A towed array 3D seismic fine-grained detection method for dam and road hazards
By designing a gravity detector and a drag array three-dimensional seismic observation system that is adapted to hardened pavement, combined with single-cannon data inverse scattering wavefield rapid imaging and common-received point-related superposition methods, the rapid drag and high-precision shallow three-dimensional on urban hardened pavement are solved. The problem of seismic detection achieves high-resolution near-surface superposition imaging.
Patent Information
- Application Number
- CN202111665242.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-12-31
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2041-12-31
AI Technical Summary
The existing technology is difficult to drag and advance quickly on urban hardened roads, and high-precision shallow three-dimensional seismic detection cannot be achieved. The traditional inverse scattering imaging method relies on surface wave velocity and source wavelets, which can easily lead to inconsistent imaging results and strong multi-solvency. question.
A towed array three-dimensional seismic fine detection method was designed, including the development of a gravity detector adapted to hardened road surfaces, the design of a towed array three-dimensional seismic observation system, the rapid imaging calculation of single-cannon data inverse scattering wavefield and the use of multiple coverage observation system The common receiving point correlation superimposition imaging method.
It realizes rapid dragging and advancement on urban hardened roads, improves the accuracy and coverage of shallow three-dimensional seismic detection, overcomes the impact of dynamic correction speed of traditional methods, does not lose shallow information, and realizes high-resolution superposition imaging at the near-surface.
Smart Images

Figure CN114252919B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of seismic detection and engineering investigation of dam hidden dangers, and in particular to a towed array three-dimensional seismic refined detection method for dam hidden dangers. Background Art
[0002] 1. Research significance and current status of gravity detectors
[0003] The development of geophones has emerged with the development of seismology. At the current stage, geophone technology can be mainly divided into velocity geophones and acceleration geophones. Velocity geophones are mostly used for medium-deep oil seismic exploration and conventional surface wave exploration. Acceleration geophones have wide bandwidth and small size, which makes them more advantageous in urban shallow surface road surveys (Ren Ligang, 2018; Zhang et al., 2019; Wen Xuekang et al., 2020). At this stage, geophones are mainly cone-type geophones, and there are also geophones specifically adapted to hardened roads. Cone-type geophones have low construction efficiency and cannot be applied to urban hardened roads. Geophones for hardened roads are large in size, heavy in equipment, and the base design is not conducive to rapid towing.
[0004] The present invention aims to design a lightweight device that can be quickly dragged on hardened urban roads. It can be used with portable seismic source equipment such as electromagnetic shock seismic sources and pneumatic seismic sources to collect broadband and high-resolution seismic wave signals.
[0005] 2. Research significance and current status of observation systems
[0006] Due to the limitations of the work site and working time of dams and urban roads, it is impossible to carry out shallow three-dimensional seismic detection work on a large scale. The detection results of a single two-dimensional survey line cannot give the spatial structure and distribution range of the shallow surface structure, and cannot meet the needs of high-precision detection of dam road hidden dangers. The wide-line observation system allows the common center point gather to have a certain azimuth, which can better suppress the lateral reflection waves and shallow scattering noise, protect the effective signal, and improve the signal-to-noise ratio. The wide-line observation system can collect seismic data with a high signal-to-noise ratio, and the exploration cost is not much higher than that of two-dimensional exploration. It is a cost-effective urban road seismic detection system.
[0007] A reasonable observation system is the key to the accuracy of shallow surface structure detection. Most of the existing wide-line multiple coverage observation system schemes are aimed at medium-deep detection targets (Table 1). There is no research on the observation system specifically for ultra-shallow structures with hidden dangers in dam channels, and there is a lack of experience in parameters such as the number of survey lines, number of shot points, track spacing, line spacing, and number of coverages.
[0008]
[0009] Due to the limitation of working space and working time of dams and urban roads, it is impossible to deploy 3D geophones over a large area to collect seismic data, and the 2D observation system has low coverage of shallow surface layers and few effective channels during superposition operation. Nowadays, most towed cable systems use linear observation systems, and it is necessary to study wide-line multiple coverage observation systems suitable for dam and road hidden dangers, increase the coverage of shallow surface targets, increase the effective channels during superposition operation, and finally achieve high-resolution imaging of shallow parts.
[0010] 3. Research significance and current status of inverse scattering imaging
[0011] The propagation law of seismic waves near the surface shows that when surface waves propagate near the surface and encounter vertical and inclined interfaces, they will generate strong backscattered waves, whose energy and quantity are much higher than the energy of reflected waves (Harmankaya et al., 2013; Phan et al., 2018; Wang et al., 2019). The backscattered wave field in surface wave data has the advantages of a large number of wave fields and high vertical resolution, and can provide information on the location and depth changes of shallow surface abnormal structures (Craig et al., 2015). Many scholars have studied the backscattering imaging method, made great breakthroughs in theory, and have many successful cases. Yu et al (2014) proposed to use backscattered surface waves to estimate the distribution of near-surface faults, and superimpose the separated backscattered surface waves along a linear trajectory with the surface wave velocity as the offset velocity. However, the imaging accuracy of this method still depends on the velocity model. Guo et al (2015) implemented offset imaging of backscattered surface wave data to determine the steep-angle structure near the surface. Abdullah et al (2016) proposed a natural migration method that can achieve the migration of inverse scattered surface waves without the need for a velocity model. The simulation results show that the natural migration method can effectively reflect the heterogeneity of the near-surface. However, the calculation parameters of this method require source wavelets. The existing inverse scattering imaging methods require parameters such as surface wave velocity and source wavelets. These parameters will affect the imaging accuracy and easily cause problems such as non-unique imaging results and strong multi-solution.
[0012] In view of the difficulty in detecting hidden dangers in dam channels, the present invention proposes a rapid imaging method of shallow surface seismic inverse scattering data to achieve rapid diagnosis of near-surface abnormal structures.
[0013] 4. Research significance and current status of single shot data stacking imaging
[0014] Shallow reflection profiles can provide various near-surface information of depth and resolution. In the seismic data processing steps, dynamic correction stacking is the most commonly used stacking method. Miller et al (1998) analyzed the impact of dynamic correction distortion on shallow information. Bachrach et al (2001) applied the pre-stack Kirchhoff migration method to achieve the imaging of the target body, and pointed out that the traditional dynamic correction stacking method distortion is difficult to provide an ideal stacking profile under the conditions of high coverage and strong impedance interface. Tian et al (2006) believed that the removal of dynamic correction stretching distortion will reduce the reflection gather of 0.1s, resulting in reduced reflection information in the shallow layer. Dynamic correction stacking imaging is generally considered to be unsatisfactory for data with low coverage and low signal-to-noise ratio. Later, the common reflection surface stacking method was proposed, which can significantly improve the quality of the stacking profile (Bergler et al., 2002; Baykulov et al., 2009; Ahmed et al., 2015).
[0015] Traditional NMO and common reflection surface stacking methods are widely used for imaging medium-deep targets, but they are not ideal for imaging targets within 20 m (Steeples et al., 1998; Sloan et al., 2008; Pugin et al., 2013). The seismic data of dam channel hazards are characterized by small offset, short arrangement and small trace spacing. The target layer depth for dam channel hazard detection is less than 20 m, or even 15 m, and the recording time is less than 0.1 s. For near-surface data, traditional NMO and common reflection surface stacking methods still have the following problems: (1) The shallow layer has poor stratification, the wave impedance value of the two interfaces is small, and the energy of the reflected wave is weak. The reflected wave is easily masked by high-energy surface waves and random noise. It is difficult to identify the reflected wave on the common center point gather and pick up the appropriate gather on the velocity spectrum to implement NMO stacking. (2) Traditional NMO stacking requires the removal of the NMO stretching distortion gather, which will reduce the coverage and lose the shallow layer information. (3) The coverage of the traditional stacking method is relatively low.
[0016] The present invention aims to propose a common receiving point correlation stacking method for realizing high-resolution stacking imaging of near-surface layers. Model data and case data show that the proposed common receiving point correlation stacking method has the following advantages: (1) the stacking imaging accuracy is not affected by the dynamic correction speed; (2) no shallow layer information is lost. Summary of the invention
[0017] In view of at least one defect of the prior art, the purpose of the present invention is to provide a towed array three-dimensional seismic fine-grained detection method for dam hidden dangers, and propose a common receiving point correlation stacking method for realizing high-resolution stacking imaging of the near-surface layer for dam hidden danger detection.
[0018] In order to achieve the above object, the present invention adopts the following technical solution: a towed array three-dimensional seismic fine detection method for dam road hidden dangers, comprising the following steps:
[0019] Step A: Develop a gravity detector that can be adapted to hardened road surfaces;
[0020] Step B: Design a towed array 3D seismic observation system;
[0021] Step C: Rapid imaging calculation of inverse scattering wave field of single shot data;
[0022] Step D: Use a multiple coverage observation system and correlate the single shot data with the common receiving points for stacking imaging.
[0023] In step A, the process of developing a gravity detector that can adapt to hardened road surfaces includes the following: a base is set at the bottom of the detector, the bottom of the base is flat, the bottom of the base is in contact with the ground, and is coupled to the ground by gravity.
[0024] The bottom surface size of the base is 132 mm×90 mm.
[0025] The bottom of the base is designed with a groove extending backwards, a cone is hinged in the groove, the large end of the cone is hinged to the groove, and the small end of the cone can extend vertically downward from the groove or rotate backwards to a horizontal state. When the small end of the cone rotates backwards to a horizontal state, the lower part of the cone does not expose the groove, which does not affect the dragging forward on the hardened road surface.
[0026] In the step B, the gravity geophone is provided with a hanging structure, and a plurality of gravity geophones are connected by hanging to form a towed array type three-dimensional seismic observation system.
[0027] In step B, designing a towed array three-dimensional seismic observation system includes the following contents:
[0028] On the basis of step A, the present invention designs a towed array type 3D seismic observation system, which has 48 gravity detectors with a channel spacing of 0.5 m. Five types of towed array type 3D seismic observation systems are designed, namely, a single row 48 channel observation system, a double row 24 channel observation system, a three row 16 channel observation system, a four row 12 channel observation system and a six row 8 channel observation system. The towed array type 3D seismic observation system is formed by the above five methods.
[0029] In step C, the single shot data inverse scattered wave field rapid imaging calculation includes the following contents:
[0030] A rapid imaging method of inverse scattering wave field is proposed for diagnosing shallow abnormal structures. The specific steps are:
[0031] Step C1: Apply automatic gain control, channel equalization, abnormal channel removal, and bandpass filtering methods to preprocess all single-shot data;
[0032] Step C2: Apply the FK method to separate the inverse scattered wave field from the single shot data;
[0033] Step C3: Apply the FK transformation method to transform the inverse scattering data from the time domain to the frequency-wavenumber domain to obtain the FK spectrum;
[0034] Step C4: Use the principal component analysis (PCA) algorithm to reduce the dimension of the FK spectrum and extract the eigenvector corresponding to the maximum eigenvalue;
[0035] Step C5: repeating step C2, step C3 and step C4 in sequence to obtain a characteristic vector of the inverse scattering data;
[0036] Step C6: Smoothing the changing characteristics of the feature vector to obtain data imaging results.
[0037] In the step D, single shot data common receiving point correlation stacking imaging includes the following contents:
[0038] In order to realize high-resolution stacking imaging of the dam road hidden danger near the surface, the present invention proposes a common receiving point correlation stacking imaging method; multiple coverage observation systems are used for shallow surface seismic detection; the arrangement length is K, the number of arrangement movements is M, and the arrangement stake number is 1-K+M-1. Assume that the recording time length of each movement acquisition is t1, t2,......., t M-1 ,t M , the total recording time of multiple coverage is The total observed field value at position x is U(x,T)=[u(x,t1),u(x,t2),....,u(x,t M )], i = 1, 2, ....., M; u(x, ti) is the observed field value at the position x in the time period ti, i = 1, 2, ....., M.
[0039] If the set window length is w, the number of iterations in the common receiving point superposition calculation is N = T / w, and the derived calculation formula is:
[0040]
[0041] Among them, l=1,2,...,N, s=w*l,(w+1)*l,...,(2w-1)*l,2*w*l; ψ a (x, t) is the function of multiplying the observation record U(x, s) and U(x, s) at the position x, and then iterating N times and normalizing.
[0042] The number of coverage times nr of the common receiving point correlation superposition method is:
[0043]
[0044] F is the number of receiving channels in the array, which is equal to 1 when S is single-sided excitation and equal to 2 when S is double-sided excitation; d is the number of channels of the detection point distance that the array moves forward after one excitation. S is the number of excitation sides.
[0045] Significant effect: The present invention provides a method for fine-tuning detection of dam-road hidden dangers by towed array 3D seismic, and proposes a common receiving point correlation stacking method for realizing high-resolution stacking imaging of near-surface layers; it is used for detecting dam-road hidden dangers. It has the following advantages: (1) the stacking imaging accuracy is not affected by the dynamic correction speed; (2) no loss of shallow layer information. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] Figure 1 . Gravity detector: (a) and (b) are structural diagrams; (c) is a physical diagram;
[0047] Figure 2 .Structure diagram of the self-weight coupling detector: (a) front view; (b) side view;
[0048] Figure 3 . Detector connection diagram: (a) disconnected state; (b) connected state; (c) moving state;
[0049] Figure 4 .Single row 48-channel observation system;
[0050] Figure 5 .Double-row 24-channel observation system.;
[0051] Figure 6 .Three rows of 16 observation systems;
[0052] Figure 7 .Four rows of 12 observation systems;
[0053] Figure 8 .Six rows and eight channels of observation system;
[0054] Fig. 9 .Processing flow of inverse scattering data imaging method;
[0055] Fig.10 .Multiple coverage of earthquake observation systems;
[0056] Fig.11 .Drag array seismic active source data acquisition diagram, the observation system is three rows of 16 channels;
[0057] Fig.12 .(a) Original data record; (b) Spectrum curve; (c) Surface wave spectrum;
[0058] Fig.13 .Drag array seismic data acquisition diagram;
[0059] Fig.14 .Original wavefield: (a) shot-1; (b) shot-18;
[0060] Fig.15 .Imaging profiles: (a) surface wave inversion profile; (b) surface wave relative velocity profile; (c) inverse scattering profile; (d) stacking profile;
[0061] Fig.16 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION
[0062] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.
[0063] like Figure 1-Figure 16 As shown, the present invention relates to a towed array three-dimensional seismic fine-grained detection method for dam road hidden dangers.
[0064] The main contents include: 1) Development of high-resolution gravity detectors that can be adapted to hardened pavement; 2) Design of an array-type three-dimensional seismic observation system; 3) Application of single-shot inverse scattering data to achieve rapid lateral positioning of near-vertical structures such as near-surface karst caves, cavities, pipelines, and air-raid shelters; 4) Proposing a related stacking method for single-shot reflection wave data to achieve high-resolution stacking imaging of the near-surface layer.
[0065] The present invention provides a method for the general survey and detailed inspection of dam channel hidden dangers. The main advantages are as follows: 1) Compared with the traditional cone-type geophone, the gravity geophone simplifies the construction steps and significantly improves the construction efficiency; 2) Compared with the single-row observation system, the observation system with two rows of 24 channels, three rows of 16 channels, four rows of 12 channels, and six rows of 8 channels has a wider detection coverage range and higher imaging accuracy for the dam channel hidden danger body; 3) The imaging calculation of single-shot inverse scattering data can quickly and accurately locate near-vertical structures such as caves, cavities, pipelines, and air-raid shelters, and quickly determine the lateral position of the dam channel hidden danger; 4) The common receiving point correlation stacking method can overcome the traditional common center point stacking method in the shallow layer. The number of effective channels is small, the coverage is low, and it is susceptible to errors such as dynamic correction stretching distortion and inaccurate stacking speed. It can achieve shallow high-resolution stacking imaging, which is also the imaging method with the highest resolution for refined detection.
[0066] 1. A towed array three-dimensional refined detection method for dam road hidden dangers, comprising the following contents:
[0067] Content 1): Develop a high-resolution gravity detector that can be used on hardened roads;
[0068] Content 2): Design a towed array 3D seismic observation system suitable for different work sites and detection needs;
[0069] Content 3): Rapid imaging calculation of inverse scattering wave field of single shot data;
[0070] Content 4): Using a multiple coverage observation system, single shot data is correlated and stacked to form common receiving point imaging.
[0071] In the above content 1, the process of developing a high-resolution gravity detector that can be adapted to hardened road surfaces includes the following:
[0072] The present invention designs a first-generation gravity geophone that is adaptable to hardened road surfaces and can be moved quickly. Figure 1 (a) is the top view of the designed detector, with a plane size of 132 mm × 90 mm. Figure 1 (b) is the side view of the designed detector, the height from the base to the top is 62mm. The base is in contact with the ground and coupled to the ground by gravity. Figure 1 (c) is a physical picture of the designed detector. The weight of the whole detector is 1689g. The designed detector is named as self-weight detector. The base of the self-weight detector can move forward and backward, and its application area is mainly in urban hardened roads.
[0073] In the urban environment, it is inevitable to encounter flower beds, lawns and other environments. In order to adapt to various complex environments, the second-generation gravity detector is designed. A cone structure (A) is designed under the first-generation gravity detector, and the cone height is 80mm. When placing the detector on the hardened road surface, the cone is laid flat and the extension length is 32mm. Figure 2 As shown in (b)B.
[0074] like Figure 2 a and Figure 2 As shown in b, the bottom of the base is designed with a groove extending backwards. Figure 2 Not shown in the figure, a cone is hinged in the groove, the large end of the cone is hinged to the groove, and the small end of the cone can extend vertically downward from the groove or rotate backward to a horizontal state. When the small end of the cone rotates backward to a horizontal state, the lower part of the cone does not expose the groove, which does not affect the dragging forward on the hardened road surface. When encountering a flower bed or lawn, the cone is rotated forward so that the cone is vertically inserted into the flower bed or lawn, which enhances the coupling between the detector and the ground, making it easier for the detector to obtain seismic wave signals.
[0075] The towed seismograph is designed with 24 detectors, which are connected by hooks to facilitate quick assembly and disassembly. Figure 3 Shown is the detector connection diagram. Figure 3 (a) and (b) are disconnected and connected states. Figure 3(c) The assembled 24-channel detector is being dragged on the asphalt road. Because the base of the detector is made of metal, the friction with the asphalt road is small. The arrangement of 24 detectors can be easily moved quickly. The arrangement of fixed track spacing has obvious advantages: 1) improving on-site construction efficiency; 2) avoiding errors caused by uneven track spacing when transforming the FV spectrum, ensuring the accuracy of the FV spectrum. The self-weight coupling method solves the coupling method between the detector and the ground, which is especially suitable for hardened roads.
[0076] In the content 2, a towed array three-dimensional seismic observation system is designed, which is suitable for different work sites and detection needs, and mainly includes the following contents:
[0077] Based on step 1, the present invention designs a towed array 3D seismic observation system, which can be applied to different work sites and detection requirements. There are 48 gravity geophones with a track spacing of 0.5m, and 5 types of observation systems can be designed, which correspond to the following figures.
[0078] Single row 48 channel observation system Figure 4 As shown in the figure. The arrangement length is the longest and the detection depth is the deepest, but there is only a single survey line and the coverage is the narrowest. Increasing the number of arrangements can increase the coverage. Reducing the number of channels will reduce the detection depth. According to the requirements of coverage and detection depth, the observation system should be reasonably selected.
[0079] In the content 3, the single shot data inverse scattered wave field rapid imaging calculation mainly includes the following contents:
[0080] The present invention proposes a rapid imaging method of inverse scattering wave field for diagnosing shallow abnormal structures. The specific steps are:
[0081] 1) Apply automatic gain control, channel equalization, abnormal channel removal, bandpass filtering and other methods to realize the preprocessing of all single shot data;
[0082] 2) Apply the FK method to separate the inverse scattered wave field from the single shot data;
[0083] 3) Apply the FK transformation method to transform the inverse scattering data from the time domain to the frequency-wavenumber domain to obtain the FK spectrum;
[0084] 4) Use the principal component analysis (PCA) algorithm to reduce the dimension of the FK spectrum and extract the eigenvector corresponding to the maximum eigenvalue;
[0085] 5) Repeat steps 2), 3) and 4) in sequence to obtain the feature vectors of the inverse scattering data at different positions;
[0086] 6) Smooth the changing characteristics of the feature vectors at different positions to obtain data imaging results.
[0087] The detailed processing flow is as follows Fig. 9 shown.
[0088] In the content 4, the single shot data common receiving point correlation stacking imaging mainly includes the following contents:
[0089] In order to achieve high-resolution stacking imaging of dam hidden dangers near the surface, the present invention proposes a common receiving point correlation stacking imaging method. Multiple coverage observation systems are used for shallow surface seismic detection. Fig.10 As shown, the length of the array is K, the number of array movements is M, and the array pile number is 1-K+M-1. Assume that the length of the recording time for each array movement is t1, t2, ......., t M-1 ,t M , the total recording time of multiple coverage is The total observed field value at position x is U(x,T)=[u(x,t1),u(x,t2),....,u(x,t M )], i = 1, 2, ....., M. u(x, ti) is the observed field value at the position x in the time period ti, i = 1, 2, ....., M.
[0090] If the set window length is w, the number of iterations in the common receiving point superposition calculation is N = T / w, and the derived calculation formula is:
[0091]
[0092] Among them, l=1,2,...N,, s=w*l, (w+1)*l,..., (2w-1)*l, 2*w*l. ψ a (x, t) is the function obtained by multiplying the observation record U(x, s) at the point x by U(x, s), iterating N times and normalizing it.
[0093] The number of coverage times nr of the common receiving point correlation superposition method is:
[0094]
[0095] F is the number of receiving channels in the array, which is equal to 1 when S is single-sided excitation and 2 when S is double-sided excitation; d is the number of channels of the detection point distance moved forward after one excitation. S is the number of excitation sides. The number of stacking times nc of the common center point stacking of the conventional reflection seismic method is
[0096]
[0097] It can be seen that the number of superpositions of the spatial autocorrelation superposition method is twice that of the common center point method.
[0098] In view of the above problems, the present invention provides a three-dimensional seismic fine-grained detection system for dam channel hidden dangers.
[0099] The three-dimensional seismic detection system for dam road hidden dangers mainly includes:
[0100] (1) A portable gravity acceleration detector is proposed. The detector base is optimized and the friction between the base and the ground is reduced to achieve rapid movement of the detector. The detector has hook components at the front and back, which can be quickly assembled and disassembled at the construction site to improve construction efficiency. At the same time, different arrangements and combinations of detectors can be used to design various types of observation systems, realizing multiple detection targets and multi-purpose urban near-surface detection.
[0101] The structural design of the base and component hook of the portable gravity acceleration detector falls within the scope of patent protection.
[0102] (2) An observation system suitable for efficient and high-precision census of dam hidden dangers is proposed. Based on the superior performance of the detector, a multi-purpose dam hidden danger observation system is proposed, mainly including single-row 48 channels, double-row 24 channels, triple-row 16 channels, quad-row 12 channels, and six-row 8 channels. In vehicle-mounted mode, the most widely used are three-row 16 channels and four-row 12 channels.
[0103] The three rows of 16 channels, four rows of 12 channels and six rows of 8 channels in the observation system are within the scope of patent protection.
[0104] The refined earthquake detection system for dam channel hidden dangers mainly includes:
[0105] (3) A single-shot backscattering rapid imaging method is proposed. When seismic waves propagate on the surface, they will generate reverse, high-energy, and large-volume backscattering wave fields when they encounter near-vertical structures such as cavities, karst caves, pipelines, air-raid shelters, and collapsed settlement areas below the dam road hidden dangers.
[0106] The algorithm for inverse scattering rapid imaging is covered by patent protection.
[0107] (4) A common receiving point correlation stacking method is proposed. Based on the single-shot seismic data of dam road hidden dangers collected by the multiple coverage observation system, the common receiving point correlation stacking imaging method is implemented to obtain a high-resolution stacking profile near the surface (≤20m deep).
[0108] The algorithm of common receiving point correlation superposition imaging falls within the scope of patent protection.
[0109] The above invention contents are currently mainly applied in urban roads and dam structures. However, the application scenarios of the inverse scattering fast imaging method and the common receiving point correlation superposition imaging algorithm are not limited to the detection of hidden dangers in dam roads.
[0110] 1. Develop a high-resolution gravity detector that can be used on hardened roads;
[0111] The present invention designs a first-generation gravity geophone that is adaptable to hardened road surfaces and can be moved quickly. Figure 1 (a) is the top view of the designed detector, with a plane size of 132 mm × 90 mm. Figure 1 (b) is the side view of the designed detector, the height from the base to the top is 62mm. The base is in contact with the ground and coupled to the ground by gravity. Figure 1 (c) is a physical picture of the designed geophone. The weight of the whole geophone is 1689g. The designed geophone is named gravity geophone. The base of gravity geophone can move forward and backward, and its application area is mainly in urban hardened roads.
[0112] In the urban environment, it is inevitable to encounter flower beds, lawns and other environments. In order to adapt to various complex environments, the second-generation gravity detector is designed. A cone structure (A) is designed under the first-generation gravity detector, and the cone height is 80mm. When placing the detector on the hardened road surface, the cone is laid flat and the extension length is 32mm. Figure 2 As shown in (b)B.
[0113] The towed seismograph is designed with 24 detectors, which are connected by hooks to facilitate quick assembly and disassembly. Figure 3 Shown is the detector connection diagram. Figure 3 (a) and (b) are disconnected and connected states. Figure 3 (c) The assembled 24-channel detector is being dragged on the asphalt road. Because the base of the detector is made of metal, the friction with the asphalt road is small. The arrangement of 24 detectors can be easily moved quickly. The arrangement of fixed track spacing has obvious advantages: 1) improving on-site construction efficiency; 2) avoiding errors caused by uneven track spacing when transforming the FV spectrum, ensuring the accuracy of the FV spectrum. The self-weight coupling method solves the coupling method between the detector and the ground, which is especially suitable for hardened roads.
[0114] 2. Design a towed array 3D seismic observation system suitable for different work sites and detection needs:
[0115] Based on step 1, the present invention designs a towed array 3D seismic observation system, which can be applied to different work sites and detection requirements. There are 48 gravity geophones with a track spacing of 0.5m, and 5 types of observation systems can be designed, which correspond to the following figures.
[0116] Single row 48 channel observation system Figure 4 As shown in the figure. The arrangement length is the longest and the detection depth is the deepest, but there is only a single survey line and the coverage is the narrowest. Increasing the number of arrangements can increase the coverage. Reducing the number of channels will reduce the detection depth. According to the requirements of coverage and detection depth, the observation system should be reasonably selected.
[0117] 3. Rapid imaging calculation of inverse scattering wave field of single shot data:
[0118] The present invention proposes a rapid imaging method of inverse scattering wave field for diagnosing shallow abnormal structures. The specific steps are:
[0119] 1) Apply automatic gain control (AGC), channel equalization, abnormal channel removal, bandpass filtering and other methods to pre-process all single shot data;
[0120] 2) Apply the FK method to separate the inverse scattered wave field from the single shot data;
[0121] 3) Apply the FK transformation method to transform the inverse scattering data from the time domain to the frequency-wavenumber domain to obtain the FK spectrum;
[0122] 4) Use the PCA algorithm to reduce the dimension of the FK spectrum and extract the eigenvector corresponding to the maximum eigenvalue;
[0123] 5) Repeat steps 2), 3) and 4) in sequence to obtain the feature vectors of the inverse scattering data at different positions;
[0124] 6) Smooth the changing characteristics of the feature vectors at different positions to obtain data imaging results.
[0125] The detailed processing flow is as follows Fig. 9 shown.
[0126] 4. Single shot data common receiving point correlation stacking imaging:
[0127] In order to achieve high-resolution stacking imaging of dam hidden dangers near the surface, the present invention proposes a common receiving point correlation stacking imaging method. Multiple coverage observation systems are used for shallow surface seismic detection. Fig.10 As shown, the length of the array is K, the number of array movements is M, and the array pile number is 1-K+M-1. Assume that the length of the recording time for each array movement is t1, t2, ......., t M-1 ,t M , the total recording time of multiple coverage is The total observed field value at position x is U(x,T)=[u(x,t1),u(x,t2),....,u(x,t M )], i=1,2,.....,M.
[0128] If the set window length is w, the number of iterations in the common receiving point superposition calculation is N = T / w, and the derived calculation formula is:
[0129]
[0130] Among them, l=1,2,...,N, s=w*l, (w+1)*l,..., (2w-1)*l, 2*w*l. ψ a (x, t) is the function obtained by multiplying the observation record U(x, s) at the point x by U(x, s), iterating N times and normalizing it.
[0131] The number of coverage times nr of the common receiving point correlation superposition method is:
[0132]
[0133] F is the number of receiving channels in the array, which is equal to 1 when S is single-sided excitation and 2 when S is double-sided excitation; d is the number of channels of the detection point distance moved forward after one excitation. S is the number of excitation sides. The number of stacking times nc of the common center point stacking of the conventional reflection seismic method is
[0134]
[0135] It can be seen that the number of superpositions of the spatial autocorrelation superposition method is twice that of the common center point method.
[0136] Engineering Application
[0137] 1. Towed gravity geophone seismic acquisition
[0138] In order to verify the reliability of gravity detector signals, signal acquisition experiments were carried out at the dam channel hidden dangers. Detectors can be freely combined into different arrangements, such as Fig.11 The figure shows the application of the towed geophone in road detection in Shenzhen. A 24-pound sledgehammer is used as the source. The offset is 3m, the track spacing is 0.5m, and the acquisition record length is 170ms. Fig.12 (a) shows the original seismic record, A is the direct wave, B is the surface wave, and C is the hyperbola-like reflection wave. Fig.12 (b) is the frequency spectrum curve, with the main frequency width being 1Hz–200Hz. Fig.12 (c) is the surface wave spectrum, which can pick up a very continuous dispersion curve. The gravity detector can collect reflected waves and surface waves, and these waveforms can be used to image the underground structure using corresponding data processing methods.
[0139] 2. Fine-grained imaging of earthquake near the surface of dam road hidden dangers;
[0140] 1.1. Case 2: Shenzhen
[0141] Data collection
[0142] In July 2021, towed seismic was used in Shenzhen to carry out stability detection of dam road hazards. The surface layer is an asphalt layer with a thickness of about 0.3m. The bottom layer is backfill clay. When collecting towed seismic data, an array observation system is deployed, with a total of 36 detectors, and 12 detectors on each survey line. The detector type is a single-component acceleration type with a bandwidth of 1Hz to 5000Hz. During data collection, the offset distance is 3.0m, the track spacing is 0.5m, the sampling interval is 41.7us, and the number of sampling points is 4096. The source is a 24-pound sledgehammer, and each shot point is hammered once. Field data collection is as follows Fig.13 As shown. A part of the 26m long survey line was selected for detailed survey experiments, and the shot spacing was set to 1.0m. Ground penetrating radar detection work had been carried out in the detection area before, and abnormal structures such as cavities were found and repaired. This article only shows the data analysis results of the middle survey line L2.
[0143] like Fig.14 It is the original drag seismic record. Symbol A is the direct wave, B is the surface wave, and C1 and C2 are considered to be reflection waves, but they are not easy to identify at close offset distances.
[0144] Comprehensive analysis
[0145] The shear wave velocity profile, relative velocity profile, inverse scattering profile and superposition profile are calculated in sequence, such as Fig.15 As shown. The four profiles show that there are abnormal responses at the horizontal position of 3.5m to 7.0m (A), the shear wave velocity profile shows a relatively low velocity (A1), the relative velocity profile shows a relatively high velocity (A2), the inverse scattering profile shows strong energy (A3), the stacked profile shows a diffraction wave (A4), and A5 is a multiple wave. The positioning accuracy of the abnormal body A in the inverse scattering profile is the highest. The shear wave velocity profile, the relative velocity profile and the stacked profile can reflect the continuous interface inside the base layer (B), among which the results of the relative velocity profile and the stacked profile are closest to the actual situation. Due to the multi-solution and uncertainty of surface wave inversion, the inversion results will have false anomalies. There are low-velocity anomalies at the horizontal position of 6m to 9m and the depth of 0m to 2m, and high-velocity anomalies at the depth of 4m to 8m (D), which are considered unreliable. When the surface wave dispersion curve is inverted and calculated, a low-velocity layer will appear near the surface in order to fit the curve. In fact, there is a backfill clay layer under the asphalt layer. There is no low-velocity layer inside the backfill clay layer, so the low-velocity layer can be considered a false anomaly. The thickness of the asphalt layer is 0.3m. The current seismic resolution is not enough to distinguish it, but the reflection interface (B) inside the base layer at a depth of 1m can be identified. Combining the inverse scattering profile and the stacked profile, the anomaly A is the location of the drainage pipe, with a depth of about 6m.
[0146] Finally, it should be noted that the above examples are only specific implementation examples of the present invention. Of course, those skilled in the art can make changes and modifications to the present invention. If these modifications and variations fall within the scope of the claims of the present invention and their equivalent technologies, they should be considered to be within the scope of protection of the present invention.
Claims
1. A towed array three-dimensional seismic fine-grained detection method for dam road hidden dangers, characterized in that: The following steps are involved: Step A: Develop a gravity detector that can be adapted to hardened road surfaces; Step B: Design a towed array 3D seismic observation system; Step C: Rapid imaging calculation of inverse scattering wave field of single shot data; In step C, the single shot data inverse scattered wave field rapid imaging calculation includes the following contents: A rapid imaging method of inverse scattering wave field is proposed for diagnosing shallow abnormal structures. The specific steps are: Step C1: Apply automatic gain control, channel equalization, abnormal channel removal, and bandpass filtering methods to preprocess all single-shot data; Step C2: Apply the FK method to separate the inverse scattered wave field from the single shot data; Step C3: Apply the FK transformation method to transform the inverse scattering data from the time domain to the frequency-wavenumber domain to obtain the FK spectrum; Step C4: Use the principal component analysis (PCA) algorithm to reduce the dimension of the FK spectrum and extract the eigenvector corresponding to the maximum eigenvalue; Step C5: repeating step C2, step C3 and step C4 in sequence to obtain a characteristic vector of the inverse scattering data; Step C6: Smoothing the changing characteristics of the feature vector to obtain the data imaging result; Step D: Use a multiple coverage observation system to correlate and stack single shot data at common receiving points for imaging; In the step D, single shot data common receiving point correlation stacking imaging includes the following contents: In order to realize high-resolution stacking imaging of the dam road hidden danger near the surface, the present invention proposes a common receiving point correlation stacking imaging method; multiple coverage observation systems are used for shallow surface seismic detection; the arrangement length is K, the number of arrangement movements is M times, and the arrangement stake number is 1-K+M-1; it is assumed that the recording time length of each movement acquisition is t1, t2,......., t M-1 ,t M , the total recording time of multiple coverage is The total observed field value at position x is U(x,T)=[u(x,t1),u(x,t2),....,u(x,t M )], i = 1, 2, ....., M; u(x, ti) is the observed field value at the position x in the time period ti; If the set window length is w, the number of iterations in the common receiving point superposition calculation is N = T / w, and the derived calculation formula is: Among them, l=1,2,...,N, s=w*l,(w+1)*l,...,(2w-1)*l,2*w*l; ψ a (x, t) is the function of multiplying the observation record U(x, s) and U(x, s) at the position x, and then iterating N times and normalizing. The number of coverage times nr of the common receiving point correlation superposition method is: F is the number of receiving channels in the arrangement, which is equal to 1 when S is single-sided excitation and equal to 2 when S is double-sided excitation; d is the number of channels of the detection point distance that the arrangement moves forward after one excitation.
2. The method for fine-tuning dam-track hidden danger towing array three-dimensional seismic detection according to claim 1 is characterized by: In step A, the process of developing a gravity detector that can be adapted to hardened road surfaces includes the following: A base is arranged at the bottom of the detector. The bottom of the base is flat and in contact with the ground, and is coupled with the ground by gravity.
3. The method for fine-tuning dam-track hidden danger towing array three-dimensional seismic detection according to claim 2 is characterized by: The bottom surface size of the base is 132 mm×90 mm.
4. The method for fine-tuning dam-track hidden danger towing array three-dimensional seismic detection according to claim 2 is characterized by: The bottom of the base is designed with a groove extending backwards, a cone is hinged in the groove, the large end of the cone is hinged to the groove, and the small end of the cone can extend vertically downward from the groove or rotate backwards to a horizontal state.
5. The method for fine-tuning dam-track hidden danger towing array three-dimensional seismic detection according to claim 1 is characterized by: In the step B, the gravity geophones are provided with hanging structures, and the gravity geophones are connected by hanging to form a towed array type three-dimensional seismic observation system.
6. The method for fine-tuning dam-track hidden danger towing array three-dimensional seismic detection according to claim 1 is characterized by: In step B, designing a towed array three-dimensional seismic observation system includes the following contents: Based on step A, the present invention designs a towed array three-dimensional seismic observation system, which has a total of 48 gravity detectors with a channel spacing of 0.5m. Five towed array three-dimensional seismic observation systems are designed, namely a single-row 48-channel observation system, a double-row 24-channel observation system, a three-row 16-channel observation system, a four-row 12-channel observation system and a six-row 8-channel observation system.
