JP7049902B2 - Blasting optimization method in mountain tunnels - Google Patents

Description

The present invention optimizes blasting in mountain tunnels so that the appropriateness of blasting conditions is judged not by relying on the skill, experience, and judgment of skilled workers, but based on the data of scattered shavings obtained by three-dimensional shape measurement. Regarding the method.

Conventionally, in the construction of mountain tunnels by blasting excavation method, there are bench cut methods such as full section excavation method, upper half section advanced method, long bench cut method, short bench cut method, mini bench cut method, etc. depending on the excavation method. However, in any case, the tunnel construction procedure involves the steps of drilling, charging, blasting, slipping, primary spraying, construction of support work, secondary spraying, and rock bolt placement in order and step by step. As a result, excavation is being carried out. It should be noted that the construction of support works, spraying and rock bolt placement may be omitted or the order may be changed depending on the ground conditions, tunnel construction method, excavation method, etc.

The blasting is a method of loading explosives such as dynamite into the drilling and exploding it to cut down the ground, but this is a detailed number and position of drilling at the planning stage based on the geological survey conducted in advance. Blasting conditions such as spacing, drilling length, and charge amount are determined, but in the implementation work, the blasting conditions become inappropriate due to geological non-uniformity, discontinuity, etc., and as shown in FIG. In many cases, it is a weak or overloaded drug.

Normally, blasting and slipping work are dangerous work that is off limits, so staff often do not look at the situation after blasting in detail, and as a result, the personal judgment of the blasting expert is prioritized in determining the appropriateness of blasting. It was often done. It is often necessary to rely on the skills, experience, and judgment of blasting skilled workers, but in order to deal with the problem of shortage of skilled workers in recent years and in the future, optimization of blasting conditions based on objective data and measured values. It has become important to try to achieve this.

For example, in Patent Document 1 below, when a predetermined hole is drilled in a rock mass by a rock drill, the drilling data collecting unit provided in the rock drill collects the drilling data of the hole and drills the hole. The drilling data of each hole was input to the calculation unit mounted on the rock drill to calculate the drilling energy value of each hole, and these drilling energy values and the storage unit of the calculation unit were stored. Compare with the rock strength or rock classification data to confirm the appropriateness of the preset charge amount of the explosive used for each hole, and correct the charge amount according to the collected drilling energy value to the bedrock. A method for optimizing the rock rupture work based on the drilling energy evaluation value of the rock drill has been proposed so that the amount of explosive used for rupture is set to an appropriate amount corresponding to the rock strength or the rock classification data.

Further, in Patent Document 2 below, a grid pattern is used as the rupture pattern, parallel core punching is standard as the core punching, and the explosive is used as a dense filling agent at the bottom of the drilling hole so that the drilling diameter is full, and smooth. Design data for designing each rupture by searching the basic database based on the search data in the control rupture expert system that assists rupture by the rupture method using blasting as standard and using MS electric thunder tube for core removal. By the tunnel basic data input unit to be obtained, the resistance line length / hole spacing design section that obtains the resistance line length and hole spacing by dividing into general holes and outer peripheral holes based on the design data, and the resistance line length / hole spacing design section. A rupture pattern design unit that obtains a rupture pattern by a predetermined procedure based on information on the required resistance wire length and hole spacing and data in the basic database, information on the rupture pattern, the resistance wire length and the hole spacing, and the basics. Based on the data in the database, the charge amount calculation unit that clearly separates the bottom charge and the column charge to obtain the charge amount, the information on the rupture pattern, the resistance line length and the hole spacing, the charge amount, and the above. Based on the data of the basic database, the countermeasure engineering department that obtains the optimum rupture conditions for vibration and noise countermeasures by fuzzy inference, the information on the amount of charge used in the rupture, the information on the on-site measurement, and the information on this rupture. An expert system for control breaking has been proposed, which is equipped with a construction / measurement data input unit that saves information on problems and countermeasures as a file.

Japanese Unexamined Patent Publication No. 7-208060 Japanese Patent No. 284002

The method according to Patent Document 1 evaluates the bedrock strength characteristics based on the drilling data by utilizing the fact that the energy required for drilling increases proportionally according to the strength of the bedrock to be drilled. By setting an appropriate amount of explosives according to the rock strength characteristics, it is possible to eliminate weak charges and overcharges and improve excavation accuracy. However, even if the presence of a fragile part can be partially detected during drilling, it is difficult to predict how the vulnerable part will affect blasting, and the impact will be evaluated when joints are present. There were problems such as difficulty.

In the method according to Patent Document 2, in the construction / measurement data input unit, information on the amount of charge used in blasting, information on on-site measurement, and information on problems and countermeasures in this blasting are saved as a file. By doing so, it is possible to feed back the blasting specifications from the next time onward, but the above-mentioned "defect in this blasting" is not specifically shown, and it is clear what kind of feedback will be given based on what. do not.

Therefore, the main subject of the present invention is not to rely on the skill, experience, and judgment of skilled workers to judge the appropriateness of blasting conditions, but to objectively judge the appropriateness of blasting conditions based on the results of three-dimensional shape measurement such as the scattering / accumulation state of slips and the excavation shape of the face. The purpose is to provide a method for optimizing blasting in mountain tunnels.

In order to solve the above-mentioned problems, the present invention according to claim 1 is a method for optimizing blasting in a mountain tunnel in which appropriateness of blasting conditions is determined by three-dimensional shape measurement before and after blasting.
Before blasting, the first step of measuring the three-dimensional shape of the planned scattering range of the face and the slag,
After the blasting, the second step of measuring the three-dimensional shape of the accumulated state of the scattered debris and measuring the excavated shape of the face after the debris is removed.
The third step to accumulate data that quantifies the relationship between the index value of the blasting situation set based on the scattering / accumulation state of the shavings and the excavation shape of the face and the quality of the blasting based on the judgment of the blasting expert.
Based on the accumulated data, it is analyzed by a statistical processing method and / or a method by artificial intelligence, and from the stage when a certain amount of data is obtained, the quality of blasting is judged based on the index value of the blasting situation. Provided is a method for optimizing blasting in a mountain tunnel, which comprises a fourth step of modifying the blasting conditions as necessary.

In the invention according to claim 1, the three-dimensional shape measurement is performed on the planned scattering range of the face and the debris before blasting (first step), and the three-dimensional shape measurement is performed on the accumulated state of the debris scattered after the blasting. After blasting, three-dimensional shape measurement is performed on the excavated shape of the face (second step).

Next, by performing image subtraction processing between the three-dimensional shape data obtained in the first step and the three-dimensional shape data obtained in the second step, the three-dimensional shape of the accumulated state of the scattered debris and the excavated shape of the face can be obtained. Try to get. Then, based on these, the index value of the blasting situation is arbitrarily set, and the relationship between the quality of the blasting and the quality of the blasting based on the judgment of the blasting skilled worker is quantified and accumulated (third step).

Next, based on the above-mentioned accumulated data, analysis is performed by statistical processing method and / or learning by artificial intelligence, and from the stage when a certain amount of data is obtained, the quality of blasting is based on the index value of the blasting situation. The judgment is made and the blasting conditions are corrected as necessary (4th step).

Through the above process, the appropriateness of blasting conditions, which used to rely on the skills, experience, and judgment of skilled workers, is objectively judged based on the results of three-dimensional shape measurement such as the scattering / accumulation state of shavings and the excavation shape of the face. You will get it.

According to the second aspect of the present invention, there is provided the method for optimizing bursting in a mountain tunnel according to claim 1, wherein the three-dimensional shape measurement is performed by a digital camera, a video camera, a three-dimensional scanner or a stereo camera.

In the invention according to claim 2, as a specific method for measuring the three-dimensional shape, a digital camera, a digital video camera, a three-dimensional scanner, or a stereo camera is used. Even with a two-dimensional photographing device such as a digital camera or a video camera, image data (two-dimensional images) taken at a plurality of points can be made three-dimensional by performing soft processing. At the same time, by having the coordinate data, it is possible to quantitatively grasp the scattering / accumulation state of the slip and the excavation shape of the face. By applying image processing to the acquired 3D point cloud data, the 3D scanner can grasp the state of scattering / accumulation of slips and the excavation shape of the face. In the stereo camera, multiple image pickup devices such as CCD cameras whose optical axes are parallel or whose intersections are known are arranged, and the coordinates of the target points are specified by the principle of triangulation, so that the scattering / accumulation state and the face of the slips are specified. It is possible to grasp the excavation shape of. In recent years, even 3D measuring devices such as high-speed, high-precision 3D scanners and stereo cameras are commercially available from optical equipment manufacturers at relatively low prices and can be easily obtained.

As the present invention according to claim 3, the method for optimizing bursting in a mountain tunnel according to claim 1 is provided, wherein the three-dimensional shape measurement is performed by a small unmanned aircraft equipped with a digital camera, a video camera, a three-dimensional scanner or a stereo camera. Will be done.

In the invention according to claim 3, the three-dimensional shape measurement is performed by a small unmanned aircraft equipped with a digital camera, a video camera, a three-dimensional scanner, or a stereo camera. Since blasting and slipping work are dangerous work, it becomes possible to perform measurement more safely and quickly by performing three-dimensional shape measurement using a small unmanned aerial vehicle.

As described in detail above, according to the present invention, the appropriateness of blasting conditions is not determined by the skill, experience, and judgment of skilled workers, but is based on the results of three-dimensional shape measurement such as the scattering / deposition state of slips and the excavation shape of the face. You will be able to do it objectively.

It is a tunnel vertical sectional view which shows the construction procedure of a mountain tunnel. It is a flow chart of the blasting optimization method which concerns on this invention. It is a tunnel vertical sectional view which shows the 3D shape measurement procedure in 1st step. It is a tunnel vertical sectional view which shows the 3D shape measurement procedure in 2nd step. An example of the index value of the blasting situation is shown, (A) is a cross-sectional view of the tunnel, and (B) is a vertical cross-sectional view of the tunnel. It is a conceptual diagram of a neural network of a method by artificial intelligence. It is a figure which shows the calculation processing concept in the intermediate layer unit of a neural network. It is a perspective view of a small unmanned aerial vehicle (drone) equipped with a three-dimensional measuring device 4. It is a figure which shows the blasting situation, (A) is a weak charge, (B) is appropriate, and (C) is an example of an overcharge.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

As shown in FIG. 1, when excavating a mountain tunnel by the blasting method, heavy equipment for tunnel construction such as a drill jumbo 1, a spraying machine 2, and a wheel loader 3 is arranged in the vicinity of the face S, for example, all with an auxiliary bench. After drilling and charging rock bolts in the upper and lower halves by the cross-section method and the upper half-section method, the upper and lower halves are blasted at once with explosives, and then slipped out → scraped / hit. → Primary sprayed concrete → Steel support → Secondary sprayed concrete → Rock bolt placement procedure, excavation is carried out every predetermined section length (every cycle). The excavation speed is approximately 4-5 cycles per day.

In the rock bolt drilling / charging work in the above series of excavation steps, the blasting conditions (number of drilling, drilling position / interval, drilling angle, loading amount, simultaneous charge amount, number of detonator stages, explosive type, etc.) are set as the ground. Due to the discontinuity of the mountain, it is important to make sequential corrections while considering the hardness, stratification, and node of the geology in order to realize efficient excavation.

In the present invention, the appropriateness of blasting conditions is judged not by relying on the skill, experience, and judgment of skilled workers, but based on the results of three-dimensional shape measurement such as the state of scattering and accumulation of shavings and the excavation shape of the face. Is. Although this determination may be made for each cycle, it should be made at a frequency of about once in a plurality of cycles (for example, once / one day to a plurality of days) in consideration of the balance with the progress speed of excavation. You may.

The blasting optimization method according to the present invention will be specifically described in detail. The blasting optimization method according to the present invention is performed by the procedure of the first step to the fourth step described later. Hereinafter, the details will be described in step order.

<First step>
First, before blasting, a three-dimensional shape measurement is performed on the planned scattering range of the face and the slip.

Specifically, as shown in FIG. 3A, a 3D measuring device 4 such as a 3D scanner or a stereo camera is installed in front of the face to measure the 3D shape of the face S, and FIG. 3 As shown in (B), the three-dimensional measuring device 4 is installed at a position near the top of the mine, and the range including the planned scattering range of the slip (the range where the slip is expected to be scattered by rupture) is three-dimensional. It is possible to measure the shape.

The three-dimensional shape measurement can be performed by using a two-dimensional photographing device such as a digital camera or a video camera in addition to the three-dimensional measuring device 4 such as the three-dimensional scanner or the stereo camera. Here, the three-dimensional shape measurement of the face shape is performed in order to quantitatively grasp the excavation shape of the face by comparison with the three-dimensional shape measurement of the face after blasting (image subtraction processing), and the scatter of slips. The reason why the planned range is measured in 3D shape is to quantitatively grasp the scattering / accumulation state of the slip due to blasting by comparing with the data obtained by measuring the same range in 3D shape after blasting (image subtraction processing).

In recent years, even two-dimensional photography equipment such as digital cameras and video cameras can be easily made three-dimensional by performing soft processing on image data (two-dimensional images) taken at multiple points. Can be done. At the same time, by having the coordinate data, it is possible to quantitatively grasp the scattering / accumulation state of the slip and the excavation shape of the face. Coordinate data can be assigned by shooting with two reference points whose coordinates are known in the captured image, or by making the coordinates of the installation point of the photographing device known. ..

The three-dimensional scanner (laser radar device) sets the horizontal range angle and the vertical range angle so that the object is included inside, and evenly arranges an arbitrary number of actual measurement points within the rectangular range specified by the horizontal range angle and the vertical range angle. Then, by measuring the distance from the 3D scanner to each measured point, the 3D data (3D point cloud data) of the object is acquired from the distance data and the angle data, and the acquired 3D point cloud data. A three-dimensional model is generated by performing various processes to specify the shape from.

In the stereo camera, image pickup devices such as a plurality of CCD cameras whose optical axes are parallel or whose crossing angles are known are arranged, and the coordinates of the target point are specified by the principle of triangulation. These are used as measuring instruments for shape specification or coordinate specification in a three-dimensional space.

The two-dimensional photographing device or the three-dimensional measuring device 4 may be set on an installation table suspended from the ceiling surface, or may be installed on the floor surface in the mine by a tripod. You may hold it in your hand and shoot. The data measured by the two-dimensional photographing device and the three-dimensional measuring device 4 are input to the personal computer 5.

<Second step>
When the three-dimensional shape measurement is completed, blasting is performed. Then, after the completion of blasting, a three-dimensional shape measurement is performed on the accumulated state of the scattered debris.

Specifically, as shown in FIG. 4 (A), a three-dimensional measuring device 4 is installed near the top of the mine to measure the accumulated state of scattered dust, and then the wheel. After carrying out the slip by the loader 3 or the dump truck, as shown in FIG. 4B, the three-dimensional measuring device 4 is installed in front of the face to measure the three-dimensional shape of the face S.

<Third step>
In the third step, the relationship between the index value of the blasting condition set based on the scattering / accumulation state of the shavings and the excavation shape of the face and the quality of the blasting based on the judgment of the blasting expert is accumulated.

Specifically, if the scattering / deposition state of the slip and the excavation shape of the face are obtained from the three-dimensional shape measurement data measured in the first step and the second step, the blasting status is determined based on these data. Set the index value. That is, by performing image subtraction processing on the three-dimensional shape measurement data of the planned scattering range of the shavings performed before blasting and the three-dimensional shape measurement data of the accumulated state of the scattered shavings performed after blasting, the scattered shavings. 3D shape data is obtained only for the deposition state of. Similarly, for the face, by performing image subtraction processing on the three-dimensional shape measurement data of the face performed before blasting and the three-dimensional shape measurement data of the face performed after blasting, the tertiary shape of the face excavation is performed. Get the original shape data.

Then, once the scattered / accumulated state of the shavings and the excavated shape of the face are obtained, the index value of the blasting situation is arbitrarily set based on these.

Various factors can be considered as the index value of the blasting situation, and an example thereof is shown in FIG. 5 and Table 1 below. The index values are the measured values obtained from the three-dimensional shape measurement data before and after blasting and the calculated values obtained by performing four arithmetic operations on these measured values.

On the other hand, we will accumulate data that quantifies the relationship between the quality of blasting and the quality of blasting based on the judgment of skilled blasting workers regarding this blasting excavation. The quality of blasting based on the judgment of the blasting skilled worker is determined by interviewing the blasting skilled worker and conducting a questionnaire after the blasting, for example, as shown in Table 2 below.

<4th step>
In the fourth step, based on the above-mentioned accumulated data, analysis is performed by learning by a statistical processing method and / or artificial intelligence, and from the stage when a certain amount of data is obtained, based on the index value of the blasting situation. The quality of the blasting is judged, and the blasting conditions are corrected as necessary (see the right column of Table 1 for the procedure for correcting the next blasting).

As mentioned above, the number of accumulated data that quantifies the relationship between the index value of blasting status and the quality of blasting based on the judgment of blasting skilled workers regarding this blasting excavation is 50 or more in order to improve the judgment accuracy. The number is preferably 100 or more.

The relationship between the two can be analyzed by statistical processing methods or artificial intelligence methods.

In the former statistical processing method, the index value serves as an explanatory variable, and the pass / fail judgment of blasting based on the judgment of a blasting expert becomes the objective variable. Both data are plotted on the coordinates, and simple regression analysis or multiple regression analysis is performed. , Derivation of the correlation equation and the correlation coefficient. Index values that are considered to have a high correlation will be adopted, and index values that are considered to have a low correlation will not be adopted. In order to find index values with higher correlation, it is also necessary to review and try sequential index values.

In the latter case, the artificial intelligence method is performed by a neural network method that imitates the human brain. This neural network models the functions of neural circuits (neurons) in the human brain and their connections, and is formed by a network in which a large number of units are connected. The feature is that while conventional computers perform sequential series information processing and logical inference, neural networks perform more intuitive inference than parallel distributed information processing, and teachers. It can be mentioned that learning is performed so that the correct answer can be obtained by giving data, that is, self-organization by learning is performed.

Specifically, as shown in FIG. 6, in the neural network, as input factors in the input layer, L: stepping stone distance (m), H: stepping stone height (m), h: stepping stone height difference (m). ), T: Circumferential extra-digging thickness (m), D: Tunnel axial extra-drilling length (m), V: Slip amount (blurring volume) (m3), α = H / L (slip flatness), β = There are 8 items of T * D * 100 (excessive excavation coefficient), and the output layer is evaluated on a 5-point scale of excellent and inferior judgment (1: excellent, 2: good, 3: ordinary, 4: inferior, 5: bad). And.

Synapses connecting each neuron element (pointing to a circle in FIG. 6, hereinafter also referred to as a unit) (a line connecting neuron elements and corresponding to a neurotransmission network) have a unique weight coefficient (coupling) for each synapse. As shown in FIG. 7, in each neuron element in the intermediate layer, the threshold θ _i possessed by itself is calculated from the sum of the input value y _i in the input layer multiplied by the coupling weight ω _ji . Its own output value Z _ji is calculated by subtracting it and using a response function f, for example, a staircase function having an output of 0 or 1, or a sigmoid function in which the output value is continuously changed between 0 and 1. When the number of input layers is 8 as in this example, the number of neurons in the middle layer is, for example, 9 to 16. Generally, the number of neurons in the middle layer cannot be uniquely determined, but if the number of neurons is small, the learning described later will not be completed, or if it is too large, the number of learnings will be large and the output will be unstable. Problems such as becoming will occur.

The learning method is performed by the backpropagation method (error back propagation method). In this backpropagation method, an input signal is given to the input layer, and when this signal comes out as an output signal from the output layer via the intermediate layer, this output signal is compared with the teacher signal so that this difference becomes small. The learning signal of each element of the output layer is obtained, and the synaptic load entering the output layer is corrected based on this learning signal. When this is repeated for various input signals and the corresponding set of output signals. It means that the output signal becomes closer to the teacher signal and is learned. In order to prevent overlearning and to have versatility, the learning is to finish the learning before the mean square error of the unlearned data starts to increase.

In the initial stage of learning, learning is performed so that a correct answer can be obtained by giving a blasting quality judgment based on the judgment of the blasting expert as teacher data, and from the stage when a certain amount of data is input, the blasting situation is changed. The quality of blasting is judged by using the index value as an input item. Then, when the output is excellent to normal of 1 to 3, the current blasting conditions are continued, and when the output is inferior to 4, the current blasting conditions are partially improved, and the output is 5 bad. In that case, the blasting conditions should be modified as necessary, such as significantly improving the current blasting conditions.

Even with this artificial intelligence, if the output cannot be accurate, the index values will be reviewed and tried sequentially, the number of intermediate layers will be increased (deep learning), and irrelevant connections will be discarded (convolution). It is important to improve the accuracy by performing).

[Examples of other forms]
(1) In the above embodiment, the three-dimensional shape measurement is performed by a two-dimensional photographing device such as a digital camera or a digital video camera, or a three-dimensional measuring device 4 such as a three-dimensional scanner or a stereo camera. A small unmanned aircraft (drone) is equipped with a two-dimensional photographing device such as a digital camera or a digital video camera, and a three-dimensional measuring device 4 such as a three-dimensional scanner or a stereo camera to perform three-dimensional shape measurement. You may.

Flight inside a tunnel is in a non-GPS environment, but in recent years, advances in drone technology have made it possible to fly stably in a non-GPS environment. This technology enables stable flight without relying on GPS by fusing an IMU sensor (inertial measurement unit including an acceleration sensor and a gyro sensor) with an ultra-wide-angle stereo camera. , Used for inspections in factories, inspections of underground pipelines such as sewage pipes, etc.

It can also be applied to the blasting optimization method of the present invention. However, in this case, since the image is acquired while flying, two reference points are set in the tunnel mine and in the camera field of view in order to match the images of the first step and the second step. It is preferable to match these reference points with the video of the first step and the video of the second step to perform image matching.

As a small unmanned aerial vehicle (drone), as shown in FIG. 8, a multi-rotor helicopter can be preferably used. This is a structure in which a motor 7c and a rotor 7d are provided at the tips of three or more radially extending arms 7b, 7b ... In the illustrated example, which are substantially horizontal from the main body 7a in the center of the machine body. ..

1 ... drill jumbo, 2 ... sprayer, 3 ... wheel loader, 4 ... 3D measuring device, 5 ... personal computer, 6 ... slip, 7 ... small unmanned aerial vehicle (drone)

Claims (3)

It is a blasting optimization method in a mountain tunnel where the appropriateness of blasting conditions is judged by three-dimensional shape measurement before and after blasting.
Before blasting, the first step of measuring the three-dimensional shape of the planned scattering range of the face and the slag,
After the blasting, the second step of measuring the three-dimensional shape of the accumulated state of the scattered debris and measuring the excavated shape of the face after the debris is removed.
The third step to accumulate data that quantifies the relationship between the index value of the blasting situation set based on the scattering / accumulation state of the shavings and the excavation shape of the face and the quality of the blasting based on the judgment of the blasting expert.
Based on the accumulated data, it is analyzed by a statistical processing method and / or a method by artificial intelligence, and from the stage when a certain amount of data is obtained, the quality of blasting is judged based on the index value of the blasting situation. A method for optimizing blasting in a mountain tunnel, which comprises the fourth step of modifying the blasting conditions as necessary. The method for optimizing bursting in a mountain tunnel according to claim 1, wherein the three-dimensional shape measurement is performed by a digital camera, a video camera, a three-dimensional scanner, or a stereo camera. The method for optimizing rupture in a mountain tunnel according to claim 1, wherein the three-dimensional shape measurement is performed by a small unmanned aircraft equipped with a digital camera, a video camera, a three-dimensional scanner, or a stereo camera.

Images