System and method for evaluating skid resistance of road surface in real time
Through a system combining laser sensor and vibration sensor, a variety of data processing technologies are used to filter and correct the pavement texture, solving the problem of insufficient detection accuracy and efficiency of pavement anti-slip performance in the existing technology, and achieving efficient and accurate pavement texture Evaluation and anti-slip performance analysis.
Patent Information
- Application Number
- CN202411156365.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-22
- Publication Date
- 2024-11-29
AI Technical Summary
The existing pavement anti-slip performance detection methods are difficult to balance between test accuracy and test efficiency. The traditional method has low testing efficiency and is difficult to evaluate the anti-slip performance at the micro-level pavement surface. Emerging methods have shortcomings in both accuracy and efficiency.
A system that combines laser sensors and vibration sensors is used to obtain three-dimensional point cloud data on the road through filtering processing, vibration and tilt correction, missing data filling and three-dimensional average structure depth calculation, combining statistical filtering, CSF ground filtering, high-pass filtering and balance height difference to achieve high-precision evaluation of pavement texture.
It realizes efficient and accurate evaluation of pavement texture in a field environment, can continuously test macroscopic and microscopic textures, improves detection efficiency and accuracy, and the processing results are highly correlated with pavement anti-slip performance.
Smart Images

Figure CN119046572A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of road surface anti-skid performance detection equipment and methods, and in particular to a system and method for real-time evaluation of road surface anti-skid performance. Background Art
[0002] The anti-skid performance test of the road surface is of great significance for ensuring traffic safety. A large number of studies have shown that the anti-skid performance of the road surface is closely related to the three-dimensional texture morphology of its surface. The quality of the anti-skid performance of the road surface directly depends on the composition and distribution of the three-dimensional texture morphology of the road surface. Among them, the macro texture and micro texture of the road surface are the main texture morphologies that affect the anti-skid performance of the road surface. Therefore, testing the texture morphology of the road surface is one of the important ways to evaluate the anti-skid performance of the road surface. At present, the traditional anti-skid performance test method of the road surface is mainly based on the sand spreading method. Although this method is simple to operate, it has the problem of low test efficiency and it is difficult to effectively evaluate the anti-skid performance of the road surface at the micro level. In recent years, the emerging methods such as CT scanning, vehicle-mounted laser scanning and close-range image recognition and processing often find it difficult to achieve an ideal balance between test accuracy and test efficiency. Specifically, the methods that can achieve continuous and rapid testing often have low accuracy and cannot meet the accuracy requirements of 0.5mm for micro-texture morphology; while the methods that can evaluate the micro level often cannot meet the needs of large-scale rapid testing on site. Therefore, in order to comprehensively and accurately test the actual pavement texture on site to evaluate the pavement's anti-skid performance, we propose a three-dimensional point cloud data testing device for pavement texture.
[0003] The point cloud data obtained based on laser scanning technology will be affected by various factors, resulting in different results. According to the different characteristics of the laser scanning data, it is necessary to take appropriate processing according to these data characteristics. The point cloud data obtained after processing is the basis for calculating the texture evaluation index. Therefore, the quality of data processing is directly related to the texture evaluation index. Summary of the invention
[0004] In view of the problems existing in the prior art, the present invention provides a system for real-time evaluation of the anti-skid performance of a road surface, the system comprising a laser sensor for collecting three-dimensional texture information of the road surface to be tested; a vibration sensor for collecting vibration data in the x, y, and z directions during the process of the laser sensor collecting the road surface texture information; a receiving and processing system for receiving information collected by the laser sensor and the vibration sensor, and processing the received information to output the three-dimensional average structural depth SMTD of the road surface. The vibration sensor and the laser sensor collect information at the same frequency;
[0005] On the basis of the above solution, the system further comprises a first photoelectric encoder for controlling the laser sensor to collect evenly spaced three-dimensional texture information of the road surface to be tested.
[0006] Based on the above scheme, the method for processing information by the receiving and processing system is as follows:
[0007] S1 performs filtering processing on the point cloud data of the three-dimensional texture collected by the laser sensor;
[0008] S1-1 uses statistical filtering technology to solve outlier noise points in road texture point cloud data;
[0009] S1-2 uses CSF ground filtering technology to solve the sharp burr noise points in the road texture point cloud data;
[0010] S2 performs vibration and tilt correction on the filtered data;
[0011] The vibration tilt correction includes: firstly using high-pass filtering to filter out the low-frequency noise signal generated by vibration; then using a vibration sensor to assist in correcting the influence of system vibration; and finally correcting the tilt in the horizontal axis direction of the road surface by using a balance correction method;
[0012] Filling in missing data in S3
[0013] Calculation of S4 3D mean structural depth SMTD
[0014]
[0015]
[0016] Where F0 represents the simulated sand paving plane, D is the integration area, F(x, y) represents the spatial function formed by the pavement texture point cloud data, A is the area of the integration area D, and dx and dy represent the integration symbols.
[0017] Based on the above solution, in step S2, when high-pass filtering is performed, the parameters used are: sampling rate fs =
[0018] 2000; cut-off frequency fc=40; filter order N=40.
[0019] On the basis of the above scheme, in step S2, the specific method of the balance calibration method is:
[0020] First, read the road segment data, then calculate the number of cross-section points, and extract the mean of 100,000 points on both sides of the road as the endpoints of the horizontal line. The coordinates of the two endpoints of the horizontal line are: (x k ,y k ) and (x n ,y n ), left endpoint (x k ,y k ) is below the X-axis, and the right endpoint (xn ,y n ) is above the X-axis; in the inclined road contour line, points at different positions have different height differences with the X-axis, and the farther away from the origin, the greater the height difference; the right endpoint (x n ,y n )The height difference from the X axis is h n , that is, y n ; At any point on the right side of the cross section (x m ,y m ), the height difference between the straight line and the X-axis is h m ,h m It can be calculated by the following formula:
[0021]
[0022] Any point on the right side of the cross section (x m ,y m ) The corrected coordinates are (x' m ,y' m ), x' m The value is equal to the original horizontal coordinate x m , y' m The value of is obtained by subtracting the height difference from the original ordinate. The specific formula is shown below. The height difference correction principle on the left side of the cross section is the same as that on the right side, except that the ordinate on the left side needs to be added with the height difference.
[0023] x′ m =x m
[0024]
[0025] Based on the above solution, the method used to fill the missing data in step S3 is cubic spline interpolation.
[0026] The present application also provides a method for real-time evaluation of the anti-skid performance of a road surface, using the above system, comprising the following steps:
[0027] (1) Data Collection
[0028] Collecting 3D texture information of the road surface using a system for real-time evaluation of the road surface's anti-skid performance;
[0029] (2) Data processing and calculation of 3D mean structural depth SMTD
[0030] The collected three-dimensional texture information is processed using a receiving and processing system, including the following steps:
[0031] S1 performs filtering on the point cloud data of the 3D texture collected by the laser sensor
[0032] S1-1 uses statistical filtering technology to solve outlier noise points in road texture point cloud data;
[0033] S1-2 uses CSF ground filtering technology to solve the sharp burr noise points in the road texture point cloud data;
[0034] S2 corrects vibration and tilt of the filtered data
[0035] The vibration tilt correction includes: firstly using high-pass filtering to filter out the low-frequency noise signal generated by vibration; then using a vibration sensor to assist in correcting the influence of system vibration; and finally correcting the tilt in the horizontal axis direction of the road surface by using a balance correction method;
[0036] Filling in missing data in S3
[0037] Calculation of S4 3D mean structural depth SMTD
[0038] (3) Evaluation model construction
[0039] The pavement structure depth MTD value is tested by using a sand spreading method on the pavement where the laser sensor (1) collects three-dimensional texture information, and a correlation analysis is performed between the evaluation index SMTD calculated in step (2) and the pavement structure depth MTD measured by the sand spreading method to obtain a correlation model formula;
[0040] (4) Real-time evaluation
[0041] The method in step (1) is used to collect the three-dimensional texture information of the road surface to be tested; the method in step (2) is used to process the data and obtain the three-dimensional average structural depth SMTD of the road surface to be tested; finally, the model formula in step (3) is used to obtain the actual structural depth MTD value of the road surface to be tested.
[0042] On the basis of the above scheme, in step (1), during collection, the first photoelectric encoder (3) of the system controls the laser sensor (1) to collect evenly spaced three-dimensional texture information of the road surface to be measured; at the same time, the vibration sensor (2) of the system and the laser sensor (1) collect information at the same frequency.
[0043] Based on the above scheme, the correlation model formula obtained in step (3) is:
[0044] y=1.7749x-0.16011
[0045] Among them, y is the three-dimensional average structural depth SMTD value of the road surface to be tested; x is the actual structural depth MTD value of the road surface to be tested.
[0046] Compared with the prior art, the advantages of the present invention are:
[0047] 1. The present invention provides a three-dimensional point cloud data testing system and a corresponding data processing method for road surface texture in actual on-site environments. Compared with traditional road surface anti-skid testing methods, the equipment is more efficient and has a wider detection range. By directly processing the three-dimensional texture of the road surface, it can restore the structural state of the road surface itself to the maximum extent, making its anti-skid evaluation more reasonable and accurate.
[0048] 2. The system of the present invention, while taking into account the test efficiency, demonstrates the ability to continuously test the macro and micro textures of the road surface in an actual on-site environment while maintaining high precision of the test results.
[0049] 3. The three-dimensional point cloud data obtained by testing with the system of this application can accurately reflect the real texture morphology of the road surface after a series of processing, and the three-dimensional texture evaluation index calculated by this method shows a high correlation with the anti-skid performance of the road surface.
[0050] 4. The data processing method of the present invention firstly targets the data noise problem and adopts a classification processing method, which can filter out data anomalies generated by different test conditions and road surface types more accurately. In view of the influence of vibration in field tests, high-pass filtering is selected to filter out low-frequency noise signals generated by large-scale vibrations, and vibration sensors are used to assist in correcting the tiny vibrations generated by the equipment, which comprehensively covers the complex vibration problem from two aspects. For the problem of missing data, cubic spline interpolation is used for processing. This interpolation method has good curve smoothness and can better reflect the changing trend of road surface texture. BRIEF DESCRIPTION OF THE DRAWINGS
[0051] Figure 1 It is a schematic diagram of the stereoscopic viewing angle structure of the system of the present invention;
[0052] Figure 2 It is a schematic diagram of the main viewing angle structure of the system of the present invention;
[0053] Figure 3 It is a schematic diagram of the left-viewing perspective structure of the system of the present invention;
[0054] Figure 4 It is a schematic diagram of the structure of the system of the present invention when viewed from an upward angle;
[0055] Figure 5 The comparison diagram before and after using the statistical filtering method to process the road texture point cloud data, a is the original road point cloud, b is the road point cloud after statistical filtering;
[0056] Figure 6 This is a comparison diagram of the effects before and after CSF ground filtering processing used in the present invention, a is the original road point cloud, and b is the road point cloud after CSF ground filtering;
[0057] Figure 7Schematic diagram of the balance height difference method adopted by the present invention;
[0058] Figure 8 Example of a point cloud of inclined road texture;
[0059] Fig. 9 Figure 8 Comparison of the texture point cloud of the medium-sloped road surface before and after correction by the balance height difference method, a is before correction, b is after correction;
[0060] Fig.10 The effect diagram of the road point cloud with missing data before and after using the cubic spline interpolation method, where (a) the road point cloud with missing data and (b) the road point cloud after interpolation;
[0061] Fig.11 This is the result of correlation analysis in the use example of the present invention. DETAILED DESCRIPTION
[0062] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be pointed out that the embodiments described below are intended to facilitate the understanding of the present invention and do not have any limiting effect on the present invention.
[0063] Example 1
[0064] like Figure 1-5 As shown, the present application provides a system for real-time evaluation of the anti-skid performance of a road surface, the system comprising: a laser sensor 1 for collecting three-dimensional texture information of a road surface to be tested; a vibration sensor 2 for collecting vibration data in the x, y, and z directions during the process of the laser sensor 1 collecting the road surface texture information; and a receiving and processing system 6 for receiving the information collected by the laser sensor 1 and the vibration sensor 2, and processing the received information to output the three-dimensional average structural depth SMTD of the road surface. The vibration sensor 2 and the laser sensor 1 collect information at the same frequency.
[0065] The laser sensor 1 can use the models in the following table:
[0066]
[0067] The vibration sensors 2 can use the models listed in the table below:
[0068]
[0069]
[0070] When in use, the laser sensor 1 in the system collects three-dimensional texture information of the road surface to be tested at uniform intervals. The vibration sensor 2 collects its vibration data in the x, y, and z directions during the process of the laser sensor 1 collecting the road surface texture information. Then, the receiving and processing system 6 receives the collected information and processes the information to obtain the three-dimensional average structural depth SMTD value of the road surface.
[0071] In order to collect evenly spaced three-dimensional texture information of the road surface, the system of the present application also includes a first photoelectric encoder 3 for controlling the laser sensor 1 to collect evenly spaced three-dimensional texture information of the road surface to be measured. The measurement adopts an automatic data saving mode, and the length of each saved data segment is set to 2m and the width is not less than 200mm. The receiving and processing system 6 can be installed in a computer. The laser sensor 1 is provided with two functional interfaces, namely a P / E cable interface and an I / O cable interface. The P / E cable interface is used to connect a power supply and a computer, and the I / O cable interface is used to connect a first photoelectric encoder.
[0072] The system of the present application can be an automatic uniform speed walking system, for example, an automatic walking trolley, on which the laser sensor 1, the vibration sensor 2, the first photoelectric encoder 3 and the receiving and processing system 6 are all arranged. Figure 1-3 In the manpower-powered trolley, in order to control the speed of the trolley, a second photoelectric encoder 4 for real-time monitoring of the speed of the device during the test can be used, and the monitoring result can be displayed on a display screen, which can be set at the end of the handlebar (the handlebar and the display screen are no longer numbered in the figure). Whether it is an automatic trolley or a manpower-powered trolley, it is collectively referred to as a vehicle body 5 in this application.
[0073] The first photoelectric encoder 3 and the second photoelectric encoder 4 can use the models in the following table:
[0074]
[0075]
[0076] In addition, other functional components can be set in the above-mentioned system of the present application, such as a level: used to adjust the device so that the upper plane of the laser sensor is parallel to the measured ground; a shading plate 6 is also provided at the bottom of the vehicle body 5 to prevent interference from external ambient light when the laser sensor 1 collects information, and a 24V30AH high-performance lithium-ion battery is placed in the vehicle body 5 as a power supply (not shown in the figure).
[0077] Example 2
[0078] Based on the system of Example 1, the present application also provides a method for processing the three-dimensional texture information collected by the laser sensor 1 by the receiving and processing system, and the method comprises the following steps:
[0079] S1 performs filtering processing on the point cloud data of the three-dimensional texture collected by the laser sensor 1;
[0080] S1-1 uses statistical filtering technology to solve outlier noise points in road texture point cloud data. This type of noise point is separated from the road texture and has irregular form. The number of point clouds that constitute the noise points varies, from a few points to a cluster of point clouds. Figure 5 This is a comparison chart of the effects before and after statistical filtering. It can be seen from the figure that statistical filtering is more effective in removing outliers, and through calculation, it is found that statistical filtering filters out about 96% of noise points.
[0081] S1-2 uses CSF ground filtering technology to solve the sharp burr-like noise points in the road texture point cloud data.
[0082] CSF (Cloth Simulation Filter) point cloud ground point filtering is also called cloth simulation filtering algorithm. The basic principle is to flip the side with spike noise in the road point cloud downward, and then assume that a piece of cloth falls from above, the falling cloth will be embedded in the road texture structure, forming a road texture contour without spike noise. Figure 6 This is a comparison of the effects before and after CSF ground filtering. From the figure, it can be seen that after the road texture point cloud is processed by CSF ground filtering technology, the sharp burr noise points are filtered out.
[0083] When performing CSF ground filtering on the point cloud data of the three-dimensional texture of this application, relevant parameters such as cloth resolution, cloth stiffness, classification threshold and maximum number of iterations are very important. Cloth resolution refers to the size of the cloth grid covering the texture, representing the horizontal distance between two adjacent particles. Cloth stiffness determines the material compactness of the cloth. This parameter is applicable to three different scenarios, namely steep slopes with large terrain undulations, relatively gentle slopes, and flat terrain. The classification threshold is a distance parameter for classifying the point cloud into ground points and non-ground points. The maximum number of iterations refers to the maximum number of terrain filtering simulations. After many experiments, the optimal setting of the parameters was obtained. When the cloth resolution is set to 2.2, the cloth stiffness is selected as a steep slope, the classification threshold is set to 0.5, and the number of iterations is 600, the noise point elimination effect is optimal.
[0084] S2 performs vibration and tilt correction on the filtered data;
[0085] The vibration tilt correction includes: firstly using high-pass filtering to filter out the low-frequency noise signal generated by vibration; then using a vibration sensor to assist in correcting the influence of system vibration; and finally correcting the tilt in the horizontal axis direction of the road surface by using a balance correction method;
[0086] S2-1 selects high-pass filtering to filter out low-frequency noise signals generated by vibration
[0087] When high-pass filtering, the parameters used are: sampling rate fs = 2000; cut-off frequency fc = 40; filter order N = 40;
[0088] A high-pass filter refers to a filter that allows frequency information above a certain threshold to pass through and filters out frequency information below this threshold, thereby greatly attenuating low frequencies.
[0089] The vibration signal generated by the large-scale vibration caused by the undulation of the road surface is a low-frequency signal compared to the road surface texture. Therefore, high-pass filtering can be used to attenuate or suppress the low-frequency components in the point cloud data, allowing only the high-frequency components to pass through, thereby obtaining a more accurate road surface texture. If this step is removed, the road surface texture information obtained by scanning will be mixed with interference information caused by vibration.
[0090] S2-2 Use vibration sensor 2 to assist the correction system in collecting road surface 3D texture information due to the influence of vibration
[0091] The system of the present application is arranged on a vehicle body 5. During the information collection process, the unevenness of the road surface may cause the vehicle body 5 to vibrate, thereby causing the laser sensor 1 that is collecting information to vibrate.
[0092] The system of this application will inevitably vibrate when encountering cracks in the road surface, foreign objects such as stones, or unstable tire pressure during driving, and the vibration will be reflected in the scanned data in real time. When processing the road texture, the accuracy of the evaluation index calculation of the road texture cannot be guaranteed. The degree of influence of vibration on the measurement results can be simply divided into small deformations at the microscopic level and long-distance and large-scale fluctuations at the macroscopic level. The different degrees of influence are related to many factors such as the undulations of the road surface itself and the measurement speed of the equipment.
[0093] In order to solve this technical problem, a vibration sensor 2 is provided in the system of the present application. The vibration sensor 2 is used to collect vibration data in the x, y and z directions when the laser sensor 1 collects road texture information. The most important thing is that the vibration sensor 2 collects information at the same frequency as the laser sensor 1. During calibration, the data measured by the vibration sensor 2 is combined with the data processed by high-pass filtering to compensate for the offset caused by vibration.
[0094] S2-3 Balance height difference calibration
[0095] When the laser sensor 1 collects the three-dimensional texture information of the road surface (road point cloud data), the wheels of the vehicle body 5 are not on the horizontal plane due to insufficient road flatness or transverse slope of the road surface, resulting in the texture data obtained being tilted at a certain angle in the cross section. In order to solve this technical problem, the present application adopts the original balance height difference method for processing.
[0096] The principle of the balance height difference method is to use the middle of the section as the reference and level both sides towards the middle, similar to a balance.
[0097] like Figure 7 As shown, the solid line represents the cross section of the road surface. The middle of the cross section of the road surface is at the origin, and the left end point (x k ,y k ) is below the X-axis, and the right endpoint (x n ,y n ) is above the X-axis. In the inclined road contour line, points at different positions have different height differences with the X-axis. The farther from the origin, the greater the height difference. n ,y n )The height difference from the X axis is h n , that is, y n ; At any point on the right side of the cross section (x m ,y m ), the height difference between the straight line and the X-axis is h m , according to the geometric relationship in the figure, h m It can be calculated by the following formula (2):
[0098]
[0099] Any point on the right side of the cross section (x m ,y m ) The corrected coordinates are (x' m ,y' m ), x' m The value is equal to the original horizontal coordinate x m , y' m The value of is obtained by subtracting the height difference from the original ordinate. The specific formulas are shown in (3) and (4). The height difference correction principle on the left side of the cross section is the same as that on the right side, except that the ordinate on the left side needs to be added with the height difference. Figure 7 The dotted line in the figure represents the result after pavement section correction.
[0100] x′ m =x m (3)
[0101]
[0102] The balance height difference calibration is performed in MATLAB. The specific method is: first read the road section data, then calculate the number of cross-section points (the number of cross-section points can be identified by MATLAB), and extract the mean of 100,000 points on both sides of the road as the endpoints of the horizontal line ( Figure 7 The straight line connecting the two ends of the inclined road contour line with black dashed lines is the horizontal line, and the endpoints are (X k ,Y k ) and (X n ,Y n ), calculate and extract the mean elevation of the left point and the mean elevation of the right point through formula (2), and calculate the cross-section inclination (the tangent value of the inclination can be calculated through the endpoint coordinates); the x-axis coordinate of each point does not change during the tilt correction, so X m =X' m The tangent value of the cross-section inclination angle is used to calculate the correction position of the y-axis ordinate of each point using formula (4), thereby performing inclination correction. To ensure the accuracy of the inclination correction, when taking the elevation difference on both sides, instead of taking a point on the edge of both sides, an average value is calculated by taking points within a certain range on both sides.
[0103] The present invention selects a section of inclined road surface texture as an example to verify the correction effect of the balance height difference method. Figure 8 As shown, the left and right elevations of the road profile section differ by 4.0746 mm, the correction angle α is 0.053 rad, the length is 76.8, the left elevation is 1.6547, and the right elevation is -2.4199.
[0104] Fig. 9 The comparison of the effects before and after the balance height difference method is shown. Fig. 9 It can be seen that the inclination of the road section tends to be horizontal after correction by the balance height difference method. The elevation values of the original point cloud data are distributed in the range of -4mm-2mm, and the highest and lowest elevation values differ by 6mm. After correction, the elevation values are distributed between 0-3mm, and the highest and lowest elevation values differ by only 3mm. The inclination problem of the road texture scanning results has been effectively improved.
[0105] S3 missing data filling
[0106] Due to the characteristics of laser scanning, when the system of the present invention collects some relatively rough road surface textures or is affected by other environmental factors, point cloud data may be missing.
[0107] Because the road surface 3D point cloud data is composed of many laser lines, when the laser line has a breakpoint, the missing data can be filled by the idea of completing the laser line. The missing of some laser points on a laser line can be solved by interpolation. The present invention adopts the cubic spline interpolation method, which can be implemented by calling the spline function in MATLAB. Fig.10 This is the effect diagram of the road point cloud with missing data before and after processing using cubic spline interpolation method. Fig.10 (a) is a road point cloud with missing data. The black tiny gaps represent the data missing areas. After interpolation using the cubic spline interpolation method, these tiny gaps almost disappear (e.g. Fig.10 (as shown in (b)).
[0108] The present invention adopts a cubic spline interpolation method to fill in the missing data. When there are a large number of adjacent known points around the data missing point, the error between the interpolation result and the true value is extremely small.
[0109] S4 calculation evaluation index: three-dimensional average structural depth SMTD
[0110] Based on the anti-sliding evaluation index mean structural depth MTD measured by the traditional sand spreading method, a three-dimensional evaluation parameter was proposed using the volume integration idea: three-dimensional mean structural depth SMTD.
[0111] The algorithm is designed based on the basic principle of measuring the pavement texture depth by the sand-laying method to calculate the theoretical texture depth value of the pavement point cloud data. To distinguish, the parameter calculated by the present invention after using laser scanning texture data is called the pavement three-dimensional average texture depth SMTD (Surface mean texture depth). The testing principle of this indicator is to simulate an imaginary sand-laying plane on the collected three-dimensional texture data of the pavement, and then calculate the volume between the simulated sand-laying plane and the pavement texture. The ratio of the volume to the simulated plane area is SMTD. In order to make the calculated value of SMTD closer to the real value, the simulated sand-laying plane only needs to cover part of the pavement texture point cloud, so that the high points of the pavement texture are above the simulated sand-laying plane.
[0112] The premise of SMTD calculation is to regard the road surface texture as a spatial surface composed of many point cloud functions. The calculation formula of the point cloud function is shown in the following formula.
[0113] Z=F(x,y)
[0114] The volume V between the simulated sand paving plane and the road surface texture can be calculated using quadratic integration, and the calculation formula is as follows.
[0115]
[0116] After the volume is calculated, SMTD can be calculated using the following formula.
[0117]
[0118] Where F0 represents the simulated sand paving plane, D is the integration area, F(x, y) represents the spatial function formed by the pavement texture point cloud data, A is the area of the integration area D, and dx and dy represent the integral signs of the integration area composed of the X and Y directions.
[0119] The data processing method of the present invention firstly targets the data noise problem and adopts a classification processing method, which can filter out data anomalies generated by different test conditions and road surface types more accurately. In view of the influence of vibration in field tests, high-pass filtering is selected to filter out low-frequency noise signals generated by large-scale vibrations, and vibration sensors are used to assist in correcting the tiny vibrations generated by the equipment, which comprehensively covers the complex vibration problem from two aspects. For the problem of missing data, the cubic spline interpolation method is used for processing. This interpolation method has good curve smoothness and can better reflect the changing trend of road surface texture.
[0120] Example 3
[0121] The system in Example 1 is combined with the method in Example 2 to evaluate the anti-skid performance of a real road surface. The laser sensor 1 in the system is used to collect three-dimensional texture information of the road surface at 30 locations. During the collection, the vehicle's driving speed is 85 m / min. The collected information is received and processed by the processing system 6 in the system (see Example 2 for the processing method). The SMTD values corresponding to the 30 locations are calculated respectively. At the 30 locations, the traditional sand-laying method is used to test the MTD value of the road surface structure depth at this location. The results are shown in the following table.
[0122]
[0123] The evaluation index SMTD obtained by the system of the present invention is correlated with the road surface structure depth MTD measured by the sand spreading method. The analysis results are as follows: Fig.11 As shown, from Fig.11It can be seen that the three-dimensional macro texture evaluation index SMTD obtained by the system of the present invention has a high correlation with the traditional anti-skid test result MTD, and the correlation between SMTD and MTD is 0.979, which proves the accuracy of the system of the present invention in data collection and analysis and processing. It also verifies the accuracy and stability of the calculated three-dimensional texture evaluation index SMTD in evaluating the macro texture level. The root mean square error between SMTD and MTD is 0.0657, which means that SMTD is also closer to the structural depth measured by the sand spreading method. At the same time, the formula between the SMTD value and the MTD value calculated by the system of the present invention is obtained: y=1.7749x-0.16011. This lays the foundation for the next step of using the system of the present invention to conduct real-time evaluation of the anti-skid performance of the road surface.
[0124] Example 4
[0125] The relationship between the SMTD value and the MTD value constructed in Example 3 was used to perform real-time evaluation of the anti-skid performance of the road surface. During the evaluation, the manner of collecting and processing information was the same as in Example 3. The specific results are shown in the following table, in which the first 6 groups of data are from the same road section as in Example 3:
[0126] Point number SMTD value measured by the system Calculated MTD value Measured MTD value 1 1.1253 0.7242 0.76 2 1.4257 0.8935 0.85 3 1.0248 0.6676 0.71 4 0.7683 0.5231 0.49 5 1.1102 0.7157 0.67 6 0.8625 0.5762 0.58 7 0.6451 0.4537 0.49 8 1.1214 0.7220 0.67 9 0.8634 0.5767 0.62 10 1.7142 1.0560 1.04 11 1.6108 0.9978 0.95 12 1.0742 0.6954 0.73
[0127] It can be seen from the data in the table that the error between the calculated MTD value and the measured MTD value is about 6%, indicating the accuracy of the prediction formula obtained in Example 3.
[0128] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.
Claims
1. A system for real-time evaluation of road skid resistance, characterized in that: include: A laser sensor (1) for collecting three-dimensional texture information of the road surface to be measured; A vibration sensor (2) for collecting vibration data in three directions, x, y and z, when the laser sensor (1) is collecting road surface texture information; the vibration sensor (2) and the laser sensor (1) have the same frequency for collecting information; A receiving and processing system (6) is used for receiving information collected by a laser sensor (1) and a vibration sensor (2), and processing the received information to output a three-dimensional average structural depth SMTD of a road surface.
2. The system for real-time evaluation of road surface anti-skid performance according to claim 1, characterized in that: It also includes a first photoelectric encoder (3) for controlling the laser sensor (1) to collect evenly spaced three-dimensional texture information of the road surface to be measured.
3. The system for real-time evaluation of road surface anti-skid performance according to claim 1, characterized in that: The method of processing information by the receiving and processing system (6) is as follows: S1 performs filtering processing on the point cloud data of the three-dimensional texture collected by the laser sensor (1); S1-1 uses statistical filtering technology to solve outlier noise points in road texture point cloud data; S1-2 uses CSF ground filtering technology to solve the sharp burr noise points in the road texture point cloud data; S2 performs vibration and tilt correction on the filtered data; The vibration tilt correction includes: firstly using high-pass filtering to filter out the low-frequency noise signal generated by vibration; then using a vibration sensor to assist in correcting the influence of system vibration; and finally correcting the tilt in the horizontal axis direction of the road surface by using a balance correction method; Filling in missing data in S3 Calculation of S4 3D mean structural depth SMTD Where F0 represents the simulated sand paving plane, D is the integration area, F(x, y) represents the spatial function formed by the pavement texture point cloud data, and A is the area of the integration area D.
4. The system for real-time evaluation of road surface anti-skid performance according to claim 3, characterized in that: In step S2, when high-pass filtering is performed, the parameters used are: sampling rate fs=2000; cut-off frequency fc=40; filter order N=40.
5. The system for real-time evaluation of road surface anti-skid performance according to claim 3, characterized in that: In step S2, the specific method of the balance calibration method is: First, read the road segment data, then calculate the number of cross-section points, and extract the mean of 100,000 points on both sides of the road as the endpoints of the horizontal line. The coordinates of the two endpoints of the horizontal line are: (x k ,y k ) and (x n ,y n ), left endpoint (x k ,y k ) is below the X-axis, and the right endpoint (x n ,y n ) is above the X-axis; in the inclined road contour line, points at different positions have different height differences with the X-axis, and the farther away from the origin, the greater the height difference; the right endpoint (x n ,y n )The height difference from the X axis is h n , that is, y n ; At any point on the right side of the cross section (x m ,y m ), the height difference between the straight line and the X-axis is h m ,h m It can be calculated by the following formula: Any point on the right side of the cross section (x m ,y m ) The corrected coordinates are (x' m ,y' m ), x' m The value is equal to the original horizontal coordinate x m , y' m The value of is obtained by subtracting the height difference from the original ordinate. The specific formula is shown below. The height difference correction principle on the left side of the cross section is the same as that on the right side, except that the ordinate on the left side needs to be added with the height difference.
6. The system for real-time evaluation of road surface anti-skid performance according to claim 3, characterized in that: The method used to fill in the missing data in step S3 is cubic spline interpolation.
7. A method for real-time evaluation of road surface anti-skid performance, characterized in that: Using the system according to any one of claims 1 to 6 comprises the following steps: (1) Data Collection Collecting 3D texture information of the road surface using a system for real-time evaluation of the road surface's anti-skid performance; (2) Data processing and calculation of 3D mean structural depth SMTD The three-dimensional texture information collected is processed using a receiving and processing system (6), including the following steps: S1 performs filtering processing on the point cloud data of the three-dimensional texture collected by the laser sensor (1) S1-1 uses statistical filtering technology to solve outlier noise points in road texture point cloud data; S1-2 uses CSF ground filtering technology to solve the sharp burr noise points in the road texture point cloud data; S2 corrects vibration and tilt of the filtered data The vibration tilt correction includes: firstly using high-pass filtering to filter out the low-frequency noise signal generated by vibration; then using a vibration sensor to assist in correcting the influence of system vibration; and finally correcting the tilt in the horizontal axis direction of the road surface by using a balance correction method; Filling in missing data in S3 Calculation of S4 3D mean structural depth SMTD Where F0 represents the simulated sand paving plane, D is the integration area, F(x, y) represents the spatial function formed by the pavement texture point cloud data, and A is the area of the integration area D; (3) Evaluation model construction The pavement structure depth MTD value is tested by using a sand spreading method on the pavement where the laser sensor (1) collects three-dimensional texture information, and a correlation analysis is performed between the evaluation index SMTD calculated in step (2) and the pavement structure depth MTD measured by the sand spreading method to obtain a correlation model formula; (4) Real-time evaluation Using the method in step (1) to collect three-dimensional texture information of the road surface to be tested; The method in step (2) is used to process the data and obtain the three-dimensional average structural depth SMTD of the road surface to be tested; finally, the model formula in step (3) is used to obtain the actual structural depth MTD value of the road surface to be tested.
8. The method for real-time evaluation of road surface anti-skid performance according to claim 7, characterized in that: In step (1), during collection, the first photoelectric encoder (3) of the system controls the laser sensor (1) to collect evenly spaced three-dimensional texture information of the road surface to be measured; at the same time, the vibration sensor (2) of the system and the laser sensor (1) collect information at the same frequency.
9. The method for real-time evaluation of road surface anti-skid performance according to claim 7, characterized in that: In step S2, the specific method of the balance calibration method is: First, read the road segment data, then calculate the number of cross-section points, and extract the mean of 100,000 points on both sides of the road as the endpoints of the horizontal line. The coordinates of the two endpoints of the horizontal line are: (x k ,y k ) and (x n ,y n ), left endpoint (x k ,y k ) is below the X-axis, and the right endpoint (x n ,y n ) is above the X-axis; in the inclined road contour line, points at different positions have different height differences with the X-axis, and the farther away from the origin, the greater the height difference; the right endpoint (x n ,y n )The height difference from the X axis is h n , that is, y n ; At any point on the right side of the cross section (x m ,y m ), the height difference between the straight line and the X-axis is h m ,h m It can be calculated by the following formula: Any point on the right side of the cross section (x m ,y m ) The corrected coordinates are (x' m ,y' m ), x' m The value is equal to the original horizontal coordinate x m , y' m The value of is obtained by subtracting the height difference from the original ordinate. The specific formula is shown below. The height difference correction principle on the left side of the cross section is the same as that on the right side, except that the ordinate on the left side needs to be added with the height difference. x′ m =x m 10. The method for real-time evaluation of road surface anti-skid performance according to claim 7, characterized in that: The correlation model formula obtained in step (3) is: y=1.7749x-0.16011 Among them, y is the three-dimensional average structural depth SMTD value of the road surface to be tested; x is the actual structural depth MTD value of the road surface to be tested.
Citation Information
Cited By
Anti-skid performance evaluation method and system for anti-skid texture structure on surface of ice body
CN122087385A
