Rapid detection method of road surface roughness based on lidar

Abstract

The present invention relates to the field of surface measurement technology, specifically to a rapid detection method for road surface smoothness based on laser radar, comprising the following steps: collecting vehicle posture data to form a three-dimensional vector, mapping laser data to a spatial coordinate system, identifying projection difference and adjusting the trajectory, detecting disturbances to generate a distribution area, translating direction lines according to lateral distribution, constructing reference lines through unified projection, extracting echoes and time indexes to reorganize time series segments, analyzing point number trends to screen stable segments, extracting symmetrical points and comparing angles to generate closed zones. In the present invention, by obtaining the posture vectors of the vehicle in roll, pitch, and yaw states, and mapping the laser echo ranging data to an absolute spatial coordinate system, the angle difference between the ranging direction and the ground direction is accurately calculated, the spatial matching accuracy of the measurement path is significantly improved, the longitudinal profile analysis result is closer to the actual road state, and the spatial integrity, trajectory smoothness, and profile structure rationality of the entire measurement process are comprehensively optimized.

Description

Technical Field

[0001] The present invention relates to the field of surface measurement technology, and in particular to a method for quickly detecting road surface flatness based on laser radar. Background Art

[0002] The field of surface measurement technology includes technical methods for acquiring, processing and analyzing the geometric characteristics of the surface of an object. Its core content is to use various types of sensing equipment and data acquisition devices to perform high-precision three-dimensional measurement of the target surface, and then conduct quantitative analysis of its parameters such as contour, texture, roughness and flatness. This technical field is widely used in industries such as manufacturing, construction, transportation and geology, and is particularly important in infrastructure maintenance and road quality assessment. In recent years, with the development of laser scanning and non-contact measurement technology, surface measurement is gradually evolving towards high efficiency, high precision and automation. The measurement methods involved include laser triangulation, phase comparison, structured light imaging and other technical paths, and are combined with positioning systems, time synchronization technology and high-performance data recording and computing platforms to form a systematic dynamic measurement solution.

[0003] Among them, the rapid detection method of road surface smoothness by LiDAR refers to the use of LiDAR sensors to collect spatial data on the road surface, synchronously obtain the vehicle motion trajectory through the cooperation of a high-precision positioning system, continuously obtain road surface elevation information during vehicle movement, and reconstruct a continuous road surface longitudinal profile. The technical matters targeted by this method include the real-time collection of road surface elevation while the vehicle is moving, the establishment of a reference trajectory, and the restoration of road surface profile curves. Specifically, the ground echo distance data is obtained by vertically arranging the LiDAR, and the attitude angle information provided by the inertial navigation system is combined with the use of geometric solution to convert the radar data into spatial coordinate data. Then, the actual road longitudinal change curve is constructed through trajectory fitting and measurement point projection methods for the calculation of road surface smoothness parameters. Through the above-mentioned measurement and coordinate conversion methods, the rapid reconstruction of the road surface spatial morphology is achieved.

[0004] Existing technologies rely on LiDAR and inertial navigation systems to obtain ranging and attitude data, constructing spatial coordinate sequences through geometric calculations. However, during dynamic measurement, attitude fluctuations caused by vehicle movement are not fully accounted for, which can easily cause offsets between trajectory points and the laser ranging direction, compromising the true reproduction of the spatial point cloud. If unstable perturbations exist in the measurement path, existing technologies struggle to identify continuous perturbation regions or perform cluster analysis, leading to the incorporation of some interfering points into the main path data, compromising the accuracy of profile construction. The unified processing of trajectory points is also relatively crude, lacking a stable trend in the main axis direction, resulting in unrepresentative reference curves. Existing solutions do not deeply mine time-series data, and periods of uneven measurement point density disrupt the coherence of subsequent profiles. Without symmetry verification in boundary profile regions, structural gaps are likely to appear. For example, during vehicle acceleration and bumpy sections, the path fluctuates significantly, yet current solutions still use a uniform sampling density, failing to mitigate the effects of these anomalies. These shortcomings lead to a decline in overall data quality, distorted path construction, and highly volatile measurement results, hindering accurate road roughness assessment. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and propose a rapid detection method for road surface flatness based on laser radar.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a rapid road surface roughness detection method based on laser radar, comprising the following steps:

[0007] S1: Collect the roll, pitch, and yaw angle data during vehicle operation to form a three-dimensional attitude vector. Map the laser vector to the absolute space coordinate system, identify the projection difference between the laser ranging direction and the trajectory point in the main axis and lateral direction, and generate an attitude linkage correction trajectory sequence.

[0008] S2: Based on a local window formed by three consecutive points in the posture linkage correction trajectory sequence, obtain the vertical length of the line from the middle point to the two ends, and determine whether the vertical length change trend constitutes a trajectory disturbance. Project and tile the position of the disturbance point to generate a distribution interval of the disturbance aggregation point;

[0009] S3: calling the distribution interval of the disturbance aggregation point, identifying the translation operation of the main axis direction line according to the horizontal coordinate value, performing an interpolation operation in the path sequence according to the adjusted direction line, uniformly projecting the original trajectory points onto the direction line, and generating a fitting flatness detection reference line;

[0010] S4: Using the fitting flatness to detect the spatial position of the path points in the reference line, extracting laser echo data and time index information, dividing the reference line into several time periods according to time continuity, counting the number of path points in the time period, and generating a time-stable segment sequence of laser measurement points.

[0011] As a further solution of the present invention, the posture linkage correction trajectory sequence includes the absolute trajectory point coordinates after posture rotation, the spatial projection error parameters in the main horizontal axis direction, and the adjusted path point position set; the disturbance concentration point distribution interval includes multiple disturbance point concentrated position blocks, trajectory path disturbance concentrated belt position marks, and distribution segment numbers under two-dimensional coordinate mapping; the fitting flatness detection reference line includes the path direction baseline, the direction line projection coordinate sequence, and the interference path interpolation positioning segment; the laser measurement point time series stable section sequence includes the continuous time period number, the number of laser points per unit time period, and the point segment under the stable driving state.

[0012] As a further solution of the present invention, the steps for obtaining the posture linkage correction trajectory sequence are specifically as follows:

[0013] S111: Collect roll, pitch, and yaw angle data during vehicle operation, construct a three-dimensional attitude vector consisting of three-axis angles, call the time tag corresponding to each moment in the three-dimensional attitude vector, match it with the time tag in the laser echo data, identify the mapping direction of the laser ranging direction under the influence of the three-dimensional attitude and the absolute space coordinate system, and obtain a sequence of laser ranging direction angle values;

[0014] S112: Calculate the cosine value of the angle between the laser ranging direction and the gravity direction based on the laser ranging direction angle value sequence, extract the corresponding trajectory point coordinate information based on the difference between the projection position of the vehicle trajectory point in the main axis and the lateral direction, adjust the original trajectory point position, and obtain the posture linkage correction trajectory sequence.

[0015] As a further solution of the present invention, the step of obtaining the distribution interval of the disturbance accumulation point is specifically as follows:

[0016] S211: Based on a local window formed by three consecutive points in the posture linkage correction trajectory sequence, calculate the vertical length of the line from the middle point to the two ends, call the spatial coordinate difference of the starting point, the end point, and the middle point in the window to generate a vertical distance value set;

[0017] S212: Calling the vertical distance value set, identifying the continuous change trend based on the vertical lengths corresponding to the middle points in the front and rear windows, screening the change segments whose fluctuation amplitude exceeds the disturbance threshold, calculating the disturbance fluctuation difference of the midpoints in the windows, and obtaining a disturbance trend mutation point sequence;

[0018] S213: Calling the disturbance trend mutation point sequence, performing two-dimensional plane projection according to the distribution position of the disturbance point in the trajectory coordinate space, using the coordinate values ​​converted by the spatial mapping, counting the concentrated distribution area of ​​the disturbance point in the continuous plane interval, and obtaining the distribution interval of the disturbance concentration point.

[0019] As a further solution of the present invention, the step of obtaining the fitted flatness detection reference line is specifically as follows:

[0020] S311: Recalling the quantitative relationship of point groups on both sides of the path in the distribution interval of the disturbance aggregation points, comparing the quantitative distribution of the point groups on the left and right sides, identifying the density difference in the horizontal coordinate, and evaluating the degree of interference tendency on the main axis of the path based on the quantitative distribution difference of the point groups on the left and right sides, to obtain the disturbance weight difference;

[0021] S312: Based on the disturbance weight difference, the offset trend of the main axis direction line is identified according to the transverse coordinate value, the point distribution within the transverse coordinate extreme value interval of the path sequence is called, and the spatial direction of the main axis direction line is adjusted in combination with the center coordinate set between the main axis starting point and the end point to obtain the direction line offset adjustment value;

[0022] S313: According to the direction line offset adjustment value, the horizontal and vertical coordinates of the path points are extracted from the path sequence, and an interpolation operation is performed to obtain the coordinates of the equidistant interpolation points on the direction line. The coordinate projection transformation relationship is identified, and the original trajectory points are uniformly projected onto the adjusted direction line. The projection distance offset is calculated, and the projection result is linearly smoothed to generate a fitting flatness detection reference line.

[0023] As a further solution of the present invention, the steps for obtaining the laser point measurement time-series stable segment sequence are specifically as follows:

[0024] S411: Based on the spatial positions of the path points in the fitted flatness detection reference line, extract the corresponding laser echo data and time index information, arrange the path points in ascending order according to the time index, and generate a time-series spatial path data set;

[0025] S412: Calling the temporal spatial path dataset, calculating the time difference between path points based on the time index values ​​of adjacent path points, dividing the time periods according to the time continuity condition, recording the number of path points in each period, and obtaining a segmented path point number sequence;

[0026] S413: Using the segmented path point number sequence, extract the change values ​​of the path point number in adjacent time periods, construct an absolute difference value sequence based on the continuity and change amplitude of the change value, perform a joint analysis based on the average offset value of the change trend, the change density and the paragraph number weight factor, calculate the path point number change amplitude index, screen the paragraph intervals with stable path point number changes in the time period, and obtain a laser measurement point time series stable paragraph sequence.

[0027] As a further solution of the present invention, the method further includes step S5:

[0028] S5: calling the start and end positions of the dense point segment in the laser point measurement time sequence stable segment sequence, extracting symmetrical point groups along both sides of the fitted flatness detection reference line in the segment, comparing the directions of the spatial vectors between the symmetrical points, and if the angles tend to be consistent, determining it as a closed boundary segment, and generating a closed zone of the flatness detection profile;

[0029] The flatness detection profile closed zone includes a set of structural boundary points on both sides, a sideband direction vector sequence, and a closed path constituting a segment.

[0030] As a further solution of the present invention, the steps for obtaining the closed zone of the flatness detection profile are specifically as follows:

[0031] S511: calling the start and end positions of the dense point segment in the laser point measurement time sequence stable paragraph sequence, and synchronously locating the same number of equally spaced measurement points along both sides of the fitted flatness detection reference line in the paragraph according to the start and end positions to generate a set of symmetrical point pairs;

[0032] S512: Based on the set of symmetrical point pairs, the spatial vector direction between each pair of points is calculated, the angle between two adjacent sets of spatial vector directions is calculated, and the difference between the angle values ​​is compared with the angle value of the previous set, thereby selecting a sequence of point pairs with stable direction changes and generating a sequence of point groups with convergent directions.

[0033] S513: Calling the directional convergence point group sequence, performing continuity identification on the index positions of the directional convergence points in the paragraph sequence, extracting the overall coverage interval, and marking the interval range on both sides of the fitted flatness detection reference line to generate a flatness detection profile closed zone.

[0034] Compared with the prior art, the advantages and positive effects of the present invention are:

[0035] In this method, the vehicle's attitude vectors are acquired during roll, pitch, and yaw, and the laser echo ranging data is mapped to an absolute spatial coordinate system. The angle difference between the ranging direction and the ground direction is accurately calculated. Combined with adjustments to the principal axis and lateral projection of the trajectory points, the coordinates of the original trajectory points can be dynamically corrected, significantly improving the spatial matching accuracy of the measurement path. By constructing a local window to calculate the vertical distance trend from the point to the trajectory line, and based on this, clusters of disturbing points are identified, enabling the identification and suppression of track interference segments and enhancing the ability to control the continuity of the measurement trajectory. After identifying the distribution interval of the disturbance cluster, the directional line is translated using the lateral point cluster distribution relationship. The original trajectory is then uniformly projected to establish a reference line that conforms to the actual principal axis trend, ensuring the consistency of the longitudinal profile. Using this reference line, the laser echo and time tag data are reconstructed in a time series. The sampling density within the time period is analyzed to effectively screen for stable segments of the measurement point time series. Within these stable segments, symmetrical points with consistent directional angles are further extracted to form a boundary closure detection zone, providing spatial closure assurance for high-fidelity profile reconstruction. Multi-level posture correction, disturbance recognition, direction reconstruction and timing stability screening significantly enhance the spatial consistency of data and significantly improve the anti-disturbance capability. The longitudinal profile analysis results are closer to the actual road conditions, and the spatial integrity, trajectory smoothness and profile structure rationality of the entire measurement process are comprehensively optimized. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a schematic diagram of the main steps of the present invention;

[0037] Figure 2 This is a flow chart for obtaining the posture linkage correction trajectory sequence of the present invention;

[0038] Figure 3 This is a flow chart for obtaining the distribution interval of disturbance accumulation points of the present invention;

[0039] Figure 4 This is a flow chart for obtaining the fitted flatness detection reference line of the present invention;

[0040] Figure 5 This is a flow chart for obtaining a stable segment sequence of laser point measurement timing according to the present invention;

[0041] Figure 6 This is a flow chart for obtaining the closed zone of the flatness detection profile of the present invention. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0043] In the description of the present invention, it should be understood that the terms "length," "width," "up," "down," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inside," "outside," and the like, indicating positions or relationships, are based on the positions or relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They do not indicate or imply that the devices or elements referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, in the description of the present invention, "plurality" means two or more, unless otherwise expressly and specifically defined.

[0044] See also Figure 1 The present invention provides a technical solution, a method for rapid detection of road surface flatness based on laser radar, comprising the following steps:

[0045] S1: Collect roll, pitch, and yaw angle data during vehicle operation to form a three-dimensional attitude vector. Map the laser vector to the absolute space coordinate system. By matching the time tag with the laser echo data point by point, calculate the angle relationship between the ranging direction of each point and the ground direction. Call the coordinates of the vehicle's motion trajectory point, identify the projection difference between the laser ranging direction and the trajectory point in the main axis and lateral direction, adjust the coordinate position of the original trajectory point, and generate an attitude linkage correction trajectory sequence.

[0046] S2: Based on the local window formed by three consecutive points in the posture linkage correction trajectory sequence, the vertical length of the line connecting the middle point to the two ends is obtained, and the vertical length change trend is judged to determine whether it constitutes a trajectory disturbance. The position of the disturbance point is projected and tiled to generate the distribution interval of the disturbance cluster point;

[0047] S3: Call the number relationship of point groups on both sides of the path in the distribution interval of the disturbance cluster points, identify the translation operation of the main axis direction line according to the horizontal coordinate value, perform interpolation operation in the path sequence according to the adjusted direction line, uniformly project the original trajectory points onto the direction line, and generate a fitting flatness detection reference line;

[0048] S4: Use fitting flatness to detect the spatial position of path points in the reference line, extract laser echo data and time index information, arrange spatial coordinates in chronological order, divide the reference line into several time periods based on time continuity, count the number of path points in each time period, analyze the trend of point count changes in continuous time periods, and generate a time-stable segment sequence of laser measurement points;

[0049] S5: Call the start and end positions of the dense point segment in the laser point measurement time sequence stabilization segment sequence, extract symmetrical point groups along both sides of the fitted flatness detection reference line in the segment, and compare the directions of the spatial vectors between the symmetrical points. If the angles tend to be consistent, it is determined to be a closed boundary segment, and a closed zone of the flatness detection profile is generated;

[0050] The posture linkage correction trajectory sequence includes the absolute trajectory point coordinates after posture rotation, the spatial projection error parameters in the main transverse axis direction, and the adjusted path point position set. The disturbance aggregation point distribution interval includes multiple disturbance point concentrated position blocks, trajectory path disturbance concentrated belt position marks, and distribution fragment numbers under two-dimensional coordinate mapping. The fitted flatness detection reference line includes the path direction baseline, the direction line projection coordinate sequence, and the interference path interpolation positioning segment. The laser measurement point time series stable section sequence includes the continuous time period number, the number of laser points per unit time period, and the point segment under the stable driving state. The flatness detection profile closed zone includes the two side structure boundary point sets, the sideband direction vector sequence, and the closed path constituent segment.

[0051] See also Figure 2 ,The specific steps for obtaining the posture linkage correction trajectory sequence are:

[0052] S111: Collect roll, pitch, and yaw angle data during vehicle operation, construct a three-dimensional attitude vector consisting of three-axis angles, call the time tag corresponding to each moment in the three-dimensional attitude vector, match it with the time tag in the laser echo data, identify the mapping direction of the laser ranging direction under the influence of the three-dimensional attitude and the absolute space coordinate system, and obtain a sequence of laser ranging direction angle values;

[0053] The collection of attitude angle information during vehicle operation requires continuous recording with high-precision inertial measurement equipment. The vehicle can be installed with an integrated attitude sensing unit, including a high-frequency gyroscope and accelerometer combination module, to detect and output attitude change parameters such as roll angle, pitch angle and yaw angle. The device operating frequency is set to 200Hz to ensure the continuity and accuracy of attitude information in rapidly changing scenarios. The original angular velocity signal is integrated and converted through the internal calculation module to obtain angle information and convert it into an attitude vector structure. Each set of attitude vectors contains three angle parameters and is bound to the current acquisition time through a timestamp. The roll angle is set when the vehicle enters the turning phase. The data gradually changes from 0 to 5 degrees, the pitch angle changes from 2 degrees to 4 degrees due to the undulation of the road surface, and the yaw angle changes from 0 degrees to 30 degrees due to the change of vehicle direction. At the same time, the corresponding time point is recorded as 10:15:30 on May 13, 2025. The time tag is extracted from the laser echo data synchronously collected by the lidar and matched with the above time tag. The laser time and the attitude time are aligned using the nearest neighbor matching method. The matched attitude vector data is selected to adjust the laser ranging direction. The spatial rotation relationship at the current moment is constructed according to the attitude angle parameters to correct the difference between the original laser direction and the actual attitude of the vehicle, and obtain the laser ranging direction angle value sequence.

[0054] S112: Calculate the cosine value of the angle between the laser ranging direction and the gravity direction based on the laser ranging direction angle value sequence. Extract the corresponding trajectory point coordinate information based on the difference between the projection position of the vehicle trajectory point in the main axis and the lateral direction. Adjust the original trajectory point position to obtain the posture linkage correction trajectory sequence.

[0055] The angle between the laser ranging direction at a given time node and the direction of gravity can be determined by the geometric relationship between the direction change and the ground direction. During the calculation process, the laser ranging direction at each moment is compared with the standard gravity direction to evaluate the degree of inclination of the angle between them. When the vehicle is climbing, the laser direction is set to deviate forward and upward, with a significant inclination angle with the vertical gravity direction. Based on the trajectory points corresponding to the vehicle's current posture, the difference between the projection of the trajectory point in the main axis direction and the lateral direction in the vehicle coordinate system is extracted. The difference reflects the offset behavior caused by the vehicle's posture change. In the case of continuous left turns, the lateral projection of the trajectory point will show a continuous change trend, while the main axis direction will change with vehicle speed fluctuations. Correction vectors are constructed based on these changes and the original trajectory point coordinates are adjusted by superposition. For example, if the yaw angle of a trajectory point increases during a left turn, the point needs to be corrected to the left in the lateral direction by a certain distance. The adjusted trajectory better matches the actual vehicle's operating path. All trajectory points after posture correction are recombined to form a new trajectory sequence. The trajectory restoration process under posture linkage is completed by reordering them in chronological order, and the posture linkage corrected trajectory sequence is obtained.

[0056] See also Figure 3 , the specific steps for obtaining the distribution interval of disturbance accumulation points are:

[0057] S211: Based on a local window formed by three consecutive points in the posture linkage correction trajectory sequence, calculate the vertical length of the line from the middle point to the two ends, call the spatial coordinate difference of the starting point, end point, and middle point in the window, and generate a vertical distance value set;

[0058] It is necessary to construct a local window by three consecutive points, namely the starting point, end point and middle point of the window. By calculating the perpendicular length between the middle point and the line connecting the two ends, the vertical distance of the local window can be obtained. For each local window, the coordinates of three points are taken for calculation. The vertical distance from the middle point to the line connecting the two ends is determined using the distance formula from a point to a line in geometry. This process involves specific spatial coordinate calculations. By comparing the vertical distances in adjacent windows, the stability of the trajectory is further evaluated. A trajectory path is set, and the spatial coordinates of a certain trajectory point can be obtained through a GPS device. By calculating the distance difference between three consecutive trajectory points, it can be determined whether there is a sudden offset in the trajectory segment. If the calculated vertical length exceeds the set tolerance range, it indicates that there is a large fluctuation in the trajectory segment. It is not only suitable for traffic path monitoring, but also widely used in path planning, logistics distribution and other fields. By processing continuous data points, abnormal changes in the trajectory can be quickly determined, providing a basis for subsequent trajectory optimization and correction. In this way, monitoring can capture changes in the trajectory in real time, provide a basis for decision-making, and make timely adjustments when anomalies are found, generating a set of vertical distance values.

[0059] S212: Call the vertical distance value set, identify the continuous change trend based on the vertical length corresponding to the middle point in the front and back windows, and filter the change segments whose fluctuation amplitude exceeds the disturbance threshold, using the formula:

[0060] ;

[0061] Calculate the disturbance fluctuation difference at the midpoint of the window to obtain the disturbance trend mutation point sequence;

[0062] in, Representative The difference of disturbance fluctuations at the midpoint of a local window, 、 、 are the vertical lengths of the midpoints of adjacent local windows, is the vertical length of the midpoint and the reference point set The mean absolute deviation between is the ratio of the disturbance threshold to the real-time vertical length, For the Point disturbance determination threshold, is the number of reference points, For the A reference midpoint length value;

[0063] Parameter meaning and formula calculation derivation process:

[0064] The calculation logic comprehensively evaluates the change trend of a local point and its deviation from a standard reference to identify segments with significant changes in a trajectory or vertical length sequence. The calculation is divided into two parts: the first part expresses the local fluctuation amplitude through the second-order difference between the center point and its adjacent points, reflecting the spatial perturbation intensity of the point relative to the preceding and following adjacent points, and adjusting the balance weight to prevent misjudgment caused by local noise; the second part averages the absolute deviation between the center point and each value in the reference set to reflect the degree of deviation of the point from the standard trajectory. At the same time, the ratio of the perturbation value to the actual length is introduced to consider the sensitivity of the vertical length itself to offset. The overall logic takes into account both local trend judgment and global deviation detection, improving the accuracy and robustness of abnormal mutation point identification.

[0065] 、 、 :Represents the , No. Hedi The vertical length of the local window is the distance from the middle point to the line connecting the two ends. It is calculated according to the previous method. The vertical length data is obtained through monitoring equipment. The specific value will be updated in real time according to the coordinate changes of each point in the trajectory.

[0066] :It is an adjustment factor used to balance the calculation of the disturbance difference. The value is set to This factor is set to optimize the calculation results of disturbance fluctuations and avoid misjudgment caused by excessive numerical fluctuations. It is based on adjustments made during multiple experiments and data analysis. Its function is to balance the influence of adjacent points and prevent the calculation from being too dependent on local changes.

[0067] : Indicates the current midpoint With reference point set The mean absolute deviation reflects the difference between the current point and the reference point. It is based on trajectory data or a set of standard trajectory data points. is the number of reference points. By calculating the difference between the current point and the reference point, the deviation degree of the trajectory can be evaluated. is 5 meters, and the reference point set is , then the deviation is calculated as:

[0068] ;

[0069] The deviation value is used to reflect the closeness of the current point to the standard trajectory;

[0070] :Indicates the The disturbance threshold of a point is used to determine whether the current point can be considered a disturbance point. The threshold is set based on data statistics or actual application scenarios. For example, if the threshold of a disturbance point is set to 0.5 meters, when the current vertical length changes by more than this value, the point is considered to be disturbed.

[0071] :Indicates the The vertical length of the middle point of a local window is obtained by calculating the vertical distance from the current point to the line connecting the two ends, reflecting the degree of change of the current trajectory point;

[0072] Set in a certain trajectory detection, The vertical length of the point is meters, and the vertical lengths of the front and rear points are rice, Meters, reference point set , and the first Perturbation threshold of point rice;

[0073] Calculate the vertical length difference:

[0074] ;

[0075] Calculate the deviation part:

[0076] ;

[0077] Calculation disturbance judgment part:

[0078] ;

[0079] Substitute into the formula for calculation:

[0080] ;

[0081] The calculation results show that the The disturbance difference of the point is 0.033 meters. The disturbance fluctuation difference is used to identify the intensity of the change trend of the window center point. It comprehensively considers the geometric deformation trend (such as sudden rise / fall) between the current point and adjacent points and the degree of deviation between the current point and the typical structure to evaluate the disturbance degree of the current trajectory. The larger the value, the more significant the trajectory disturbance of the point. Since this value is less than the set threshold of 0.5 meters, it can be determined that the point is not considered a disturbance point. It helps to further screen areas with abnormal fluctuations in the trajectory and provide data support for subsequent analysis.

[0082] S213: Calling the disturbance trend mutation point sequence, performing two-dimensional plane projection according to the distribution position of the disturbance point in the trajectory coordinate space, using the coordinate values ​​converted by the spatial mapping, counting the concentrated distribution area of ​​the disturbance point in the continuous plane interval, and obtaining the distribution interval of the disturbance concentration point;

[0083] The disturbance points are spatially projected. The purpose of projection is to map the disturbance points from the actual spatial coordinate system of the trajectory onto a two-dimensional plane, facilitating subsequent cluster analysis and spatial distribution research. To achieve this goal, geometric transformation methods, especially plane coordinate transformation and linear transformation, are used. The disturbance points are mapped to the new coordinate system based on the reference points of the original coordinate system through the transformation matrix. The distribution of the disturbance points in the plane can identify the concentrated areas of abnormal changes in the trajectory, further helping to locate potential problem areas. In traffic flow monitoring, the projection of the disturbance points can help determine the specific areas where traffic congestion or path deviation occurs. By statistically analyzing the concentration of the projected points, it is possible to determine whether there is a long-term path distortion or severe traffic bottleneck. The projected data not only helps to monitor anomalies in real time, but also provides a quantitative basis for decision-making such as path optimization and traffic light control. It is not only applicable in the transportation field, but can also be widely used in location tracking in the Internet of Things, such as drone path adjustment and transportation distribution route optimization. By analyzing the distribution interval of the disturbance points, directional suggestions can be provided for path optimization.

[0084] See also Figure 4 , the specific steps for obtaining the fitting flatness detection reference line are:

[0085] S311: Recall the quantitative relationship of point groups on both sides of the path in the distribution interval of the disturbance aggregation points, compare the quantitative distribution of the point groups on the left and right sides, identify the density difference in the horizontal coordinate, and evaluate the degree of interference tendency on the main axis of the path based on the quantitative distribution difference of the point groups on the left and right sides to obtain the disturbance weight difference;

[0086] Divide the path on both sides based on the center axis of the path, by setting the center value of the horizontal coordinate axis Divide the points in the path sequence into the left point group and the right point group, and obtain the number of points on each side of the path point set and , taking the trajectory of the unmanned vehicle during construction as an example, if the horizontal coordinate value range is Meter, center value , then the points with abscissas less than 0 are placed in the left point group, and the points with abscissas greater than 0 are placed in the right point group. In the sample path, there are 126 point groups on the left and 98 point groups on the right. Compare the number of point groups on both sides and calculate the difference between the left and right points. , if the difference is greater than the disturbance judgment reference threshold , it is judged that the disturbance on the left is more concentrated, and then the horizontal density is calculated based on the point distribution density in each side point group. and The density can be calculated by the average spacing between the points in each group of points. For example, the average spacing on the left is 0.15 meters, and on the right is 0.21 meters. The density on the left is higher, indicating that the interference trend is biased to the left. Combined with the difference Density ratio Perform normalized weight judgment according to the set disturbance tendency factor , you can construct a calculation formula , substitute 、 , we can conclude , obtain the disturbance interference weight difference; the result shows that the disturbance direction is biased to the left, and the density change enhances the disturbance perception amplitude, and its numerical value can be used as a basis for subsequent direction line correction.

[0087] S312: Based on the disturbance weight difference, the offset trend of the main axis direction line is identified according to the transverse coordinate value, the point distribution within the transverse coordinate extreme value interval of the path sequence is called, and the spatial direction of the main axis direction line is adjusted in combination with the center coordinate set between the main axis starting point and the end point to obtain the direction line offset adjustment value;

[0088] According to the horizontal coordinate value, the deviation trend of the main axis direction line is identified. The starting point and the end point need to be selected from the original path. The starting point coordinate is set to , the end point is , the coordinates of its center point are , then select the horizontal extreme point from the path, such as the maximum horizontal coordinate is 2.1 meters and the minimum is -2.3 meters, sample the number of points in the interval where the extreme point is located, and calculate the offset angle of the linear trend of the path center axis in this area , set the offset angle to 5°, and the corresponding offset direction is based on the previous step Indicates that the offset direction is to the left. The corresponding direction line needs to be based on the current main axis and rotate 5° to the left around the center point of the path. After the rotation, the slope of the main axis direction line changes from Adjust to new slope , calculated , a new linear expression is constructed for the direction line, which is used for subsequent path interpolation operations to obtain the direction line offset adjustment value. The result is used to clarify the position expression of the direction line after rotation.

[0089] S313: According to the direction line offset adjustment value, the horizontal and vertical coordinates of the path points are extracted from the path sequence, and the interpolation operation is performed to obtain the coordinates of the equidistant interpolation points on the direction line. The coordinate projection transformation relationship is identified, and the original trajectory points are uniformly projected onto the adjusted direction line using the formula:

[0090] ;

[0091] Calculate the projection distance offset, perform linear smoothing on the projection result, and generate a fitting flatness detection reference line;

[0092] in, Represents the projection distance offset, 、 The original trajectory point is The horizontal and vertical coordinates of the position, 、 The original trajectory point is The horizontal and vertical coordinates of the uniform projection of the position on the direction line, For the The symmetry factor of the site, is the total number of path sequence points;

[0093] The benefit of this formula is that it constructs an absolute value expression by combining the Euclidean distance between the original trajectory point and its projection point on the fitted direction line with a symmetry factor. This not only reflects the degree of trajectory deviation, but also takes into account the symmetry of the path point distribution relative to the center line, thereby generating the projection distance deviation degree.

[0094] The projection distance offset is used to measure the degree of alignment offset between the original trajectory point and its alignment on the unified direction line after the projection adjustment. The calculation logic is based on the Euclidean distance formula, combined with the direction correction factor after the coordinate point projection, to evaluate the geometric difference before and after the trajectory adjustment. The formula uses the original trajectory point as the Its projection point on the same direction line The distance between them is taken as the core calculation content, and the horizontal coordinate symmetry adjustment factor is introduced , perform lateral correction on the projection offset of the path points to enhance the sensitivity to the lateral deviation trend of the path. The offset of the path points is summed and averaged to obtain the comprehensive offset of the entire path after the direction line projection adjustment. The smaller the value, the better the overall alignment of the path with the direction line and the smaller the offset. Conversely, it indicates that the original path has strong asymmetry or local abnormal points. This indicator can be used for deviation detection and accuracy evaluation before path smoothing, and assist in judging the rationality and symmetry characteristics of trajectory adjustment;

[0095] According to the direction line offset adjustment value obtained in the previous step, the original coordinates of all track points need to be extracted from the path point set, and the path is set to contain points, corresponding to the following array of original coordinates: , based on the slope of the direction line Build direction line representation ,in is the intercept of the direction line at the center point of the path, substitute the point have to , the direction line formula is , project all original trajectory points vertically onto the direction line and obtain the coordinates of the projection points , and then perform numerical calculations according to the formula;

[0096] set up And all projection points are , then the first term is calculated as: ;

[0097] ;

[0098] Calculate and sum the six trajectory points separately, setting the values ​​to 0.4784, 0.39, 0.33, 0.25, 0.29, and 0.24, and the corresponding total is 1.9784. Calculate:

[0099] ;

[0100] The projected distance offset is a quantitative measure of the geometric deviation of a trajectory point from the adjusted direction line. Its significance lies in multiple dimensions, including geometric deviation measurement, trajectory rationality assessment, and spatial symmetry judgment. This result indicates that the current path has an overall offset value of 0.3297 from the adjusted direction line, which can be used to generate a reference baseline for path deviation correction.

[0101] See also Figure 5 The specific steps for obtaining the stable segment sequence of laser measurement points are as follows:

[0102] S411: Based on the spatial positions of the path points in the fitted flatness detection reference line, the corresponding laser echo data and time index information are extracted, and the path points are arranged in ascending order according to the time index to generate a time series spatial path dataset;

[0103] Based on the spatial position of the path points in the fitting flatness detection reference line, the path point data are extracted one by one, and a structure array in index order is established. Each structure contains the three-dimensional coordinates of the path point and its corresponding timestamp. The three groups of path points with path point indexes of 0, 1, and 2 in the paragraph correspond to timestamps of 1000ms, 1020ms, and 1045ms respectively. The spatial coordinates are X=0.0, 1.2, and 2.5m, Y=0.0, 0.8, and 1.5m, and Z=0.0, 0.5, and 0.9m respectively. In actual application, the path point data can be extracted one by one through the laser. The measuring device continuously scans the trajectory path. Each time a set of echo data is obtained, it is bound to a time stamp to form a group-by-group accumulated path point data set. After the data is imported, it is arranged in ascending order by timestamp, forming a time-series linear structure of path points with time as the main sequence. The path point set in any time interval is quickly retrieved and located through the array index structure or hash structure. At the software implementation level, the timestamp data is used as the primary key index, and XYZ space vector information is attached to each path point. The derived parameters such as the path segment length or slope distribution are calculated through vector difference to generate a time-series space path data set.

[0104] S412: Calling the temporal spatial path dataset, calculating the time difference between path points based on the time index values ​​of adjacent path points, dividing the time periods according to the time continuity condition, recording the number of path points in each period, and obtaining a segmented path point number sequence;

[0105] Extract the time difference between adjacent path points. Set the time difference between path point index 1 and index 0 to 20ms, and between index 2 and index 1 to 25ms. Set the continuity threshold to 30ms. Then the first two segments meet the continuity condition and are included in the same time period during the division. If the next time difference exceeds the threshold, it is divided into a new segment. In this way, the entire data set is divided into time periods according to the time difference value. Count the number of path points in each segment and generate a path point number array. Set that if the path point index in a certain time period is 5, 6, 7, and 8, the number of path points in this segment is 4. Generate a number sequence such as [4, 3, 5]. The path point continuity judgment criterion is that the difference between any two adjacent timestamps is less than the set time continuity threshold. The threshold is determined by the echo frequency. If the laser sampling period is 20ms, set the continuity threshold to no more than 30ms. In this example, the value is 30ms. Complete the time period division and count the number of path points through difference-by-difference judgment to obtain a segmented path point number sequence.

[0106] S413: Using the segmented path point number sequence, extract the change values ​​of the path point number in adjacent time periods, and construct an absolute difference value sequence based on the continuity and amplitude of the change values. Combined with the average offset value of the change trend, the intensity of the change, and the weight of the number of sections, a joint analysis is performed using the formula:

[0107] ;

[0108] Calculate the index of the change amplitude of the number of path points, select the section interval with stable change of the number of path points in the time period, and obtain the section sequence with stable laser measurement point timing;

[0109] in, Indicates the index of the change in the number of path points. Representative The difference in the number of path points between a segment and its adjacent segments, represents the mean of the difference in the number of path points during each time period, Indicates the The cumulative sum of the time index differences between a segment and its adjacent segments. Indicates the The number of waypoints in a segment and the offset weight for the average number, Indicates the total number of paragraphs;

[0110] The benefit of the formula is that by introducing the number of path points to offset the weight , time difference indicator The absolute value measurement of the difference in change amplitude comprehensively describes the temporal continuity and point density change characteristics within a paragraph and between adjacent paragraphs, providing a structured indicator for path stability judgment. The offset weight in the denominator provides a regularity constraint on the distribution of the number of paragraphs.

[0111] Indicator of the change in the number of path points It is a quantitative measure used to evaluate the trend of the number of points in a path segment. It aims to capture the local fluctuation characteristics of the number of path points in the time series and identify relatively stable or sudden change areas. The indicator is calculated by comparing the difference in the number of points in each segment with its adjacent segments. and the mean of the difference The absolute deviation is statistically analyzed, combined with the average deviation and weight factors , calculate the contribution of the change intensity of each segment. The numerator is the sum of the weighted correction change deviations, emphasizing the sudden change of the number of points; the denominator is adjusted with the offset weight to avoid the imbalance of the results caused by the number of points, thereby improving the sensitivity and robustness of fluctuation identification. The larger the value, the more unstable the change in the number of path points, while the smaller the value, the more stable the path structure distribution at the paragraph level. This indicator can be widely used in laser point cloud path, image path or spatial trajectory data to assist in locating areas with significant changes in landing point density for subsequent filtering, streamlining or anomaly detection.

[0112] Referring to the above parameter examples, Table 1 lists the sample data involved in the formula calculation:

[0113] Table 1 Laser measurement point path section index table

[0114] ;

[0115] The experimental data listed in Table 1 are used as the quantitative basis for the specific operations in the formula. The number of paragraph path points and the time index characteristics are combined for analysis. The results show that paragraph 1 and paragraph 2 have different change indicators. The above shows low-amplitude fluctuations, so it is identified as a temporally stable segment;

[0116] For the sequence of segmented path point numbers, define the change value of the path point number between segments, calculate the absolute change value sequence by sequence number, set the sequence [20, 18, 22] to [|20-18|, |18-22|] or [2, 4], and take the average change value For each paragraph, the change density metric is set to the sum of the time difference between the previous and next paragraphs , if the time between paragraphs is 20ms and 25ms respectively, then , the quantity offset weight is set as the square difference of the number of deviations from the average number in the paragraph. For example, if the average number of path points is 20, the offset weight is , substitute each parameter into the formula;

[0117] Taking paragraphs 1 and 2 as an example, there are , , , ,but:

[0118] ;

[0119] ;

[0120] The path point number variation index is a composite index that measures the intensity of variation and the stability of fluctuations in the number of path points in each segment of the time series path data. It evaluates the local structural characteristics of the path in different time periods or spatial segments by the amplitude of the change in the number of points and the trend consistency. The results show that the path point number variation index is 2.06, and the path segment stability threshold is set to 0.8 times the average variation amplitude value, that is, ,Will When compared with this value and the result is lower than the threshold, the segment is identified as a stable segment. The segment sequence whose variation index is lower than the average offset value is screened out to obtain the stable segment sequence of the laser measurement point time series.

[0121] See also Figure 6 The specific steps for obtaining the closed zone of the flatness detection profile are as follows:

[0122] S511: calling the start and end positions of the dense point segment in the laser point measurement time sequence stabilization segment sequence, and synchronously locating the same number of equally spaced measurement points along both sides of the fitted flatness detection reference line in the segment according to the start and end positions to generate a set of symmetrical point pairs;

[0123] According to the start and end positions, the same number of equally spaced measuring points are synchronously positioned on both sides of the fitted flatness detection reference line in the segment. This process requires identifying dense point segments in the complete time series of measuring points, scanning the laser measuring point time series point by point, and extracting the time interval variation pattern between measuring points. If the average time interval in a certain continuous interval is significantly lower than the average time interval of the entire sequence, and the number of consecutive points exceeds 15, it is marked as a dense point segment, and the first and last time points of the segment are the start and end positions; taking a certain trajectory segment as an example, the interval in the measuring point interval from time 0.35 seconds to 1.10 seconds is 0.01 seconds. , which is significantly shorter than the average interval of 0.03 seconds, and the total number of points is 76, it can be determined as a dense point segment; after obtaining the start and end positions, a fitting flatness detection reference line is constructed on this segment of the trajectory, and a center reference line is established by curve fitting; measurement point layout nodes are set at intervals of 0.2 meters on both sides of the reference line, and the number of nodes is evenly divided based on the full length of the start and end segments; a symmetrical measurement point coordinate is determined at each layout node on the left and right sides respectively, and a paired measurement point set can be established in the segment. If the total length of the segment is 2 meters, a total of 10 groups of measurement point pairs can be divided to generate a symmetrical point pair set.

[0124] S512: Based on the set of symmetrical point pairs, the spatial vector direction between each pair of points is calculated. The angle between two adjacent sets of spatial vector directions is calculated. The difference between the angle values ​​is compared with the angle value of the previous set. A sequence of point pairs with stable direction changes is screened out to generate a sequence of point groups with convergent directions.

[0125] Calculate the spatial vector direction between each pair of points, calculate the angle value for the two adjacent sets of spatial vector directions, and compare the difference between the angle values ​​with the previous set of angle values. Read the coordinates of each pair of left and right measuring points from the set of symmetrical point pairs in turn, and use the vector difference method to point each set of left measuring points to the right measuring point to define the spatial vector , the unit direction form of the calculated vector is used to compare the angles between adjacent vectors. Let Group vector and The angle between the group vectors is , and its calculation formula is:

[0126] ;

[0127] in, is the vector modulus, and the angle value in angle unit is obtained after angle conversion. Set the third and fourth groups of vectors to be 、 , then the included angle is calculated as follows:

[0128] ;

[0129] ;

[0130] ;

[0131] ;

[0132] After completing the calculation of each group of angle values ​​in turn, the angle value sequence is processed by first-order difference to determine whether the adjacent angle differences are continuously within a certain small range (set within 5°). If there are multiple measuring point sections with continuous differences stable in this range in the angle value sequence, the direction change trend of the measuring point pairs in this section tends to be consistent. The measuring point group number sequence with stable direction (such as measuring point group numbers 3 to 9) is screened out to generate a sequence of point groups with convergent directions.

[0133] S513: calling the direction convergence point group sequence, performing continuity identification on the index positions of the direction convergence points in the paragraph sequence, extracting the overall coverage interval, and marking the interval range on both sides of the fitted flatness detection reference line to generate a flatness detection profile closed zone;

[0134] Based on the index position of the directional convergence point pairs in the paragraph sequence, the continuity is identified, the overall coverage interval is extracted, and the interval range is marked on both sides of the fitted flatness detection reference line. The start and end positions corresponding to each measuring point number in the directional convergence point group sequence are confirmed, and the measuring point coordinate range is obtained from the previous symmetrical point pair set. The segmentation is based on the continuity of the measuring point numbers. If there is a sequence segment with continuous numbers (such as numbers 3 to 9), the left coordinate of the starting point pair and the right coordinate of the ending point pair are used as the interval boundaries to extract the spatial coverage range. The left measuring point of number 3 is set to (0.50, 1.19) and the right measuring point of number 9 is set to (3.00, 1.12). The spatial interval of this directional stable segment is (0.50, 1.19) to (3.00, 1.12). Based on this interval, the profile range is delineated on both sides of the fitted flatness detection reference line to form a closed detection area with left and right symmetry. The closed contour boundary wireframe is constructed to generate the flatness detection profile closed zone.

[0135] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A rapid detection method for road surface flatness based on laser radar, characterized in that: The following steps are involved: S1: Collect the roll, pitch, and yaw angle data during vehicle operation to form a three-dimensional attitude vector. Map the laser vector to the absolute space coordinate system, identify the projection difference between the laser ranging direction and the trajectory point in the main axis and lateral direction, and generate an attitude linkage correction trajectory sequence. S2: Based on a local window formed by three consecutive points in the posture linkage correction trajectory sequence, obtain the vertical length of the line from the middle point to the two ends, and determine whether the vertical length change trend constitutes a trajectory disturbance. Project and tile the position of the disturbance point to generate a distribution interval of the disturbance aggregation point; S3: calling the distribution interval of the disturbance aggregation point, identifying the translation operation of the main axis direction line according to the horizontal coordinate value, performing an interpolation operation in the path sequence according to the adjusted direction line, uniformly projecting the original trajectory points onto the direction line, and generating a fitting flatness detection reference line; S4: Using the fitting flatness to detect the spatial position of the path points in the reference line, extracting laser echo data and time index information, dividing the reference line into several time periods according to time continuity, counting the number of path points in the time period, and generating a time-stable segment sequence of laser measurement points.

2. The rapid detection method of road surface flatness based on laser radar according to claim 1 is characterized in that: The posture linkage correction trajectory sequence includes the absolute trajectory point coordinates after posture rotation, the spatial projection error parameters in the main transverse axis direction, and the adjusted path point position set; the disturbance concentration point distribution interval includes multiple disturbance point concentrated position blocks, trajectory path disturbance concentrated belt position marks, and distribution segment numbers under two-dimensional coordinate mapping; the fitting flatness detection reference line includes the path direction baseline, the direction line projection coordinate sequence, and the interference path interpolation positioning segment; the laser measurement point time series stable section sequence includes the continuous time period number, the number of laser points per unit time period, and the point segment under the stable driving state.

3. The rapid detection method of road surface flatness based on laser radar according to claim 1, characterized in that: The steps for obtaining the posture linkage correction trajectory sequence are specifically as follows: S111: Collect roll, pitch, and yaw angle data during vehicle operation, construct a three-dimensional attitude vector consisting of three-axis angles, call the time tag corresponding to each moment in the three-dimensional attitude vector, match it with the time tag in the laser echo data, identify the mapping direction of the laser ranging direction under the influence of the three-dimensional attitude and the absolute space coordinate system, and obtain a sequence of laser ranging direction angle values; S112: Calculate the cosine value of the angle between the laser ranging direction and the gravity direction based on the laser ranging direction angle value sequence, extract the corresponding trajectory point coordinate information based on the difference between the projection position of the vehicle trajectory point in the main axis and the lateral direction, adjust the original trajectory point position, and obtain the posture linkage correction trajectory sequence.

4. The rapid detection method of road surface flatness based on laser radar according to claim 3 is characterized in that: The steps for obtaining the distribution interval of the disturbance accumulation points are specifically as follows: S211: Based on a local window formed by three consecutive points in the posture linkage correction trajectory sequence, calculate the vertical length of the line from the middle point to the two ends, call the spatial coordinate difference of the starting point, the end point, and the middle point in the window to generate a vertical distance value set; S212: Calling the vertical distance value set, identifying the continuous change trend based on the vertical lengths corresponding to the middle points in the front and rear windows, screening the change segments whose fluctuation amplitude exceeds the disturbance threshold, calculating the disturbance fluctuation difference of the midpoints in the windows, and obtaining a disturbance trend mutation point sequence; S213: Calling the disturbance trend mutation point sequence, performing two-dimensional plane projection according to the distribution position of the disturbance point in the trajectory coordinate space, using the coordinate value after spatial mapping conversion, counting the concentrated distribution area of ​​the disturbance point in the continuous plane interval, and obtaining the distribution interval of the disturbance concentration point.

5. The rapid detection method of road surface flatness based on laser radar according to claim 4 is characterized in that: The steps for obtaining the fitting flatness detection reference line are specifically as follows: S311: Recalling the quantitative relationship of point groups on both sides of the path in the distribution interval of the disturbance aggregation points, comparing the quantitative distribution of the point groups on the left and right sides, identifying the density difference in the horizontal coordinate, and evaluating the degree of interference tendency on the main axis of the path based on the quantitative distribution difference of the point groups on the left and right sides, to obtain the disturbance weight difference; S312: Based on the disturbance weight difference, the offset trend of the main axis direction line is identified according to the transverse coordinate value, the point distribution within the transverse coordinate extreme value interval of the path sequence is called, and the spatial direction of the main axis direction line is adjusted in combination with the center coordinate set between the main axis starting point and the end point to obtain the direction line offset adjustment value; S313: According to the direction line offset adjustment value, the horizontal and vertical coordinates of the path points are extracted from the path sequence, and an interpolation operation is performed to obtain the coordinates of the equidistant interpolation points on the direction line. The coordinate projection transformation relationship is identified, and the original trajectory points are uniformly projected onto the adjusted direction line. The projection distance offset is calculated, and the projection result is linearly smoothed to generate a fitting flatness detection reference line.

6. The rapid detection method of road surface flatness based on laser radar according to claim 5 is characterized in that: The steps for obtaining the laser measurement point time sequence stable segment sequence are specifically as follows: S411: Based on the spatial positions of the path points in the fitted flatness detection reference line, extract the corresponding laser echo data and time index information, arrange the path points in ascending order according to the time index, and generate a time-series spatial path data set; S412: Calling the temporal spatial path dataset, calculating the time difference between path points based on the time index values ​​of adjacent path points, dividing the time periods according to the time continuity condition, recording the number of path points in each period, and obtaining a segmented path point number sequence; S413: Using the segmented path point number sequence, extract the change values ​​of the path point number in adjacent time periods, construct an absolute difference value sequence based on the continuity and change amplitude of the change value, perform a joint analysis based on the average offset value of the change trend, the change density and the paragraph number weight factor, calculate the path point number change amplitude index, screen the paragraph intervals with stable path point number changes in the time period, and obtain a laser measurement point time series stable paragraph sequence.

7. The rapid detection method of road surface flatness based on laser radar according to claim 1, characterized in that: The method further comprises step S5: S5: calling the start and end positions of the dense point segment in the laser point measurement time sequence stable segment sequence, extracting symmetrical point groups along both sides of the fitted flatness detection reference line in the segment, comparing the directions of the spatial vectors between the symmetrical points, and if the angles tend to be consistent, determining it as a closed boundary segment, and generating a closed zone of the flatness detection profile; The flatness detection profile closed zone includes a set of structural boundary points on both sides, a sideband direction vector sequence, and a closed path constituting a segment.

8. The rapid detection method of road surface flatness based on laser radar according to claim 7 is characterized in that: The steps for obtaining the closed zone of the flatness detection profile are specifically as follows: S511: calling the start and end positions of the dense point segment in the laser point measurement time sequence stable paragraph sequence, and synchronously locating the same number of equally spaced measurement points along both sides of the fitted flatness detection reference line in the paragraph according to the start and end positions to generate a set of symmetrical point pairs; S512: Based on the set of symmetrical point pairs, the spatial vector direction between each pair of points is calculated, the angle between two adjacent sets of spatial vector directions is calculated, and the difference between the angle values ​​is compared with the angle value of the previous set, thereby selecting a sequence of point pairs with stable direction changes and generating a sequence of point groups with convergent directions. S513: Calling the directional convergence point group sequence, performing continuity identification on the index positions of the directional convergence points in the paragraph sequence, extracting the overall coverage interval, and marking the interval range on both sides of the fitted flatness detection reference line to generate a flatness detection profile closed zone.

Images