Pavement structure layer thickness full-section nondestructive detection and data processing method based on three-dimensional ground penetrating radar

Abstract

The invention discloses a pavement structure layer thickness full-section nondestructive detection and data processing method based on a three-dimensional ground penetrating radar, and the method comprises the steps: dividing a to-be-detected region into 1.5 m wide detection channels, and achieving the full-section scanning through a specific antenna array; reasonable detection parameters are set, for example, sampling spacing is set according to antenna center frequency, and detection time windows are set according to detection depth. After detection, core sample calibration is drilled, radar data noise suppression, Fourier transform, clutter removal and other processing are carried out, the dielectric constant is inversely calculated, and the thickness is calculated. And giving coordinates to the detection points by utilizing MATLAB, combining a full-section thickness distribution matrix, converting the full-section thickness distribution matrix into an RGB color distribution matrix, and drawing a thickness distribution cloud chart and a pie chart. The method solves the problem that existing detection representativeness is insufficient, can comprehensively and accurately detect the thickness of the pavement structure layer, has the advantages of high-precision data processing and visual result presentation, and can effectively improve the pavement quality and avoid early damage.

Description

Technical Field

[0001] The present invention relates to the technical field of road engineering detection, in particular to a method for full-section non-destructive detection of pavement structure layer thickness and data processing based on three-dimensional ground penetrating radar. Background Art

[0002] Due to factors such as poor flatness of the lower layer top surface and poor adjustment of paving equipment during paving construction, the various structural layers of newly built pavements are prone to a high proportion of areas with insufficient thickness during construction. Insufficient structural layer thickness will affect the overall strength and function of the pavement, increase the risk of early damage to the pavement structure in this area, and form weak areas of the pavement. To improve the pass rate of the pavement structural layer thickness distribution, in addition to improving the flatness of the lower layer and strengthening the commissioning of paving equipment, it is also necessary to implement thickness testing methods to obtain the full-section thickness distribution of the lower layer after the lower layer construction is completed. During construction, measures such as increasing the paving thickness of the upper layer can be taken to address thin areas in the lower layer. This ensures that the overall thickness of the pavement meets the requirements for acceptance and avoids weak areas. Full-section thickness testing technology for the pavement structural layer is key.

[0003] Currently, the main methods for measuring the thickness of pavement structural layers include the core drilling method, two-dimensional ground-penetrating radar (GPR), and three-dimensional GPR. The core drilling method is not suitable for measuring the thickness of large-scale pavement structural layers due to its significant damage to asphalt pavements, limited sample points, and lack of representativeness. The two-dimensional GPR method has the advantages of high detection efficiency and no damage to the pavement structure. However, a single detection can only collect one longitudinal profile, which cannot achieve full-section coverage scanning. It also suffers from the obvious lack of representativeness. The detection principle of three-dimensional GPR is the same as that of two-dimensional GPR. It uses a multi-channel antenna, and the single detection width range is close to covering an entire lane, which can achieve full-section scanning of asphalt pavement thickness. It has become a new method suitable for rapid non-destructive testing of large-scale pavement thickness in recent years. Its application to structural layer thickness detection can enable thickness detection to make the leap from "line" to "surface".

[0004] Currently, domestic research on structural layer thickness detection methods primarily focuses on two-dimensional ground-penetrating radar (GPR), with relatively little research on three-dimensional GPR for structural layer thickness detection. Applying three-dimensional GPR to structural layer detection requires addressing the following challenges: first, how to conduct on-site testing; second, how to process the data collected by stepped-frequency three-dimensional GPR, which is a three-dimensional frequency-domain data matrix, and convert it into a usable two-dimensional time-domain data matrix; third, how to calibrate the dielectric constant; and fourth, how to accurately, intuitively, and effectively present the test results, given that thickness detection achieves full cross-section coverage and increases the amount of data by an order of magnitude.

[0005] Existing technical problems

[0006] 1. Main technical issues:

[0007] Three-dimensional ground-penetrating radar is used to carry out full-section and large-area thickness detection of pavement structure layers. Currently, there is a lack of complete sets of technologies such as on-site detection methods, data calibration, data processing, and expression of detection results.

[0008] 2. Secondary issues

[0009] (1) Current thickness detection methods, whether single-point core drilling or single-line two-dimensional ground penetrating radar, cannot achieve full-section coverage detection of asphalt pavement, resulting in insufficient representativeness of the test results.

[0010] (2) During the construction process of a newly built pavement, each structural layer is prone to having areas of insufficient thickness. It is necessary to obtain the thickness distribution of the entire cross-section of the lower structural layer after the construction of the lower layer is completed. During the construction process, treatment measures should be taken for the thin areas of the lower layer to ensure that the overall thickness of the pavement meets the requirements during the acceptance inspection. Summary of the Invention

[0011] The purpose of the present invention is to provide a full-section non-destructive detection and data processing method for the thickness of the pavement structure layer based on three-dimensional ground penetrating radar, so as to realize the thickness detection of the pavement structure layer over the entire section and a large area, solve the shortcomings of the existing detection methods, improve the qualified rate of the pavement structure layer thickness distribution, and avoid the appearance of weak areas on the road surface.

[0012] The present invention is achieved through the following technical solutions:

[0013] A method for full-section non-destructive detection of pavement structure layer thickness and data processing based on three-dimensional ground penetrating radar comprises the following steps:

[0014] Step S1: Divide the inspection area into inspection lanes with a width of 1.5 m. First, measure the width of the inspection area (w meters). Calculate and record the distance (x meters) between the inspection lane and the central median and the number of inspection lanes (n) using Equations 1 and 2. Using a marker pen, divide the inspection area into n inspection lanes with a width of 1.5 m, starting from x meters from the right edge of the inspection area. Number the inspection lanes 1, 2, ..., n, from right to left.

[0015]

[0016] x=(w-1-1.5×n) / 2 (2)

[0017] Step S2: Setting detection parameters:

[0018] 3D GPR is used to detect the thickness of pavement structure layers and provide high-resolution underground images. The appropriate radar frequency and scanning mode must be selected based on the specific conditions of the detection object. During the process of detecting the thickness of the structure layer, the 3D GPR antenna must set three antenna parameters: sampling interval, detection time window, and the number of repeated scans (N).

[0019] (1) Sampling interval Δx

[0020] The sampling interval is the horizontal distance between two sampling points in the direction of antenna travel. The sampling interval is set according to the antenna center frequency and the dielectric constant of the medium. The calculation formula is as follows:

[0021]

[0022] Where Δx is the sampling interval, unit: m, c is the speed of light in vacuum, unit: m / s, f c is the center frequency of the antenna, unit: Hz, μ r is the magnetic permeability, ε r is the dielectric constant;

[0023] (2) Detection time window W

[0024] The detection time window is the time that the antenna continues to receive electromagnetic waves after transmitting electromagnetic waves. The detection time window is set according to the required detection depth and the electromagnetic properties of the medium. The calculation formula is as follows:

[0025]

[0026] Where h is the depth of the detection target, unit: m, W is the detection time window, unit: s;

[0027] (3) Repeat scan times N

[0028] The 3D ground penetrating radar averages the A-scans collected repeatedly at the sampling points and repeatedly collects N A-scans to reduce the noise amplitude. times, improving the signal-to-noise ratio by 10log 10 N times;

[0029] A roller rangefinder was used to accurately measure a 50m section to calibrate the vehicle-mounted DMI system. The test lanes were scanned one by one at a speed of 10km / h, fully covering the test area. The 3D frequency domain data matrix was saved and output, and the test lane number, starting stake number, and ending stake number corresponding to each data matrix were recorded.

[0030] Step S3, drilling calibration core samples; after the test is completed, drill a calibration core sample at the starting point, and then drill a calibration core sample every 1 km, numbering and marking the core samples in sequence according to 1, 2...m, and record the number of each calibration core sample, the corresponding pile number and the distance from the central median;

[0031] Step S4, data preprocessing: noise suppression is performed on the radar data, and a suppression threshold is set; the three-dimensional frequency domain data matrix is ​​converted into a time domain signal using an inverse discrete Fourier transform. The conversion equation is shown in Equation 3:

[0032]

[0033] Where, f c is the center frequency of the equivalent pulse of the transmitted signal, B is the bandwidth of the equivalent pulse of the transmitted signal, and the other parameters are constants.

[0034] The background clutter is removed by the average method, and the amplitude data is processed using Equation 4:

[0035]

[0036] Where A lmi The amplitude data of the signal in the mth row and ith column in the data collected by antenna number l;

[0037] The automatic line tracking tool is used to track the continuous interlayer boundary lines with the largest reflection signal amplitudes at the top and bottom of the structural layer in the radar image, and the time domain values ​​at the interlayer boundary lines at the top and bottom of the structural layer are exported to form a time domain data matrix.

[0038] Step S5, generating a thickness distribution matrix; combining the core sample thickness calibration data, back-calculating the dielectric constant of the structure layer of the test section according to Formula 3; using the calibrated dielectric constant, calculating the thickness of the structure layer of the test section according to Formula 4;

[0039]

[0040] Where h is the thickness of the asphalt layer (unit: m), h0 is the thickness of the calibration core sample (unit: m), and c is the speed of light (3×10 8 m / s), ε is the relative dielectric constant, t1 is the propagation time of the reflected electromagnetic wave to the ground (unit: s), and t2 is the propagation time of the reflected electromagnetic wave to the bottom of the asphalt layer (unit: s). t1 and t2 are obtained by tracking the reflected signals from the top and bottom of the asphalt structure layer in the radar image.

[0041] Step S6, synthesizing a full-section thickness distribution matrix; using MATLAB to compile a program based on the detection section stake number and detection lane number information recorded in each detection data, assigning the detection point in the detection data two-dimensional coordinates (x, y), where x is the stake number and y is the distance from the central median; based on the principle of arranging the two-dimensional coordinates (x, y) from small to large, readjusting the row and column positions of each point in each detection data; and combining the detection lane data from small to large according to the number of each detection lane data into a full-section thickness distribution matrix;

[0042] Step S7: Convert the full-section thickness distribution matrix into an RGB color distribution matrix; according to the qualified upper and lower limits of the structural layer thickness, the thickness values ​​below the qualified lower limit in the thickness distribution matrix are corresponding to (255, 0, 0) (red), and the thickness values ​​above the upper limit are corresponding to (0, 0, 255) (blue). The thickness values ​​between the two are converted by calculating the three primary color ratio according to the function shown in Formula 5: if the adjusted ratio is less than 0, 0 is used;

[0043]

[0044] Where h is the thickness of the asphalt layer (unit: cm), h d is the designed thickness of the asphalt layer (unit: cm), h dmin is the lower limit of the qualified value of the structural layer thickness (unit: cm), h dmax The upper limit of the qualified value of the structural layer thickness (unit: cm);

[0045] Step S8, drawing a thickness distribution cloud map; generating a corresponding coordinate axis area according to the detection area, and drawing an intuitive thickness distribution cloud map in the coordinate axis area according to the coordinate values ​​and RGB color values ​​of each point in the RGB color distribution matrix. According to the color changes in the map, the thickness change trend and thickness situation can be intuitively obtained;

[0046] Step S9, draw a pie chart of the proportions of each interval; count the area ratios of regions where the thickness values ​​are lower than the lower limit of the qualified value, higher than the upper limit, and between the two, and draw them as a pie chart. Based on the ratio below the lower limit of the qualified value, the ratio of weak areas with insufficient thickness of the structural layer can be accurately obtained.

[0047] As a further improvement to the technical solution of the present invention, the road thickness compensation method:

[0048] The calculation method of asphalt layer thickness compensation value is as follows:

[0049] (1) Divide the detection section units into 200m intervals

[0050] First, divide the inspection section into units and adjust the upper layer paving thickness value by unit;

[0051] (2) Calculate the thickness compensation value based on the thickness representative value requirement

[0052] The amount of full-section thickness data collected by radar is large, with each unit containing 563,380 points. The representative unit thickness value is calculated using the t-distribution model. When the degrees of freedom are greater than 120, the difference with the mean thickness value is less than 1‰, and can be considered equal. The thickness compensation value based on the representative value of the asphalt layer thickness can be calculated as follows:

[0053]

[0054] Where: ΔX L is the thickness compensation value required based on the thickness representative value; X d is the design value of the thickness of the lower asphalt layer; P1 is the allowable deviation of the representative value of the total thickness of the asphalt layer; X t is the design value of the total thickness of the asphalt layer; is the average thickness of the lower asphalt layer.

[0055] (3) Calculate the thickness compensation value based on the minimum thickness requirement.

[0056] The thickness compensation value based on the minimum thickness requirement is to ensure that the total thickness of the asphalt layer after the upper layer is added meets the minimum value of the specification acceptance standard, and is calculated as follows:

[0057] ΔX m =X d -P2×X t -X dmin

[0058] Where: ΔX m is the thickness compensation value based on the minimum thickness requirement (cm); P2 is the minimum allowable deviation of the total thickness of the asphalt layer; X dmin is the minimum thickness of the lower asphalt layer;

[0059] (4) Calculate the asphalt layer thickness compensation value.

[0060] The asphalt layer should meet the requirements of thickness representative value and thickness minimum value at the same time. The thickness compensation value should be the larger value of the thickness compensation value calculated based on the representative value and the minimum value, and calculated according to the following formula:

[0061] ΔX=max(ΔX L ,ΔX m )

[0062] Where: ΔX is the thickness compensation value (cm).

[0063] As a further improvement to the technical solution of the present invention, a method for dynamically adjusting the paving thickness is provided:

[0064] Based on the asphalt layer thickness compensation value, adjust the on-site paving thickness based on the designed thickness of the upper bearing layer, compensate for the asphalt layer thickness, and calculate the upper layer thickness adjustment value of each unit. The calculation method is:

[0065] (1) Calculate the upper layer thickness adjustment value of the detection path

[0066] The upper layer thickness adjustment value should meet the requirements of the required compensation thickness and the design thickness at the same time, and can be calculated as follows:

[0067] X u =max(ΔX+X ut ,Xut )

[0068] Where: X u is the thickness adjustment value of the upper layer of the detection path (cm); X ut Design thickness of the upper asphalt layer (cm);

[0069] (2) Calculate the upper layer thickness adjustment value of the unit

[0070] The paving thickness of the same cross section of asphalt layer paving construction is the same, and the thickness adjustment value of the unit needs to be determined according to the thickness adjustment value of each test lane; the width of the test lane is 1.5m, and the paving surface width of asphalt layer paving construction is generally greater than 1.5m. It is necessary to set up two or more test lanes to achieve full coverage detection of the paving surface. In order to ensure that the asphalt layer thickness of each test lane in the unit meets the requirements after adjustment, the upper layer thickness adjustment value of the unit is the maximum value of the upper layer thickness adjustment values ​​of each test lane, and is calculated according to the following formula:

[0071] X uw =max(X u1 ,X u2 ,...X ui ,...X un )

[0072] Where: X uw is the thickness adjustment value of the upper layer of the unit (cm), the same below; X ui is the thickness adjustment value of the upper layer of detection lane number i (cm);

[0073] (3) Calculate the adjustment value of the upper layer paving thickness of the unit

[0074] During actual paving construction, the loose paving coefficient of the upper layer should be determined according to the type of mixture and compaction process, and the adjustment value of the upper layer paving thickness of the unit should be calculated as follows:

[0075] X=K×X

[0076] X=K×X uw

[0077] Where: X is the adjustment value of the upper layer paving thickness of the unit (cm); K is the loose paving coefficient used in paving construction.

[0078] As a further improvement of the technical solution of the present invention, in step S1, the three-dimensional ground penetrating radar antenna array consists of 21 antennas, the spacing between each antenna is 0.75m, and the width of the area that can be covered by a single scan is 1.5m. The wider area to be inspected is divided into multiple adjacent detection channels according to the antenna detection width to ensure that there is no overlapping area in the scanning area and the entire section of the entire area to be inspected is covered.

[0079] As a further improvement to the technical solution of the present invention, in step S2, a roller rangefinder is used to accurately measure an area with a length of 50m to calibrate the DMI system, and its error is less than 1‰; to ensure positioning accuracy, during the detection process, the three-dimensional ground penetrating radar data pile number is recorded every 1km to ensure that the positioning error is less than 1m.

[0080] As a further improvement to the technical solution of the present invention, in step S2, an infrared rangefinder is used to adjust the distance between the inspection vehicle and the edge during the forward movement to ensure that the lateral distance that the inspection vehicle deviates from the inspection track is less than 0.3m.

[0081] As a further improvement to the technical solution of the present invention, in step S3, a core sample is drilled every 1 km.

[0082] As a further improvement to the technical solution of the present invention, in step S4, a threshold method is used to suppress noise, and an average method is used to remove background clutter.

[0083] As a further improvement to the technical solution of the present invention, in step S4, the three-dimensional frequency domain data matrix is ​​converted into a time domain data matrix by inverse discrete Fourier transform.

[0084] As a further improvement to the technical solution of the present invention, in step S4, by specifying the reflection signal symbols of the top and bottom of the structural layer and the time domain range in which they are located, an automatic line tracing tool is used to track the continuous inter-layer boundary lines with the largest reflection signal amplitudes at the top and bottom of the structural layer in the radar map.

[0085] As a further improvement to the technical solution of the present invention, in step S5, the dielectric constant of the coordinate structure layer can be inversely calculated based on the core sample thickness and the time domain data under the corresponding coordinates, and the dielectric constant value used for calculating the thickness of the calibration area is used. The time domain data matrix is ​​converted into a thickness data matrix. After calibration, the detection accuracy of the pavement structure layer reaches 0.002m.

[0086] As a further improvement to the technical solution of the present invention, in step S6, the two-dimensional coordinates (x, y) of the detection point in the detection data are assigned, and the row and column positions of each point in each channel of detection data are readjusted according to the principle of arranging the two-dimensional coordinates from small to large; and the data of each detection channel are combined into a full-section thickness distribution matrix according to the principle of arranging the numbers from small to large.

[0087] As a further improvement to the technical solution of the present invention, in step S7, the full-section thickness distribution matrix is ​​converted into an RGB color distribution matrix, thicknesses exceeding the upper or lower limit are represented by red or blue, and thickness values ​​between the two are represented by lighter gradient colors, and a thickness distribution cloud map is drawn using this RGB color distribution matrix.

[0088] As a further improvement to the technical solution of the present invention, in step S8, the coordinate axis corresponding to the detection area is generated, and an intuitive thickness distribution cloud map is drawn in the coordinate axis area according to the coordinate value and RGB color value of each point in the RGB color distribution matrix, so that the thickness change trend and thickness distribution of the entire cross-section of the structural layer can be intuitively obtained.

[0089] As a further improvement to the technical solution of the present invention, in step S9, the area ratio of regions below the lower limit of the qualified value, above the upper limit, and between the two is counted, so as to accurately obtain the proportion of weak areas with insufficient thickness of the structural layer and evaluate the overall distribution of the thickness of the structural layer.

[0090] In summary, the present invention has the following beneficial effects:

[0091] Comprehensive and accurate detection: By dividing the area to be inspected into detection lanes of appropriate width and using the multi-channel antenna of the three-dimensional ground penetrating radar to achieve full-section coverage scanning, more comprehensive and accurate information on the thickness of the pavement structure layer can be obtained, solving the problem of insufficient representativeness of existing detection methods.

[0092] High-precision data processing: Reasonable setting of detection parameters and the use of scientific data pre-processing methods, including noise suppression, background clutter removal, inverse discrete Fourier transform, etc., as well as accurate dielectric constant calibration and thickness calculation methods, improve the accuracy and reliability of data processing and make the thickness detection accuracy reach 0.002m.

[0093] Intuitive result presentation: Converting the full-section thickness distribution matrix into an RGB color distribution matrix and drawing a thickness distribution cloud map and a pie chart of each interval ratio can intuitively display the thickness change trend of the pavement structure layer, the thickness distribution, and the proportion of weak areas, providing an intuitive basis for subsequent corresponding treatment measures.

[0094] Improve road surface quality: Based on accurate test results, the road surface thickness can be effectively compensated and adjusted, the qualified rate of the road surface structure layer thickness distribution can be improved, early damage to the road surface due to insufficient thickness can be avoided, and the overall quality and service life of the road surface can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0096] Figure 1 This is a thickness distribution cloud diagram of an embodiment of the present invention;

[0097] Figure 2 A pie chart showing the proportions of each interval in an embodiment of the present invention;

[0098] Figure 3 A three-dimensional ground penetrating radar field operation diagram according to an embodiment of the present invention;

[0099] Figure 4 The result of adjusting the upper layer paving thickness of each unit in the road section 1 according to an embodiment of the present invention;

[0100] Figure 5 A thickness distribution cloud diagram of an asphalt pavement section (unit 3) using a thickness adjustment value according to an embodiment of the present invention;

[0101] Figure 6 This is a thickness distribution cloud diagram of a conventionally constructed asphalt pavement (unit 16) according to an embodiment of the present invention;

[0102] Figure 7 is the ratio of unqualified areas of asphalt layer thickness in the embodiment of the present invention;

[0103] Figure 8 The change in the mean unit thickness before and after paving of each road section in the embodiment of the present invention;

[0104] Figure 9 The thickness mean variation coefficient of the embodiment of the present invention is changed

[0105] Figure 10 This is a flow chart of a method for full-section non-destructive detection of pavement structure layer thickness and data processing based on three-dimensional ground penetrating radar according to an embodiment of the present invention. DETAILED DESCRIPTION

[0106] In order to make the purpose, features, and advantages of this application more obvious and easy to understand, the technical solutions in the embodiments of this application are clearly and completely described. Obviously, the embodiments described below are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0107] Reference Figure 10 A method for full-section non-destructive detection and data processing of pavement structure layer thickness based on three-dimensional ground penetrating radar includes the following steps:

[0108] (1) Divide the area to be inspected into inspection lanes with a width of 1.5 m. First, measure the width of the area to be inspected, w meters. Calculate and record the distance x meters between the inspection lane and the central median and the number of inspection lanes, n, using Formulas 1 and 2. Starting from x meters from the right edge of the area to be inspected, use a marker pen to divide the area to be inspected into n inspection lanes with a width of 1.5 m. Number the inspection lanes 1, 2, ..., n from right to left.

[0109]

[0110] x=(w-1-1.5×n) / 2 (2)

[0111] (2) Detection parameter settings:

[0112] 3D GPR is used to detect the thickness of pavement structures and provide high-resolution underground images. The appropriate radar frequency and scanning mode must be selected based on the specific conditions of the inspection object. When measuring structural layer thickness, the 3D GPR antenna primarily sets three parameters: sampling interval, detection time window, and the number of repeated scans (N). These parameters significantly influence the antenna's operation.

[0113] (1) Sampling interval Δx

[0114] The sampling interval refers to the horizontal distance between two sampling points in the direction of antenna travel. If a 3D ground penetrating radar uses an excessively large sampling interval, there will be "blank areas" between adjacent sampling points that are not covered by electromagnetic waves, resulting in missing valid information. If the sampling interval is too small, electromagnetic wave signals will overlap and interfere with each other. The sampling interval should be appropriately set based on the antenna center frequency and the dielectric constant of the medium. The calculation formula is as follows:

[0115]

[0116] Where Δx is the sampling interval, unit: m, c is the speed of light in vacuum, unit: m / s, f c is the center frequency of the antenna, unit: Hz, μ r is the magnetic permeability, ε r is the dielectric constant.

[0117] (2) Detection time window W

[0118] The detection window refers to the time that the antenna continues to receive electromagnetic waves after transmitting them. It is closely related to the antenna's detection depth. The longer the reception time, the greater the detection depth. However, using an excessively large detection window will reduce the antenna's detection speed. The detection window should be appropriately set based on the desired detection depth and the electromagnetic properties of the medium. The calculation formula is as follows:

[0119]

[0120] Where h is the depth of the detection target, unit: m, W is the detection time window, unit: s.

[0121] (3) Repeat scan times N

[0122] 3D GPR can effectively reduce signal noise and improve signal-to-noise ratio by averaging the A-scans collected repeatedly at the sampling points. Repeated collection of N A-scans can reduce the noise amplitude. times, improving the signal-to-noise ratio by 10log 10N times. However, when N increases to a certain level, the improvement in signal-to-noise ratio decreases rapidly. Also, the frequency of antenna A-scan acquisition is limited, typically 100-1000 kHz. Excessive repeated scans significantly reduce antenna acquisition efficiency. Typically, N is set between 8 and 32.

[0123] A roller rangefinder was used to accurately measure a 50m section to calibrate the vehicle-mounted DMI system. The test lanes were scanned one by one at a speed of 10km / h, fully covering the test area. The 3D frequency domain data matrix was saved and output, and the corresponding test lane number, starting stake number, and ending stake number for each data matrix were recorded.

[0124] (3) Drill calibration core samples. After the test is completed, drill a calibration core sample at the starting point, and then drill a calibration core sample every 1 km. Number and mark the core samples in sequence of 1, 2, ... m, and record the number of each calibration core sample, the corresponding pile number, and the distance from the central median strip;

[0125] (4) Data preprocessing. Noise suppression is performed on radar data, and the suppression threshold is generally set to 10 dB. The three-dimensional frequency domain data matrix is ​​converted into a time domain signal using the inverse discrete Fourier transform. The conversion equation is shown in Equation 3:

[0126]

[0127] Where, f c is the center frequency of the equivalent pulse of the transmitted signal, B is the bandwidth of the equivalent pulse of the transmitted signal, and the other parameters are constants.

[0128] The background clutter is removed by the average method, and the amplitude data is processed using Equation 4:

[0129]

[0130] Where A lmi The amplitude data of the signal in the mth row and ith column in the data collected by antenna number l.

[0131] The automatic line tracking tool is used to track the continuous interlayer boundary lines with the largest reflection signal amplitudes at the top and bottom of the structural layer in the radar image, and the time domain values ​​at the interlayer boundary lines at the top and bottom of the structural layer are exported to form a time domain data matrix.

[0132] (5) Generate thickness distribution matrix. Combined with the core sample thickness calibration data, the dielectric constant of the structural layer of the test section is calculated according to formula 3; using the calibrated dielectric constant, the thickness of the structural layer of the test section is calculated according to formula 4;

[0133]

[0134] Where h is the thickness of the asphalt layer (unit: m), h0 is the thickness of the calibration core sample (unit: m), and c is the speed of light (3×10 8 m / s), ε is the relative dielectric constant, t1 is the propagation time of the reflected electromagnetic wave to the ground (unit: s), and t2 is the propagation time of the reflected electromagnetic wave to the bottom of the asphalt layer (unit: s). t1 and t2 are obtained by tracking the reflected signals from the top and bottom of the asphalt structure layer in the radar image.

[0135] (6) Synthesize the full-section thickness distribution matrix. Based on the information such as the detection road section pile number and detection lane number recorded in each detection data, use MATLAB to compile a program to assign the two-dimensional coordinates (x, y) of the detection points in the detection data (x is the pile number, y is the distance from the central dividing strip). According to the principle of arranging the two-dimensional coordinates (x, y) from small to large, readjust the row and column positions of each point in each detection data. Based on the number of each detection lane data, combine the detection lane data from small to large to form a full-section thickness distribution matrix;

[0136] (7) Convert the full-section thickness distribution matrix into an RGB color distribution matrix. According to the upper and lower limits of the structural layer thickness, the thickness values ​​below the lower limit of the qualified value in the thickness distribution matrix correspond to (255, 0, 0) (red), and the thickness values ​​above the upper limit correspond to (0, 0, 255) (blue). The thickness values ​​between the two are converted by calculating the ratio of the three primary colors according to the function shown in Formula 5 (when the adjusted ratio is less than 0, 0 is used);

[0137]

[0138] Where h is the thickness of the asphalt layer (unit: cm), h d is the designed thickness of the asphalt layer (unit: cm), h dmin is the lower limit of the qualified value of the structural layer thickness (unit: cm), h dmax It is the upper limit of the qualified value of the structural layer thickness (unit: cm).

[0139] (8) Draw a thickness distribution cloud map, such as Figure 1 As shown, the corresponding coordinate axis area is generated according to the detection area. According to the coordinate value and RGB color value of each point in the RGB color distribution matrix, an intuitive thickness distribution cloud map is drawn in the coordinate axis area. According to the color change in the map, the trend of thickness change and thickness situation can be intuitively obtained;

[0140] (9) Draw a pie chart of the proportions of each interval, such as Figure 2 The area ratios of thickness values ​​below the lower limit of the qualified value, above the upper limit, and between the two are statistically analyzed and plotted as a pie chart. Based on the ratio below the lower limit of the qualified value, the proportion of weak areas with insufficient thickness in the structural layer can be accurately obtained.

[0141] Pavement thickness compensation method:

[0142] Based on the accurate distribution of the underlying layer thickness, the thickness of the upper layer can be appropriately increased to compensate for any thinner underlying layer, based on design compliance requirements, to ensure the total thickness meets the requirements. This method is known as the pavement thickness compensation method. This method is applicable not only to asphalt pavements, but also to various structural layers of other pavement types. The asphalt layer thickness compensation value is calculated as follows:

[0143] (1) The detection section units are divided into 200m intervals.

[0144] First, the inspection section is divided into units, and the upper layer paving thickness is adjusted in each unit. When dividing the units, it is considered that frequent adjustments to the paving thickness increase the difficulty of paving construction operations, and that a too short construction working surface is not conducive to adjusting the road surface flatness and segregation. Therefore, the upper layer paving thickness should not be adjusted too frequently. However, if the unit division length is too long, the upper layer paving thickness in longer areas needs to be adjusted to compensate for the thickness of the asphalt structural layer in thinner areas, which makes it difficult to ensure economic efficiency. Taking all these factors into consideration and combining them with the characteristics of the on-site construction, adjusting the paving construction thickness in units of 200m can better balance the quality and economic efficiency of the asphalt layer construction.

[0145] (2) Calculate the thickness compensation value based on the thickness representative value requirement.

[0146] The radar collects a large amount of full-section thickness data, with each unit (200m) containing 563,380 points. The representative unit thickness value is calculated using the t-distribution model. When the degrees of freedom are extremely high (>120), the difference with the mean thickness value is minimal (<1‰), and can be considered equal. The thickness compensation value based on the required representative asphalt layer thickness value can be calculated as follows:

[0147]

[0148] Where: ΔX L is the thickness compensation value required based on the thickness representative value; X d is the design value of the thickness of the lower asphalt layer; P1 is the allowable deviation of the representative value of the total thickness of the asphalt layer; X t is the design value of the total thickness of the asphalt layer; is the average thickness of the lower asphalt layer.

[0149] (3) Calculate the thickness compensation value based on the minimum thickness requirement.

[0150] The thickness compensation value based on the minimum thickness requirement is to ensure that the total thickness of the asphalt layer after the upper layer is added meets the minimum value of the specification acceptance standard, and is calculated as follows:

[0151] ΔX m =X d-P2×X t -X dmin

[0152] Where: ΔX m is the thickness compensation value based on the minimum thickness requirement (cm); P2 is the minimum allowable deviation of the total thickness of the asphalt layer; X dmin It is the minimum thickness of the lower asphalt layer.

[0153] (4) Calculate the asphalt layer thickness compensation value.

[0154] The asphalt layer should meet the requirements of thickness representative value and thickness minimum value at the same time. The thickness compensation value should be the larger value of the thickness compensation value calculated based on the representative value and the minimum value, and calculated according to the following formula:

[0155] ΔX=max(ΔX L ,ΔX m )

[0156] Where: ΔX is the thickness compensation value (cm).

[0157] Dynamic adjustment method of paving thickness:

[0158] Based on the asphalt layer thickness compensation value, adjust the on-site paving thickness based on the designed thickness of the upper bearing layer, compensate for the asphalt layer thickness, and calculate the upper layer thickness adjustment value of each unit. The calculation method is:

[0159] (1) Calculate the thickness adjustment value of the upper layer of the detection path.

[0160] The upper layer thickness adjustment value should meet the requirements of the required compensation thickness and the design thickness at the same time, and can be calculated as follows:

[0161] X u =max(ΔX+X ut ,X ut )

[0162] Where: X u is the thickness adjustment value of the upper layer of the detection path (cm); X ut Design thickness of upper asphalt layer (cm).

[0163] (2) Calculate the thickness adjustment value of the upper layer of the unit.

[0164] The same paving thickness is applied to the same cross section of asphalt layer paving construction, and the thickness adjustment value of the unit needs to be determined based on the thickness adjustment value of each test lane. The width of a single test lane is 1.5m, and the paving surface width of asphalt layer paving construction is generally greater than 1.5m. It is necessary to set up two or more test lanes to achieve full coverage inspection of the paving surface. To ensure that the asphalt layer thickness of each test lane in the unit meets the requirements after adjustment, the upper layer thickness adjustment value of the unit is the maximum value of the upper layer thickness adjustment values ​​of each test lane, and is calculated as follows:

[0165] X uw =max(X u1 ,X u2 ,...X ui ,...X un )

[0166] Where: X uw is the thickness adjustment value of the upper layer of the unit (cm), the same below; X ui The thickness adjustment value of the upper layer of detection lane number i (cm).

[0167] (3) Calculate the adjustment value of the upper layer paving thickness of the unit.

[0168] During actual paving construction, the loose paving coefficient of the upper layer should be determined according to the type of mixture and compaction process, and the adjustment value of the upper layer paving thickness of the unit should be calculated as follows:

[0169] X=K×X uw

[0170] Where: X is the adjustment value of the upper layer paving thickness of the unit (cm); K is the loose paving coefficient used in paving construction.

[0171] According to the calculation process of the adjustment value of the upper layer paving thickness of the unit, in theory, it can ensure that the representative value within the unit area and the minimum thickness of all measuring points meet the requirements.

[0172] Case:

[0173] 1. Project Overview

[0174] A newly constructed expressway in Guangdong Province adopts the standards of a two-way four-lane expressway. The specific structure of the asphalt layer in the inspection section is: 4.5cm GAC-16C upper layer; 5.5cm GAC-20C middle layer; and 8cm GAC-25C lower layer. The design requires a 0.9cm (5%) tolerance for the representative value of the total asphalt layer thickness, which is required to be greater than 17.1cm; the tolerance for the total asphalt layer thickness at a single inspection point is 1.8cm (10%), which is required to be greater than 16.2cm. After the completion of the middle layer construction, the thickness of the upper layer paving will be adjusted. The upper layer mixture will be paved using a full-width construction method with a loose paving coefficient of 1.24.

[0175] 2. Detection plan

[0176] Before and after the top layer was paved, asphalt layer thickness was measured using 3D ground-penetrating radar on the right section (K0+000-K6+000), a total length of 6 km. The first half (Section 1) employed dynamic thickness adjustment technology to adjust the top layer thickness. This 3 km section was divided into 15 sections, numbered 1-15. The second half (Section 2) underwent normal construction, with the top layer paved conventionally according to the designed thickness. This 3 km section was divided into 15 sections, numbered 16-30.

[0177] The cross-section width of a single lane is 7.5m, with 5 detection lanes set up to perform full cross-section coverage scanning. The sampling interval of the scanning parameters is set to 2.5cm and the time window is 35ns. The on-site detection is shown in the figure below. Figure 3 shown.

[0178] 3. Calculation of paving thickness adjustment value

[0179] Using a radar signal processing program, the continuous trajectory of the ground signal and the continuous trajectory of the reflected signal from the bottom of the asphalt layer are automatically tracked. The signal time values ​​t1 and t2 are derived. Combined with the core sample calibration data, the dielectric constant is calibrated according to equations (3-1) and (3-2). Taking unit 13 as an example, the calibration data for this section is shown in Table 3-1 below.

[0180] Table 3-1 Core sample calibration data summary

[0181]

[0182] Therefore, the unit's dielectric constant was calibrated to 5.64, and the asphalt layer thickness in all areas of the unit was calculated using this calibrated dielectric constant. To verify the accuracy and precision of the asphalt layer thickness test results, four core samples were drilled within the test section for comparison and verification.

[0183] Table 3-2 Core sample data comparison and verification

[0184]

[0185] Comparison and verification of core sample data revealed an average absolute deviation of 0.1 cm and a maximum of 0.2 cm, indicating that the asphalt layer thickness test results had an error of less than 2 mm, essentially meeting the accuracy requirements for calculating the upper layer paving thickness. The thickness of the lower asphalt layer and the adjustment value for the upper layer paving thickness were calculated for unit 13. The results are shown in Table 3-3.

[0186] Table 3-3 Summary of calculation results of unit 13 test data

[0187]

[0188] The upper layer thickness adjustment value of unit 13 is 6.3cm. The thickness compensation value and upper layer thickness adjustment value of units 1-15 are calculated according to the calculation method shown in unit 13. The paving thickness adjustment results of each unit in section 1 are as follows Figure 4 shown.

[0189] Depend on Figure 4 The average thickness compensation value for all units in Section 1 is 0.4 cm, with a maximum of 0.6 cm and a minimum of -0.3 cm. The average paving thickness adjustment value is 6.1 cm, with a maximum of 6.3 cm. Of the 15 units in Section 1, 14 require an increase in the upper paving thickness to compensate for the total asphalt layer thickness. Unit 11 has a compensation value of -0.3 cm, but to ensure the upper layer thickness meets the design requirements, the upper layer thickness adjustment value is retained at the design value of 4.5 cm, corresponding to a paving thickness of 5.6 cm.

[0190] In each unit section, the recommended upper layer thickness adjustment (not considering the loose paving coefficient) is a maximum of 5.1 cm, with a compensation value of 0.6 cm. The "Highway Engineering Quality Inspection and Assessment Standard" (JTG F80 / 1-2004) requires that the elevation deviation of the longitudinal section of the highway pavement be less than 1.5 cm. Therefore, adjusting the paving thickness of the upper layer will affect the longitudinal section elevation within the allowable range.

[0191] 4. Engineering application effect evaluation

[0192] Element thickness analysis

[0193] According to the design requirements, the design value of the total thickness of the asphalt layer is 18cm, and the allowable deviation of the extreme value of a single measuring point is 1.8cm, that is, the thickness value of all measuring points is required to be greater than 16.2cm, and the allowable deviation of the representative value of the regional thickness is 0.9cm, that is, it is required to be greater than 17.1cm.

[0194] The design value of the thickness of the underlying asphalt layer is 13.5cm. For the sake of convenience, the qualified thickness of the underlying asphalt layer is determined by the same deviation standard as the total thickness of the asphalt layer, that is, the thickness values ​​of all measuring points are required to be greater than 11.7cm, and the representative value of the regional thickness is required to be greater than 12.6cm.

[0195] The thickness distribution of asphalt layer is studied from the single unit thickness distribution cloud map. The thickness distribution cloud maps of asphalt layer before and after paving of unit 3 in section 1 are as follows: Figure 5 The asphalt layer thickness distribution cloud diagrams before and after paving of unit 16 in section 2 are as follows: Figure 6 shown.

[0196] The thickness compensation value of unit 3 is 0.5cm, and the upper layer is paved according to the upper layer paving thickness adjustment value of 6.2cm. Unit 16 is paved according to the paving thickness corresponding to the design thickness of 5.6cm.

[0197] The asphalt layer thickness distribution cloud map shows that the thickness distribution of the lower asphalt pavement in Units 3 and 16 shows significant variability, mainly reflected in the appearance of a thin transverse thickness band (red area) along the driving direction, which is mainly related to factors such as the paving equipment used in on-site construction and the flatness of the base layer.

[0198] Unit 16 paved the upper structure according to the designed thickness. The thickness distribution of the formed asphalt layer structure still showed obvious horizontal strip-like fluctuations. Comparison before and after construction showed that the thinner areas of the lower layer were still unqualified, and there was even a trend of diffusion in some local areas. The unqualified rate increased slightly from 8.6% to 9.1%. The representative value of the lower layer thickness was 13.6cm, and the representative value of the upper layer thickness after paving was 18.1cm, both meeting the design requirements.

[0199] After the recommended paving thickness value was used to adjust the pavement construction of Unit 3, the thin strip-shaped areas of the asphalt structural layer were significantly improved, becoming qualified areas slightly lower than the design thickness but still meeting the design requirements. The overall qualified rate was significantly improved, and the unqualified rate dropped sharply from 8.3% to 0.6%; the representative value of the lower layer thickness was 13.4cm, and the representative value of the upper layer thickness was 18.4cm, both of which met the design requirements.

[0200] 5. Analysis of road section thickness qualification rate

[0201] Compare the changes in the qualified rate of thickness of each unit before and after the upper layer paving of Section 1 and Section 2, and evaluate the improvement of the thin thickness areas of Section 1 and Section 2, such as Figure 7 shown.

[0202] Before construction of the upper layer, the average percentage of unqualified pavement thickness in each unit of Section 1 was 7.9%, with a maximum of 12.0% and a minimum of 0%. After the upper layer was constructed using the recommended thickness adjustment values, the average percentage of unqualified areas was 1.3%, with a maximum of 1.9% and a minimum of 0%. The qualified pavement thickness rate reached over 98%, a 6.6% increase compared to the qualified rate of the lower layer thickness before construction.

[0203] Before construction of the upper layer, the average percentage of unqualified pavement thickness in each unit of Section 2 was 6.8%, with a maximum of 10.8% and a minimum of 3.9%. After construction using the conventional design thickness, the average percentage of unqualified areas was 6.9%, with a maximum of 10.9% and a minimum of 3.4%. The pavement thickness compliance rate before and after construction did not change significantly, falling short of 95%.

[0204] Comparing the two construction methods clearly demonstrates that the dynamic adjustment of the upper layer thickness effectively improves the pavement construction thickness qualification rate, ensuring that the qualified percentage of the total asphalt layer thickness reaches over 98%, meeting project requirements and preventing premature failure of the asphalt pavement due to insufficient asphalt layer thickness. Furthermore, thickness compensation is not required for sections with good thickness distribution, which helps ensure project economic efficiency.

[0205] 6. Evaluation of road section thickness uniformity

[0206] Calculate the average thickness of the asphalt structure layer of each unit in each section before and after the upper layer of Section 1 and Section 2 is paved, draw the thickness fluctuation curve, and compare them. Figure 8 shown.

[0207] The analysis shows that:

[0208] (1) The average thickness of the upper layer before paving in Section 1 is 13.5 cm, with a range of 13.1 to 14.0 cm; the average thickness of the upper layer before paving in Section 2 is 13.5 cm, with a range of 13.1 to 13.9 cm. This indicates that before paving the upper layer, each unit in Sections 1 and 2 meets the thickness representative value requirements, and the thickness average, variation range and other indicators are relatively close, indicating that the paving construction technology levels are similar.

[0209] (2) The average thickness of the upper layer after paving in Section 1 was 18.5 cm, with a range of 18.2 to 18.6 cm. The average thickness of the upper layer after paving in Section 2 was 18.1 cm, with a range of 17.7 to 18.4 cm. The average thickness of Section 1 was slightly higher than that of Section 2, mainly due to the thickness compensation of each unit. The variation range of the average thickness of each unit in Section 1 was smaller than that of each unit in Section 2, indicating that after the upper layer was paved, the average thickness of each unit in Section 1 was more stable than that in Section 2.

[0210] In order to further study the stability of the mean value of each unit thickness, the coefficient of variation of the mean value of each unit thickness of Section 1 and Section 2 before and after the upper layer is paved is calculated, as shown in the following example: Figure 9 shown.

[0211] Before the top layer was laid, the coefficient of variation of the thickness of each unit in Section 1 was 1.9%, essentially the same as that of Section 2. After the top layer was laid, the coefficient of variation of the thickness of the asphalt structural layer in Section 1 was approximately 0.8%, significantly lower than the 1.9% coefficient of variation in Section 2 after construction. This demonstrates that the overall uniformity of the pavement thickness distribution in Section 1 is superior to that in Section 2 and further demonstrates that the use of dynamic thickness adjustment technology can effectively reduce the variability of asphalt pavement thickness distribution during construction and improve overall thickness uniformity.

[0212] Specifically, in the present embodiment, in step (1), the three-dimensional ground penetrating radar antenna array consists of 21 antennas, the spacing between each antenna is 0.75m, and the width of the area that can be covered by a single scan is 1.5m. The wider area to be inspected is divided into multiple adjacent detection channels according to the antenna detection width to ensure that there is no overlapping area in the scanning area and the entire section of the entire area to be inspected is covered.

[0213] Specifically, in this embodiment, in step (2), a roller rangefinder is used to accurately measure a 50m area to calibrate the DMI system. Its ranging accuracy is high, with an error of less than 1‰. To ensure positioning accuracy, the 3D ground penetrating radar data stake number is recorded every 1km during the detection process to ensure that the positioning error is less than 1m.

[0214] Specifically, in the present embodiment, in step (2), an infrared rangefinder is used to adjust the distance between the detection vehicle and the edge during the forward movement, which can ensure that the lateral distance of the detection vehicle deviating from the detection path is less than 0.3m, thereby greatly reducing the omission and overlap of the scanning area caused by the lateral deviation of the detection vehicle from the detection path.

[0215] Specifically, in the present embodiment, in step (3), a core sample is drilled every 1 km to better ensure the representativeness of the calibration core sample, avoid too many calibration core samples in the same section and excessive damage to the pavement structure, and prevent insufficient calibration core samples in some sections, poor representativeness of the calibration dielectric constant, and large deviations in the thickness calculation results.

[0216] Specifically, in the embodiment, in step (4), a threshold method is used to suppress noise, and an average method is used to remove background clutter, which can remove most of the noise and clutter interference signals and improve the signal-to-noise ratio.

[0217] Specifically, in the present embodiment, in step (4), the three-dimensional frequency domain data matrix can be converted into an intuitive and easy-to-understand time domain data matrix through inverse discrete Fourier transform.

[0218] Specifically, in the present embodiment, in step (4), by specifying the sign (positive or negative) of the reflection signal at the top and bottom of the structural layer and the time domain range in which they are located, an automatic line tracking tool is used to track the continuous inter-layer boundary line with the largest reflection signal amplitude at the top and bottom of the structural layer in the radar map. The process is easy to operate, accurate, and highly automated.

[0219] Specifically, in the present embodiment, in step (5), the dielectric constant of the coordinate structure layer can be inversely calculated based on the core sample thickness and the time domain data at the corresponding coordinates, and the dielectric constant value used for calculating the thickness of the calibration area is used. The time domain data matrix is ​​converted into a thickness data matrix. After calibration, the detection accuracy of the pavement structure layer thickness reaches 0.002m.

[0220] Specifically, in the present embodiment, in step (6), the two-dimensional coordinates (x, y) of the detection point in the detection data are assigned (x, y) (x is the stake number, y is the distance from the central dividing strip), and the row and column positions of each point in each channel of detection data are readjusted according to the principle of arranging the two-dimensional coordinates from small to large. According to the principle of arranging the numbers from small to large, the data of each detection channel are combined into a full-section thickness distribution matrix. The process is completed by computer with a high degree of automation, providing basic data for the subsequent drawing of the full-section thickness distribution cloud map of the structural layer.

[0221] Specifically, in the present embodiment, in step (7), the full-section thickness distribution matrix is ​​converted into an RGB color distribution matrix, and thicknesses exceeding the upper or lower limit are represented by red or blue, and thickness values ​​between the upper and lower limits are represented by lighter gradient colors. The thickness distribution cloud map drawn using this RGB color distribution matrix has the advantages of obvious thickness change trends and highlighting the over-limit values.

[0222] Specifically, in the present embodiment, in step (8), a coordinate axis corresponding to the detection area is generated, and an intuitive thickness distribution cloud map is drawn in the coordinate axis area according to the coordinate values ​​and RGB color values ​​of each point in the RGB color distribution matrix, so that the thickness change trend and thickness distribution of the entire cross-section of the structural layer can be intuitively obtained.

[0223] Specifically, in the present embodiment, in step (9), the area ratios of regions below the lower limit of the qualified value, above the upper limit, and between the upper limit and the lower limit are counted, so as to accurately obtain the ratio of weak regions with insufficient thickness of the structural layer and evaluate the overall distribution of the thickness of the structural layer.

[0224] In summary, compared with the prior art, the present invention has the following beneficial effects:

[0225] Comprehensive and accurate detection: By dividing the area to be inspected into detection lanes of appropriate width and using the multi-channel antenna of the three-dimensional ground penetrating radar to achieve full-section coverage scanning, more comprehensive and accurate information on the thickness of the pavement structure layer can be obtained, solving the problem of insufficient representativeness of existing detection methods.

[0226] High-precision data processing: Reasonable setting of detection parameters and the use of scientific data pre-processing methods, including noise suppression, background clutter removal, inverse discrete Fourier transform, etc., as well as accurate dielectric constant calibration and thickness calculation methods, improve the accuracy and reliability of data processing and make the thickness detection accuracy reach 0.002m.

[0227] Intuitive result presentation: Converting the full-section thickness distribution matrix into an RGB color distribution matrix and drawing a thickness distribution cloud map and a pie chart of each interval ratio can intuitively display the thickness change trend of the pavement structure layer, the thickness distribution, and the proportion of weak areas, providing an intuitive basis for subsequent corresponding treatment measures.

[0228] Improve road surface quality: Based on accurate test results, the road surface thickness can be effectively compensated and adjusted, the qualified rate of the road surface structure layer thickness distribution can be improved, early damage to the road surface due to insufficient thickness can be avoided, and the overall quality and service life of the road surface can be improved.

[0229] The technical solutions provided by the embodiments of the present invention are introduced in detail above. Specific examples are used herein to illustrate the principles and implementation methods of the embodiments of the present invention. The description of the above embodiments is only applicable to help understand the principles of the embodiments of the present invention. At the same time, for those skilled in the art, according to the embodiments of the present invention, there may be changes in the specific implementation methods and application scopes. In summary, the contents of this specification should not be understood as limiting the present invention.

Claims

1. A method for full-section non-destructive detection and data processing of pavement structure layer thickness based on three-dimensional ground penetrating radar, characterized in that: The following steps are involved: Step S1: Divide the area to be inspected into inspection lanes with a width of 1.5 m. First, measure the width of the area to be inspected (w meters), and calculate and record the distance (x meters) between the inspection lane and the central median strip and the number of inspection lanes (n) according to Formulas 1 and 2. Starting from x meters away from the right edge of the area to be inspected, use a marker to divide the area to be inspected into n inspection lanes with a width of 1.5 meters. Number the inspection lanes 1, 2, ..., n from right to left. x=(w-1-1.5×n) / 2 (2) Step S2: Setting detection parameters: 3D GPR is used to detect the thickness of pavement structure layers and provide high-resolution underground images. The appropriate radar frequency and scanning mode must be selected based on the specific conditions of the detection object. During the process of detecting the thickness of the structure layer, the 3D GPR antenna must set three antenna parameters: sampling interval, detection time window, and the number of repeated scans (N). (1) Sampling interval Δx The sampling interval is the horizontal distance between two sampling points in the direction of antenna travel. The sampling interval is set according to the antenna center frequency and the dielectric constant of the medium. The calculation formula is as follows: Where Δx is the sampling interval, unit: m, c is the speed of light in vacuum, unit: m / s, f c is the center frequency of the antenna, unit: Hz, μ r is the magnetic permeability, ε r is the dielectric constant; (2) Detection time window W The detection time window is the time that the antenna continues to receive electromagnetic waves after transmitting electromagnetic waves. The detection time window is set according to the required detection depth and the electromagnetic properties of the medium. The calculation formula is as follows: Where h is the depth of the detection target, unit: m, W is the detection time window, unit: s; (3) Repeat scan times N The 3D ground penetrating radar averages the A-scans collected repeatedly at the sampling points and repeatedly collects N A-scans to reduce the noise amplitude. times, improving the signal-to-noise ratio by 10log 10 N times; A roller rangefinder was used to accurately measure a 50m section to calibrate the vehicle-mounted DMI system. The test lanes were scanned one by one at a speed of 10km / h, fully covering the test area. The 3D frequency domain data matrix was saved and output, and the test lane number, starting stake number, and ending stake number corresponding to each data matrix were recorded. Step S3, drilling calibration core samples; after the test is completed, drill a calibration core sample at the starting point, and then drill a calibration core sample every 1 km, numbering and marking the core samples in sequence according to 1, 2...m, and record the number of each calibration core sample, the corresponding pile number and the distance from the central median; Step S4, data preprocessing: noise suppression is performed on the radar data, and a suppression threshold is set; the three-dimensional frequency domain data matrix is ​​converted into a time domain signal using an inverse discrete Fourier transform. The conversion equation is shown in Equation 3: Where, f c is the center frequency of the equivalent pulse of the transmitted signal, B is the bandwidth of the equivalent pulse of the transmitted signal, and the other parameters are constants. The background clutter is removed by the average method, and the amplitude data is processed using Equation 4: Where A lmi The amplitude data of the signal in the mth row and ith column in the data collected by antenna number l; The automatic line tracking tool is used to track the continuous interlayer boundary lines with the largest reflection signal amplitudes at the top and bottom of the structural layer in the radar image, and the time domain values ​​at the interlayer boundary lines at the top and bottom of the structural layer are exported to form a time domain data matrix. Step S5, generating a thickness distribution matrix; combining the core sample thickness calibration data, back-calculating the dielectric constant of the structure layer of the test section according to Formula 3; using the calibrated dielectric constant, calculating the thickness of the structure layer of the test section according to Formula 4; Where h is the thickness of the asphalt layer (unit: m), h0 is the thickness of the calibration core sample (unit: m), and c is the speed of light (3×10 8 m / s), ε is the relative dielectric constant, t1 is the propagation time of the reflected electromagnetic wave to the ground (unit: s), and t2 is the propagation time of the reflected electromagnetic wave to the bottom of the asphalt layer (unit: s). t1 and t2 are obtained by tracking the reflected signals from the top and bottom of the asphalt structure layer in the radar image. Step S6, synthesizing a full-section thickness distribution matrix; using MATLAB to compile a program based on the detection section stake number and detection lane number information recorded in each detection data, assigning the detection point in the detection data two-dimensional coordinates (x, y), where x is the stake number and y is the distance from the central median; based on the principle of arranging the two-dimensional coordinates (x, y) from small to large, readjusting the row and column positions of each point in each detection data; and combining the detection lane data from small to large according to the number of each detection lane data into a full-section thickness distribution matrix; Step S7: Convert the full-section thickness distribution matrix into an RGB color distribution matrix; according to the qualified upper and lower limits of the structural layer thickness, the thickness values ​​below the qualified lower limit in the thickness distribution matrix are corresponding to (255, 0, 0) (red), and the thickness values ​​above the upper limit are corresponding to (0, 0, 255) (blue). The thickness values ​​between the two are converted by calculating the three primary color ratio according to the function shown in Formula 5: if the adjusted ratio is less than 0, 0 is used; Where h is the thickness of the asphalt layer (unit: cm), h d is the designed thickness of the asphalt layer (unit: cm), h dmin is the lower limit of the qualified value of the structural layer thickness (unit: cm), h dmax The upper limit of the qualified value of the structural layer thickness (unit: cm); Step S8, drawing a thickness distribution cloud map; generating a corresponding coordinate axis area according to the detection area, and drawing an intuitive thickness distribution cloud map in the coordinate axis area according to the coordinate values ​​and RGB color values ​​of each point in the RGB color distribution matrix. According to the color changes in the map, the thickness change trend and thickness situation can be intuitively obtained; Step S9, draw a pie chart of the proportions of each interval; count the area ratios of regions where the thickness values ​​are lower than the lower limit of the qualified value, higher than the upper limit, and between the two, and draw them as a pie chart. Based on the ratio below the lower limit of the qualified value, the ratio of weak areas with insufficient thickness of the structural layer can be accurately obtained.

2. The method for full-section nondestructive detection and data processing of pavement structure layer thickness based on three-dimensional ground penetrating radar according to claim 1 is characterized in that: Also includes: Pavement thickness compensation method: The calculation method of asphalt layer thickness compensation value is as follows: (1) Divide the detection section units into 200m intervals First, divide the inspection section into units and adjust the upper layer paving thickness value by unit; (2) Calculate the thickness compensation value based on the thickness representative value requirement The amount of full-section thickness data collected by radar is large, with each unit containing 563,380 points. The representative unit thickness value is calculated using the t-distribution model. When the degrees of freedom are greater than 120, the difference with the mean thickness value is less than 1‰, and can be considered equal. The thickness compensation value based on the representative value of the asphalt layer thickness can be calculated as follows: Where: ΔX L is the thickness compensation value required based on the thickness representative value; X d is the design value of the thickness of the lower asphalt layer; P1 is the allowable deviation of the representative value of the total thickness of the asphalt layer; X t is the design value of the total thickness of the asphalt layer; X d is the average thickness of the lower asphalt layer. (3) Calculate the thickness compensation value based on the minimum thickness requirement. The thickness compensation value based on the minimum thickness requirement is to ensure that the total thickness of the asphalt layer after the upper layer is added meets the minimum value of the specification acceptance standard, and is calculated as follows: ΔX m =X d -P2×X t -X dmin Where: ΔX m is the thickness compensation value based on the minimum thickness requirement (cm); P2 is the minimum allowable deviation of the total thickness of the asphalt layer; X dmin is the minimum thickness of the lower asphalt layer; (4) Calculate the asphalt layer thickness compensation value. The asphalt layer should meet the requirements of thickness representative value and thickness minimum value at the same time. The thickness compensation value should be the larger value of the thickness compensation value calculated based on the representative value and the minimum value, and calculated according to the following formula: ΔX=max(ΔX L ,ΔX m ) Where: ΔX is the thickness compensation value (cm).

3. The method for full-section nondestructive detection and data processing of pavement structure layer thickness based on three-dimensional ground penetrating radar according to claim 1 is characterized in that: Also includes: Dynamic adjustment method of paving thickness: Based on the asphalt layer thickness compensation value, adjust the on-site paving thickness based on the designed thickness of the upper bearing layer, compensate for the asphalt layer thickness, and calculate the upper layer thickness adjustment value of each unit. The calculation method is: (1) Calculate the upper layer thickness adjustment value of the detection path The upper layer thickness adjustment value should meet the requirements of the required compensation thickness and the design thickness at the same time, and can be calculated as follows: X u =max(ΔX+X ut ,X ut ) Where: X u is the thickness adjustment value of the upper layer of the detection path (cm); X ut Design thickness of the upper asphalt layer (cm); (2) Calculate the upper layer thickness adjustment value of the unit The paving thickness of the same cross section of asphalt layer paving construction is the same, and the thickness adjustment value of the unit needs to be determined according to the thickness adjustment value of each test lane; the width of the test lane is 1.5m, and the paving surface width of asphalt layer paving construction is generally greater than 1.5m. It is necessary to set up two or more test lanes to achieve full coverage detection of the paving surface. In order to ensure that the asphalt layer thickness of each test lane in the unit meets the requirements after adjustment, the upper layer thickness adjustment value of the unit is the maximum value of the upper layer thickness adjustment values ​​of each test lane, and is calculated according to the following formula: X uw =max(X u1 ,X u2 ,...X ui ,...X un ) Where: X uw is the thickness adjustment value of the upper layer of the unit (cm), the same below; X ui is the thickness adjustment value of the upper layer of detection lane number i (cm); (3) Calculate the adjustment value of the upper layer paving thickness of the unit During actual paving construction, the loose paving coefficient of the upper layer should be determined according to the type of mixture and compaction process, and the adjustment value of the upper layer paving thickness of the unit should be calculated as follows: X=K×X Where: X is the adjustment value of the upper layer paving thickness of the unit (cm); K is the loose paving coefficient used in paving construction.

4. The method for full-section nondestructive detection and data processing of pavement structure layer thickness based on three-dimensional ground penetrating radar according to claim 1 is characterized by: In step S1, the three-dimensional ground penetrating radar antenna array consists of 21 antennas, the spacing between each antenna is 0.75m, and the width of the area that can be covered by a single scan is 1.5m. The wider area to be inspected is divided into multiple adjacent detection channels according to the antenna detection width to ensure that there is no overlapping area in the scanning area and the entire section of the entire area to be inspected is covered.

5. The method for full-section nondestructive detection and data processing of pavement structure layer thickness based on three-dimensional ground penetrating radar according to claim 4 is characterized by: In step S2, a roller rangefinder is used to accurately measure an area of ​​50m in length to calibrate the DMI system, with an error of less than 1‰; to ensure positioning accuracy, during the detection process, the three-dimensional ground penetrating radar data stake number is recorded every 1km to ensure that the positioning error is less than 1m; in step S2, an infrared rangefinder is used to adjust the distance between the detection vehicle and the edge during the forward process to ensure that the lateral distance of the detection vehicle deviating from the detection path is less than 0.3m.

6. The method for full-section nondestructive detection and data processing of pavement structure layer thickness based on three-dimensional ground penetrating radar according to claim 5, characterized in that: In step S3, a core sample is drilled every 1 km; In step S4, a threshold method is used to suppress noise, and an average method is used to remove background clutter; In step S4, the three-dimensional frequency domain data matrix is ​​converted into a time domain data matrix through inverse discrete Fourier transform; in step S4, by specifying the reflection signal symbols of the top and bottom of the structural layer and the time domain range in which they are located, an automatic line tracing tool is used to trace the continuous inter-layer boundary lines with the largest reflection signal amplitudes at the top and bottom of the structural layer in the radar map.

7. The method for full-section nondestructive detection and data processing of pavement structure layer thickness based on three-dimensional ground penetrating radar according to claim 6, characterized in that: In step S5, the dielectric constant of the coordinate structure layer can be inversely calculated based on the core sample thickness and the time domain data at the corresponding coordinates, and used as the dielectric constant value used for calculating the thickness of the calibration area. The time domain data matrix is ​​converted into a thickness data matrix. After calibration, the detection accuracy of the pavement structure layer thickness reaches 0.002m.

8. The method for full-section nondestructive detection and data processing of pavement structure layer thickness based on three-dimensional ground penetrating radar according to claim 7, characterized in that: In step S6, the two-dimensional coordinates (x, y) of the detection points in the detection data are assigned, and the row and column positions of each point in each channel of detection data are readjusted according to the principle of arranging the two-dimensional coordinates from small to large; and the data of each detection channel are combined into a full-section thickness distribution matrix according to the principle of arranging the numbers from small to large.

9. The method for full-section nondestructive detection and data processing of pavement structure layer thickness based on three-dimensional ground penetrating radar according to claim 8, characterized in that: In step S7, the full-section thickness distribution matrix is ​​converted into an RGB color distribution matrix, where thicknesses exceeding the upper or lower limit are represented by red or blue, and thickness values ​​between the upper and lower limits are represented by lighter gradient colors. The thickness distribution cloud map is drawn using this RGB color distribution matrix.

10. The method for full-section nondestructive detection and data processing of pavement structure layer thickness based on three-dimensional ground penetrating radar according to claim 9, characterized in that: In step S8, a coordinate axis corresponding to the detection area is generated, and an intuitive thickness distribution cloud map is drawn in the coordinate axis area according to the coordinate values ​​and RGB color values ​​of each point in the RGB color distribution matrix, so that the thickness change trend and thickness distribution of the entire cross-section of the structural layer can be intuitively obtained; in step S9, the area ratio of the area below the lower limit of the qualified value, above the upper limit, and between the two is counted, so as to accurately obtain the proportion of weak areas with insufficient thickness of the structural layer and evaluate the overall distribution of the thickness of the structural layer.

Images