Anchor rod shear strength prediction method and device for rock mass reinforcement

Abstract

The invention provides an anchor rod shear strength prediction method and device for rock mass reinforcement, and relates to the technical field of data processing, and the method comprises the steps: constructing an anchor rod stress theoretical model based on anchor rod design data and anchor rod displacement data, and constructing the anchor rod stress theoretical model and a strength criterion to obtain a shear strength physical model; carrying out finite difference processing according to anchor rod design data and anchor rod historical monitoring data to obtain an information data sample, inputting the information data sample into a neural network for training, and carrying out optimization processing on a training result and parameters in the neural network by an improved sparrow optimization algorithm to obtain a final neural network; and predicting the shear strength of the joint surface of the rock mass reinforced by the anchor rod according to the optimization parameters and historical monitoring data in the final neural network to obtain a shear strength prediction model, and inputting anchor rod design data into the shear strength prediction model for prediction. According to the method, the problems of various parameters and complex mutual coupling among the parameters are solved, and the safety and the stability of an engineering structure are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method and a device for predicting the shear strength of an anchor used for rock reinforcement. Background Art

[0002] In the existing field of data processing technology, in order to solve the problem of anchor failure in rock joints under the combined action of shear force and axial force, the current industry mainly relies on a series of indoor rock direct shear tests when calculating the shear strength performance of anchors anchored to joint surfaces. These tests not only analyze the type of anchor material and the size of the anchoring angle, but also introduce the yield criterion of anisotropic materials to further improve the anchor force analysis model. However, when calculating the shear strength of the anchored joint surface, the process is significantly affected by human factors, involving many parameters and complex mutual coupling between the parameters, leading to the risk of structural failure.

[0003] Therefore, there is an urgent need for a method and device for predicting the shear strength of anchor rods for rock reinforcement to solve the problem of numerous parameters and complex coupling between parameters, thereby ensuring the safety and stability of engineering structures. Summary of the invention

[0004] The object of the present invention is to provide a method and device for predicting the shear strength of an anchor for rock reinforcement to improve the above-mentioned problem. In order to achieve the above-mentioned object, the technical solution adopted by the present invention is as follows:

[0005] In a first aspect, the present application provides a method for predicting the shear strength of an anchor for rock reinforcement, comprising:

[0006] Obtain anchor design data and anchor historical monitoring data;

[0007] Based on the anchor design data and the preset anchor displacement data, a theoretical model of anchor force is constructed, and a shear strength physical model is obtained by constructing the theoretical model of anchor force and the preset strength criterion;

[0008] Performing finite difference processing on the anchor design data and the anchor historical monitoring data to obtain information data samples, and obtaining a training set by inputting the information data samples into the shear strength physical model for calculation;

[0009] Inputting the training set into a preset neural network for training to obtain training results, optimizing the training results and the parameters in the neural network using a preset improved sparrow optimization algorithm to obtain a final neural network;

[0010] The shear strength of the rock mass joint surface reinforced with anchor rods is predicted based on the optimized parameters in the final neural network and the historical monitoring data to obtain a shear strength prediction model, and the prediction is performed by inputting the anchor rod design data into the shear strength prediction model.

[0011] In a second aspect, the present application also provides a device for predicting the shear strength of an anchor for rock reinforcement, comprising:

[0012] An acquisition module is used to acquire anchor design data and anchor historical monitoring data;

[0013] A construction module, used to construct an anchor rod force theory model based on the anchor rod design data and preset anchor rod displacement data, and to obtain a shear strength physical model by constructing the anchor rod force theory model and preset strength criteria;

[0014] A processing module, used for performing finite difference processing on the anchor design data and the anchor historical monitoring data to obtain information data samples, and obtaining a training set by inputting the information data samples into the shear strength physical model for calculation;

[0015] A training module, used for inputting the training set into a preset neural network for training to obtain training results, and optimizing the training results and the parameters in the neural network by a preset improved sparrow optimization algorithm to obtain a final neural network;

[0016] The prediction model is used to predict the shear strength of the rock joint surface reinforced by anchor rods according to the optimization parameters in the final neural network and the historical monitoring data, and obtain a shear strength prediction model, and the prediction is performed by inputting the anchor rod design data into the shear strength prediction model.

[0017] In a third aspect, the present application also provides an anchor shear strength prediction device for rock reinforcement, comprising:

[0018] Memory for storing computer programs;

[0019] A processor is used to implement the steps of the method for predicting the shear strength of anchor rods for rock reinforcement when executing the computer program.

[0020] In a fourth aspect, the present application further provides a readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the above-mentioned method for predicting the shear strength of anchor rods for rock reinforcement are implemented.

[0021] The beneficial effects of the present invention are:

[0022] The present invention establishes a theoretical model that can accurately describe the stress of anchor rods through precise mathematical analysis and physical simulation, and then constructs a shear strength physical model through the theoretical model of anchor rod stress to reflect the deformation of anchor rods in the actual stress process. The stress state and deformation behavior of anchor rods under complex geological conditions are accurately simulated by numerical processing of finite difference method, and a series of highly accurate and practical information data samples are generated. The information data samples not only reflect the response characteristics of anchor rods under different working conditions, but also contain potential shear strength related information. The information data samples are then processed by the final neural network, key features are extracted, and the parameters in the neural network are adjusted and optimized to obtain a shear strength prediction model. By inputting the anchor rod design data into the shear strength prediction model for prediction, the shear strength prediction value of the rock joint surface after anchor rod reinforcement under specific geological and working conditions is automatically calculated. This prediction value provides an important reference basis for engineers, helping them to more accurately evaluate the reinforcement effect of anchor rods, optimize the design scheme, and ensure the safety and stability of engineering structures.

[0023] In summary, the present invention solves the problem of numerous parameters and complex mutual coupling between parameters, thereby ensuring the safety and stability of the engineering structure.

[0024] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or be understood by implementing the embodiments of the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0026] Figure 1 It is a schematic flow chart of a method for predicting the shear strength of an anchor used for rock reinforcement described in an embodiment of the present invention;

[0027] Figure 2 It is a schematic diagram of the structure of the anchor shear strength prediction device for rock reinforcement described in an embodiment of the present invention.

[0028] Markings in the figure: 800, anchor shear strength prediction device for rock reinforcement; 801, processor; 802, memory; 803, multimedia component; 804, I / O interface; 805, communication component. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0030] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings. At the same time, in the description of the present invention, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0031] Embodiment 1:

[0032] This embodiment provides a method for predicting the shear strength of an anchor for rock reinforcement.

[0033] See also Figure 1 , the figure shows that the method includes steps S1 to S6, including:

[0034] S1: Obtain anchor design data and anchor historical monitoring data;

[0035] Preferably, the anchor rod is a BFRP anchor rod.

[0036] S2: constructing a theoretical model of anchor force based on the anchor design data and preset anchor displacement data, and obtaining a shear strength physical model by constructing the theoretical model of anchor force and preset strength criteria;

[0037] In order to clarify the specific method of obtaining the shear strength physical model, step S2 includes S21 to S25, which are:

[0038] S21: constructing based on the anchor angle of the anchor rod in the anchor rod design data and the preset anchor rod displacement data to obtain a deformation component, and calculating the anchor rod force based on the deformation component to obtain an anchor rod force theoretical model;

[0039] In this step, the joint surface of the rock mass will undergo shear displacement and normal displacement perpendicular to the joint surface under shear load and shear dilation effect, which will cause the anchor rod to produce axial displacement and tangential displacement, making the anchor rod form an "S"-like deformation component near the joint surface. By calculating the shear deformation component of the anchor rod and the anchor rod force, the theoretical model of the anchor rod force during the shear process is obtained;

[0040] Wherein, the axial displacement and the tangential displacement are anchor displacement data, the anchor force includes the axial force of the anchor and the shear force of the anchor, and the anchor shear deformation component includes the axial deformation component of the anchor and the lateral deformation component of the anchor perpendicular to the axial direction;

[0041] Calculating according to the anchor angle of the anchor rod, the shear displacement of the preset joint surface, and the normal displacement of the preset joint surface in the anchor rod design data to obtain the axial deformation component of the anchor rod and the lateral deformation component of the anchor rod perpendicular to the axial direction;

[0042] The axial deformation component of the anchor rod is:

[0043] u 0 =0.5(U 0 cosα+U d sinα) (1)

[0044] In the above formula (1): u 0 represents the axial deformation component of the anchor rod; U 0 represents the shear displacement of the anchored joint surface; U d represents the normal displacement of the anchored joint surface; α represents the anchor angle of the anchor bolt;

[0045] The axial deformation component of the anchor rod is:

[0046] v 0 =0.5(U 0 sinα-U d cosα) (2)

[0047] In the above formula (2): v 0 It represents the transverse deformation component of the anchor bolt perpendicular to the axial direction; U 0 represents the shear displacement of the anchored joint surface; U d represents the normal displacement of the anchored joint surface; α represents the anchor angle of the anchor bolt;

[0048] Calculating based on the axial deformation component of the anchor rod and the transverse deformation component of the anchor rod perpendicular to the axial direction, to obtain the axial force of the anchor rod and the shear force of the anchor rod;

[0049] The shear force of the anchor rod is:

[0050]

[0051] In the above formula (2): Q 0 represents the axial force of the anchor rod; v 0 The lateral deformation component of the anchor bolt perpendicular to the axial direction; E represents the elastic modulus of the anchor bolt; D b Indicates the diameter of the anchor rod; P u represents the ultimate reaction force per unit length of rock mass; b represents the Poisson's ratio of the anchor; π represents the circumference of a circle, which is 3.14159;

[0052] The axial force of the anchor rod is:

[0053]

[0054] In the above formula (4): N 0 Indicates the axial force of the anchor rod; u 0 represents the axial deformation component of the anchor rod; E represents the elastic modulus of the anchor rod; D b Indicates the diameter of the anchor rod; P u represents the ultimate reaction force per unit length of rock mass; b represents the Poisson's ratio of the anchor rod; π represents the circumference of a circle, which is 3.14159; Q 0 Represents the axial force of the anchor rod.

[0055] S22: judging whether the anchor rod force theory model satisfies a preset strength yield condition, and obtaining a strength criterion result when the anchor rod force theory model satisfies the strength yield condition;

[0056] In this step, during the shearing process, as the shear displacement of the anchored joint surface increases, the axial force of the anchor rod and the shear force of the anchor rod gradually increase and gradually reach a yield state;

[0057] The strength criterion is:

[0058]

[0059] In the above formula (5), σ x represents the axial normal stress of the anchor rod; σ y represents the transverse normal stress of the anchor rod; σ s represents the shear stress in the anchor plane; X represents the axial strength of the anchor; Y represents the transverse strength of the anchor; S represents the shear strength in the anchor plane; N o Indicates the axial force of the anchor rod; Q o represents the shear force of the anchor rod; p u Indicates the ultimate reaction force per unit length of rock mass; D b Indicates the diameter of the anchor rod; A b represents the cross-sectional area of ​​the anchor rod; π represents the ratio of pi to 3.14159;

[0060] Among them, when the anchor rod stress theoretical model does not meet the strength yield condition, the shear displacement of the initial anchor joint surface and the normal displacement of the initial anchor joint surface are increased, and the axial deformation component of the anchor rod and the lateral deformation component of the anchor rod perpendicular to the axial direction are recalculated until the anchor rod yields.

[0061] S23: constructing a shear strength physical model based on the strength criterion result and the anchor rod force theoretical model.

[0062] In this step, the shear strength of the anchor-reinforced rock joint surface in the shear strength physical model is divided into three components: the first part is the shear strength of the joint surface itself, the second part is the shear strength converted from the axial force of the anchor at the joint surface, and the third part is the shear strength converted from the shear force of the anchor at the joint surface; the shear strength physical model is used to reflect the deformation of the anchor during the actual force process.

[0063] The shear strength physical model is:

[0064]

[0065] In the above formula (6), τ ba represents the shear strength converted from the anchor axial force; τ bt represents the shear strength converted from the shear force of the anchor at the joint surface; τ j Indicates the shear strength of the joint surface itself; c j represents the cohesion of the joint surface; φ represents the internal friction angle of the joint surface; σ j Represents the historical monitoring data of the rock mass joint surface reinforced by anchor bolts. The normal stress of the rock mass joint surface reinforced at each preset point; N o Indicates the axial force of the anchor rod; Q o represents the shear force of the anchor bolt; α represents the anchor angle of the anchor bolt; ψ represents the shear dilation angle of the joint surface; A b represents the cross-sectional area of ​​the anchor rod; τ represents the shear strength of the rock joint surface reinforced by the anchor rod.

[0066] S3: performing finite difference processing on the anchor design data and the anchor historical monitoring data to obtain information data samples, and obtaining a training set by inputting the information data samples into the shear strength physical model for calculation;

[0067] In order to clarify the specific method of obtaining the training set, step S3 includes S31 to S33, which are:

[0068] S31: performing image processing by finite difference software according to the anchor design data and the anchor historical monitoring data to obtain a random field image;

[0069] S32: extracting based on the random field image, calculating by combining the preset anchor rod under the heterogeneity of the rock mass and the irregularity of the preset joint surface, to obtain information data samples, wherein the information data samples are no less than two;

[0070] S33: inputting all the information data samples into the shear strength physical model for calculation to obtain a plurality of shear strength values, and integrating all the shear strength values ​​and all the information data samples to obtain a training set.

[0071] In this step, by accurately simulating the stress state and deformation behavior of anchor rods under complex geological conditions, a series of highly accurate and practical information data samples were generated. The information data samples not only reflect the response characteristics of anchor rods under different working conditions, but also contain potential shear strength related information. Through complex physical equations and numerical calculation methods, the shear strength values ​​of anchor rods under different conditions are simulated and calculated, thus forming a training set with rich training information, ensuring that the subsequent machine learning model can learn the core physical mechanism of anchor rod shear strength.

[0072] S4: inputting the training set into a preset neural network for training to obtain training results, optimizing the training results and the parameters in the neural network by a preset improved sparrow optimization algorithm to obtain a final neural network;

[0073] In order to clarify the specific method of obtaining the final neural network, step S4 includes S41 to S46, which are:

[0074] S41: Obtaining the initial sparrow population;

[0075] S42: inputting the training set into a preset neural network for training to obtain a training result;

[0076] In this step, the training set is trained to optimize the parameters of the preset neural network so that it has higher prediction and generalization capabilities.

[0077] S43: initializing the training results, the penalty parameters and the kernel parameters in the neural network based on a preset improved sparrow optimization algorithm, and uniformly distributing the initial sparrow population through a preset chaotic mapping method to obtain a sparrow population distribution;

[0078] In this step, the preset improved sparrow optimization algorithm is used to initialize the training results, the penalty parameters and the kernel parameters in the neural network, the parameters of the improved sparrow optimization algorithm are set, and the sin-Tent-Cosine chaotic mapping method is introduced to uniformly distribute the initial sparrow population to obtain the sparrow population distribution;

[0079] The sparrows are mutated and reproduced through the chaotic mapping method, and the chaotic mapping method is introduced into the original sparrow search algorithm to obtain a new swarm intelligence algorithm. The new swarm intelligence algorithm increases the algorithm's operating efficiency and global optimization ability by improving the diversity of the initialized population.

[0080] The parameters of the improved sparrow optimization algorithm are the maximum number of iterations of the algorithm, the proportion of discoverers, the warning value, the proportion of scouts, the population size, and a sparrow population with a population size of n is randomly generated.

[0081] The sin-Tent-Cosine chaotic mapping method is:

[0082]

[0083] In the above formula (7), cos, x(i+1) represents the updated position of the population distribution after the chaotic mapping is improved; x(i) represents the initial position of the initial sparrow population; π represents the pi value of 3.14159; r represents the chaotic orbit state value;

[0084] The updated position of the population distribution after the improved chaotic mapping represents the sparrow population distribution.

[0085] S44: Screening the initial sparrow population based on the sparrow population distribution to obtain optimal sparrow population information;

[0086] In order to clarify the specific method of obtaining the optimal sparrow population information, step S44 includes S441 to S445, which are:

[0087] S441: Get the current number of iterations;

[0088] S442: Screening the initial sparrow population based on the sparrow population distribution and a preset greedy selection strategy to obtain an excellent sparrow population;

[0089] In this step, the initial sparrow population position and the preset target sparrow population position are compared and screened based on the sparrow population distribution and a preset greedy selection strategy to determine whether they can become the next generation of sparrow population position iteration individuals.

[0090] The judgment and screening expression is:

[0091]

[0092] In the above formula (8), X i+1 (t+1) represents the position of the sparrow population in the next iteration; U i (t+1) represents the initial position of the sparrow population; X i(t) represents the location of the target sparrow population; f represents the evaluation function; Other represents other situations;

[0093] S443: Calculating a fitness function based on the excellent sparrow population to obtain a fitness value;

[0094] In this step, the fitness function is:

[0095]

[0096] In the above formula (9), f i (t) represents the fitness value of the excellent sparrow population fi at the tth iteration; y i Represents the output result of the neural network model; Q i Indicates expected value; It means that all samples i from 1 to n are summed;

[0097] Among them, the excellent sparrow population f i The fitness value is used to indicate the position of the excellent sparrow population after screening at the tth iteration.

[0098] S444: judging the current number of iterations according to a preset maximum number of iterations to obtain an iteration result;

[0099] S445: Select and calculate the position information in the initial sparrow population according to the iteration result and the fitness value to obtain the optimal sparrow population information.

[0100] In order to clarify the specific method of obtaining the optimal sparrow population information, step S445 includes S4451 to S4455, which are specifically:

[0101] S4451: Obtain the current position of the follower and the current position of the alerter;

[0102] S4452: selecting different strategies according to the iteration results, calculating the position information in the initial sparrow population by using the different strategies, and obtaining the position information of the finder after iteration;

[0103] In this step, when the current number of iterations is less than 0.5 times the maximum number of iterations, the sine-cosine strategy perturbation is used to calculate the position information after the finder iteration;

[0104] The sine-cosine strategy update position expression is:

[0105]

[0106] In the above formula (10), Represents the position information of the discoverer after the t+1th iteration under the sine-cosine strategy; represents the location information of the discoverer at the tth iteration; r 1 Represents a random number; r 2 Indicates the alert value; the value range is [0.5,1]; represents the global optimal position in the tth iteration; t represents the current iteration number; T max Represents the maximum number of iterations; R 2 represents the alert value; St represents the safety threshold of the sparrow population location; ω represents the inertia weight;

[0107] When the current number of iterations is greater than or equal to 0.5 times the maximum number of iterations, the Levy flight perturbation mutation strategy is used to calculate the position information of the finder after iteration;

[0108] The updated position expression of the Levy flight disturbance mutation strategy is:

[0109]

[0110] In the above formula (11), represents the position information of the discoverer after the t+1th iteration under the Levy flight perturbation mutation strategy; Indicates the global optimal position found in the tth iteration; represents the global worst position found by the discoverer in the tth iteration; λ represents the random number between [0,1] generated by Cauchy mutation; t represents the current iteration number; Lrvy(d) represents the Levy flight step length of the problem dimension d; d represents the dimension of the problem to be solved; T max Indicates the maximum number of iterations.

[0111] Among them, the different strategies selected by the iteration result are the sine-cosine strategy and the Levy flight disturbance variation strategy; the position information in the initial sparrow population is the position information of the discoverer in the tth iteration.

[0112] S4453: Calculate the current position of the early warning person and the optimal position found by the preset finder according to the fitness value to obtain the position information of the early warning person after iteration;

[0113] In this step, the position information of the early warning person after iteration is:

[0114]

[0115] In the above formula (12), Represents the location information of the warner after the t+1th iteration; Represents the location information of the warner at the tth iteration; represents the global worst position found by the finder in the tth iteration; represents the global optimal position in the tth iteration; β represents a random number that follows a standard normal distribution and is used as a step size control parameter; K represents a random number that is a motion orientation control parameter; ε represents a minimum constant; fi represents the fitness value of the excellent sparrow population; fg represents the global optimal fitness value; gw represents the threshold of the global optimal fitness value;

[0116] The position information of the warner at the tth iteration indicates the current position of the warner; the global optimal position at the tth iteration indicates the optimal position found by the discoverer;

[0117] S4454: Calculate according to the current position of the follower and the optimal position found by the preset finder to obtain the position information of the follower after iteration;

[0118] In this step, the position information of the follower after iteration is:

[0119]

[0120] In the above formula (13), Represents the position information of the follower after the t+1th iteration; Represents the current position information of the follower at the tth iteration; represents the global worst position found by the finder in the tth iteration; represents the optimal position found by the discoverer in the t+1th iteration; t represents the current iteration number; Q represents a random number that obeys the normal distribution; exp represents the exponential function; L represents the unit vector; A represents a 1*d matrix whose elements are randomly assigned 1 or -1; i represents the individual index; and n represents the total number of individuals in the sparrow population.

[0121] Among them, when When the position update rule of the follower involves the global worst position and current location The difference is i 2 After scaling, it is adjusted by an exponential function and a normally distributed random number Q; when When the position update rule involves the optimal position and current location The difference is A + After scaling with the unit vector L, the difference is Add.

[0122] The current position information of the follower at the t-th iteration represents the current position of the follower; the optimal position found by the discoverer in the t+1-th iteration represents the optimal position found by the discoverer.

[0123] S4455: Optimize and update based on the iterative position information of the discoverer, the iterative position information of the follower, and the iterative position information of the warner to obtain the optimal sparrow population information.

[0124] S45: Optimizing the penalty parameter and the kernel parameter of the neural network according to the optimal sparrow population information and a preset simulated annealing algorithm to obtain an optimized penalty parameter and an optimized kernel parameter;

[0125] In order to clarify the specific method of obtaining the optimization penalty parameter and the optimization core parameter, step S45 includes S451 to S456, which are:

[0126] S451: Obtain the current solution;

[0127] S452: judging a preset optimal value according to the optimal sparrow population information and a preset simulated annealing algorithm to obtain a judgment result, wherein the optimal value is obtained by sorting the fitness values;

[0128] S453: generating an initial solution for the optimal sparrow population information based on the judgment result, and performing random perturbation on the current solution through the initial solution to obtain a new perturbation solution;

[0129] S454: performing calculation based on the new perturbed solution and the initial solution to obtain a function value increment;

[0130] In this step, the function value increment is:

[0131] df=f(S 2 )-f(S 1 ) (14)

[0132] In the above formula (14), df represents the function value increment between the initial solution and the perturbed new solution; f(S 2 ) represents the function value of the perturbed new solution; f(S 1 ) represents the function value of the initial solution.

[0133] S455: judging the acceptance probability of the new disturbance solution based on the function value increment, when the acceptance value of the new disturbance solution is less than a preset acceptance threshold, the new disturbance solution replaces the initial solution to obtain the current optimal value;

[0134] In this step, the acceptance probability expression of the new perturbation solution is:

[0135]

[0136] In the above formula (15), exp represents the exponential function; df represents the function value increment between the initial solution and the perturbed new solution; T represents the parameter controlling the acceptance probability; rand represents a random number uniformly distributed in the interval (0,1);

[0137] S456: The current optimal value is judged according to a preset Meltrov criterion. When the current optimal value does not satisfy the Meltrov criterion, the penalty parameters and kernel parameters in the neural network are optimized by the simulated annealing algorithm to obtain optimized penalty parameters and optimized kernel parameters.

[0138] S46: Assign values ​​based on the optimized penalty parameter and the optimized kernel parameter to obtain a final neural network.

[0139] In order to clarify the specific method of obtaining the final neural network, step S46 includes S461 to S465, which are:

[0140] S461: Inputting the optimized penalty parameter and the optimized kernel parameter into the neural network for linear construction to obtain a linear regression model;

[0141] In this step, the linear regression model is:

[0142] f(x)=w T x+a (16)

[0143] In the above formula (16), f(x) represents the predicted output; w represents the weight vector; w T represents the transpose of the weight vector; x represents the input feature vector; a represents the bias term.

[0144] S462: Assigning values ​​to the neural network according to the linear regression model to obtain an assigned neural network;

[0145] S463: constructing a loss function for the linear regression model, and obtaining a minimized loss function by minimizing the loss function;

[0146] In this step, the loss function is:

[0147]

[0148] In the above formula (17), f(w,b) represents the loss function; N represents the number of samples; y g represents the actual observed value of the g-th sample; x g represents the input feature vector of the g-th sample; w T represents the transpose of the weight vector; w represents the weight vector; a represents the bias term.

[0149] The loss function represents a function of the weight vector w and the bias term a.

[0150] S464: adjusting the partial derivative of the bias term in the linear regression model according to the minimization loss function to obtain an optimized parameter;

[0151] In this step, the partial derivatives of the weight vector w and the bias term a are adjusted to zero according to the minimization loss function, and the partial derivatives of the weight vector w and the bias term a are solved to obtain the optimization parameters;

[0152] S465: Training the assigned neural network based on the training set and the optimization parameters to obtain a final neural network.

[0153] In this step, the assigned neural network is trained based on the training set and the optimization parameters, and the optimization penalty parameters and the optimization kernel parameters are updated to determine whether the updated optimization penalty parameters and the optimization kernel parameters meet the preset end conditions. When the end conditions are met, the final neural network model is generated; when the end conditions are not met, the process will go back to step S461 to reconstruct the neural network until the end conditions are met to generate the final neural network.

[0154] S5: predicting the shear strength of the rock mass joint surface reinforced with anchor rods according to the optimization parameters in the final neural network and the historical monitoring data to obtain a shear strength prediction model, and performing prediction by inputting the anchor rod design data into the shear strength prediction model.

[0155] In this step, the optimized parameters in the final neural network are used to predict the shear strength of the rock joints reinforced with anchor bolts. The parameters of the stepwise optimization model can accurately predict the shear strength of the rock joints reinforced with anchor bolts. The prediction is made based on the historical monitoring data to verify the performance of the model.

[0156] The present invention automatically calculates the predicted value of shear strength of rock joints after anchor reinforcement under specific geological and working conditions. This predicted value provides an important reference for engineers, helping them to more accurately evaluate the reinforcement effect of anchors, optimize design solutions, and ensure the safety and stability of engineering structures. The prediction can dynamically obtain the shear strength of anchor joints and promptly discover potential safety hazards.

[0157] Embodiment 2:

[0158] This embodiment provides a device for predicting the shear strength of an anchor for rock reinforcement, the device comprising:

[0159] An acquisition module is used to acquire anchor design data and anchor historical monitoring data;

[0160] A construction module, used to construct an anchor rod force theory model based on the anchor rod design data and preset anchor rod displacement data, and to obtain a shear strength physical model by constructing the anchor rod force theory model and preset strength criteria;

[0161] To clarify the specific ways to obtain the building blocks, there are:

[0162] A first construction unit is used to construct based on the anchor angle of the anchor rod in the anchor rod design data and the preset anchor rod displacement data to obtain a deformation component, and obtain a theoretical model of anchor rod force by calculating the anchor rod force based on the deformation component;

[0163] The second construction unit is used to determine whether the anchor rod force theory model satisfies a preset strength yield condition, and obtain a strength criterion result when the anchor rod force theory model satisfies the strength yield condition;

[0164] The third construction unit is used to construct a shear strength physical model based on the strength criterion result and the anchor force theoretical model.

[0165] A processing module, used for performing finite difference processing on the anchor design data and the anchor historical monitoring data to obtain information data samples, and obtaining a training set by inputting the information data samples into the shear strength physical model for calculation;

[0166] A training module, used for inputting the training set into a preset neural network for training to obtain training results, and optimizing the training results and the parameters in the neural network by a preset improved sparrow optimization algorithm to obtain a final neural network;

[0167] To clarify the specific ways to obtain the training modules, they are:

[0168] A first acquisition unit is used to acquire an initial sparrow population;

[0169] A first training unit is used to input the training set into a preset neural network for training to obtain a training result;

[0170] The second training unit is used to initialize the training results, the penalty parameters and the kernel parameters in the neural network based on a preset improved sparrow optimization algorithm, and evenly distribute the initial sparrow population through a preset chaotic mapping method to obtain a sparrow population distribution;

[0171] A third training unit is used to screen the initial sparrow population based on the sparrow population distribution to obtain optimal sparrow population information;

[0172] To clarify the specific ways to obtain the third training unit, they are as follows:

[0173] A first acquisition subunit is used to obtain the current iteration number;

[0174] A first training subunit is used to screen the initial sparrow population based on the sparrow population distribution and a preset greedy selection strategy to obtain an excellent sparrow population;

[0175] A second training subunit is used to calculate a fitness function based on the excellent sparrow population to obtain a fitness value;

[0176] A third training subunit is used to determine the current number of iterations according to a preset maximum number of iterations to obtain an iteration result;

[0177] The fourth training subunit is used to select and calculate the position information in the initial sparrow population according to the iteration result and the fitness value to obtain the optimal sparrow population information.

[0178] A fourth training unit is used to optimize the penalty parameters and kernel parameters of the neural network according to the optimal sparrow population information and a preset simulated annealing algorithm to obtain optimized penalty parameters and optimized kernel parameters;

[0179] The fifth training unit is used to assign values ​​based on the optimized penalty parameter and the optimized kernel parameter to obtain a final neural network.

[0180] The prediction model is used to predict the shear strength of the rock joint surface reinforced by anchor rods according to the optimization parameters in the final neural network and the historical monitoring data, and obtain a shear strength prediction model, and the prediction is performed by inputting the anchor rod design data into the shear strength prediction model.

[0181] It should be noted that, regarding the device in the above embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment of the method, and will not be elaborated here.

[0182] Embodiment 3:

[0183] Corresponding to the above method embodiment, this embodiment also provides an anchor shear strength prediction device for rock reinforcement. The anchor shear strength prediction device for rock reinforcement described below and the anchor shear strength prediction method for rock reinforcement described above can be referenced to each other.

[0184] Figure 2 FIG. 8 is a block diagram of an anchor shear strength prediction device 800 for rock reinforcement according to an exemplary embodiment. Figure 2As shown, the anchor shear strength prediction device 800 for rock reinforcement may include: a processor 801, a memory 802. The anchor shear strength prediction device 800 for rock reinforcement may also include one or more of a multimedia component 803, an I / O interface 804, and a communication component 805.

[0185] The processor 801 is used to control the overall operation of the anchor shear strength prediction device 800 for rock reinforcement to complete all or part of the steps in the above-mentioned anchor shear strength prediction method for rock reinforcement. The memory 802 is used to store various types of data to support the operation of the anchor shear strength prediction device 800 for rock reinforcement, and these data may include, for example, instructions for any application or method operating on the anchor shear strength prediction device 800 for rock reinforcement, as well as application-related data, such as contact data, sent and received messages, pictures, audio, video, etc. The memory 802 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, disk or optical disk. The multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touch screen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signal may be further stored in the memory 802 or sent via the communication component 805. The audio component also includes at least one speaker for outputting audio signals. The I / O interface 804 provides an interface between the processor 801 and other interface modules, and the above-mentioned other interface modules can be keyboards, mice, buttons, etc. These buttons can be virtual buttons or physical buttons. The communication component 805 is used for wired or wireless communication between the anchor shear strength prediction device 800 for rock reinforcement and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G or 4G, or a combination of one or more of them, so the corresponding communication component 805 can include: Wi-Fi module, Bluetooth module, NFC module.

[0186] In an exemplary embodiment, the anchor shear strength prediction device 800 for rock reinforcement can be implemented by one or more application specific integrated circuits (Application Specific Integrated Circuit, referred to as ASIC), digital signal processors (Digital Signal Processor, referred to as DSP), digital signal processing devices (Digital Signal Processing Device, referred to as DSPD), programmable logic devices (Programmable Logic Device, referred to as PLD), field programmable gate arrays (Field Programmable Gate Array, referred to as FPGA), controllers, microcontrollers, microprocessors or other electronic components to execute the above-mentioned anchor shear strength prediction method for rock reinforcement.

[0187] Embodiment 4:

[0188] Corresponding to the above method embodiment, a readable storage medium is also provided in this embodiment. The readable storage medium described below and the anchor shear strength prediction method for rock reinforcement described above can refer to each other.

[0189] A readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the method for predicting the shear strength of an anchor used for rock reinforcement in the above method embodiment.

[0190] The readable storage medium may specifically be a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or other readable storage medium that can store program codes.

[0191] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0192] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for predicting the shear strength of an anchor for rock reinforcement, characterized in that: include: Obtain anchor design data and anchor historical monitoring data; Based on the anchor design data and the preset anchor displacement data, a theoretical model of anchor force is constructed, and a shear strength physical model is obtained by constructing the theoretical model of anchor force and the preset strength criterion; Performing finite difference processing on the anchor design data and the anchor historical monitoring data to obtain information data samples, and obtaining a training set by inputting the information data samples into the shear strength physical model for calculation; Inputting the training set into a preset neural network for training to obtain training results, optimizing the training results and the parameters in the neural network using a preset improved sparrow optimization algorithm to obtain a final neural network; The shear strength of the rock mass joint surface reinforced with anchor rods is predicted based on the optimized parameters in the final neural network and the historical monitoring data to obtain a shear strength prediction model, and the prediction is performed by inputting the anchor rod design data into the shear strength prediction model.

2. The method for predicting the shear strength of anchor bolts for rock reinforcement according to claim 1, characterized in that: Based on the anchor design data and the preset anchor displacement data, a theoretical model of anchor force is constructed, and the shear strength physical model is obtained by constructing the theoretical model of anchor force and the preset strength criterion, including: Based on the anchor angle of the anchor rod in the anchor rod design data and the preset anchor rod displacement data, a deformation component is obtained, and the anchor rod force is calculated based on the deformation component to obtain the anchor rod force theoretical model; Determining whether the anchor rod force theory model satisfies a preset strength yield condition, and obtaining a strength criterion result when the anchor rod force theory model satisfies the strength yield condition; A shear strength physical model is constructed based on the strength criterion results and the anchor force theoretical model.

3. The method for predicting the shear strength of anchor bolts for rock reinforcement according to claim 1, characterized in that: The training set is input into a preset neural network for training to obtain a training result, and the training result and the parameters in the neural network are optimized by a preset improved sparrow optimization algorithm to obtain a final neural network, including: Get the initial sparrow population; Inputting the training set into a preset neural network for training to obtain a training result; Initializing the training results, the penalty parameters and the kernel parameters in the neural network based on a preset improved sparrow optimization algorithm, uniformly distributing the initial sparrow population through a preset chaotic mapping method, and obtaining a sparrow population distribution; Screening the initial sparrow population based on the sparrow population distribution to obtain optimal sparrow population information; Optimizing the penalty parameters and kernel parameters of the neural network according to the optimal sparrow population information and a preset simulated annealing algorithm to obtain optimized penalty parameters and optimized kernel parameters; The optimization penalty parameter and the optimization kernel parameter are assigned values ​​to obtain a final neural network.

4. The method for predicting the shear strength of anchor bolts for rock reinforcement according to claim 3, characterized in that: The initial sparrow population is screened based on the sparrow population distribution to obtain optimal sparrow population information, including: Get the current iteration number; The initial sparrow population is screened based on the sparrow population distribution and a preset greedy selection strategy to obtain an excellent sparrow population; Calculating a fitness function based on the excellent sparrow population to obtain a fitness value; The current number of iterations is determined according to a preset maximum number of iterations to obtain an iteration result; The position information in the initial sparrow population is selected and calculated according to the iteration result and the fitness value to obtain the optimal sparrow population information.

5. The method for predicting the shear strength of anchor bolts for rock reinforcement according to claim 4, characterized in that: Optimizing the penalty parameters and kernel parameters of the neural network according to the optimal sparrow population information and a preset simulated annealing algorithm to obtain optimized penalty parameters and optimized kernel parameters, including: Get the current solution; The preset optimal value is judged according to the optimal sparrow population information and the preset simulated annealing algorithm to obtain a judgment result, and the optimal value is obtained by sorting the fitness values; Based on the judgment result, an initial solution is generated for the optimal sparrow population information, and the current solution is randomly disturbed by the initial solution to obtain a new disturbed solution; Perform calculation based on the perturbed new solution and the initial solution to obtain a function value increment; The acceptance probability of the new perturbation solution is judged based on the function value increment, and when the acceptance value of the new perturbation solution is less than a preset acceptance threshold, the new perturbation solution replaces the initial solution to obtain the current optimal value; The current optimal value is judged according to a preset Meltrov criterion. When the current optimal value does not meet the Meltrov criterion, the penalty parameters and kernel parameters in the neural network are optimized by the simulated annealing algorithm to obtain optimized penalty parameters and optimized kernel parameters.

6. The method for predicting the shear strength of anchor bolts for rock reinforcement according to claim 3, characterized in that: Assigning values ​​based on the optimization penalty parameter and the optimization kernel parameter to obtain a final neural network includes: Based on the optimized penalty parameter and the optimized kernel parameter, the optimized penalty parameter and the optimized kernel parameter are input into the neural network for linear construction to obtain a linear regression model; Assigning values ​​to the neural network according to the linear regression model to obtain the assigned neural network; Constructing a loss function for the linear regression model, and obtaining a minimized loss function by minimizing the loss function; Adjusting the partial derivative of the bias term in the linear regression model according to the minimization loss function to obtain an optimized parameter; The assigned neural network is trained based on the training set and the optimization parameters to obtain a final neural network.

7. An anchor shear strength prediction device for rock reinforcement, characterized in that: include: An acquisition module is used to acquire anchor design data and anchor historical monitoring data; A construction module, used to construct an anchor rod force theory model based on the anchor rod design data and preset anchor rod displacement data, and to obtain a shear strength physical model by constructing the anchor rod force theory model and preset strength criteria; A processing module, used for performing finite difference processing on the anchor design data and the anchor historical monitoring data to obtain information data samples, and obtaining a training set by inputting the information data samples into the shear strength physical model for calculation; A training module, used for inputting the training set into a preset neural network for training to obtain training results, and optimizing the training results and the parameters in the neural network by a preset improved sparrow optimization algorithm to obtain a final neural network; The prediction model is used to predict the shear strength of the rock joint surface reinforced by anchor rods according to the optimization parameters in the final neural network and the historical monitoring data, and obtain a shear strength prediction model, and the prediction is performed by inputting the anchor rod design data into the shear strength prediction model.

8. The anchor shear strength prediction device for rock reinforcement according to claim 7, characterized in that: Building blocks, including: A first construction unit is used to construct based on the anchor angle of the anchor rod in the anchor rod design data and the preset anchor rod displacement data to obtain a deformation component, and obtain a theoretical model of anchor rod force by calculating the anchor rod force based on the deformation component; The second construction unit is used to determine whether the anchor rod force theory model satisfies a preset strength yield condition, and obtain a strength criterion result when the anchor rod force theory model satisfies the strength yield condition; The third construction unit is used to construct a shear strength physical model based on the strength criterion result and the anchor force theoretical model.

9. The anchor shear strength prediction device for rock reinforcement according to claim 7, characterized in that: Training modules include: A first acquisition unit is used to acquire an initial sparrow population; A first training unit is used to input the training set into a preset neural network for training to obtain a training result; The second training unit is used to initialize the training results, the penalty parameters and the kernel parameters in the neural network based on a preset improved sparrow optimization algorithm, and evenly distribute the initial sparrow population through a preset chaotic mapping method to obtain a sparrow population distribution; A third training unit is used to screen the initial sparrow population based on the sparrow population distribution to obtain optimal sparrow population information; A fourth training unit is used to optimize the penalty parameters and kernel parameters of the neural network according to the optimal sparrow population information and a preset simulated annealing algorithm to obtain optimized penalty parameters and optimized kernel parameters; The fifth training unit is used to assign values ​​based on the optimized penalty parameter and the optimized kernel parameter to obtain a final neural network.

10. The method for predicting the shear strength of anchor bolts for rock reinforcement according to claim 9, characterized in that: The third training unit includes: A first acquisition subunit is used to obtain the current iteration number; A first training subunit is used to screen the initial sparrow population based on the sparrow population distribution and a preset greedy selection strategy to obtain an excellent sparrow population; A second training subunit is used to calculate the fitness function based on the excellent sparrow population to obtain a fitness value; A third training subunit is used to determine the current number of iterations according to a preset maximum number of iterations to obtain an iteration result; The fourth training subunit is used to select and calculate the position information in the initial sparrow population according to the iteration result and the fitness value to obtain the optimal sparrow population information.

Images