Tunnel smooth blasting effect evaluation method based on the distribution characteristics of remaining holes around the perimeter holes

Abstract

This application relates to the field of underground engineering technology, and specifically includes a method for evaluating the smooth blasting effect of tunnels based on the distribution characteristics of peripheral residual holes. The method includes: obtaining the coordinate information of peripheral residual holes; dividing the heading face into multiple regions according to a preset partition angle, and sequentially collecting the residual hole images in each region based on the coordinate information of the peripheral residual holes; performing image segmentation on the residual hole images and performing a repair operation of linear fitting on the segmentation results after image segmentation to obtain the residual hole traces after linear repair; extracting the characteristic parameters of the residual hole traces, and calculating the actual length and adjacent trace spacing of the residual hole traces based on the characteristic parameters; establishing an evaluation index according to the actual length and adjacent trace spacing of the residual hole traces, and then generating an evaluation result of the blasting effect. It can improve the accuracy and efficiency of tunnel smooth blasting evaluation, reduce tunnel overbreak and underbreak, save construction costs, and promote the intelligent development of tunnel smooth blasting design and construction.

Description

Technical Field

[0001] This application relates to the technical field of underground engineering, and particularly to a method for evaluating the smooth blasting effect of tunnels based on the distribution characteristics of peripheral hole remaining holes. Background Art

[0002] During the tunnel construction process, the quality of smooth blasting excavation greatly affects the construction cost, safety, and progress of the project. At present, on-site, the remaining hole quantity and its proportion of peripheral holes are generally manually visually counted to roughly evaluate the blasting excavation effect of this cycle, and then the blasting parameters of the next cycle are adjusted relying on experience. This process not only easily produces large errors in actual operation, but on the other hand, it is also difficult to accurately evaluate the blasting effect of each remaining hole position, resulting in the inability to quantitatively and finely adjust the blasting parameters. This not only affects the tunnel excavation efficiency and blasting effect, increases the construction cost, but also hinders the intelligent development of blasting design and construction. Summary of the Invention

[0003] This application provides a method for evaluating the smooth blasting effect of tunnels based on the distribution characteristics of peripheral hole remaining holes, so as to improve the accuracy and efficiency of tunnel smooth blasting evaluation, reduce overbreak and underbreak, save construction costs, and promote the intelligent development of tunnel smooth blasting design and construction.

[0004] The method for evaluating the smooth blasting effect of tunnels based on the distribution characteristics of peripheral hole remaining holes provided by the embodiments of this application includes:

[0005] In the tunnel face area after blasting, the coordinate information of peripheral remaining holes is obtained through an image acquisition module fixed on the tunnel central axis;

[0006] The tunnel face is divided into multiple regions according to a preset partition angle, and the remaining hole images in each region are sequentially collected based on the coordinate information of the peripheral remaining holes;

[0007] Semantic segmentation is performed on the remaining hole images, and repair operations such as connected component grouping and linear fitting are performed on the remaining hole images after semantic segmentation to obtain the repaired remaining hole trace lines;

[0008] The characteristic parameters of the remaining hole trace lines are extracted, and the actual length of the remaining hole traces and the adjacent trace spacing are calculated based on the characteristic parameters;

[0009] An evaluation result of the blasting effect is generated according to the actual length of the remaining hole traces and the adjacent trace spacing.

[0010] In the above implementation, by integrating image acquisition, intelligent analysis, and mathematical modeling, the accuracy and efficiency of the evaluation of tunnel smooth blasting effect are significantly improved. Traditional manual evaluation relies on empirical visual statistics, which has problems such as strong subjectivity and fragmented data. In this solution, through the semantic segmentation technology driven by deep learning, the residual hole traces under the complex surrounding rock background can be accurately identified, and the broken traces can be fitted and repaired by means of connected component grouping and gradient descent method, reducing the measurement error of the residual hole length. On this basis, based on the calculation of the actual residual hole length and spacing through calibration ratio conversion, indicators such as half-hole rate, overbreak and underbreak, and trace parallelism can be quantitatively generated, directly relating to the rationality of blasting parameters. It can improve the accuracy and efficiency of tunnel smooth blasting evaluation, reduce overbreak and underbreak, save construction costs, and promote the intelligent development of tunnel smooth blasting design and construction.

[0011] In some embodiments, the preset partition angle is 60 degrees, and the tunnel face is divided into multiple regions according to the preset partition angle. Collecting the residual hole images in each region based on the coordinate information of the peripheral residual holes includes:

[0012] Partitioning at 60 degrees with the tunnel symmetry point as the reference, and the multiple regions are sequentially the left peripheral residual hole area, the arch peripheral residual hole area, and the right peripheral residual hole area.

[0013] In some embodiments, performing semantic segmentation on the residual hole images, performing connected component grouping and linear fitting repair operations on the semantically segmented residual hole images, and obtaining the repaired residual hole trace line includes:

[0014] Performing semantic segmentation on the residual hole images based on the semantic segmentation network U-Net to identify and label the residual hole traces in the residual hole images;

[0015] Processing the labeled residual hole images by using a connected region labeling algorithm to obtain multiple groups of grouped residual hole traces;

[0016] Calculating the slope and intercept of each group of residual hole traces based on the gradient descent method, and linearly repairing the disconnected residual hole traces through threshold control to obtain the repaired residual hole trace line.

[0017] In the above implementation process, the high-precision recognition and repair of residual hole traces can be achieved through the collaboration of multi-stage algorithms. The semantic segmentation network based on U-Net, with its encoder-decoder structure and skip connection mechanism, can effectively distinguish residual hole traces in the complex background of the tunnel face after blasting. By pixel-level segmentation, it extracts the residual hole contour with sub-pixel accuracy, reducing the recognition error. The connected component labeling algorithm logically groups the discrete residual hole regions after segmentation according to the connectivity criterion, clustering the fractured traces belonging to the same residual hole into multiple groups of fitted line segments, which can solve the problem of trace fracture caused by uneven blasting energy or surrounding rock occlusion, providing a structured data basis for subsequent mathematical modeling. Through the gradient descent method, by iteratively optimizing the slope and intercept parameters of the fitted line, combined with the fracture determination rule of double thresholds of slope difference and intercept difference, the fractured traces that meet linear continuity can be automatically connected, restoring the original design axis of the residual hole and improving the calculation accuracy of the length of the residual hole after repair.

[0018] In some embodiments, the labeled residual hole image is processed using the connected component labeling algorithm to obtain multiple groups of residual hole traces after grouping, including:

[0019] In the connected component labeling algorithm, the residual hole image is grouped based on the 4-connectivity criterion or the 8-connectivity criterion to obtain multiple groups of residual hole traces.

[0020] In some embodiments, the slope and intercept of each group of residual hole traces are calculated based on the gradient descent method, and the disconnected residual hole traces are linearly repaired through threshold control to obtain the repaired residual hole trace line, including:

[0021] For the first fitted line and the second fitted line calculated based on the gradient descent method, calculate the slope difference and intercept difference between the first fitted line and the second fitted line. When both the slope difference and intercept difference between the first fitted line and the second fitted line are not greater than the preset threshold, determine that the first fitted line and the second fitted line are traces of the same residual hole.

[0022] In some embodiments, the slope and intercept of each group of residual hole traces are calculated based on the gradient descent method, and the disconnected residual hole traces are linearly repaired through threshold control to obtain the repaired residual hole trace line, and it further includes:

[0023] Extract the abscissas of multiple endpoints of the first fitted line and the second fitted line, sort the absolute values of all abscissas, and obtain the two coordinates with the largest and smallest absolute values in the sequence;

[0024] Connect the endpoints corresponding to the two abscissas into a straight line to obtain the repaired residual hole trace line.

[0025] In some embodiments, the characteristic parameters include the midpoint coordinates of the residual hole trace and the length ratio parameter, and the length ratio is obtained by converting the actual size of the calibration object in the residual hole image.

[0026] In some embodiments, the evaluation results of the blasting effect include the residual hole length ratio and the residual hole inclination rate; wherein, the residual hole length ratio is the ratio of the length of the residual hole trace to the length of the marking line.

[0027] Generating the evaluation result of the blasting effect according to the actual length of the residual hole trace and the adjacent trace spacing includes:

[0028] If the residual hole length ratio is greater than or equal to 90% and the residual hole inclination rate is less than 5%, then the blasting effect level of the residual hole is A.

[0029] If the residual hole length ratio is less than 90% and greater than or equal to 80%, and the residual hole inclination rate is greater than or equal to 5% and less than 10%, then the blasting effect level of the residual hole is B.

[0030] If the residual hole length ratio is less than 80% and greater than or equal to 60%, and the residual hole inclination rate is greater than or equal to 10% and less than 20%, then the blasting effect level of the residual hole is C.

[0031] If the residual hole length ratio is less than 60% and the residual hole inclination rate is greater than 15%, then the blasting effect level of the residual hole is D.

[0032] In some embodiments, generating the evaluation result of the blasting effect according to the actual length of the residual hole trace and the adjacent trace spacing further includes:

[0033] Determining the evaluation result of the smooth blasting of the tunnel according to the ratio of the number of residual holes in each evaluation level.

[0034] In some embodiments, finally determining the evaluation result of the smooth blasting of the tunnel according to the ratio of the number of residual holes in each evaluation level includes:

[0035] In the case where the surrounding rock of the tunnel is hard rock, if the proportion of the number of residual holes of grade A is greater than or equal to 80%, then the evaluation result of the smooth blasting of the tunnel is excellent; if the total proportion of the number of residual holes of grade A and grade B is greater than or equal to 80%, then the evaluation result of the smooth blasting of the tunnel is good; if the total proportion of the number of residual holes of grade A, grade B, and grade C is greater than or equal to 80%, then the evaluation result of the smooth blasting of the tunnel is qualified; if the proportion of the number of residual holes of grade D is greater than or equal to 20%, then the evaluation result of the smooth blasting of the tunnel is unqualified.

[0036] In the case where the surrounding rock of the tunnel is medium rock, if the proportion of the remaining hole quantity of grade A is greater than or equal to 70%, the evaluation result of the smooth blasting effect of the tunnel is excellent; if the total proportion of the remaining hole quantity of grade A and grade B is greater than or equal to 70%, the evaluation result of the smooth blasting effect of the tunnel is good; if the total proportion of the remaining hole quantity of grade A, grade B and grade C is greater than or equal to 70%, the evaluation result of the smooth blasting effect of the tunnel is qualified; if the proportion of the remaining hole quantity of grade D is greater than or equal to 30%, the evaluation result of the smooth blasting effect of the tunnel is unqualified;

[0037] In the case where the surrounding rock of the tunnel is soft rock, if the proportion of the remaining hole quantity of grade A is greater than or equal to 50%, the evaluation result of the smooth blasting effect of the tunnel is excellent; if the total proportion of the remaining hole quantity of grade A and grade B is greater than or equal to 50%, the evaluation result of the smooth blasting effect of the tunnel is good; if the total proportion of the remaining hole quantity of grade A, grade B and grade C is greater than or equal to 50%, the evaluation result of the smooth blasting effect of the tunnel is qualified; if the proportion of the remaining hole quantity of grade D is greater than or equal to 50%, the evaluation result of the smooth blasting effect of the tunnel is unqualified.

[0038] Compared with the prior art, the beneficial effects of the present application are as follows: by integrating image acquisition, intelligent analysis and mathematical modeling, the accuracy and efficiency of the evaluation of the smooth blasting effect of the tunnel are significantly improved. Traditional manual evaluation relies on empirical visual statistics, which has problems of strong subjectivity and fragmented data. In this solution, through the semantic segmentation technology driven by deep learning, the remaining hole traces under the complex surrounding rock background can be accurately identified, and the broken traces can be fitted and repaired by means of connected domain grouping and gradient descent method, reducing the measurement error of the remaining hole length. On this basis, based on the calculation of the actual remaining hole length and spacing by calibration ratio conversion, indexes such as half hole rate, overbreak and underbreak amount, and trace parallelism can be quantitatively generated, directly correlating with the rationality of blasting parameters. It can improve the accuracy and efficiency of the evaluation of the smooth blasting of the tunnel, reduce the overbreak and underbreak of the tunnel, save the construction cost, and promote the intelligent development of the design and construction of the smooth blasting of the tunnel. Description of the Drawings

[0039] Figure 1 It is a step schematic diagram of the method for evaluating the smooth blasting effect of the tunnel based on the distribution characteristics of the remaining holes of the peripheral holes provided by the embodiment of the present application.

[0040] Figure 2 It is a positioning schematic diagram of the image acquisition module provided by the embodiment of the present application.

[0041] Figure 3 It is a schematic diagram of collecting images in segments of the heading face provided by the embodiment of the present application.

[0042] Figure 4 It is a schematic diagram of the on-site image and the corresponding Labelme annotation provided by the embodiment of the present application.

[0043] Figure 5 Schematic diagram of the prediction result of the residual hole trace provided by the embodiment of the present application.

[0044] Figure 6 Schematic diagram of the connectivity criterion provided by the embodiment of the present application.

[0045] Figure 7 Schematic diagram of grouping the prediction map of the residual hole trace based on the connected component labeling algorithm provided by the embodiment of the present application.

[0046] Figure 8 Schematic diagram of the comparison before and after the linear repair of the residual hole trace provided by the embodiment of the present application.

[0047] Figure 9 Schematic diagram of the result after automatic numbering provided by the embodiment of the present application.

[0048] Figure 10 Schematic diagram of calculating the distance between adjacent traces provided by the embodiment of the present application. Detailed implementation manners

[0049] The present application will be further described in detail below in combination with test examples and specific implementation manners. However, it should not be understood that the scope of the above-mentioned subject matter of the present application is limited to the following embodiments. Any technology implemented based on the content of the present application belongs to the scope of protection of the present application.

[0050] In the description of the specific embodiments of the present application, without special explanation, the expression terms of the orientation or positional relationship indicated by "upper", "lower", "left", "right", "center", "inner", "outer", "side", etc. are all based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship when the product / device / device is usually used. These terms of orientation or positional relationship are only for the convenience of describing the solution of the present application or simplifying the description in the specific embodiments, so as to facilitate technicians to quickly understand the solution, rather than indicating or implying that a specific device / component / element must have a specific orientation, or be constructed and operated in a specific positional relationship. Therefore, it should not be construed as a limitation to the present application.

[0051] In the description of the embodiments of the present application, technical terms such as "first" and "second" only distinguish one entity or operation from another entity or operation, and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity, specific order or primary and secondary relationship of the indicated technical features. In the description of the embodiments of the present application, "a plurality of" means two or more, unless otherwise specifically defined.

[0052] References to "embodiments" in this specification mean that the particular features, structures, or characteristics described in connection with the embodiments can be included in at least one embodiment of the present application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0053] Please refer to Figure 1 , Figure 1 which is a schematic diagram of the steps of the tunnel smooth blasting effect evaluation method based on the distribution characteristics of peripheral hole remaining holes provided by the embodiments of the present application. The tunnel smooth blasting effect evaluation method based on the distribution characteristics of peripheral hole remaining holes may include:

[0054] S1. In the tunnel face area after blasting, obtain the coordinate information of the peripheral remaining holes through an image acquisition module fixed on the tunnel central axis.

[0055] Among them, the peripheral holes refer to the remaining holes arranged along the tunnel contour in smooth blasting, and the remaining holes are the part of the holes that are not completely destroyed after blasting. The distribution characteristics of peripheral hole remaining holes refer to the distribution of the holes remaining on the tunnel face after blasting. Tunnel smooth blasting is a controlled blasting technology, aiming to form a smooth tunnel contour surface after blasting, reduce overbreak and underbreak, and improve the quality of the tunnel wall surface. The tunnel face refers to the exposed rock or soil surface at the forefront of the working face of the tunnel under construction during the tunnel excavation process. It is the direct working face of the tunnel tunneling operation and also the main area for construction activities such as blasting, drilling, and support.

[0056] The image acquisition module can be an image acquisition device fixed on the tunnel central axis, such as a high-definition camera, a laser scanner, etc., for obtaining the image and spatial coordinates of the remaining holes on the face after blasting.

[0057] Please refer to Figure 2 , Figure 2 which is a schematic diagram of the positioning of the image acquisition module provided by the embodiments of the present application. Among them, 1 is the image acquisition module, 2 is the tunnel face, and 3 is the tunnel central axis. The way to fix the image acquisition module can be that after the blasting excavation and ventilation and dust removal of the previous cycle of the tunnel face are completed, the image acquisition module is positioned by a total station or a laser locator, and it is ensured that it is placed on the tunnel central axis, with a longitudinal distance of 3 - 5 m from the tunnel face; it can also be installed on a full computer three-arm rock drilling jumbo and positioned through the positioning system of the full computer three-arm rock drilling jumbo, so as to obtain the coordinate information of the peripheral hole remaining holes through positioning.

[0058] S2. Divide the face into multiple regions according to a preset partition angle, and sequentially collect the remaining hole images in each region based on the coordinate information of the peripheral remaining holes.

[0059] Among them, the contour line of the heading face can be partitioned at 60° with the symmetry point O of the tunnel as the reference. The multiple regions are successively the left peripheral remaining hole region, the arch peripheral remaining hole region, and the right peripheral remaining hole region. Please refer to Figure 3 , Figure 3 which is a schematic diagram of the segmented acquisition image of the heading face provided by the embodiment of the present application. In the figure, point O is the symmetry point of the tunnel, 1 is the image acquisition module, 4 is the left peripheral remaining hole region, 5 is the arch peripheral remaining hole region, and 6 is the right peripheral remaining hole region.

[0060] S3. Perform semantic segmentation on the remaining hole image, perform connected domain grouping and linear fitting repair operations on the segmentation result after semantic segmentation, and obtain the repaired remaining hole trace line.

[0061] Exemplarily, S3 may specifically include:

[0062] Perform semantic segmentation on the remaining hole image based on the semantic segmentation network U-Net to identify and label the remaining hole traces in the remaining hole image;

[0063] Process the labeled remaining hole image using the connected region labeling algorithm to obtain multiple groups of remaining hole traces after grouping;

[0064] Calculate the slope and intercept of each group of remaining hole traces based on the gradient descent method, and perform linear repair on the disconnected remaining hole traces through threshold control to obtain the repaired remaining hole trace line.

[0065] Among them, the input remaining hole image can be processed through a convolutional neural network. First, extract the local features of the image through a series of convolutional layers and pooling layers, introduce non-linearity through the activation function, and gradually abstract higher-level features; at the same time, combine the low-level features in the encoder with the high-level features in the decoder through skip connections to retain the position information, and use the upsampling layer in the decoder to restore the feature map to the original image resolution; finally, generate a feature map with the same number of categories through a 1x1 convolutional layer, and apply the Softmax function to classify each pixel, so as to achieve precise classification of each pixel in the image and output a segmentation map with semantic information. Therefore, it is first necessary to make relevant data sets, including the original image and the segmentation annotation map.

[0066] Exemplarily, a high-definition camera can be used to take the on-site image after blasting, and the Labelme annotation tool can be used to perform pixel-level annotation on the remaining hole traces to generate a training data set. Please refer to Figure 4 , Figure 4 which is a schematic diagram of the on-site image and the corresponding Labelme annotation provided by the embodiment of the present application. Among them, the left side is the on-site image, and the right side is the Labelme annotation. In the embodiment of the present application, an artificially placed leveling staff is used as the reference line.

[0067] The semantic segmentation network U-Net is trained on the above constructed training set, and the real residual hole images are predicted. U-Net is a deep learning network architecture specifically for image segmentation, and its structure consists of two main parts: downsampling (encoder) and upsampling (decoder). In the downsampling stage, the network extracts image features through convolution and pooling operations and gradually reduces the resolution; while in the upsampling stage, the network gradually restores the image resolution through deconvolution operations, and at the same time uses skip connections to combine the feature maps of the corresponding layers in the encoder with the feature maps in the decoder to retain detailed information. U-Net can assign class labels to each pixel and is usually trained using cross-entropy loss. Its high efficiency and accuracy have led to remarkable achievements in many fields such as remote sensing image analysis, making it an important benchmark model in this field. Please refer to Figure 5 , Figure 5 which is a schematic diagram of the prediction result of the residual hole trace provided by the embodiment of the present application. Among them, the left side is the on-site image, and the right side is the prediction result of the residual hole trace.

[0068] After automatically identifying the residual hole traces using semantic segmentation technology, in order to further analyze the distribution law and blasting effect of the residual holes, it is necessary to group the identified residual holes. The connected component labeling algorithm is a method in image processing for determining which pixels in an image belong to the same target region. Its basic principle is to merge adjacent pixels or sub-regions into larger regions through an iterative method according to pre-defined connectivity criteria. Specifically, the algorithm usually starts from a seed point, and according to the connectivity criteria, determines whether the pixels around the seed point belong to the same region as the seed point. If so, these pixels are added to the current region and used as new seed points to continue expanding outward until no more pixels can be added to the region.

[0069] Please refer to Figure 6 , Figure 6 which is a schematic diagram of the connectivity criterion provided by the embodiment of the present application. In the embodiment of the present application, the residual hole image can be grouped into connected components based on the 4-connectivity criterion or the 8-connectivity criterion to obtain multiple groups of residual hole traces. Among them, the left side is a schematic of the 4-connectivity criterion, and the right side is a schematic of the 8-connectivity criterion.

[0070] Please refer to Figure 7 , Figure 7 which is a schematic diagram of grouping the residual hole trace prediction map based on the connected component labeling algorithm provided by the embodiment of the present application. Among them, the left side is the residual hole trace prediction map, and the right side is the corresponding grouping and clustering result based on the connected component labeling algorithm. Among them, lines of different colors represent different grouping and clustering results.

[0071] Assume that there are data points and form a data set ( , ). The goal is to calculate the slope and intercept of the linear equation of a single residual hole trace. .

[0072] (1)Set the initial values of the slope and intercept to 0, and set the initial value of the learning rate to 0.1. The learning rate is a hyperparameter that controls the magnitude of each parameter update, thus affecting the speed and stability of the gradient descent algorithm training.

[0073] (2)Define the mean squared error (MSE) as the loss function . The loss function is used to measure the gap between the predicted value and the actual value of the gradient descent algorithm. The calculation formula is as follows:

[0074]

[0075] where, is the loss function, is the slope, is the intercept, is the number of data points in the data set, is the number of a single data in the data set.

[0076] (3)Calculate the partial derivatives of the loss function with respect to the slope and intercept :

[0077] The partial derivative with respect to the slope :

[0078]

[0079] The partial derivative with respect to the intercept :

[0080]

[0081] (4)Update the gradient values of the slope and intercept according to the calculated gradient update values:

[0082]

[0083]

[0084] (5)Repeat the calculation process of steps (2) to (4) until the loss function The change in one iteration process is less than 0.001, or the maximum number of iterations reaches 10,000 times. The above iteration stop condition is formulated based on past engineering experience. After the iteration stops, the updated value of the slope gradient calculated is , and the updated value of the intercept gradient are the slope and intercept for the target solution.

[0085] Based on the gradient descent method, calculate the slope and intercept of each set of residual hole traces, and linearly repair the disconnected residual hole traces through threshold control to obtain the repaired residual hole trace line, which may include:

[0086] For the first fitting line and the second fitting line calculated based on the gradient descent method, calculate the slope difference and intercept difference between the first fitting line and the second fitting line. When both the slope difference and intercept difference between the first fitting line and the second fitting line are not greater than the preset threshold, determine that the first fitting line and the second fitting line are the same residual hole trace.

[0087] Through the calculation results of the above steps, the slope and intercept of each residual hole trace can be obtained. Further, by calculating the slope difference and intercept difference between adjacent residual hole traces, the purpose of checking whether two adjacent residual hole traces are the same residual hole trace after being identified and broken can be achieved. Taking the residual hole trace of the first fitting line and the residual hole trace of the second fitting line as an example, the calculation method is as follows:

[0088]

[0089]

[0090] Among them, is the slope calculated for the residual hole trace of the first fitting line , is the intercept calculated for the residual hole trace of the first fitting line , is the slope calculated for the residual hole trace of the second fitting line , is the intercept calculated for the residual hole trace of the second fitting line , is the slope difference between the two residual hole traces and , is the intercept difference between the two residual hole traces and .

[0091] Exemplarily, when any two residual hole traces and simultaneously satisfy that the slope difference is less than or equal to the threshold 0.05, the intercept difference When it is less than or equal to the threshold value of 0.05, it is considered that the two residual hole traces are the results of the disconnection after the recognition of the same residual hole trace, and the disconnected residual hole traces are linearly repaired.

[0092] Based on the gradient descent method, calculate the slope and intercept of each group of residual hole traces, and linearly repair the disconnected residual hole traces through threshold control to obtain the repaired residual hole trace line. It may also include:

[0093] Extract the abscissas of multiple endpoints of the first fitting line and the second fitting line, sort the absolute values of all abscissas, and obtain the two coordinates with the largest and smallest absolute values in the sequence;

[0094] Connect the endpoints corresponding to the two abscissas into a straight line to obtain the repaired residual hole trace line.

[0095] Exemplarily, two residual hole traces determined to be disconnected after the recognition of the same residual hole trace can be extracted and the abscissas of the 4 endpoints of are sorted from largest to smallest in absolute value, and further the two coordinates with the largest and smallest absolute values in the sequence are extracted. Finally, erase the two residual hole traces and recognized as the same residual hole trace, and connect the endpoints corresponding to the two abscissas extracted to form a new residual hole trace. Please refer to Figure 8 Figure 8 which is a comparison schematic diagram of the residual hole trace before and after the linear repair provided by the embodiment of the present application. Among them, the left side is the prediction diagram of the residual hole trace before the linear repair, and the right side is the prediction diagram of the residual hole trace after the linear repair.

[0096] S4. Extract the characteristic parameters of the residual hole trace line, and calculate the actual length and adjacent trace spacing of the residual hole trace based on the characteristic parameters.

[0097] Among them, the characteristic parameters may include the midpoint coordinates and length ratio parameters of the residual hole trace line, and the length ratio is obtained by converting the actual size of the calibration object in the residual hole image.

[0098] Exemplarily, extract the midpoint coordinates of each repaired residual hole trace, and set the midpoint coordinates of any residual hole trace as , and further automatically sort according to the ordinate of the midpoint of each residual hole trace. The number with the smallest absolute value is numbered 0, and so on. Please refer to Figure 9 Figure 9 which is the schematic diagram of the result after automatic numbering provided by the embodiment of the present application. Among them, the left side is the prediction diagram of the residual hole trace before automatic numbering, and the right side is the prediction diagram of the residual hole trace after automatic numbering. In the prediction diagram of the residual hole trace on the right side, 0-10 are the automatic numbers.

[0099] Calculating the length of the residual hole trace and the distance between the ends of adjacent residual hole traces may specifically include:

[0100] When taking images, a marking line with a certain length is set at the bottom residual hole by using a total station for calibration or by manually placing a leveling rod, and it is numbered 0. Then, according to the length of the marking line, the length of each residual hole trace is calculated proportionally. The length of the marking line is determined by the designed cyclic excavation advance. For example, if the cyclic excavation advance is 3.5 meters, the set length of the marking line is 3.5 meters. Then, the length ratio of each residual hole trace in the residual hole picture relative to the marking line can be calculated. The calculation method is as follows:

[0101]

[0102]

[0103] Among them, is the length of the marking line in the residual hole picture, is the length of any one residual hole trace in the residual hole picture, is the residual hole length ratio, that is, the ratio of the length of any one residual hole trace to the length of the marking line, is the actual length of any one residual hole trace calculated.

[0104] Calculating the distance between the ends of adjacent residual hole traces with numbers: Calculate the linear equation of any one residual hole trace numbered in the residual hole picture as follows:

[0105]

[0106] Among them, A, B, and C are the coefficients of the linear equation of the residual hole trace ; , are the coordinate values of the endpoints of the residual hole trace. Calculate the residual hole trace in the residual hole picture that is adjacent to the residual hole trace numbered and is numbered +1. The endpoint with the smaller absolute value of the abscissa has coordinates , and the endpoint with the larger absolute value of the abscissa has coordinates .

[0107] Calculate the residual hole trace +1, the shortest distance between the endpoint with the smaller absolute value of the abscissa and the residual hole trace , and the calculation formula is as follows:

[0108]

[0109] Calculate the residual hole trace The endpoint with a larger absolute value of the abscissa, and the shortest distance to the residual hole trace is , and the calculation formula is as follows:

[0110]

[0111] Please refer to Figure 10 , Figure 10 , which is a schematic diagram for calculating the adjacent trace spacing provided by the embodiment of the present application. By determining and , the adjacent trace spacing can be determined.

[0112] S5. Generate an evaluation result of the blasting effect according to the actual length of the residual hole trace and the adjacent trace spacing.

[0113] Among them, the evaluation result of the blasting effect may include the residual hole length ratio and the residual hole inclination rate ; among them, the residual hole length ratio is the ratio of the actual length of the residual hole trace to the length of the marking line. Among them, can be determined according to the steps in the above description content, and the residual hole inclination rate can be calculated based on the following formula:

[0114]

[0115] Evaluate and grade the blasting effect of each residual hole, calculate the evaluation index of the th ( starting from 1) residual hole, and then evaluate and grade according to Table 1. Please refer to Table 1, and Table 1 is the evaluation and grading table of the blasting effect.

[0116] Table 1 Evaluation and grading table of blasting effect

[0117]

[0118] If the residual hole length ratio is greater than or equal to 90% and the residual hole inclination rate is less than 5%, the grade of the blasting effect of the residual hole is A.

[0119] If the residual hole length ratio is less than 90% and greater than or equal to 80%, and the residual hole inclination rate is greater than or equal to 5% and less than 10%, the grade of the blasting effect of the residual hole is B.

[0120] If the residual hole length ratio is less than 80% and greater than or equal to 60%, and the residual hole inclination rate is greater than or equal to 10% and less than 20%, the grade of the blasting effect of the residual hole is C.

[0121] If the residual hole length ratio is less than 60% and the residual hole inclination rate is greater than 15%, the grade of the blasting effect of the residual hole is D.

[0122] Generate an evaluation result of the blasting effect based on the actual length of the remaining hole traces and the spacing between adjacent traces. It may also include:

[0123] Determine the evaluation result of the smooth blasting effect of the tunnel according to the ratio of the number of remaining holes in each evaluation grade.

[0124] Exemplarily, count the evaluation grades of all remaining holes, and then according to the ratio of the number of remaining holes in different evaluation grades to the total number of remaining holes under different surrounding rock conditions f Evaluate the overall blasting effect of the heading face, and there is:

[0125]

[0126] Among them, 、 、 、 are the ratios of the number of remaining holes with blasting effect grades A, B, C, and D respectively. When the surrounding rock of the tunnel is hard rock, if the ratio of the number of remaining holes of grade A is greater than or equal to 80%, the evaluation result of the smooth blasting effect of the tunnel is excellent; if the total ratio of the number of remaining holes of grade A and grade B is greater than or equal to 80%, the evaluation result of the smooth blasting effect of the tunnel is good; if the total ratio of the number of remaining holes of grade A, grade B, and grade C is greater than or equal to 80%, the evaluation result of the smooth blasting effect of the tunnel is qualified; if the ratio of the number of remaining holes of grade D is greater than or equal to 20%, the evaluation result of the smooth blasting effect of the tunnel is unqualified.

[0127] When the surrounding rock of the tunnel is medium rock, if the ratio of the number of remaining holes of grade A is greater than or equal to 70%, the evaluation result of the smooth blasting effect of the tunnel is excellent; if the total ratio of the number of remaining holes of grade A and grade B is greater than or equal to 70%, the evaluation result of the smooth blasting effect of the tunnel is good; if the total ratio of the number of remaining holes of grade A, grade B, and grade C is greater than or equal to 70%, the evaluation result of the smooth blasting effect of the tunnel is qualified; if the ratio of the number of remaining holes of grade D is greater than or equal to 30%, the evaluation result of the smooth blasting effect of the tunnel is unqualified.

[0128] When the surrounding rock of the tunnel is soft rock, if the ratio of the number of remaining holes of grade A is greater than or equal to 50%, the evaluation result of the smooth blasting effect of the tunnel is excellent; if the total ratio of the number of remaining holes of grade A and grade B is greater than or equal to 50%, the evaluation result of the smooth blasting effect of the tunnel is good; if the total ratio of the number of remaining holes of grade A, grade B, and grade C is greater than or equal to 50%, the evaluation result of the smooth blasting effect of the tunnel is qualified; if the ratio of the number of remaining holes of grade D is greater than or equal to 50%, the evaluation result of the smooth blasting effect of the tunnel is unqualified.

[0129] Through the above calculations, the evaluation of the overall blasting effect of the tunnel can be completed from the grading evaluation of each remaining hole. Refer to Table 2 for the grading of the evaluation of the overall blasting effect of the tunnel. Please refer to Table 2, which is the grading table of the evaluation results of the overall blasting effect.

[0130] Table 2 Grading Table of Evaluation Results of Overall Blasting Effect

[0131]

[0132] In the above implementation process, by integrating image acquisition, intelligent analysis and mathematical modeling, the accuracy and efficiency of the evaluation of the smooth blasting effect of the tunnel can be significantly improved. Traditional manual evaluation relies on empirical visual statistics, which has problems such as strong subjectivity and fragmented data. In this solution, through the semantic segmentation technology driven by deep learning, the remaining hole traces under the complex surrounding rock background can be accurately identified, and the broken traces can be fitted and repaired by means of connected domain grouping and gradient descent method to reduce the measurement error of the remaining hole length. On this basis, based on the calculation of the actual remaining hole length and spacing by calibration ratio conversion, indicators such as semi-hole rate, overbreak and underbreak amount, and trace parallelism can be quantitatively generated, which are directly related to the rationality of blasting parameters. It can improve the accuracy and efficiency of the evaluation of the smooth blasting of the tunnel, reduce the overbreak and underbreak of the tunnel, save construction costs, and promote the intelligent development of the smooth blasting design and construction of the tunnel.

[0133] The above-described embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for evaluating the smooth blasting effect of a tunnel based on the distribution characteristics of remaining holes in peripheral holes, characterized in that Including: In the area of the tunnel face after blasting, obtain the coordinate information of the surrounding remaining holes through an image acquisition module fixed on the tunnel central axis; Divide the tunnel face into multiple areas according to a preset partition angle, and sequentially collect the remaining hole images in each area based on the coordinate information of the surrounding remaining holes; Perform semantic segmentation on the remaining hole images, perform connected component grouping and linear fitting repair operations on the remaining hole images after semantic segmentation to obtain the repaired remaining hole trace line; Extract the characteristic parameters of the remaining hole trace line, and calculate the actual length of the remaining hole trace and the adjacent trace spacing based on the characteristic parameters; Generate an evaluation result of the blasting effect according to the actual length of the remaining hole trace and the adjacent trace spacing; Among them, the performing semantic segmentation on the remaining hole images, performing connected component grouping and linear fitting repair operations on the remaining hole images after semantic segmentation to obtain the repaired remaining hole trace line includes: Perform semantic segmentation on the remaining hole images based on the semantic segmentation network U-Net to identify and label the remaining hole traces in the remaining hole images; Process the labeled remaining hole images using a connected region labeling algorithm to obtain multiple groups of remaining hole traces after grouping; Calculate the slope and intercept of each group of remaining hole traces based on the gradient descent method, and perform linear repair on the disconnected remaining hole traces through threshold control to obtain the repaired remaining hole trace line.

2. The method according to claim 1, characterized in that, The preset partition angle is 60 degrees. The dividing the tunnel face into multiple areas according to the preset partition angle and sequentially collecting the remaining hole images in each area based on the coordinate information of the surrounding remaining holes includes: Partition at 60 degrees with the tunnel symmetry point as the reference. The multiple areas are sequentially the left-side surrounding remaining hole area, the arch surrounding remaining hole area, and the right-side surrounding remaining hole area.

3. The method according to claim 1, wherein The processing the labeled remaining hole images using a connected region labeling algorithm to obtain multiple groups of remaining hole traces after grouping includes: In the connected component labeling algorithm, perform connected component grouping on the remaining hole images based on the 4-connected criterion or the 8-connected criterion to obtain multiple groups of remaining hole traces.

4. The method according to claim 1, wherein The calculating the slope and intercept of each group of remaining hole traces based on the gradient descent method, and performing linear repair on the disconnected remaining hole traces through threshold control to obtain the repaired remaining hole trace line includes: For the first fitting line and the second fitting line calculated based on the gradient descent method, calculate the slope difference and intercept difference between the first fitting line and the second fitting line. When both the slope difference and intercept difference between the first fitting line and the second fitting line are not greater than a preset threshold, determine the first fitting line and the second fitting line as the same remaining hole trace.

5. The method according to claim 4, characterized in that, The calculating the slope and intercept of each group of remaining hole traces based on the gradient descent method, and performing linear repair on the disconnected remaining hole traces through threshold control to obtain the repaired remaining hole trace line further includes: Extract the abscissas of multiple endpoints of the first fitting line and the second fitting line, sort the absolute values of all abscissas, and obtain the two coordinates with the largest and smallest absolute values in the sequence; Connect the endpoints corresponding to the two abscissas into a straight line to obtain the repaired remaining hole trace line.

6. The method according to claim 1, wherein The characteristic parameters include the midpoint coordinates of the residual hole trace and the length ratio parameter, and the length ratio is obtained by converting the actual size of the calibration object in the residual hole image.

7. The method according to claim 1, wherein The evaluation indexes of the blasting effect include the residual hole length ratio and the residual hole inclination rate; among them, the residual hole length ratio is the ratio of the length of the residual hole trace to the length of the marking line; The evaluation results obtained according to the residual hole length ratio and the residual hole inclination rate include: If the residual hole length ratio is greater than or equal to 90% and the residual hole inclination rate is less than 5%, the blasting effect level of the residual hole is A; If the residual hole length ratio is less than 90% and greater than or equal to 80%, and the residual hole inclination rate is greater than or equal to 5% and less than 10%, the blasting effect level of the residual hole is B; If the residual hole length ratio is less than 80% and greater than or equal to 60%, and the residual hole inclination rate is greater than or equal to 10% and less than 20%, the blasting effect level of the residual hole is C; If the residual hole length ratio is less than 60% and the residual hole inclination rate is greater than 15%, the blasting effect level of the residual hole is D.

8. The method according to claim 7, wherein The evaluation result of the blasting effect generated according to the actual length of the residual hole trace and the adjacent trace spacing further includes: Finally determining the evaluation result of the smooth blasting effect of the tunnel according to the ratio of the number of residual holes in each evaluation level.

9. The method according to claim 8, wherein Finally determining the evaluation result of the smooth blasting effect of the tunnel according to the ratio of the number of residual holes in each evaluation level includes: In the case where the surrounding rock of the tunnel is hard rock, if the proportion of the number of residual holes of grade A is greater than or equal to 80%, the evaluation result of the smooth blasting effect of the tunnel is excellent; if the total proportion of the number of residual holes of grade A and grade B is greater than or equal to 80%, the evaluation result of the smooth blasting effect of the tunnel is good; if the total proportion of the number of residual holes of grade A, grade B and grade C is greater than or equal to 80%, the evaluation result of the smooth blasting effect of the tunnel is qualified; if the proportion of the number of residual holes of grade D is greater than or equal to 20%, the evaluation result of the smooth blasting effect of the tunnel is unqualified; In the case where the surrounding rock of the tunnel is medium rock, if the proportion of the number of residual holes of grade A is greater than or equal to 70%, the evaluation result of the smooth blasting effect of the tunnel is excellent; if the total proportion of the number of residual holes of grade A and grade B is greater than or equal to 70%, the evaluation result of the smooth blasting effect of the tunnel is good; if the total proportion of the number of residual holes of grade A, grade B and grade C is greater than or equal to 70%, the evaluation result of the smooth blasting effect of the tunnel is qualified; if the proportion of the number of residual holes of grade D is greater than or equal to 30%, the evaluation result of the smooth blasting effect of the tunnel is unqualified; In the case where the surrounding rock of the tunnel is soft rock, if the proportion of the remaining hole quantity of grade A is greater than or equal to 50%, the evaluation result of the smooth blasting effect of the tunnel is excellent; if the total proportion of the remaining hole quantity of grade A and grade B is greater than or equal to 50%, the evaluation result of the smooth blasting effect of the tunnel is good; if the total proportion of the remaining hole quantity of grade A, grade B and grade C is greater than or equal to 50%, the evaluation result of the smooth blasting effect of the tunnel is qualified; if the proportion of the remaining hole quantity of grade D is greater than or equal to 50%, the evaluation result of the smooth blasting effect of the tunnel is unqualified.

Images