Anchorage Bolt Hole Positioning Method and System Based on Neural Network and Grey Wolf Optimization Algorithm

Abstract

The present invention relates to the technical field of machine vision and image processing, and specifically provides an anchor bolt hole positioning method and system based on a neural network and a grey wolf optimization algorithm, including: obtaining a two-dimensional image of the tunnel inner wall; evenly dividing the two-dimensional image of the tunnel inner wall to obtain a number of square regions, and establishing a coordinate system; determining whether the preset anchor bolt holes of the tunnel catenary avoid adverse conditions such as cracks and water seepage points. If so, determining the current anchor bolt hole position; if not, randomly generating artificial wolves in the square region with a coordinate system by using the grey wolf optimization algorithm; establishing a fitness function, sorting the wolf pack, and obtaining the lead wolf, scout wolf, and brute wolf; constructing a feedforward neural network, and combining the intelligent behaviors of the lead wolf, scout wolf, and brute wolf to obtain the position of the lead wolf and the fitness function value of the lead wolf; wherein, the position of the lead wolf is the optimal position of the current anchor bolt hole, and the positioning of the anchor bolt holes of the tunnel catenary is completed. The present invention provides a scientific solution for the positioning of the anchor bolt holes of the tunnel catenary.

Description

Technical Field

[0001] The present invention belongs to the technical field of machine vision and image processing, and particularly relates to a method and system for locating anchor bolt holes based on a neural network and a grey wolf optimization algorithm. Background Art

[0002] With the rapid development of railway electrification construction, the optimization of the target position of the vacant position of the catenary anchor bolt hole in the railway tunnel has become a key link to ensure the stability and construction accuracy of the catenary. The catenary anchor bolt hole in the tunnel refers to the hole drilled on the top wall of the tunnel for fixing the suspension device of the catenary system. Whether its layout is reasonable directly affects the geometric shape, stability and operation safety of the catenary. However, traditional target hole position positioning methods, such as manual measurement and rule-based optimization methods, often rely on the experience of engineers for design when determining the optimal position of the anchor bolt hole. Although it can meet the basic construction requirements, in the face of complex tunnel structures, such as irregular cross-sections, complex stress distributions, structural joints, cracks, water seepage points, etc., there are many problems such as low construction efficiency, increased construction costs, poor adaptability to complex constraints, and low positioning accuracy of the hole position.

[0003] Among the existing methods for positioning the position of the catenary anchor bolt hole in the tunnel, mainly manual design or rule-driven optimization methods are adopted, and there are the following disadvantages:

[0004] (1) High consumption of human and material resources: The manual design method needs to rely on the experience of engineers to optimize the design of the anchor bolt hole position. In the face of complex tunnel top structural joints, cracks, water seepage points and irregular mechanical constraints, etc., the design and construction efficiency is low, and it is difficult to obtain the global optimal solution.

[0005] (2) Poor adaptability and easy to fall into local optimal solutions: In a complex search space, when the individuals of the grey wolf optimization algorithm are overly focused around the position of the alpha wolf, it is easy to fall into local optimal solutions. In the case of a more complex search space with multiple local optimal solutions, this phenomenon is more obvious.

[0006] (3) Low search efficiency: The step size of the grey wolf optimization algorithm is usually fixed or lacks an effective adaptive mechanism. In the initial stage of the search, the fixed step size may not be able to quickly cover the entire search space, resulting in low exploration efficiency. In the later stage of the search, the fixed step size may not be able to perform fine local search, making the convergence speed slow.

[0007] Machine vision has been widely used in environmental detection due to its advantages such as high efficiency, high precision, and strong adaptability. Machine vision is used to collect images of the top wall of a railway tunnel and perform image segmentation to complete the environmental modeling of the tunnel top. The Grey Wolf Optimizer (GWO) has received extensive attention in recent years due to its strong global search ability and flexibility. The Grey Wolf Optimizer is an intelligent optimization algorithm inspired by the hunting behavior of grey wolf groups. It completes search and hunting by simulating the division of labor and cooperation among the lead wolf, scout wolf, and striker wolf. However, the original Grey Wolf Optimizer not only has low efficiency in dealing with target search problems but also is prone to falling into local optima and other problems. Therefore, combined with the construction requirements of positioning the target position of the catenary anchor holes in the railway tunnel, the present invention proposes a positioning method and system for tunnel catenary anchor holes based on the fusion of neural network and Grey Wolf Optimizer, so as to achieve high-efficiency positioning of the optimal position of the optimal anchor hole.

[0008] Although the Grey Wolf Optimizer has achieved certain results in the field of target search and optimization, its performance is still significantly affected by algorithm parameters. Especially parameters such as the wandering step size, sprint step size, and attack step size, whose settings during the search process will directly affect the search ability, convergence speed, and distribution quality of the solution of the algorithm. However, the traditional step size parameter settings are mostly fixed values or manual adjustments based on experience, lacking self-adaptability and being difficult to adapt to the complex and changeable tunnel top environment.

[0009] In recent years, the development of neural network technology has provided a new solution for the parameter adjustment of optimization algorithms. Neural networks have powerful non-linear modeling capabilities and can achieve adaptive adjustment of parameters by learning the feature laws in data. Therefore, combining neural networks with the wolf pack optimization algorithm and dynamically adjusting the step size parameters through training the model can significantly improve the search efficiency and accuracy of the wolf pack optimization algorithm. Summary of the Invention

[0010] To solve the problems existing in the prior art, the present invention provides an anchor hole positioning method and system based on neural network and Grey Wolf Optimizer, which combines the adaptive learning ability of the neural network and can dynamically adjust the step size to improve the search performance of the optimal position of the anchor hole in the complex tunnel top environment.

[0011] To achieve the above object, the present invention provides the following solutions:

[0012] An anchor hole positioning method based on neural network and Grey Wolf Optimizer, comprising the following steps:

[0013] Scanning the inner wall of the tunnel based on machine vision technology to obtain a two-dimensional image of the inner wall of the tunnel;

[0014] Uniformly dividing the two-dimensional image of the inner wall of the tunnel to obtain a number of square regions and establish a coordinate system;

[0015] Based on a square area with a coordinate system, determine whether the preset anchor bolt holes of the tunnel catenary avoid cracks and water seepage points. If so, determine the current anchor bolt hole position; if not, use the Grey Wolf Optimization Algorithm to randomly generate artificial wolves in the square area with a coordinate system; where the position of each artificial wolf represents an anchor bolt hole position.

[0016] Based on the preset objective function and the current position of the artificial wolf, establish a fitness function.

[0017] Based on the fitness function, conduct wolf pack ranking to obtain the lead wolf, scout wolf, and brute wolf.

[0018] Construct a feedforward neural network, combine the intelligent behaviors of the lead wolf, scout wolf, and brute wolf, and obtain the position of the lead wolf and the fitness function value of the lead wolf; where the position of the lead wolf is the optimal position of the current anchor bolt hole, and the positioning of the anchor bolt holes of the tunnel catenary is completed.

[0019] Preferably, the method for obtaining the fitness function includes:

[0020] Based on the coordinate system, obtain the position coordinates of the artificial wolf, and based on the position coordinates, obtain the position of the artificial wolf. The calculation formula for the position of the artificial wolf is as follows:

[0021] ,

[0022] where, represents the th position coordinate of the artificial wolf , and are the upper and lower bounds of the spatial search range respectively, and the position of each artificial wolf ;

[0023] Based on the position of the artificial wolf and the weight coefficients, obtain the fitness function; where the weight coefficients include a first weight coefficient for adjusting the influence ratio of the benefit in the fitness, a second weight coefficient for adjusting the proportion of the energy loss in the fitness, and a third weight coefficient for adjusting the proportion of the risk cost; the expression of the fitness function is as follows:

[0024] ,

[0025] where, represents the objective function value, measuring the optimization effect of the current position ; is the first weight coefficient, represents the distance between the current position of the artificial wolf and the preset position ; The second weight coefficient, is represented as a set of dangerous areas, is an indicator function, and is the third weight coefficient.

[0026] Preferably, the intelligent behaviors include the alpha wolf generation rule, the scout wolf wandering behavior, and the alpha wolf summoning the brute wolf behavior.

[0027] Preferably, the method for obtaining the position of the alpha wolf and the fitness function value of the alpha wolf includes:

[0028] Selecting an artificial wolf with an objective function value that meets the preset optimal requirements as the alpha wolf;

[0029] Taking the direction in which the objective function value is greater than the objective function value at the current position as the wandering direction; based on the wandering direction and the preset wandering step size, the scout wolf repeatedly executes the wandering behavior until the objective function value sensed by the scout wolf is greater than the objective function value of the current alpha wolf or the scout wolf reaches the preset number of wandering times, then updating the scout wolf as the alpha wolf and initiating a summoning behavior;

[0030] In the alpha wolf summoning the brute wolf behavior, the alpha wolf guides the brute wolf to rush towards the current position of the alpha wolf at a preset rushing step size, and obtains a feature vector based on the obstacle density, the space scale, and the distance between the brute wolf and the alpha wolf on the rushing path of the brute wolf;

[0031] Taking the feature vector as the input vector of the feedforward neural network to obtain the predicted rushing step size;

[0032] The brute wolf rushes based on the predicted rushing step size. When the objective function value sensed by the brute wolf is greater than the objective function value of the current alpha wolf or the brute wolf reaches the preset rushing condition, updating the brute wolf as the alpha wolf and obtaining the position of the updated alpha wolf and the fitness function of the alpha wolf.

[0033] Preferably, the feedforward neural network includes an input layer, a hidden layer, and an output layer;

[0034] Among them, the input layer includes 3 nodes, corresponding to the obstacle density, the space scale, and the distance between the brute wolf and the alpha wolf respectively;

[0035] The hidden layer includes 2 layers, and each layer contains 4 neurons;

[0036] The output layer includes 1 node for outputting the predicted rushing step size.

[0037] The present invention also provides an anchor bolt hole positioning system based on a neural network and a grey wolf optimization algorithm for implementing the method, including:

[0038] A two-dimensional image acquisition module for scanning the inner wall of the tunnel based on machine vision technology to obtain a two-dimensional image of the inner wall of the tunnel;

[0039] An image segmentation module, which is used to evenly segment the two-dimensional image of the tunnel inner wall, obtain a number of square regions, and establish a coordinate system;

[0040] An artificial wolf generation module, which is used to determine whether the preset anchor bolt holes of the tunnel catenary avoid cracks and water seepage points based on the square regions with a coordinate system established. If so, the current anchor bolt hole position is determined; if not, the grey wolf optimization algorithm is used to randomly generate artificial wolves in the square regions with a coordinate system established; among them, the position of each artificial wolf represents an anchor bolt hole position;

[0041] A fitness function construction module, which is used to establish a fitness function based on a preset objective function and the current artificial wolf position;

[0042] A wolf pack sorting module, which is used to sort the wolf pack based on the fitness function to obtain the lead wolf, scout wolves, and strong wolves;

[0043] An optimal anchor bolt hole positioning module, which is used to construct a feedforward neural network, and combine the intelligent behaviors of the lead wolf, scout wolves, and strong wolves to obtain the position of the lead wolf and the fitness function value of the lead wolf; among them, the position of the lead wolf is the optimal position of the current anchor bolt hole, and the positioning of the anchor bolt holes of the tunnel catenary is completed.

[0044] Preferably, the fitness function construction module includes:

[0045] An artificial wolf position acquisition unit, which is used to obtain the position coordinates of the artificial wolf based on the coordinate system, and obtain the position of the artificial wolf based on the position coordinates. The calculation formula of the artificial wolf position is as follows:

[0046] ,

[0047] where, represents the th position coordinate of the artificial wolf , and are the upper and lower bounds of the spatial search range respectively, and the position of each artificial wolf ;

[0048] A fitness function calculation unit, which is used to obtain the fitness function based on the position of the artificial wolf and the weight coefficients; among them, the weight coefficients include a first weight coefficient for adjusting the influence ratio of the benefit in the fitness, a second weight coefficient for adjusting the proportion of the energy loss in the fitness, and a third weight coefficient for adjusting the proportion of the danger cost; the expression of the fitness function is as follows:

[0049] ,

[0050] where, denotes the objective function value, measuring the optimization effect at the current position of the current position; is the first weight coefficient, denotes the current position of the artificial wolf and the preset position distance; The second weight coefficient, denotes the set of dangerous areas, is the indicator function, is the third weight coefficient.

[0051] Preferably, the intelligent behavior includes the alpha wolf generation rule, the scout wolf wandering behavior, and the alpha wolf summoning the brute wolf behavior.

[0052] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0053] (1) Saving manpower and material resources: Based on the image feature information of the tunnel top collected and analyzed by the present invention, the image information is converted into accurate mathematical models and algorithm parameters, and multiple potential optimization schemes for the anchor bolt hole positions are generated quickly and accurately, and the global optimal solution is selected from them, thereby greatly improving the design and construction efficiency and the positioning accuracy, and reducing the errors and risks caused by insufficient human experience.

[0054] (2) Improving adaptability: The present invention can fully adapt to the complex and changeable structural environment of the tunnel top, effectively preventing the population from converging to the local optimal solution prematurely, so as to ensure that the true global optimal anchor bolt hole layout can be stably explored in the complex search space.

[0055] (3) Improving the search efficiency: By means of the dynamic output of the step size by the neural network, the defects of the traditional algorithm in different search stages are effectively solved, the flexibility and accuracy of the search process are enhanced, the search efficiency and the convergence speed of the algorithm in the process of positioning the tunnel anchor bolt hole positions are greatly improved, and the high quality and reliability of the positioning result are guaranteed.

[0056] Through the above improvements, the present invention can significantly improve the global search ability, the uniformity of the solution distribution and the adaptability of the algorithm when solving complex optimization problems, and provide an efficient optimization tool for fields such as engineering design and resource allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions of the present invention, the drawings required to be used in the embodiments are briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can also obtain other drawings according to these drawings without creative efforts.

[0058] Figure 1Flowchart of the anchor bolt hole positioning method based on neural network and grey wolf optimization algorithm in the embodiments of the present invention;

[0059] Figure 2 Schematic diagram of randomly generating artificial wolves in the embodiments of the present invention. Specific implementation manners

[0060] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part rather than all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0061] To make the above objects, features and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific implementation manners.

[0062] Embodiment 1

[0063] As Figure 1 shown, the anchor bolt hole positioning method based on neural network and grey wolf optimization algorithm includes the following steps:

[0064] S1: Scan the inner wall of the tunnel based on machine vision technology to obtain a two-dimensional image of the inner wall of the tunnel. The characteristic information of the image includes structural joints, cracks, water seepage points, etc.

[0065] S2: Uniformly divide the two-dimensional image of the inner wall of the tunnel to obtain a number of square regions and establish a coordinate system. In this embodiment, the high-precision two-dimensional image obtained by pre-scanning the inner wall of the tunnel using machine vision technology is uniformly divided into a number of square regions and a coordinate system is established, as shown in the appendix Figure 2 shown. The side length of the square region is , which is equal to the diameter of the anchor bolt hole. The hollow circle represents the position of the anchor bolt hole, and the black curve represents the cracks existing on the inner wall of the tunnel.

[0066] S3: Based on the square region with a coordinate system established, determine whether the preset anchor bolt hole of the tunnel catenary avoids bad conditions such as cracks and water seepage points. If so, determine the current anchor bolt hole position; if not, use the grey wolf optimization algorithm to randomly generate artificial wolves in the square region with a coordinate system established; where the position of each artificial wolf represents the position of an anchor bolt hole, as shown by the grey rhombus in the appendix Figure 2 shown.

[0067] S4: Establish a fitness function based on the preset objective function and the current position of the artificial wolf.

[0068] A further implementation manner lies in that the method for obtaining the fitness function includes:

[0069] Based on the coordinate system, the position coordinates of the artificial wolf are obtained, and based on the position coordinates, the position of the artificial wolf is obtained. The calculation formula for the position of the artificial wolf is as follows:

[0070]

[0071] where represents the position coordinates of the th artificial wolf , and are the upper and lower bounds of the spatial search range respectively, and the position of each artificial wolf .

[0072] Based on the position of the artificial wolf and the weight coefficients, a fitness function is obtained; among them, the weight coefficients include a first weight coefficient for adjusting the influence ratio of the benefit in the fitness, a second weight coefficient for adjusting the proportion of the energy loss in the fitness, and a third weight coefficient for adjusting the proportion of the risk cost; the expression of the fitness function is as follows:

[0073]

[0074] where represents the objective function value, measuring the optimization effect of the current position ; is the first weight coefficient, represents the distance between the current position of the artificial wolf and the preset position ; the second weight coefficient, represents the set of dangerous areas, is an indicator function, if is within the dangerous area, it takes the value of 1, otherwise it is 0. is the third weight coefficient.

[0075] S5: Based on the fitness function, the wolf pack is sorted to obtain the alpha wolf, beta wolf, and delta wolf; specifically, according to the fitness value function. The wolf pack is sorted, and three wolves are selected: the alpha wolf (the leader, with the highest fitness value), the beta wolf (the deputy leader, with the second highest fitness value), and the delta wolf (the explorer, with the third highest fitness value).

[0076] S6: A feedforward neural network is constructed, and by combining the intelligent behaviors of the alpha wolf, beta wolf, and delta wolf, the position of the alpha wolf and the fitness function value of the alpha wolf are obtained; among them, the position of the alpha wolf is the optimal position of the current anchor bolt hole, and the positioning of the tunnel catenary anchor bolt hole is completed.

[0077] In a further embodiment, the intelligent behaviors include the alpha wolf generation rule, the scout wolf wandering behavior, and the alpha wolf summoning the brute wolf behavior.

[0078] In a further embodiment, the method for obtaining the position of the alpha wolf and the fitness function value of the alpha wolf includes:

[0079] Select an artificial wolf with an objective function value that meets the preset optimal requirements as the alpha wolf; specifically, in the initial solution space, by calculating the objective function value, find the individual with the optimal objective function value as the alpha wolf. During the loop, compare the value of the best wolf objective function after each loop with the value of the leader in the previous generation. If it is better than the previous one, then adjust the position of the leader. If there is more than one, then randomly select one of them as the new leader. The new leader no longer participates in the three intelligent actions but immediately starts the next loop until it is replaced by a more powerful artificial wolf.

[0080] Take the direction where the objective function value is greater than the objective function value at the current position as the wandering direction; based on the wandering direction and the preset wandering step size, the scout wolf repeatedly performs the wandering behavior until the objective function value sensed by the scout wolf is greater than the objective function value of the current alpha wolf or the scout wolf reaches the preset number of wandering times, then update the scout wolf as the alpha wolf and initiate the summoning behavior; specifically, the scout wolf searches for prey. Regard the artificial wolf with the best performance except the alpha wolf in the solution space as the scout wolf, then search for prey in the solution space and randomly select one as the proportionality factor α of the integer scout wolf. Select the direction with the strongest smell (the largest objective function value) and greater than the smell concentration at the current position and take a step forward. As shown in Equation (3): Where

[0081]

[0082] represents the position coordinates of the scout wolf after update, represents the current position coordinates of the th scout wolf. is the wandering step size, h is the wandering direction, . Repeat the above wandering behavior until the smell concentration of a certain scout wolf , at this time this scout wolf replaces the alpha wolf and initiates the summoning behavior, or reaches the number of wandering times .

[0083] In the alpha wolf summoning the brute wolf behavior, the alpha wolf guides the brute wolf to rush towards the current position of the alpha wolf at the preset rushing step size, and based on the obstacle density, spatial scale of the rushing path between the brute wolf and the alpha wolf, and the distance between the brute wolf and the alpha wolf, obtain the feature vector.

[0084] Use the feature vector as the input vector of the feedforward neural network to obtain the predicted raiding step length. Specifically, specifically, after the alpha wolf is updated, it means that a higher objective function value has been found. At this moment, the alpha wolf can start to command the task and guide the beta wolves to quickly move towards the place where the alpha wolf is currently located with a step length for raiding, as shown in formula (4):

[0085]

[0086] where is the position of the alpha wolf of the k-th generation population in space, is the raiding step length. represents the current position of the beta wolves, represents the position of the beta wolves after update. The beta wolves do not simply raid. During the raiding process, they will sense the odor concentration of the prey around them.

[0087] The obstacle density of the raiding paths of the beta wolves and the alpha wolf , the space scale , the distance between the beta wolves and the alpha wolf, as shown in formula (5):

[0088]

[0089] These input features are normalized to form the input vector of the neural network, as shown in formula (6):

[0090]

[0091] The beta wolves raid based on the predicted raiding step length. When the objective function value sensed by the beta wolves is greater than the objective function value of the current alpha wolf or the beta wolves reach the preset raiding conditions, update the beta wolves to alpha wolves to obtain the position of the updated alpha wolf and the fitness function of the alpha wolf.

[0092] A further implementation is that the feedforward neural network includes an input layer, a hidden layer, and an output layer.

[0093] Among them, the input layer includes 3 nodes, corresponding to the obstacle density, the space scale, and the distance between the beta wolves and the alpha wolf, that is, corresponding to the features .

[0094] The hidden layer includes 2 layers, each layer contains 4 neurons, and the activation function is , if the input is greater than 0, then the output is itself, otherwise the output is 0;

[0095] The output layer includes 1 node, which is used to output the predicted raiding step length . The target step length and the predicted step length The mean squared error (MSE), as shown in Equation (7):

[0096]

[0097] Since the neural network is used to dynamically output the running step length of the fierce wolf, every time the fierce wolf runs, according to the current state Predict the appropriate step length through the trained neural network , and update the position, as shown in Equation (8):

[0098]

[0099] During the running of the fierce wolf, the smell concentration of the prey is detected , triggering the alpha wolf update mechanism, thus , making this fierce wolf become the new generation of alpha wolf and replacing the original alpha wolf to conduct the command. If the fierce wolf does not perceive that the smell concentration of the prey is greater than that of the current alpha wolf during the running, it will continue to run until the maximum number of iterations is reached.

[0100] When the alpha wolf meets the optimization accuracy requirement or the algorithm reaches the maximum number of iterations, output the position of the alpha wolf and the output value of the fitness function of the alpha wolf, and use the position of the alpha wolf as the optimal position of the anchor bolt hole.

[0101] In summary, according to the fitness function Evaluate the starting points of the artificial wolves, and select the artificial wolf with the largest fitness function value as the alpha wolf . The other artificial wolves are regarded as scout wolves, responsible for implementing the wandering behavior until the objective function value perceived by one of the scout wolves exceeds the objective function value of the alpha wolf , or the set maximum number of wandering times is reached . After that, the alpha wolf performs a summoning behavior to guide the fierce wolf to run towards the direction of the alpha wolf. Take the running step length as a dynamic adjustment parameter, and use the neural network for training to make it adaptively adjust according to the environmental characteristics and the state of the wolf pack.

[0102] Embodiment 2

[0103] The present invention also provides an anchor bolt hole positioning system based on a neural network and a grey wolf optimization algorithm for implementing the method, including:[[]]

[0104] A two-dimensional image acquisition module, used to scan the inner wall of the tunnel based on machine vision technology to obtain a two-dimensional image of the inner wall of the tunnel;

[0105] An image segmentation module, used to evenly segment the two-dimensional image of the inner wall of the tunnel to obtain a number of square regions and establish a coordinate system;

[0106] An artificial wolf generation module, which is used to determine whether the preset anchor bolt holes of the tunnel catenary avoid adverse conditions such as cracks and water seepage points based on a square area with a coordinate system. If so, the position of the current anchor bolt hole is determined; if not, the grey wolf optimization algorithm is used to randomly generate artificial wolves in the square area with a coordinate system; where the position of each artificial wolf represents the position of an anchor bolt hole;

[0107] A fitness function construction module, which is used to establish a fitness function based on a preset objective function and the current position of the artificial wolf;

[0108] A wolf pack sorting module, which is used to sort the wolf pack based on the fitness function to obtain the lead wolf, scouting wolves, and fierce wolves;

[0109] An optimal positioning module for anchor bolt holes, which is used to construct a feedforward neural network and combine the intelligent behaviors of the lead wolf, scouting wolves, and fierce wolves to obtain the position of the lead wolf and the fitness function value of the lead wolf; where the position of the lead wolf is the optimal position of the current anchor bolt hole, and the positioning of the anchor bolt holes of the tunnel catenary is completed.

[0110] A further implementation manner lies in that the fitness function construction module includes:

[0111] An artificial wolf position acquisition unit, which is used to obtain the position coordinates of the artificial wolf based on the coordinate system and obtain the position of the artificial wolf based on the position coordinates. The calculation formula for the position of the artificial wolf is as follows:

[0112] ,

[0113] where, represents the position coordinates of the th artificial wolf , and are respectively the upper and lower bounds of the spatial search range, and the position of each artificial wolf ;

[0114] A fitness function calculation unit, which is used to obtain the fitness function based on the position of the artificial wolf and the weight coefficients; where the weight coefficients include a first weight coefficient for adjusting the influence ratio of the benefit in the fitness, a second weight coefficient for adjusting the proportion of the energy loss in the fitness, and a third weight coefficient for adjusting the proportion of the risk cost; the expression of the fitness function is as follows:

[0115] ,

[0116] where, represents the objective function value, which measures the optimization effect of the current position ; is the first weight coefficient, represents the current position of the artificial wolf The distance from a preset position ; The second weight coefficient is represented as a set of dangerous areas is an indicator function is the third weight coefficient.

[0117] A further embodiment lies in that the intelligent behavior includes the alpha wolf generation rule, the scout wolf wandering behavior, and the alpha wolf summoning the brute wolf behavior.

[0118] The above-described embodiments are only descriptions of the preferred embodiments of the present invention, and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. An anchor bolt hole positioning method based on a neural network and a grey wolf optimization algorithm, characterized in that It includes the following steps: Scan the inner wall of the tunnel based on machine vision technology to obtain a two-dimensional image of the inner wall of the tunnel; Uniformly divide the two-dimensional image of the inner wall of the tunnel to obtain a number of square regions, and establish a coordinate system; Based on the square regions with a coordinate system established, determine whether the preset anchor holes of the tunnel catenary avoid cracks and water seepage points. If so, determine the current anchor hole position; if not, use the grey wolf optimization algorithm to randomly generate artificial wolves in the square regions with a coordinate system established. Among them, the position of each artificial wolf represents an anchor hole position; Based on the preset objective function and the current position of the artificial wolf, establish a fitness function; Sort the wolf pack based on the fitness function to obtain the lead wolf, scout wolf, and fierce wolf; Construct a feedforward neural network, combine the intelligent behaviors of the lead wolf, scout wolf, and fierce wolf to obtain the position of the lead wolf and the fitness function value of the lead wolf. Among them, the position of the lead wolf is the optimal position of the current anchor hole, and the positioning of the anchor holes of the tunnel catenary is completed.

2. The method according to claim 1, wherein The method for obtaining the fitness function includes: Based on the coordinate system, obtain the position coordinates of the artificial wolf, and based on the position coordinates, obtain the position of the artificial wolf. The calculation formula for the position of the artificial wolf is as follows: ， Among them, represents the position coordinates of the th artificial wolf , and are the upper and lower bounds of the spatial search range respectively, and the position of each artificial wolf Based on the position of the artificial wolf and the weight coefficients, obtain the fitness function. Among them, the weight coefficients include a first weight coefficient for adjusting the influence ratio of the benefit in the fitness, a second weight coefficient for adjusting the proportion of the energy loss in the fitness, and a third weight coefficient for adjusting the proportion of the risk cost. The expression of the fitness function is as follows: ， Among them, represents the objective function value, which measures the optimization effect of the current position ; is the first weight coefficient, represents the distance between the current artificial wolf position and the preset position ; is the second weight coefficient, represents the set of dangerous areas, is the indicator function, is the third weight coefficient.

3. The method according to claim 2, wherein The intelligent behaviors include the lead wolf generation rule, the scout wolf wandering behavior, and the lead wolf summoning the fierce wolf behavior.

4. The method according to claim 3, wherein The method for obtaining the position of the lead wolf and the fitness function value of the lead wolf includes: Select an artificial wolf with an objective function value that meets the preset optimal requirements as the lead wolf; Take the direction with an objective function value greater than the objective function value of the current position as the wandering direction. Based on the wandering direction and the preset wandering step length, the scout wolf repeatedly executes the wandering behavior until the objective function value sensed by the scout wolf is greater than the objective function value of the current lead wolf or the scout wolf reaches the preset wandering times, then update the scout wolf as the lead wolf and initiate a summoning behavior; In the lead wolf summoning the fierce wolf behavior, the lead wolf guides the fierce wolf to rush towards the current position of the lead wolf at a preset rushing step length, and based on the obstacle density, spatial scale of the rushing path between the fierce wolf and the lead wolf, and the distance between the fierce wolf and the lead wolf, obtain a feature vector; Take the feature vector as the input vector of the feedforward neural network to obtain a predicted rushing step length; The fierce wolf rushes based on the predicted rushing step length. When the objective function value sensed by the fierce wolf is greater than the objective function value of the current lead wolf or the fierce wolf reaches the preset rushing condition, update the fierce wolf as the lead wolf to obtain the position of the updated lead wolf and the fitness function of the lead wolf.

5. The method according to claim 4, characterized in that The feedforward neural network includes an input layer, a hidden layer, and an output layer; Among them, the input layer includes 3 nodes, corresponding to the obstacle density, the spatial scale, and the distance between the fierce wolf and the lead wolf respectively; The hidden layer includes 2 layers, and each layer contains 4 neurons; The output layer includes 1 node for outputting the predicted rushing step length.

6. An anchor bolt hole positioning system based on a neural network and a grey wolf optimization algorithm for implementing the method according to any one of claims 1-5, characterized in that, It includes: A two-dimensional image acquisition module, which is used to scan the inner wall of the tunnel based on machine vision technology to obtain a two-dimensional image of the inner wall of the tunnel; An image segmentation module, which is used to evenly segment the two-dimensional image of the inner wall of the tunnel to obtain a number of square regions and establish a coordinate system; An artificial wolf generation module, which is used to judge whether the preset anchor bolt holes of the tunnel catenary avoid cracks and water seepage points based on the square regions with a coordinate system established. If so, determine the current anchor bolt hole position; if not, randomly generate artificial wolves in the square regions with a coordinate system established by using the grey wolf optimization algorithm; where the position of each artificial wolf represents an anchor bolt hole position; A fitness function construction module, which is used to establish a fitness function based on a preset objective function and the current artificial wolf position; A wolf pack sorting module, which is used to sort the wolf pack based on the fitness function to obtain the lead wolf, scout wolf and fierce wolf; An optimal anchor bolt hole positioning module, which is used to construct a feedforward neural network and combine the intelligent behaviors of the lead wolf, scout wolf and fierce wolf to obtain the position of the lead wolf and the fitness function value of the lead wolf; where the position of the lead wolf is the optimal position of the current anchor bolt hole, and the positioning of the anchor bolt holes of the tunnel catenary is completed.

7. The system according to claim 6, characterized in that, The fitness function construction module includes: An artificial wolf position acquisition unit, which is used to obtain the position coordinates of the artificial wolf based on the coordinate system, and obtain the position of the artificial wolf based on the position coordinates. The calculation formula of the artificial wolf position is as follows: ， Among them, represents the position coordinates of the th artificial wolf, and are the upper and lower bounds of the spatial search range respectively, and the position of each artificial wolf A fitness function calculation unit, which is used to obtain the fitness function based on the position of the artificial wolf and the weight coefficients; where the weight coefficients include a first weight coefficient for adjusting the influence proportion of the benefit in the fitness, a second weight coefficient for adjusting the proportion of the energy loss in the fitness, and a third weight coefficient for adjusting the proportion of the danger cost; the expression of the fitness function is as follows: ， Among them, represents the objective function value, which measures the optimization effect of the current position ; is the first weight coefficient, represents the distance between the current artificial wolf position and the preset position ; is the second weight coefficient, represents the set of dangerous areas, is the indicator function, is the third weight coefficient.

8. The system according to claim 6, wherein The intelligent behaviors include the lead wolf generation rule, the scout wolf wandering behavior and the lead wolf summoning fierce wolf behavior.

Images