displacement constitutive of the fracture plastic zone theory is used to describe the mechanical behavior of the joint element, and the internal force of the joint element is obtained;

[0076] When the joint element fails, the solid elements connected with the joint element form potential contact pairs, the potential contact pairs are identified by using a contact search algorithm, and contact detection is performed, finally, the contact force of all detected contact pairs is calculated, that is, the prediction data of the stress field and the damage field distribution of the rock mass are obtained.

[0077] In the preferred embodiment of the present application, the method for obtaining observation data in step 2 specifically comprises:

[0078] The electrical conductivity or resistivity is used to represent the change of the electrical property of the rock mass, and then the internal crack development and expansion change are reflected, and the internal damage of the rock is quantified;

[0079] The acoustic-electric sensor is arranged on the surface of the rock sample, and the external force is applied to the rock by using the loading device to simulate the real underground stress condition; the resistivity is calculated by collecting the resistivity signal of the rock mass through the acoustic-electric sensor;

[0080] The observation potential difference and the potential difference obtained by the forward calculation are minimized by inversion calculation; the resistivity calculation result is updated by iteration, and the objective function value is gradually reduced, so as to obtain the rock mass damage resistivity cloud picture as the observation data.

[0081] In the preferred embodiment of the present application, the method of assimilation algorithm in step 3 specifically comprises:

[0082] In the prediction stage, the prediction data is calculated by using the numerical model according to the current state variable or parameter estimation;

[0083] In the analysis / update stage, the prediction data and the measured data are compared, the difference between them is calculated, the gain matrix is used to adjust the weight of the prediction data and the observation data, and the updated state variable or parameter estimation is generated;

[0084] In the iteration process, the steps of the prediction and the analysis / update stage are repeatedly executed until the convergence condition or the iteration number is reached.

[0085] Embodiment 2

[0086] The embodiment of the present application is based on the embodiment 1, and specific algorithms and processes of each step are given. The rock fracture process prediction method and system based on data assimilation mainly includes three parts of driving numerical model, observation data and assimilation algorithm.

[0087] 1. Driving numerical model

[0088] In the study of deep underground engineering activities, one of the main research methods is to simulate the gradual fracture process of deep rock under load by numerical model, and to study the rock damage and fracture mechanism. The fracture behavior in the rock simulation process will cause changes in the model boundary conditions and physical fields. By comparing and adjusting the boundary conditions with the measured data, the rock failure process caused by damage and fracture is truly fed back, which provides sufficient experimental basis for the rock damage and fracture mechanism, and realizes the prediction of rock failure.

[0089] The numerical calculation method of rock mechanics mainly includes two types: one is the continuous method based on the continuous medium mechanics, such as finite element method and finite difference method, and the other is the non-continuous method which simulates the non-continuous mechanical behavior of materials through discrete elements and discrete particles. The continuous / non-continuous method combines the advantages of traditional continuous medium mechanics and non-continuous medium mechanics, and can simulate the brittle material from micro crack initiation to macro fracture, which is widely used in many fields. In the establishment of numerical model, the continuous / non-continuous method mainly includes two steps: first, the geometric model is divided into finite element grid to form a solid element model, and then the joint element with no thickness is inserted between adjacent solid elements to simulate the initiation and fracture of cracks and other non-continuous deformation behaviors. Based on Newton's second law, the main motion control equation of the continuous / non-continuous method can be expressed as follows:

[0090]

[0091] wherein, and are the mass matrix and damping matrix of the finite element grid; is the node displacement vector; , , and are the internal force of the solid element, the internal force of the joint element, the contact force and the combined external force, respectively.

[0092] Based on the finite element calculation method of continuous method and the elastic-plastic constitutive theory, the internal force of the solid element can be expressed as:

[0093]

[0094] wherein is a solid, is the shape function gradient matrix of the solid element node, Cauchy stress tensor of solid element.

[0095] Unlike solid element, joint element has no thickness initially, and its internal force cannot be directly analyzed by existing constitutive theory, so the stress-displacement constitutive of fracture plastic zone theory is used to describe the mechanical behavior of joint element. The stress of joint element between two adjacent solid elements can be calculated by the following formula:

[0096]

[0097]

[0098] wherein, and are the normal force and shear force of the joint element; and are the relative normal displacement and tangential displacement between two adjacent solid elements; and are the elastic limit of and ; and are the plastic fracture limit of normal displacement and tangential displacement. and are the tensile strength and shear strength of the joint element, is the material softening coefficient. Thus, the internal force of the joint element can be expressed as:

[0099]

[0100] When the joint element fails, the solid elements connected to it form potential contact pairs. Through the contact search algorithm, these potential contact pairs can be effectively identified and contact detection can be performed. Finally, all detected contact pairs are calculated for contact force. The current contact force calculation method is mainly divided into two categories: one is the geometric-based contact force calculation, and the other is the potential-based contact force calculation. As shown in Figure 3 , wherein the potential-based contact force calculation is widely used in the contact force calculation of continuous / discontinuous method due to its simplicity and stability in the calculation process. Its main principle can be expressed as:

[0101]

[0102]

[0103] wherein, is the normal contact force between the contact pairs, is the normal contact stiffness, is the contact potential gradient inside the contact block, is the contact potential gradient inside the target block.

[0104] 2. Observation data acquisition

[0105] The electrical property of rock is one of the inherent physical properties of rock, which is closely related to the porosity, fracture propagation and deformation of rock under constant environmental factors. When rock is subjected to various factors such as load, environmental change and chemical reaction, it will cause changes in the initiation, propagation and penetration of internal defects, resulting in changes in its electrical properties and affecting related electrical parameters. Therefore, the electrical property of rock can be represented by the conductivity or resistivity, which can reflect the internal fracture development and expansion and quantify the internal damage of rock.

[0106] Specifically, an acoustic-electric sensor is needed to be arranged on the surface of the rock sample, and an external force is applied to the rock by using a loading device to simulate the real underground stress condition. The acoustic-electric sensor is connected to a switch through a high-temperature and high-pressure cable, and then connected to a resistivity acquisition instrument to receive the resistivity signal, including potential difference, apparent resistivity, data acquisition error, etc. Finally, the signal is processed and analyzed by using an algorithm to extract useful signals, such as Figure 2 as shown.

[0107] The electrical measurement is performed by measuring the potential difference and current between the electrodes, and calculating the resistivity of the rock. The specific formula for calculating the resistivity depends on the electrode arrangement method used, and the common arrangement methods include the Wenner method and the Schlumberger method.

[0108] Regardless of the arrangement method used, the general formula for calculating the resistivity is:

[0109]

[0110] where: is the apparent resistivity ( · ), is the geometric factor, is the measured potential difference (V), is the injected current (A).

[0111] By inversion calculation, the observed potential difference and the calculated potential difference obtained by forward calculation are minimized, that is, the error between the observed potential difference and the calculated potential difference is minimized. The objective function is usually expressed as:

[0112]

[0113] where, is the potential difference of the th measurement, is the potential difference calculated according to the resistivity model, is the regularization term, which is used to smooth the resistivity model, is a regularization parameter that controls the weight of the regularization term.

[0114] By continuously updating the resistivity model, the objective function value is gradually reduced, and thus the rock failure resistivity cloud map is obtained as the observed value.

[0115] 3. Data assimilation algorithm

[0116] The core idea of the assimilation algorithm is to combine the actual observation data (resistivity data) and the information of the driving numerical model (predicted value) through an optimization method, to correct the initial conditions or parameters of the numerical model, so that it better meets the actual observation data, thereby improving the accuracy and reliability of the model.

[0117] The assimilation algorithm can be expressed in the following basic form:

[0118]

[0119] where, is the state variable or parameter variable (optimal estimate of the model state variable or parameter) of the th iteration, is the updated state variable or parameter vector, i.e., the optimal estimate of the th iteration, is the gain matrix of the th iteration, used to adjust the weight of the observation data and the model predicted data, is the measured observation data vector obtained by the electrode sensor, is the model operator that maps the state variable or parameter variable to the predicted value in the observation space, i.e., the model predicted data.

[0120] The main steps are as follows:

[0121] (1) Prediction stage, according to the current state variable or parameter estimate , the predicted data is calculated through the numerical model

[0122] (2) Analysis / Update stage, compare the predicted data with the measured data , calculate the difference between them, use the gain matrix to adjust the weight of the predicted data and the observation data, and generate the updated state variable or parameter estimate .

[0123] (3) Iterative process, repeatedly perform the prediction and analysis steps until the convergence condition or the number of iterations is reached.

[0124] Example 3

[0125] The rock fracture prediction system based on the resistivity data assimilation algorithm comprises a temperature and pressure control device, a loading device, an acoustic electrode sensor, an electrode switch, a high-density electrical method instrument and a data processing device.

[0126] The loading device is used for applying external force to the rock sample.

[0127] The temperature and pressure control device is connected with the loading device and is used for controlling the temperature and pressure of the loading device to simulate the real underground stress condition.

[0128] The acoustic electrode sensor is arranged on the surface of the rock sample and is used for collecting the conductivity or resistivity of the rock sample.

[0129] The electrode switch is connected with the acoustic electrode sensor at one end and is connected with the high-density electrical method instrument at the other end.

[0130] The high-density electrical method instrument is used for receiving the resistivity signal.

[0131] The data processing device is connected with the high-density electrical method instrument and is provided with a computer program for realizing the rock fracture prediction method based on the resistivity data assimilation algorithm.

[0132] It should be understood that the size of the serial number of each step in the above embodiment does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0133] It should be understood that those skilled in the art can improve or change according to the above description, and all these improvements and changes should belong to the protection scope of the appended claims of the present application.

Claims

1. A rock failure prediction method based on resistivity data assimilation algorithm, characterized in that, The method comprises the following steps: According to the boundary conditions and initial conditions of the rock mass, a numerical model for simulating the gradual failure process of the rock mass under load is established; the stress field and the distribution of the failure and damage field of the rock mass are obtained by simulating the failure process of the rock mass under force by using a numerical calculation method; the method for obtaining the prediction data comprises the following steps: A continuous / discontinuous method is used to establish a numerical model, and a finite element mesh is divided on the geometric model of the rock mass to form solid elements, and a joint element with no thickness is inserted between adjacent solid elements to simulate the initiation and fracture of the crack and the non-continuous deformation behavior; The joint element has an initial thickness, and a stress-displacement constitutive model of the fracture plastic zone theory is used to describe the mechanical behavior of the joint element to obtain the internal force of the joint element; The method for obtaining the internal force of the joint element comprises the following steps: The stress of the joint element between two adjacent solid elements is calculated by the following formula: The internal force of the joint element is represented by the following formula: wherein, and are the normal force and shear force of the joint element; and are the relative normal displacement and tangential displacement between two adjacent solid elements, respectively; and are the elastic limit of and ; and are the plastic fracture limit of the normal displacement and tangential displacement; and are the tensile strength and shear strength of the joint element, is the material softening coefficient; The failure process of the rock sample is monitored in real time by using the high-density electrical method to obtain observation data; the failure and damage distribution of the rock mass is obtained based on the observation data; the method for obtaining the observation data comprises the following steps: The electrical conductivity or resistivity is used to represent the change of the electrical property of the rock mass, and then the change of the internal crack development and expansion is reflected, and the internal damage of the rock is quantified; An acoustic-electric sensor is arranged on the surface of the rock sample, and an external force is applied to the rock by using a loading device to simulate the actual underground stress condition; the resistivity of the rock mass is calculated by collecting the resistivity signal of the rock mass by the acoustic-electric sensor; The observation potential difference and the potential difference obtained by the forward calculation are minimized by inversion calculation; the resistivity calculation result is updated by continuous iteration to gradually reduce the objective function value, so as to obtain the rock failure resistivity cloud picture as the observation data; The objective function used in the inversion calculation is represented by the following formula: wherein is the potential difference of the first is the potential difference of the second is the potential difference calculated from the resistivity, is a regularization term for smoothing the resistivity calculation, is a regularization parameter controlling the weight of the regularization term; An assimilation algorithm is used to combine the information of the observation data and the prediction data to correct the initial conditions or parameters of the numerical model; and then the corrected rock failure prediction data is obtained.

2. The resistivity data assimilation algorithm based rock failure prediction method of claim 1, wherein, The method for obtaining the prediction data comprises the following steps: When the joint element fails, the solid elements connected to the joint element form potential contact pairs, the potential contact pairs are identified by a contact search algorithm, and contact detection is performed, and finally the contact force calculation of all the detected contact pairs is performed, that is, the prediction data of the stress field and the failure and damage field distribution of the rock mass are obtained.

3. The resistivity data assimilation algorithm based rock failure prediction method of claim 2, wherein, The continuous / discontinuous method comprises the following steps: The motion control equation of the continuous / discontinuous method is represented by the following formula: wherein, and are the mass matrix and the damping matrix of the finite element mesh; is the nodal displacement vector; , , and are the internal forces of solid elements, the internal forces of joint elements, the contact forces and the resultant external forces, respectively. Based on the finite element calculation method of the continuous method and the elastoplastic constitutive theory, the internal force of the solid element is represented by the following formula: wherein is a certain entity element, is a shape function gradient matrix of the entity element node, is the Cauchy stress tensor of the entity element.

4. The resistivity data assimilation algorithm based rock fracture prediction method of claim 1, wherein, The method for performing the contact force calculation comprises the following steps: A potential-based contact force calculation method is used, which is represented by the following formula: wherein, is the contact force normal to the pair of contacts, is the normal contact stiffness, is the contact potential gradient inside the contact patch, is the contact potential gradient inside the target patch.

5. The resistivity data assimilation algorithm based rock fracture prediction method of claim 4, wherein, The formula for calculating the resistivity is represented by the following formula: wherein: is the resistivity, is the geometric factor, is the measured potential difference, is the injected current.

6. The resistivity data assimilation algorithm-based rock fracture prediction method of claim 1, wherein, The assimilation algorithm comprises the following steps: In the prediction stage, the prediction data is calculated by the numerical model according to the current state variable or parameter estimation; In the analysis / update stage, the difference between the prediction data and the measured data is calculated by comparing the prediction data with the measured data, and a gain matrix is used to adjust the weights of the prediction data and the observation data to generate updated state variables or parameter estimations; In the iteration process, the steps of the prediction and analysis / update stages are repeatedly executed until the convergence condition or the number of iterations is reached.

7. The resistivity data assimilation algorithm-based rock fracture prediction method according to claim 6, characterized in that, The assimilation algorithm is represented in the form of: in, It is the first The optimal estimate of the state variables or parameter variables in the next iteration, i.e., the optimal estimate of the model's state variables or parameters. It is the updated state variable or parameter vector, i.e., the first... The optimal estimate in the next iteration It is the first The gain matrix of the next iteration is used to adjust the weights of the observed data and the model prediction data. It is a vector of measured observation data, acquired by the electrode sensor. It is a model operator that converts state variables or parameter variables. The predicted values ​​mapped into the observation space are the observed data.

8. A rock failure prediction system based on resistivity data assimilation algorithm, characterized in that, Comprising: A temperature and pressure control device, a loading device, an acoustic-electric sensor, an electrode switch, a high-density electrical method instrument and a data processing device; wherein: The loading device is used to apply external force to the rock sample; The temperature and pressure control device is connected to the loading device, and is used to control the temperature and pressure of the loading device to simulate the real underground stress condition; The acoustic-electric sensor is arranged on the surface of the rock sample, and is used to collect the electrical conductivity or resistivity of the rock sample; The electrode switch is connected to the acoustic-electric sensor at one end and to the high-density electrical method instrument at the other end; The high-density electrical method instrument is used to receive the resistivity signal; The data processing device is connected to the high-density electrical method instrument, and a computer program is arranged in the data processing device to realize the rock fracture prediction method based on the resistivity data assimilation algorithm in any one of claims 1 to 7.

Images