Mine support parameter optimization method and system based on big data

Abstract

The invention relates to the technical field of mine supporting, in particular to a mine supporting parameter optimization method and system based on big data, and the method comprises the steps: initializing supporting parameters, and constructing a mine supporting model; finite element analysis is conducted on the mine support model, the maximum stress and the maximum deformation under the support parameters are obtained, and the minimum value in the difference value of the allowable strength limit and the maximum stress and the difference value of the allowable deformation threshold and the maximum deformation serves as the support stability; constructing a fitness function; and iteratively updating the support parameters by using a particle swarm algorithm to obtain target support parameters. According to the technical scheme, the mine support parameters can be accurately obtained.

Description

Technical Field

[0001] The present application relates to the field of mine support technology, and in particular to a mine support parameter optimization method and system based on big data. Background Art

[0002] Mine support is a key technology in mining engineering, mainly used to improve the stability and service life of tunnels, stopes or key underground engineering spaces, and to prevent geological disasters such as collapse and rock bursts. How to determine mine support parameters to meet the urgent needs of intelligent design, construction and management of mining engineering is an urgent problem to be solved.

[0003] At present, a patent application document with application publication number CN118468393A discloses a method and device, storage medium, and terminal for dynamic design of mine tunnel support, wherein the method includes: obtaining real-time disturbance microseismic monitoring data and reference support parameters of the tunnel in the execution stage; updating a refined numerical model according to the reference support parameters, and performing support stability simulation according to the real-time disturbance microseismic monitoring data and the updated refined numerical model to obtain a first simulation result; if the first simulation result does not meet the support stability condition, adjusting each parameter in the reference support parameter according to the support parameter adjustment strategy to obtain multiple first improved support parameters; matching at least one second improved support parameter from a preset support parameter mapping relationship set according to the real-time disturbance microseismic monitoring data; generating multiple groups of candidate support parameters according to the first improved support parameters and the second improved support parameters, and obtaining a second simulation result corresponding to each candidate support parameter through simulation of the updated refined numerical model; determining the target support parameter for implementing the target stage tunnel support from the candidate support parameters according to the second simulation result.

[0004] The above method adjusts each parameter in the reference support parameters according to the support parameter adjustment strategy to obtain the first improved support parameters. During the adjustment process, each parameter is adjusted individually according to a fixed strategy, ignoring the influence of each parameter in the reference support parameters on the support effect, resulting in the inability to accurately obtain the optimal support parameters, making the mine support parameters inaccurate. Summary of the invention

[0005] In order to solve the technical problem of inaccurate mine support parameters, the present application provides a mine support parameter optimization method and system based on big data, which can accurately obtain mine support parameters.

[0006] In a first aspect, the present application provides a method for optimizing mine support parameters based on big data, the optimization method comprising: initializing support parameters and constructing a mine support model, the support parameters comprising anchor parameters, steel mesh parameters and shotcrete parameters; performing finite element analysis on the mine support model to obtain the maximum stress and maximum deformation under the support parameters, and taking the difference between the allowable strength limit and the maximum stress, and the minimum value of the difference between the allowable deformation threshold and the maximum deformation as the support stability; constructing a fitness function, the fitness function is positively correlated with the support stability and negatively correlated with the support cost; iteratively updating the support parameters using a particle swarm algorithm until the fitness function of the global optimal position is greater than a preset value, or the number of iterations is greater than a preset number, obtaining the target support parameters, comprising: calculating the change amount and fitness contribution value of each support parameter in this iteration; taking the velocity gain of each support parameter after normalization as the inertia weight in this iteration, and updating the real-time position of each particle, the velocity gain is negatively correlated with the change amount, and positively correlated with the fitness contribution value.

[0007] A mine support model is constructed according to the initialized support parameters, and the maximum stress and maximum deformation of the mine support model under the support parameters are obtained with the help of finite element analysis. When the maximum stress is less than the allowable strength limit and the maximum deformation is less than the allowable deformation threshold, it means that the support stability of the support parameters is good. A fitness function is constructed based on the support stability and support cost. The better the support stability and the lower the support cost, the larger the fitness function of the support parameter. The particle swarm algorithm is further used to iteratively update the support parameters to obtain the target support parameters corresponding to the maximum value of the fitness function.

[0008] Preferably, the anchor rod parameters include anchor rod spacing and anchor rod length; the steel mesh parameters include mesh density and steel bar diameter; and the shotcrete parameters include concrete thickness.

[0009] Preferably, this iteration Medium support parameters The amount of change for: , and Respectively The iteration and Support parameters in the iteration The numerical value of .

[0010] Support parameters The amount of change The larger the value, the better the iteration. Previously, the particle swarm algorithm was used to support the parameters We have conducted extensive exploration on this topic and accurately quantified the particle swarm algorithm’s performance in supporting parameters. The scope of exploration.

[0011] Preferably, this iteration Medium support parameters The method for obtaining the fitness contribution value of includes: Support parameters in the iteration The value of Support parameters in the iteration The value of the simulated support parameters is obtained; the first support stability and the first support stability under the simulated support parameters are calculated. The second support stability under the support parameters in the iteration, and The third support stability under the support parameters in the iteration; the ratio of the difference between the first support stability and the third support stability to the difference between the second support stability and the third support stability is taken as the stability of this iteration Medium support parameters The fitness contribution value of .

[0012] Support parameters The larger the fitness contribution value is, the better the fitness is. Can effectively improve the support effect and accurately quantify the support parameters The degree of influence on the fitness function.

[0013] Preferably, this iteration Medium support parameters Speed ​​gain for: , and For this iteration Medium support parameters The fitness contribution value and change of is the Relu activation function.

[0014] If the change of any support parameter is larger, it means that a large-scale exploration was carried out on the support parameter before this iteration. In order to make the particle swarm algorithm explore all support parameters and ensure the accurate global optimal solution, the speed of the support parameter is reduced in this iteration, and the speed of other support parameters is increased; at the same time, if the fitness contribution value of any support parameter is larger, it means that the support effect can be effectively improved by changing the support parameter. In this iteration, increasing the speed of the support parameter can effectively improve the support effect by adjusting the value of the support parameter, accelerate the convergence speed of the particle swarm algorithm, and quickly obtain the target support parameter. In summary, the speed of each support parameter is adjusted according to the change of each support parameter and the fitness contribution value, while accelerating the convergence speed of the particle swarm algorithm to ensure the accurate global optimal solution.

[0015] Preferably, updating the real-time position of each particle includes: ; ; in, For this iteration A velocity gain vector, including the velocity gain of each support parameter after normalization, corresponding to the inertia weight; For the The velocity of particle m in the iteration, including the velocity components of each support parameter; for and Hadamard and and social factors of particle m, respectively. and is a random number between 0 and 1. and For this iteration The global optimal position and the individual optimal position of particle m, and Respectively iterations and this iteration The real-time position of particle m in .

[0016] Preferably, the individual factor of particle m The method for obtaining includes: in the individual optimal positions of all particles in this iteration, the outlier value of particle m is calculated using the LOF algorithm, and the difference between 1 and the outlier value is used as the individual factor of particle m. .

[0017] Individual factor of particle m Indicates the degree of influence of the individual optimal position of particle m on the exploration process. The accuracy of the individual optimal position also affects the convergence speed of the particle swarm algorithm. The accuracy of the individual optimal position is accurately judged according to the distribution of the individual optimal positions of all particles, and then the individual factor of each particle is calculated. A larger individual factor is assigned to particles with greater accuracy to ensure the convergence speed, and a smaller individual factor is assigned to particles with less accuracy to explore the global optimal solution.

[0018] Preferably, the support stability Satisfies the relationship: , and are the allowable strength limit and maximum stress, respectively. and They are the allowed deformation threshold and maximum deformation respectively.

[0019] Preferably, the fitness function Satisfies the relationship: , To maintain stability, For the support cost, is the adjustment coefficient.

[0020] In the second aspect of the present application, a mine support parameter optimization system based on big data is also provided, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the mine support parameter optimization method based on big data described in the first aspect of the present application is implemented.

[0021] The technical solution of this application has the following beneficial technical effects: A mine support model is constructed according to the initialized support parameters, and the maximum stress and maximum deformation of the mine support model under the support parameters are obtained with the help of finite element analysis. When the maximum stress is less than the allowable strength limit and the maximum deformation is less than the allowable deformation threshold, it means that the support stability of the support parameters is good. A fitness function is constructed based on the support stability and support cost. The better the support stability and the lower the support cost, the larger the fitness function of the support parameter. The particle swarm algorithm is further used to iteratively update the support parameters to obtain the target support parameters corresponding to the maximum value of the fitness function.

[0022] Furthermore, in the process of iteratively updating the support parameters using the particle swarm algorithm, the change amount and fitness contribution value of each support parameter in any iteration are calculated; if the change amount of any support parameter is larger, it means that a large amount of exploration has been carried out on the support parameter before this iteration. In order to make the particle swarm algorithm explore all support parameters and ensure the accurate global optimal solution, the speed of the support parameter is reduced in this iteration, and the speed of other support parameters is increased; at the same time, if the fitness contribution value of any support parameter is larger, it means that the support effect can be effectively improved by changing the support parameter. In this iteration, increasing the speed of the support parameter can effectively improve the support effect by adjusting the value of the support parameter, accelerate the convergence speed of the particle swarm algorithm, and quickly obtain the target support parameter. In summary, the speed of each support parameter is adjusted according to the change amount and fitness contribution value of each support parameter, while accelerating the convergence speed of the particle swarm algorithm to ensure the accurate global optimal solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a flow chart of a mine support parameter optimization method based on big data according to an embodiment of the present application.

[0024] Figure 2 It is a structural block diagram of a mine support parameter optimization system based on big data according to an embodiment of the present application. DETAILED DESCRIPTION

[0025] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0026] According to the first aspect of the present application, the present application provides a mine support parameter optimization method based on big data. Figure 1 : is a flow chart of a method for optimizing mine support parameters based on big data according to an embodiment of the present application. Figure 1 As shown, the mine support parameter optimization method based on big data includes steps S101 to S104, which are described in detail below.

[0027] S101, initializing support parameters and constructing a mine support model, wherein the support parameters include anchor parameters, steel mesh parameters and shotcrete parameters.

[0028] In one embodiment, the support parameters include anchor rod parameters, steel mesh parameters and shotcrete parameters. More specifically, the anchor rod parameters include anchor rod spacing and anchor rod length; the steel mesh parameters include mesh density and steel bar diameter; and the shotcrete parameters include concrete thickness.

[0029] Among them, after the support parameters are determined, the mine support model can be simulated based on the support parameters.

[0030] S102, perform finite element analysis on the mine support model to obtain the maximum stress and maximum deformation under the support parameters, and take the minimum value of the difference between the allowable strength limit and the maximum stress, and the minimum value of the difference between the allowable deformation threshold and the maximum deformation as the support stability.

[0031] In one embodiment, after obtaining the mine support model, the mine support model is meshed, and after applying mechanical constraints of the rock mass to the mine support model in the form of uniformly distributed loads, a finite element analysis is performed on the mine support model to obtain the maximum stress and maximum deformation of the mine support model under the support parameters.

[0032] The maximum stress and the allowable strength limit, as well as the allowable deformation threshold and the maximum deformation are compared respectively. If the maximum stress does not exceed the allowable strength limit and the maximum deformation does not exceed the allowable deformation threshold, it means that the current mine support model can effectively achieve the effect of mine support, which further indicates that the corresponding support parameters have good support stability.

[0033] Specifically, support stability Satisfies the relationship: , and are the allowable strength limit and maximum stress, respectively. and They are the allowed deformation threshold and maximum deformation respectively.

[0034] Among them, the allowable strength limit is related to the materials of anchor rods and steel mesh, and the allowable deformation threshold is related to the construction requirements of mine support.

[0035] In this way, the maximum stress and maximum deformation under the support parameters are obtained with the help of finite element analysis, and then the support stability of the support parameters is accurately evaluated.

[0036] S103, constructing a fitness function, wherein the fitness function is positively correlated with support stability and negatively correlated with support cost.

[0037] In one embodiment, the purpose of optimizing mine support parameters is to achieve maximum support stability at minimum support cost. Therefore, a fitness function is constructed, which is positively correlated with support stability and negatively correlated with support cost. The larger the fitness function, the better the support effect, which provides a basis for the subsequent use of particle swarm algorithm to obtain target support parameters.

[0038] Specifically, the support cost includes material cost and construction cost, and the fitness function Satisfies the relationship: , To maintain stability, For the support cost, is the adjustment coefficient. The value of the adjustment coefficient is 0.01.

[0039] S104, using the particle swarm algorithm to iteratively update the support parameters until the fitness function of the global optimal position is greater than a preset value, or the number of iterations is greater than a preset number, and the target support parameters are obtained.

[0040] In one embodiment, after constructing the fitness function, multiple initialized support parameters are obtained, and one support parameter corresponds to the position of one particle; illustratively, the number of particles is 30, that is, 30 different support parameters need to be obtained. The particle swarm algorithm is used to iteratively update the position of each particle. One iteration process will simultaneously update the real-time position of all particles, and the real-time position with the largest fitness function value in the historical iteration is used as the individual optimal position of the corresponding particle, and the individual optimal position corresponding to the maximum fitness function value among all individual optimal positions is used as the global optimal position.

[0041] Specifically, the iterative update of support parameters using the particle swarm algorithm includes: calculating the change amount and fitness contribution value of each support parameter in this iteration; using the velocity gain of each support parameter after normalization as the inertia weight in this iteration, and updating the real-time position of each particle, and the velocity gain is negatively correlated with the change amount and positively correlated with the fitness contribution value.

[0042] Among them, this iteration Medium support parameters The amount of change for: , and Respectively The iteration and Support parameters in the iteration The numerical value of .

[0043] Among them, this iteration Medium support parameters The method for obtaining the fitness contribution value of includes: Support parameters in the iteration The value of Support parameters in the iteration The value of the simulated support parameters is obtained; the first support stability and the first support stability under the simulated support parameters are calculated. The second support stability under the support parameters in the iteration, and The third support stability under the support parameters in the iteration; the ratio of the difference between the first support stability and the third support stability to the difference between the second support stability and the third support stability is taken as the stability of this iteration Medium support parameters The fitness contribution value of .

[0044] Understandably, the support parameters The amount of change The larger the value, the better the iteration. Previously, the particle swarm algorithm was used to support the parameters In order to make the particle swarm algorithm explore all support parameters and ensure the accurate global optimal solution, in this iteration In the case of speed, increase the support parameters The speed of other support parameters; at the same time, support parameters The larger the fitness contribution value is, the better the fitness is. It can effectively improve the support effect. Increasing the support parameters The speed can be adjusted by adjusting the support parameters The value of can effectively improve the support effect, accelerate the convergence speed of the particle swarm algorithm, and quickly obtain the target support parameters.

[0045] Therefore, this iteration Medium support parameters Speed ​​gain for: , and For this iteration Medium support parameters The fitness contribution value and change of is the Relu activation function. When the fitness contribution value When it is less than or equal to 0, it indicates the support parameter The change of will make the fitness function smaller, so the Relu activation function is used to The value of is mapped to 0, and this iteration Medium support parameters Set the speed gain to 0.

[0046] In one embodiment, for this iteration The velocity gain of each support parameter in is normalized. The specific process is: calculate this iteration Medium support parameters Speed ​​gain With this iteration The ratio of the sum of the velocity gains of all support parameters in the normalized support parameter is obtained. Speed ​​gain .

[0047] In one embodiment, for this iteration After the velocity gain of each support parameter in is normalized, the specific process of updating the real-time position of each particle is described by taking particle m as an example, including: ; ;in, For this iteration The velocity gain vector includes the velocity gain of each support parameter after normalization; For the The velocity of particle m in the iteration, including the velocity components of each support parameter; for and Hadamard and and social factors of particle m, respectively. and is a random number between 0 and 1. and For this iteration The global optimal position and the individual optimal position of particle m, and Respectively iterations and this iteration The real-time position of particle m in this iteration The velocity gain vector Corresponding to the inertia weight in the particle swarm algorithm.

[0048] in, is a multidimensional vector, including the The velocity component of each support parameter in the iteration, the larger the velocity component, the greater the change of the corresponding support parameter; similarly, It is also a multidimensional vector, including this iteration The velocity gain of each support parameter after normalization is calculated by The Hadamard product between them is used to adjust the velocity component of each support parameter, thereby adjusting the current iteration The variation of each support parameter in .

[0049] Among them, the individual factor of particle m and social factors The value of is 1.

[0050] In another embodiment, the individual factor of particle m It indicates the influence of the individual optimal position of particle m on the exploration process. The accuracy of the individual optimal position also affects the convergence speed of the particle swarm algorithm. The accuracy of the individual optimal position can be judged according to the distribution of the individual optimal positions of all particles, and then the individual factor of each particle is calculated. Specifically, the individual factor of particle m The method for obtaining includes: in the individual optimal positions of all particles in this iteration, the outlier value of particle m is calculated using the LOF algorithm, and the difference between 1 and the outlier value is used as the individual factor of particle m. .

[0051] In this way, the accuracy of the individual optimal position is judged according to the distribution of the individual optimal positions of all particles in this iteration, and then the individual factor of the particle is determined. A larger individual factor is assigned to the particle with greater accuracy to ensure the convergence speed, and a smaller individual factor is assigned to the particle with less accuracy to explore the global optimal solution.

[0052] In this way, the real-time positions of all particles are iteratively updated until the fitness function of the global optimal position is greater than the preset value, or the number of iterations is greater than the preset number, indicating that the global optimal position meets the mine support requirements, and the global optimal position at this time is used as the target support parameter.

[0053] According to the second aspect of the present application, the present application also provides a mine support parameter optimization system based on big data. Figure 2 is a structural block diagram of a mine support parameter optimization system based on big data according to an embodiment of the present application. Figure 2 As shown, the system 50 includes a processor and a memory, the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the method for optimizing mine support parameters based on big data according to the first aspect of the present application is implemented. The system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, whose settings and functions are known in the art, and therefore will not be described in detail here.

[0054] It should be pointed out that, for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present application, which all fall within the scope of protection of the present application.

Claims

1. A mine support parameter optimization method based on big data, characterized in that: The optimization method comprises: Initializing support parameters and building a mine support model, wherein the support parameters include anchor parameters, steel mesh parameters and shotcrete parameters; Finite element analysis is performed on the mine support model to obtain the maximum stress and maximum deformation under the support parameters, and the minimum value of the difference between the allowable strength limit and the maximum stress, and the minimum value of the difference between the allowable deformation threshold and the maximum deformation is taken as the support stability; Constructing a fitness function, wherein the fitness function is positively correlated with support stability and negatively correlated with support cost; The support parameters are iteratively updated using the particle swarm algorithm until the fitness function of the global optimal position is greater than a preset value, or the number of iterations is greater than a preset number, and the target support parameters are obtained, including: calculating the change amount and fitness contribution value of each support parameter in this iteration; using the velocity gain of each support parameter after normalization as the inertia weight in this iteration, and updating the real-time position of each particle, the velocity gain is negatively correlated with the change amount, and positively correlated with the fitness contribution value.

2. The mine support parameter optimization method based on big data according to claim 1 is characterized in that: The anchor rod parameters include anchor rod spacing and anchor rod length; the steel mesh parameters include mesh density and steel bar diameter; and the shotcrete parameters include concrete thickness.

3. The mine support parameter optimization method based on big data according to claim 1 is characterized in that: This iteration Medium support parameters The amount of change for: , and Respectively The iteration and Support parameters in the iteration The numerical value of .

4. The mine support parameter optimization method based on big data according to claim 1 is characterized in that: This iteration Medium support parameters The methods for obtaining the fitness contribution value include: The first Support parameters in the iteration The value of Support parameters in the iteration The numerical value of is used to obtain the simulated support parameters; Calculate the stability of the first support and the second support under the simulated support parameters The second support stability under the support parameters in the iteration, and The stability of the third support under the support parameters in the iteration; The ratio of the difference between the first support stability and the third support stability to the difference between the second support stability and the third support stability is taken as the value of this iteration. Medium support parameters The fitness contribution value of .

5. The mine support parameter optimization method based on big data according to claim 1 is characterized in that: This iteration Medium support parameters Speed ​​gain for: , and For this iteration Medium support parameters The fitness contribution value and change of is the Relu activation function.

6. The mine support parameter optimization method based on big data according to claim 1 is characterized in that: The updating of the real-time position of each particle includes: ； ； in, For this iteration A velocity gain vector, including the velocity gain of each support parameter after normalization, corresponding to the inertia weight; For the The velocity of particle m in the iteration, including the velocity components of each support parameter; for and The Hadamard and and social factors of particle m, respectively. and is a random number between 0 and 1. and For this iteration The global optimal position and the individual optimal position of particle m, and Respectively iterations and this iteration The real-time position of particle m in .

7. The method for optimizing mine support parameters based on big data according to claim 6, characterized in that: Individual factor of particle m The methods for obtaining include: In the individual optimal positions of all particles in this iteration, the outlier value of particle m is calculated using the LOF algorithm, and the difference between 1 and the outlier value is used as the individual factor of particle m. .

8. The mine support parameter optimization method based on big data according to claim 1 is characterized in that: Support stability Satisfies the relationship: , and are the allowable strength limit and maximum stress, respectively. and They are the allowed deformation threshold and maximum deformation respectively.

9. The mine support parameter optimization method based on big data according to claim 1, characterized in that: The fitness function Satisfies the relationship: , To maintain stability, For the support cost, is the adjustment coefficient.

10. The mine support parameter optimization system based on big data is characterized by: It comprises a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a mine support parameter optimization method based on big data according to any one of claims 1 to 9 is implemented.

Images