Diversion tunnel inner surface acoustic echo point extraction method

Abstract

The invention relates to the technical field of safety monitoring and health management in the service period of a diversion tunnel in water conservancy and hydropower engineering, in particular to a diversion tunnel inner surface acoustic echo point extraction method which comprises the following steps: (1) an omnidirectional sonar actively emits sound waves to irradiate the inner surface of the tunnel, and meanwhile, periodic sampling is performed to form a series of discrete echo data points; (2) filtering direct waves, multipath echoes and noise data points in the discrete echo data points, and reserving actual echo data points on the inner surface of the tunnel; (3) one frame of data in the actual echo data points is used as a processing unit, and the intensity value change rate R between the adjacent data points is calculated one by one; (4) distance information and intensity information represented by acoustic echo points on the inner surface of the diversion tunnel should be optimal values, the acoustic echo points should be distributed at positions with large reflection echo intensity values and dense reflection echo intensity values in a short-distance spherical wave propagation mode, and the distance and intensity information is considered as a whole; the next data point with the maximum intensity value change rate between the adjacent data points is taken as an acoustic echo point P of the current frame data; (5) acoustic echo point extraction is carried out on the discrete echo data points of the next frame according to the steps (3) and (4) until acoustic echo points of all data frames are extracted and stored; according to the method, the problem of multiple data points formed by echo periodic sampling can be effectively solved, tunnel inner surface acoustic echo points giving consideration to strength information and position information are extracted, and reliable data support is provided for reconstruction and disease detection and recognition of the diversion tunnel inner surface.

Description

Technical Field

[0001] The present invention relates to the technical field of safety monitoring and health management during the service of long-distance water diversion tunnels in large-scale water conservancy and hydropower projects such as the South-to-North Water Diversion Project; by actively emitting sound waves to illuminate the inner surface of the water diversion tunnel, and based on the principle of sound wave reflection, periodically collecting echoes to form a series of echo data points to construct an acoustic point cloud model of the inner surface of the tunnel, so as to achieve the purpose of acoustic three-dimensional reconstruction of the inner surface of the tunnel structure and disease detection and identification. In order to obtain a high-precision and high-resolution model of the inner surface of the tunnel structure, it is necessary to screen and extract the actual multiple echo data points of the inner surface formed by the periodic collection of echoes, so as to obtain a single acoustic echo point that can characterize the characteristic information of the inner surface of the tunnel. The specific field is a method for extracting acoustic echo points on the inner surface of a water diversion tunnel. Background Art

[0002] The maintenance and disease monitoring of water diversion tunnels and other structures in large-scale water conservancy and hydropower projects such as the South-to-North Water Diversion Project are key means to ensure their safe operation. Since the water diversion tunnel is located in a special water environment with turbid water and no natural light inside, the detection mode based on optical principles cannot achieve the expected detection results due to strong attenuation and scattering problems. On the contrary, the detection mode based on acoustic principles is currently the most effective way to detect damage to the inner surface of tunnel structures in large diameters, long distances and turbid water environments. The tunnel structure detection technology based on the principle of sound wave reflection emits sound waves of a certain width. When the sound waves are transmitted to the inner surface of the tunnel, a reflected echo is generated. The reflected echo signal is transmitted through the water body to the signal receiver and received, so that the collected reflected echo signal is used to detect and identify the inner surface of the tunnel. At present, some domestic researchers use acoustic imaging equipment to study the reconstruction and disease identification technology of the inner surface structure of the tunnel, and detect and identify diseases by introducing foreign acoustic imaging equipment and the tunnel point cloud model output by the equipment. However, since tunnel defect detection requires high accuracy of collected data, the point cloud model output by the data processing mechanism of foreign acoustic detection equipment cannot detect and identify small defects such as tunnel cracks. Therefore, it is necessary to conduct in-depth research on the internal data processing mechanism of acoustic detection equipment based on the principle of acoustic reflection.

[0003] Since the emitted sound wave has a certain time width, the echo reflected by the inner surface of the tunnel also has a certain time width. After periodic sampling, a series of discrete sampling point data are formed. After spatial relocation and splicing, the acoustic point cloud model of the inner surface of the tunnel reconstructed by these discrete data points is a three-dimensional cluster point cloud, rather than a common surface point cloud model. In order to achieve the purpose of high-precision reconstruction and disease identification of the internal surface spatial information and texture information of the tunnel, it is necessary to accurately extract a single acoustic echo point representing the inner surface of the tunnel from this series of discrete data points, and simplify the three-dimensional cluster point cloud into a surface point cloud model. At present, there are few technologies for extracting acoustic echo data points on the inner surface of the water diversion tunnel. There is a need for an acoustic echo point extraction method suitable for three-dimensional reconstruction and damage detection of the inner surface of the water diversion tunnel. Summary of the invention

[0004] The purpose of the present invention is to provide a method for extracting acoustic echo points on the inner surface of a water diversion tunnel to solve the problem of multiple echo data points that appear in the existing periodic uniform sampling based on omnidirectional sonar detection, thereby obtaining real information representing the inner surface structure of the tunnel, forming a high-precision, high-resolution digital point cloud model of the inner surface of the water diversion tunnel, realizing effective monitoring of disaster-causing factors such as cracks on the inner surface of the tunnel, and ensuring the safety of the water diversion tunnel during its service.

[0005] To achieve the above object, the present invention provides the following technical solution: a method for extracting acoustic echo points on the inner surface of a water diversion tunnel, the steps of which are as follows:

[0006] (1) The omnidirectional sonar actively emits sound waves to illuminate the inner surface of the tunnel, and periodically samples to form a series of discrete echo data points;

[0007] (2) Filter the direct wave, multipath echo and noise data points in the discrete echo data points to retain the actual echo data points on the inner surface of the tunnel;

[0008] (3) Taking one frame of data in the actual echo data point as the processing unit, the intensity value change rate R between adjacent data points is calculated one by one;

[0009] (4) The distance information and intensity information represented by the acoustic echo points on the inner surface of the water diversion tunnel should be optimal. In the short-distance spherical wave propagation mode, the acoustic echo points should be distributed at locations where the reflected echo intensity values ​​are large and dense. Considering the distance and intensity information, the next data point with the largest intensity value change rate between adjacent data points is taken as the acoustic echo point P of the current frame data;

[0010] (5) According to steps (3) and (4), the acoustic echo points of the next frame of discrete echo data points are extracted until the acoustic echo points of all data frames are extracted and stored.

[0011] Preferably, according to step (3), the algorithm needs to first determine whether there is a data frame. If there is no data frame, the algorithm ends; if there is, the intensity value and position number of each point in the actual echo data frame on the inner surface of the tunnel are read, and the intensity value change rate R between adjacent points is calculated one by one.

[0012] Preferably, the intensity value change rate R is calculated by the following formula:

[0013] R=(V b -V f ) / (S b -S f ) (1);

[0014] Among them, V f , S f Respectively represent the intensity value and position number of the previous data point; V b , S b They respectively represent the intensity value and position number of the next data point.

[0015] Preferably, according to step (4), in the short-distance spherical wave propagation mode, the acoustic echo points on the inner surface of the water diversion tunnel should be distributed at positions where the reflected echo intensity values ​​are large and dense, and the distance and intensity information factors are comprehensively considered, and the next data point with the largest intensity value change rate between adjacent data points is taken as the current acoustic echo point P. The acoustic echo point P calculation formula is:

[0016]

[0017] Where R is the rate of change of intensity value; It means rounding backward, that is, taking the point after the adjacent data point with the largest intensity change rate.

[0018] Beneficial Effects

[0019] Compared with the prior art, the beneficial effects of the present invention are: it fills the gap in the acoustic echo data point extraction method that appears in the process of detecting and identifying the defects of hydraulic structures such as water diversion tunnels using omnidirectional sonar and sound wave reflection principle detection methods. Through the invention, the real echo data points that best represent the inner surface strength and distance information of the water diversion tunnel can be obtained, thereby obtaining high-precision and high-reliability data, with a maximum average relative error of 3.4% and a total average error of 16 mm, providing data support for the visualization of the inner surface of the water diversion tunnel and the monitoring of disaster-causing factors such as cracks. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 Flowchart of the algorithm of the present invention.

[0021] Figure 2 It is a schematic diagram of the propagation of acoustic spherical waves of the present invention.

[0022] Figure 3 It is a trend diagram of actual echo intensity values ​​on the inner surface of the water diversion tunnel of the present invention.

[0023] Figure 4 It is a schematic diagram of extracting real acoustic echo points on the inner surface of a water diversion tunnel according to the present invention. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] See also Figures 1 to 4 The present invention provides a technical solution: a method for extracting acoustic echo points on the inner surface of a water diversion tunnel, the steps of which are:

[0026] (1) The omnidirectional sonar actively emits sound waves to illuminate the inner surface of the tunnel, and periodically samples to form a series of discrete echo data points;

[0027] (2) Filter the direct wave, multipath echo and noise data points in the discrete echo data points to retain the actual echo data points on the inner surface of the tunnel;

[0028] (3) Taking one frame of data in the actual echo data point as the processing unit, the intensity value change rate R between adjacent data points is calculated one by one;

[0029] (4) The distance information and intensity information represented by the acoustic echo points on the inner surface of the water diversion tunnel should be optimal. In the short-distance spherical wave propagation mode, the acoustic echo points should be distributed at locations where the reflected echo intensity values ​​are large and dense. Considering the distance and intensity information, the next data point with the largest intensity value change rate between adjacent data points is taken as the acoustic echo point P of the current frame data;

[0030] (5) According to steps (3) and (4), the acoustic echo points of the next frame of discrete echo data points are extracted until the acoustic echo points of all data frames are extracted and stored.

[0031] According to step (3), the algorithm must first determine whether there is a data frame. If there is no data frame, the algorithm ends; if there is, the intensity value and position sequence number of each point in the actual echo data frame on the inner surface of the tunnel are read, and the intensity value change rate R between adjacent points is calculated one by one.

[0032] The intensity value change rate R is calculated by the following formula:

[0033] R=(V b -V f ) / (Sb -S f ) (1);

[0034] Among them, V f , S f Respectively represent the intensity value and position number of the previous data point; V b , S b They respectively represent the intensity value and position number of the next data point.

[0035] According to step (4), in the short-distance spherical wave propagation mode, the acoustic echo points on the inner surface of the water diversion tunnel should be distributed at locations where the reflected echo intensity values ​​are large and dense. Considering the distance and intensity information factors, the next data point with the largest intensity value change rate between adjacent data points is taken as the current acoustic echo point P. The acoustic echo point P calculation formula is:

[0036]

[0037] Where R is the rate of change of intensity value; It means rounding backward, that is, taking the point after the adjacent data point with the largest intensity change rate.

[0038] Through this technical solution, Figure 1 As shown, the present invention uses omnidirectional sonar to actively transmit sound waves to illuminate the inner surface of the tunnel, and based on the principle of sound wave reflection, periodically samples the sound wave signals (including transmitted sound waves and reflected sound waves) to obtain a series of discrete data points. First, for the direct wave data points and multipath echo data points in the discrete data points, a fixed threshold method is used to perform prior filtering based on the tunnel geometry parameters and the sonar inherent parameters. Secondly, for the noise data points, a radius filtering method is used for filtering. Third, the algorithm first determines whether there is a tunnel data frame at present. If not, the algorithm ends and there is no need to extract acoustic echo points; if so, acoustic echo points are extracted. Fourth, the algorithm calculates the rate of change R of the intensity values ​​between adjacent data points in the same data frame, that is, R=(V b -V f ) / (S b -S f ), as the acoustic echo point extraction criterion for the current data frame; where V f , S f Respectively represent the intensity value and position number of the previous data point, V b , S b Respectively represent the intensity value and position number of the next data point. Fifth, the algorithm selects the next data point of the adjacent point corresponding to the largest intensity value change rate as the acoustic echo point P of the current data frame, that is, Where R is the intensity value change rate, = means rounding backward, that is, taking the point after the adjacent data point with the largest intensity change rate. Finally, repeat the fourth and fifth steps until the acoustic echo point extraction of all data frames is completed.

[0039] like Figure 2 As shown in the figure, short-distance sound propagation should be regarded as spherical wave propagation. The spherical wave with a certain time width arrives at the inner surface of the tunnel at different times, so the echo signal values ​​collected at different times are different. In theory, the acoustic wave front point a first arrives at the detection surface (first arrival wave), corresponding to the first non-zero intensity value and the value is usually small, and its echo arrival time can accurately represent the distance information; followed by the waveform front point b, c and the subsequent waveform point d arriving at the detection surface at the same time, and the sampling value should be the coherent superposition of the echoes of these three points. As the number of waveform points arriving at the detection surface continues to increase, the sampling value gradually increases to saturation after the waveform superposition, until the sampling value gradually decreases after the acoustic wave trailing edge arrives at the detection surface. Usually, the first arrival wave data point is selected as the acoustic echo point. However, due to the influence of environmental noise, reverberation and the previous frame echo, the first non-zero value collected may be the echo value of the non-point a, not the actual echo data point of the tunnel.

[0040] like Figure 3 As shown in the graph of the intensity variation trend of the echo data points reflected from the inner surface of the tunnel, it can be seen that the intensity of the echo data points on the inner surface of the tunnel shows a trend of gradually increasing and then gradually decreasing from left to right along the sampling point sequence number. The intensity value of the first arrival wave data point is small, while the high intensity value data points are more and concentrated in the middle. The distance information and intensity information corresponding to the ideal acoustic echo point should all be optimal values, but due to noise, sound propagation mode and the influence of the previous frame echo signal, the collected echo data points often cannot take both into account. The intensity value of the first arrival wave is often small and cannot represent the actual texture information of the illuminated area on the inner surface of the tunnel; and the position with the maximum intensity value is often in the middle of the sampling point sequence, rather than the first arrival wave position, which cannot represent the distance information of the illuminated area on the inner surface of the tunnel. The present invention comprehensively considers the distance and intensity information factors, and believes that the coherent superposition of several position points of the acoustic spherical wave can best reflect the intensity value of the detection area on the inner surface of the tunnel. The intensity change rate is proposed as the criterion for extracting acoustic echo points, and the overall optimal solution is obtained, as shown in formulas (1) and (2):

[0041] R=(V b -V f ) / (S b -S f ) (1)

[0042]

[0043] Among them, R represents the intensity value change rate between adjacent data points, V f , S fRespectively represent the intensity value and position number of the previous data point, V b , S b Respectively represent the intensity value and position number of the next data point; P represents the acoustic echo point, It means rounding up, that is, taking the next sampling point of the adjacent data point with the largest intensity change rate as the true echo point.

[0044] Extraction diagram as Figure 4 As shown in the figure, the upper broken line is the intensity value curve of a single frame data point. The intensity value of the broken line part contained in the dotted rectangular frame is relatively large and dense; the acoustic echo data points obtained by the above analysis should fall within this range. The lower broken line is the change rate of the intensity value of adjacent data points. The maximum change rate is at the circular dotted line mark. It can be seen that the position of the maximum change rate belongs to the area with large and dense intensity values, and is located in the front half of the rectangular dotted frame; from the above analysis, it can be concluded that the acoustic echo data points extracted by the method of the present invention not only optimize the distance position information but also take into account the texture intensity information.

[0045] A cylindrical container is used to simulate the hydraulic structure of a water diversion tunnel. The acoustic echo points of all data frames on the inner surface are extracted according to the method of the present invention to form an inner surface acoustic point cloud model. The radius of the simulated water diversion tunnel is 475 mm. Due to the presence of the surface annular pattern, the cross-sectional radius of the cylindrical container varies by ±5 mm. The experimental results are shown in Table 1:

[0046]

[0047] Randomly select 4 sections for error statistical analysis. Due to the presence of the annular pattern, the true radius of the 4 sections is 475mm, 470mm, 480mm, and 475mm respectively. The largest error occurs in Section 2, reaching 140.8mm, but its average error is only 16.2mm, indicating that the maximum error is only a few data points, mainly due to the presence of obstacles in the water area. The minimum error of the 4 sections is 0.4mm, the maximum average relative error is 3.4%, and the total average error is 16mm, which can meet the needs of high-precision 3D reconstruction of the inner surface of the water diversion tunnel and the detection of related disaster factors.

[0048] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for extracting acoustic echo points on the inner surface of a water diversion tunnel, characterized in that: The steps are: (1) The omnidirectional sonar actively emits sound waves to illuminate the inner surface of the tunnel, and periodically samples to form a series of discrete echo data points; (2) Filter the direct wave, multipath echo and noise data points in the discrete echo data points to retain the actual echo data points on the inner surface of the tunnel; (3) Taking one frame of data in the actual echo data point as the processing unit, the intensity value change rate R between adjacent data points is calculated one by one; (4) The distance information and intensity information represented by the acoustic echo points on the inner surface of the water diversion tunnel should be optimal. In the short-distance spherical wave propagation mode, the acoustic echo points should be distributed at locations where the reflected echo intensity values ​​are large and dense. Considering the distance and intensity information, the next data point with the largest intensity value change rate between adjacent data points is taken as the acoustic echo point P of the current frame data; (5) According to steps (3) and (4), the acoustic echo points of the next frame of discrete echo data points are extracted until the acoustic echo points of all data frames are extracted and stored.

2. The method for extracting acoustic echo points on the inner surface of a water diversion tunnel according to claim 1, characterized in that: According to step (3), the algorithm must first determine whether there is a data frame. If there is no data frame, the algorithm ends; if there is, the intensity value and position sequence number of each point in the actual echo data frame on the inner surface of the tunnel are read, and the intensity value change rate R between adjacent points is calculated one by one.

3. The method for extracting acoustic echo points on the inner surface of a water diversion tunnel according to claim 2, characterized in that: The intensity value change rate R is calculated by the following formula: R=(V b -V f ) / (S b -S f ) (1); Among them, V f , S f Respectively represent the intensity value and position number of the previous data point; V b , S b They respectively represent the intensity value and position number of the next data point.

4. The method for extracting acoustic echo points on the inner surface of a water diversion tunnel according to claim 1, characterized in that: According to step (4), in the short-distance spherical wave propagation mode, the acoustic echo points on the inner surface of the water diversion tunnel should be distributed at locations where the reflected echo intensity values ​​are large and dense. Considering the distance and intensity information factors, the next data point with the largest intensity value change rate between adjacent data points is taken as the current acoustic echo point P. The acoustic echo point P calculation formula is: Where R is the intensity value change rate; It means rounding backward, that is, taking the point after the adjacent data point with the largest intensity change rate.

Images