3D intelligent paving control method and system

Abstract

The application provides a 3D intelligent paving control method and system, and belongs to the field of construction road control, and comprises the following steps: S1, receiving a building information model file containing road design information, and obtaining a smooth base deviation field after processing; S2, calculating a preliminary target paving thickness by using a negative feedback compensation strategy, constraining the preliminary target paving thickness, and calculating a target elevation of a screed; S3, calculating a deviation change intensity index in a scanning area in front of a current position of a paver; S4, dynamically adjusting a preview distance and a feedforward compensation coefficient of the screed by using a nonlinear mapping function; S5, calculating a feedforward control output; calculating a feedback control output by using a proportional-integral control law, obtaining a screed height adjusting amount, and adjusting the screed height of the paver. The application dynamically adjusts the paving thickness by using an adaptive feedforward-feedback, and improves the paving layer thickness uniformity and the surface flatness.

Description

Technical Field

[0001] This invention relates to the field of construction road control technology, and in particular to a 3D intelligent paving control method and system. Background Technology

[0002] Quality control in asphalt pavement paving involves multiple parameters, among which paving thickness is the most direct and critical control objective. According to the technical specifications for highway asphalt pavement construction, the thickness deviation of the asphalt surface layer should be controlled within a certain range of the design thickness. The accuracy of thickness control directly affects the pavement's performance and service life: insufficient thickness leads to a decrease in the pavement's structural load-bearing capacity and fatigue resistance, making it prone to early failure; excessive thickness results in material waste, increased project costs, and may also affect compaction quality due to changes in heat dissipation conditions.

[0003] Asphalt pavement is a multi-layered composite structural system. The paving of the asphalt surface layer is built upon the base course, and the geometry of the base course directly determines the boundary conditions of the paving operation. In actual construction, base course elevation deviations can occur for several reasons. First, there are inherent construction errors in the base course itself. Second, uneven settlement of the base course materials can occur during curing, especially in soft soil foundations or cut-fill junctions, where settlement differences are more pronounced. Third, uneven compaction during the compaction process can cause localized bulges or depressions. If these base course elevation deviations are not effectively compensated for during the paving process, they will directly lead to uneven distribution of the asphalt surface layer thickness and deterioration of its smoothness.

[0004] Chinese patent CN108463834A discloses a control system for coordinating paving operations. This system integrates operating parameters from multiple machines, including the paver's ground speed and screed settings, and displays them to the operator through a graphical user interface to help coordinate the operation of each machine during paving. However, it lacks the ability to perceive the actual condition of the base layer in real time and cannot dynamically adjust the paving thickness according to the elevation deviation of the base layer. As a result, it is difficult to ensure the uniformity of the paving layer thickness and the surface smoothness when the base layer has poor flatness. Summary of the Invention

[0005] In view of this, the present invention proposes a 3D intelligent paving control method and system, which can solve the problem in the prior art that the paving thickness cannot be adjusted in real time according to the actual condition of the base layer. By detecting the three-dimensional contour of the base layer in real time and comparing it with the design model, the elevation deviation of the base layer is identified, and the paving thickness is dynamically adjusted by adaptive feedforward-feedback, thereby improving the uniformity of paving layer thickness and surface smoothness.

[0006] The technical solution of this invention is implemented as follows: On the one hand, the present invention provides a 3D intelligent paving control method, comprising the following steps: S1. Receive the building information model file containing road design information, generate the base layer design elevation surface function and the surface layer design thickness function based on the building information model file, and obtain the spatial position coordinates and attitude angle of the paver to calculate the deviation between the measured elevation and the design elevation of the base layer. After gridding and smoothing filtering, the smoothed base layer deviation field is obtained. S2. Based on the smoothed base course deviation field and the surface course design thickness function, the initial target paving thickness is calculated using a negative feedback compensation strategy. The initial target paving thickness is constrained to obtain the constrained target paving thickness. The loose paving thickness is calculated based on the constrained target paving thickness and the compaction coefficient. The target elevation of the screed is calculated based on the base course design elevation, the smoothed base course deviation, and the loose paving thickness. S3. Within the scanning area in front of the current position of the paver, for each lateral position, calculate the longitudinal change rate sequence of the smoothed base deviation field, and normalize it based on the standard deviation of the longitudinal change rate sequence to obtain the deviation change intensity index. S4. Based on the deviation change severity index, the anticipation distance and feedforward compensation coefficient of the ironing plate feedforward control are dynamically adjusted using a nonlinear mapping function, and the anticipation distance and feedforward compensation coefficient of the adjusted ironing plate are then subjected to lateral smoothing to obtain the processed anticipation distance and processed feedforward compensation coefficient. S5. Based on the smoothed predicted distance and the target elevation of the screed, the feedforward control output is calculated by combining the smoothed feedforward gain and the current screed elevation. The proportional-integral control law is used to calculate the feedback control output based on the deviation between the actual screed elevation and the target elevation. The feedforward control output and the feedback control output are superimposed to obtain the screed height adjustment amount. The screed height adjustment amount is coordinated and corrected according to the cross slope angle of the screed. The screed height of the paver is adjusted according to the coordinated and corrected screed height adjustment amount.

[0007] Based on the above technical solutions, preferably, step S1 specifically includes: The spatial position coordinates and attitude angles of the paver are obtained through multiple sensors; The three-dimensional geometry of the base surface in front of the paving operation is scanned by a front laser contour scanner to obtain the three-dimensional point coordinates in the local coordinate system of the sensor. The coordinates are then transformed into the engineering coordinate system and the road alignment coordinate system in turn through a homogeneous coordinate transformation matrix to obtain the mileage value and lateral offset of each point in the point cloud. For any point in the point cloud, the corresponding design elevation is queried from the base design elevation surface function, and the base deviation is obtained by subtracting the design elevation from the measured base elevation. The base deviation of each point in the point cloud is processed into a grid. The paving area is divided into regular rectangular grids in the road linear coordinate system. For each grid cell, all points falling within the grid range are searched, and the median of the elevation deviation of all these points is calculated as the representative value of the grid, thus obtaining the gridded base deviation field. Gaussian smoothing filtering is applied to the gridded base deviation field to obtain the smoothed base deviation field.

[0008] Based on the above technical solutions, preferably, the negative feedback compensation strategy specifically includes: When the actual elevation of the base course is lower than the design elevation, the paving thickness is increased to compensate for the base course deficit; when the actual elevation of the base course is higher than the design elevation, the paving thickness is decreased to correct the base course superelevation. The compensation direction is opposite to the base course deviation direction. The preliminary target paving thickness is calculated by subtracting the compensation coefficient from the product of the designed paving thickness and the smoothed base course deviation. The calculation formula is as follows: in, The initial target paving thickness, To design the paving thickness, For compensation coefficient, To smooth the base layer deviation field, and These are the grid coordinates for longitudinal mileage and lateral offset, respectively.

[0009] Based on the above technical solutions, preferably, the constraints include absolute thickness constraints and thickness gradient constraints, wherein, The absolute thickness constraint is set according to the material compaction process requirements, specifying the minimum and maximum thicknesses. The thickness gradient constraint indirectly controls the attitude angle of the ironing plate by limiting the thickness variation gradient between adjacent mesh cells, wherein the thickness gradient constraint includes longitudinal thickness gradient constraint and lateral thickness gradient constraint; When the initial target paving thickness is less than the minimum thickness or greater than the maximum thickness, or when the thickness variation gradient between adjacent grid cells exceeds the allowable thickness variation rate, the gradient projection method is used to correct the initial target paving thickness so that the corrected constrained target paving thickness satisfies the absolute thickness constraint and the thickness gradient constraint.

[0010] Based on the above technical solutions, preferably, step S3 specifically includes: For each lateral position within the scanning area in front of the current position of the paver, extract the smoothed base deviation value distributed along the longitudinal direction at that lateral position to form a longitudinal deviation sequence; For any forward grid point in the longitudinal deviation sequence, the rate of change of the longitudinal deviation at that grid point is calculated using the central difference method. Extract the longitudinal deviation rate of change sequence within the forward scanning range and calculate the standard deviation of the rate of change sequence; The standard deviation of the rate of change series is normalized to obtain the index of the severity of deviation change.

[0011] Based on the above technical solutions, preferably, the step of dynamically adjusting the prediction distance and feedforward compensation coefficient of the ironing plate feedforward control using a nonlinear mapping function specifically includes: The baseline foreseeability distance is calculated by multiplying the paving speed by the response time of the screed height adjustment unit. Based on the deviation change severity index, the hyperbolic tangent function is used to adaptively adjust the baseline prediction distance. Upper and lower limit constraints are applied to the adjusted prediction distance to obtain the adjusted prediction distance: When the deviation change severity index is higher than the severity threshold, the hyperbolic tangent function outputs a positive value, which extends the prediction distance to read the elevation information of the target ahead in advance. When the deviation change severity index is lower than the severity threshold, the hyperbolic tangent function outputs a negative value, which shortens the prediction distance; Set a baseline feedforward compensation coefficient. When the intensity index is below the intensity threshold, maintain the baseline feedforward compensation coefficient. When the intensity index is above the intensity threshold, decrease the baseline feedforward compensation coefficient according to the degree to which the intensity index exceeds the threshold.

[0012] Based on the above technical solutions, preferably, the formula for calculating the adjusted prediction distance is as follows: ; in, This is the adjusted forecast distance. Indicates the baseline foresight distance. To adjust the amplitude coefficient, Represents the hyperbolic tangent function. The intensity threshold, This is the soft threshold transition width. These are the grid coordinates for the horizontal position. Indicates horizontal position The index of the degree of deviation change at a certain point.

[0013] Based on the above technical solutions, preferably, step S5 specifically includes: The predicted position is determined based on the smoothed predicted distance. The predicted position is equal to the current position of the paver plus the smoothed predicted distance. The target elevation of the screed at the predicted position is calculated using a linear interpolation method. The difference between the target elevation of the ironing plate at the predicted location and the current elevation of the ironing plate is multiplied by the smoothed feedforward compensation coefficient to calculate the feedforward control output. The difference between the actual elevation of the ironing board and the target elevation of the ironing board at the current position is defined as the elevation tracking error, and the feedback control output is calculated using a proportional-integral control law. The height adjustment of the ironing board is obtained by superimposing the output of the feedforward control and the output of the feedback control. Calculate the actual and target cross slope angles of the screed, and adjust the screed height adjustment amount accordingly. Adjust the screed height of the paver based on the adjusted screed height adjustment amount.

[0014] More preferably, the formula for calculating the feedback control output is: ; in, For feedback control output quantity, For elevation tracking error, This is the proportionality coefficient. The integral coefficient is... To control the cycle, Index for historical moments.

[0015] On the other hand, the present invention provides a 3D intelligent paving control system, which applies a 3D intelligent paving control method as described above, including: The deviation calculation module is used to receive the building information model file containing road design information, generate the base layer design elevation surface function and the surface layer design thickness function based on the building information model file, and obtain the spatial position coordinates and attitude angle of the paver to calculate the deviation between the measured elevation of the base layer and the design elevation. After gridding and smoothing filtering, the smoothed base layer deviation field is obtained. The thickness compensation module is used to calculate the initial target paving thickness based on the smoothed base course deviation field and the surface course design thickness function, using a negative feedback compensation strategy. It then constrains the initial target paving thickness to obtain the constrained target paving thickness. Based on the constrained target paving thickness and the compaction coefficient, it calculates the loose paving thickness. Finally, it calculates the target elevation of the screed based on the base course design elevation, the smoothed base course deviation, and the loose paving thickness. The deviation analysis module is used to calculate the longitudinal rate of change sequence of the smoothed base deviation field for each lateral position within the scanning area in front of the current position of the paver, and to normalize it based on the standard deviation of the longitudinal rate of change sequence to obtain the deviation change intensity index. The parameter adjustment module is used to dynamically adjust the prediction distance and feedforward compensation coefficient of the ironing plate feedforward control according to the deviation change severity index using a nonlinear mapping function, and to perform lateral smoothing on the adjusted prediction distance and feedforward compensation coefficient of the ironing plate to obtain the processed prediction distance and processed feedforward compensation coefficient. The control output module is used to calculate the feedforward control output based on the smoothed predicted distance and the target elevation of the screed, combined with the smoothed feedforward gain and the current screed elevation; it uses a proportional-integral control law to calculate the feedback control output based on the deviation between the actual screed elevation and the target elevation; it then superimposes the feedforward control output and the feedback control output to obtain the screed height adjustment amount, and adjusts the screed height of the paver according to the screed height adjustment amount.

[0016] The 3D intelligent paving control method and system of the present invention have the following advantages over the prior art: (1) The present invention acquires the three-dimensional geometric shape of the base layer in real time through a front laser contour scanner, which can accurately perceive the actual state of the base layer and provide a reliable data basis for dynamic compensation of paving thickness; (2) The deviation field processing method combining gridding and Gaussian smoothing filtering not only ensures the spatial resolution of the base deviation data, but also effectively suppresses measurement noise and interference from local abrupt changes. (3) By adjusting the feedforward prediction distance and feedforward compensation coefficient, the control parameters are dynamically adjusted according to the degree of change of the deviation of the front base layer. In the gradual change region, the prediction distance is shortened to improve the tracking accuracy, and in the sudden change region, the prediction distance is extended and the feedforward gain is reduced to avoid over-response, thus reducing the phenomenon of thickness exceeding the limit. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a 3D intelligent paving control method according to the present invention; Figure 2 This is a schematic diagram of the base deviation field processing of a 3D intelligent paving control method according to the present invention; Figure 3 This is a block diagram of thickness compensation calculation for a 3D intelligent paving control method according to the present invention; Figure 4 This is a control system architecture diagram of a 3D intelligent paving control method according to the present invention. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0020] like Figure 1 As shown, the present invention provides a 3D intelligent paving control method, comprising the following steps: S1. Receive the building information model file containing road design information, generate the base course design elevation surface function and the surface course design thickness function based on the building information model file, and obtain the spatial position coordinates and attitude angle of the paver to calculate the deviation between the measured elevation and the design elevation of the base course. After gridding and smoothing filtering, the smoothed base course deviation field is obtained.

[0021] Understandably, Building Information Modeling (BIM) files contain information such as the three-dimensional coordinate sequence of the road centerline, road cross-section design parameters, base course design elevation, and surface course design thickness. A continuous base course design elevation surface function is generated using cubic spline interpolation. and surface layer design thickness function ,in This is the cumulative mileage along the centerline of the road, in meters. The lateral offset is perpendicular to the centerline, expressed in meters, with positive values ​​to the left and negative values ​​to the right.

[0022] This invention acquires the three-dimensional geometric shape of the base layer in real time through a front-mounted laser contour scanner, which can accurately perceive the actual state of the base layer and avoid the problem of paving thickness deviating from the design requirements due to base layer deviation, thus providing a reliable data basis for dynamic compensation of paving thickness.

[0023] like Figure 2 As shown, specifically, step S1 includes: The spatial position coordinates and attitude angles of the paver are acquired through multiple sensors; the attitude angles include the pitch angle. Roll angle and heading angle ; The three-dimensional geometry of the base surface in front of the paving operation is scanned by a front laser contour scanner to obtain the three-dimensional point coordinates in the local coordinate system of the sensor. The coordinates are then transformed into the engineering coordinate system and the road alignment coordinate system in turn through a homogeneous coordinate transformation matrix to obtain the mileage value and lateral offset of each point in the point cloud. For any point in the point cloud, query the corresponding design elevation from the base design elevation surface function. The actual measured elevation at the grassroots level Subtract design elevation Obtaining grassroots deviation A positive value indicates that the actual elevation of the base course is higher than the design value, while a negative value indicates that the actual elevation of the base course is lower than the design value; among which, Indicating the first point cloud Mileage coordinates of each point, Indicating the first point cloud The lateral offset coordinates of each point Indicating the first point cloud The measured elevation of each point at the base level. Indicating the first point cloud The base design elevation corresponding to each point location; The base deviation of each point in the point cloud is processed into a grid. The paving area is divided into regular rectangular grids in the road alignment coordinate system. For each grid cell... The algorithm searches for all points falling within the grid range and calculates the median elevation deviation of all these points as the representative value of the grid, thus obtaining the gridded base deviation field; where, , Represents the mileage coordinates of the vertical grid points. Indicates the offset coordinates of the horizontal grid points. Indicates a vertical grid index. Indicates the vertical grid spacing. Indicates the horizontal grid spacing. Indicates a horizontal grid index; Gaussian smoothing filtering is applied to the gridded base deviation field to obtain the smoothed base deviation field.

[0024] Understandably, the front-mounted laser profile scanner is installed at the front of the paver frame and works on the principle of line laser triangulation. It transforms the three-dimensional geometry of the scanner into the road alignment coordinate system through a homogeneous coordinate transformation matrix, obtaining the mileage value of each point in the point cloud. and lateral offset The coordinate transformation process involves multiple transformations from the sensor coordinate system to the paver body coordinate system, from the paver body coordinate system to the engineering coordinate system, and from the engineering coordinate system to the road alignment coordinate system. Each transformation uses a four-dimensional homogeneous coordinate matrix to achieve a unified expression of rotation and translation.

[0025] This invention employs a deviation field processing method that combines gridding and Gaussian smoothing filtering. This method ensures the spatial resolution of the base deviation data while effectively suppressing measurement noise and interference from local abrupt changes. It extracts deviation features that truly reflect the overall trend of the base layer, providing a stable and reliable input for subsequent compensation calculations and avoiding frequent erroneous adjustments caused by noise.

[0026] In one embodiment of the present invention, the formula for calculating the smoothed base layer deviation field is: in, Represents grid points Deviation of the base layer after smoothing The standard deviation parameter of the Gaussian filter is determined based on the flatness characteristics of the base layer. express, express, Let be a normalization constant, such that the sum of all weight coefficients equals one. The radius of the filtering window is 7×7 grid size. Represents grid points At the base deviation of the grid, Indicates relative position The weight coefficients of the neighboring grid points.

[0027] S2. Based on the smoothed base course deviation field and the surface course design thickness function, the initial target paving thickness is calculated using a negative feedback compensation strategy. The initial target paving thickness is constrained to obtain the constrained target paving thickness. The loose paving thickness is calculated based on the constrained target paving thickness and the compaction coefficient. The target elevation of the screed is calculated based on the base course design elevation, the smoothed base course deviation, and the loose paving thickness. like Figure 3 As shown, specifically, the negative feedback compensation strategy includes: When the actual elevation of the base course is lower than the design elevation, the paving thickness is increased to compensate for the base course deficit; when the actual elevation of the base course is higher than the design elevation, the paving thickness is decreased to correct the base course superelevation. The compensation direction is opposite to the base course deviation direction. The preliminary target paving thickness is calculated by subtracting the compensation coefficient from the product of the designed paving thickness and the smoothed base course deviation. The calculation formula is as follows: in, For grid points The initial target paving thickness at the location. For grid points The design paving thickness at this location, For compensation coefficient, For grid points After smoothing, the base layer deviation field is obtained. and These are the grid coordinates for longitudinal mileage and lateral offset, respectively.

[0028] In one embodiment of the present invention, the constraints include an absolute thickness constraint and a thickness gradient constraint, wherein, The absolute thickness constraint is set according to the material compaction process requirements, specifying the minimum and maximum thicknesses. The thickness gradient constraint indirectly controls the attitude angle of the ironing plate by limiting the thickness variation gradient between adjacent mesh cells, wherein the thickness gradient constraint includes longitudinal thickness gradient constraint and lateral thickness gradient constraint; When the initial target paving thickness is less than the minimum thickness or greater than the maximum thickness, or when the thickness variation gradient between adjacent grid cells exceeds the allowable thickness variation rate, the gradient projection method is used to correct the initial target paving thickness so that the corrected constrained target paving thickness satisfies the absolute thickness constraint and the thickness gradient constraint.

[0029] Understandably, the longitudinal thickness gradient is limited as follows: in, Represents grid points The initial target paving thickness at the location. The allowable longitudinal thickness variation rate is determined based on the paving machinery performance and paving speed.

[0030] The lateral thickness gradient is limited to: in The allowable rate of change in lateral thickness.

[0031] This invention constrains the initial target paving thickness, ensuring that the paving thickness and screed angle meet the requirements of material compaction process and mechanical properties. This avoids construction quality problems such as excessive thickness or excessive screed angle caused by overcompensation, and guarantees the operability of paving operations and the stability of construction quality.

[0032] In one embodiment of the present invention, the gradient projection method specifically includes: A1. Truncate and correct grid points that violate the absolute value constraint, that is, set grid values ​​less than the minimum thickness to the minimum thickness and grid values ​​greater than the maximum thickness to the maximum thickness. A2. Adjust adjacent mesh pairs that violate gradient constraints by distributing thickness deviations proportionally between the two mesh points. A3. Iterate through A1-A2 until all constraints are met or the maximum number of iterations is reached, and obtain the corrected target paving thickness after constraints are met. .

[0033] When the deviation of some grid cells is too large, even if the maximum or minimum thickness is used, it is still impossible to meet the design elevation requirements of the finished surface. The system records the coordinates and deviation values ​​of these locations and sends an early warning signal to the operator, indicating that the deviation of the base layer exceeds the compensation capacity of the paving process and that the base layer needs to be repaired before paving can proceed.

[0034] Furthermore, the loose paving thickness is calculated based on the constrained target paving thickness and compaction coefficient. The formula for calculating the loose paving thickness is as follows: in, Represents grid points The thickness of the paving at the point, Represents grid points The target paving thickness after constraints at the location. Indicates the compaction coefficient; The target elevation of the screed is calculated based on the base course design elevation, the smoothed base course deviation, and the loose-lay thickness. The target elevation of the screed is the actual base course elevation plus the loose-lay thickness. The calculation formula is as follows: in, Represents grid points The ironing board is high. Represents grid points The basic design elevation of the location.

[0035] Understandably, the compaction coefficient is the ratio of the compacted thickness to the loose paving thickness. For asphalt concrete mixtures, the compaction coefficient is usually between 0.77 and 0.83; for asphalt mastic macadam mixtures, the compaction coefficient is between 0.87 and 0.95.

[0036] In one embodiment of the present invention, AC-20 asphalt concrete is used, and the compaction coefficient is 0.82.

[0037] By introducing a negative feedback compensation mechanism for base course deviations, the paving thickness can be dynamically adjusted according to the actual condition of the base course. Compared with the traditional fixed-thickness paving method, this effectively compensates for the impact of base course elevation deviations on the finished surface elevation. Simultaneously, by applying process constraints, the operability of the compensation scheme and the stability of construction quality are ensured.

[0038] S3. Within the scanning area in front of the current position of the paver, for each lateral position, calculate the longitudinal change rate sequence of the smoothed base deviation field, and normalize it based on the standard deviation of the longitudinal change rate sequence to obtain the deviation change intensity index. Specifically, step S3 includes: For each lateral position within the scanning area in front of the paver's current position, extract the smoothed base course deviation values ​​distributed longitudinally at that lateral position to form a longitudinal deviation sequence. ,in , ; This indicates the mileage coordinates of the paver's current position. Indicates the position preceding the current position. One grid, Indicates the number of vertical grid points within the scanning range ahead; For any leading grid point in the longitudinal deviation sequence, the rate of change of longitudinal deviation at that grid point is calculated using the central difference method. The calculation formula is as follows: in, Indicates the grid points in front. The rate of change of longitudinal deviation at that location. Indicates the mileage coordinates of the observation position ahead. Represents grid points Deviation of the base layer after smoothing; Extract the longitudinal deviation rate of change sequence within the forward scanning range. Calculate the standard deviation of this rate of change series using the following formula: in, Indicates horizontal position The standard deviation of the longitudinal deviation rate of change sequence Indicates horizontal position The mean of the longitudinal deviation rate of change sequence; Normalizing the standard deviation of the rate of change series yields an index of the severity of deviation change, calculated using the following formula: in, Indicates horizontal position The index of the degree of deviation change at a given location. This represents the reference standard deviation.

[0039] This invention achieves a quantitative description of the spatial variation characteristics of base course deviation by using a deviation severity index based on the standard deviation of the rate of change. It can distinguish between two different modes: gradual settlement and abrupt defect, and can dynamically optimize response characteristics according to the base course condition characteristics.

[0040] S4. Based on the deviation change severity index, the predictive distance and feedforward compensation coefficient of the screed are dynamically adjusted using a nonlinear mapping function. The predicted distance and feedforward compensation coefficient of the adjusted screed are then subjected to lateral smoothing to obtain the processed predicted distance and processed feedforward compensation coefficient. The lateral smoothing is performed using a three-point weighted average method, which involves weighting the current point and its two adjacent points according to their weights. The weighting method corresponds to a normalized symmetric convolution kernel, which can suppress lateral abrupt changes while maintaining the overall trend of the data. This avoids drastic changes in control parameters at adjacent lateral positions that could cause the screed to twist and deform in the lateral direction, ensuring the coordinated action of the control points on the left and right sides of the screed and the lateral smoothness of the paving surface.

[0041] In one embodiment of the present invention, the step of dynamically adjusting the prediction distance and feedforward compensation coefficient of the ironing plate feedforward control using a nonlinear mapping function specifically includes: The baseline prediction distance is calculated by multiplying the paving speed by the response time of the screed height adjustment unit: in, Indicates the baseline foresight distance. Indicates the paving speed. This indicates the response time of the ironing board height adjustment actuator; Based on the deviation change severity index, the hyperbolic tangent function is used to adaptively adjust the baseline prediction distance. Upper and lower limit constraints are applied to the adjusted prediction distance to obtain the adjusted prediction distance: When the deviation change severity index is higher than the severity threshold, the hyperbolic tangent function outputs a positive value, which extends the prediction distance to read the elevation information of the target ahead in advance. When the deviation change severity index is lower than the severity threshold, the hyperbolic tangent function outputs a negative value, which shortens the prediction distance; Set a baseline feedforward compensation coefficient. When the intensity index is below the intensity threshold, maintain the baseline feedforward compensation coefficient. When the intensity index is above the intensity threshold, decrease the baseline feedforward compensation coefficient according to the degree to which the intensity index exceeds the threshold.

[0042] Specifically, the formula for calculating the adjusted forecast distance is as follows: in, This is the adjusted forecast distance. Indicates the baseline foresight distance. To adjust the amplitude coefficient, Represents the hyperbolic tangent function. The intensity threshold, This is the soft threshold transition width. These are the grid coordinates for the horizontal position. Indicates horizontal position The index of the degree of deviation change at a certain point.

[0043] In one embodiment of the present invention, the reference feedforward compensation coefficient is attenuated, and the calculation formula is as follows: in, Indicates horizontal position The adjusted feedforward compensation coefficient, Indicates the reference feedforward compensation coefficient. Indicates the attenuation coefficient. Indicates horizontal position The index of the degree of deviation change at a given location. Indicates the severity threshold; when hour, ;when At that time, the feedforward compensation coefficient decreases as the intensity increases, and the higher the intensity index, the greater the attenuation.

[0044] This invention dynamically adjusts control parameters based on the degree of change in the deviation of the base layer by adjusting the feedforward prediction distance and feedforward compensation coefficient. In the gradual change region, the prediction distance is shortened to improve tracking accuracy, while in the abrupt change region, the prediction distance is extended and the feedforward gain is reduced to avoid overreaction. Compared with fixed parameter control, this invention improves the adaptability to different base layer conditions and reduces the phenomenon of thickness exceeding limits.

[0045] S5. Based on the smoothed predicted distance and the target elevation of the screed, the feedforward control output is calculated by combining the smoothed feedforward gain and the current screed elevation. The proportional-integral control law is used to calculate the feedback control output based on the deviation between the actual screed elevation and the target elevation. The feedforward control output and the feedback control output are superimposed to obtain the screed height adjustment amount. The screed height adjustment amount is coordinated and corrected according to the cross slope angle of the screed. The screed height of the paver is adjusted according to the coordinated and corrected screed height adjustment amount.

[0046] like Figure 4 As shown, specifically, step S5 includes: The predicted position is determined based on the smoothed predicted distance. The predicted position is equal to the paver's current position plus the smoothed predicted distance. The target elevation of the screed at the predicted position is then calculated using a linear interpolation method. in, Indicates horizontal position The target elevation of the screed at the predicted location. Represents the linear interpolation coefficients. Represents grid points The target elevation of the ironing plate at the location. Represents grid points The target elevation of the ironing plate at the location. Indicates horizontal position The predicted location mileage coordinates, This indicates the mileage coordinates of the paver's current position. Indicates horizontal position The predicted distance after smoothing. Represents the mileage coordinates of the vertical grid points. Indicates the vertical grid spacing; The difference between the target elevation of the screed at the predicted location and the current elevation of the screed is multiplied by the smoothed feedforward compensation coefficient to calculate the feedforward control output: in, Indicates the left adjustment point at time. The feedforward control output. This represents the feedforward compensation coefficient after smoothing at the left adjustment point position. Indicates horizontal position The target elevation of the screed at the predicted location. Indicates the current elevation of the left-side screed. The difference between the actual elevation of the ironing board and the target elevation of the ironing board at the current position is defined as the elevation tracking error, and the feedback control output is calculated using a proportional-integral control law. The ironing board height adjustment amount is obtained by superimposing the feedforward control output and the feedback control output: in, Indicates the left adjustment point at time. The height adjustment range of the ironing board. Indicates the left adjustment point at time. The feedforward control output. Indicates the left adjustment point at time. Feedback control output quantity, Indicates the adjustment point on the right at time [time]. The height adjustment range of the ironing board. Indicates the adjustment point on the right at time [time]. Feedforward control output quantity Indicates the right-side adjustment point at time [time]. Feedback control output quantity; Calculate the actual and target cross slope angles of the screed, and adjust the screed height adjustment amount accordingly. Then, adjust the paver's screed height based on the adjusted screed height adjustment amount. in, This indicates the actual cross slope angle of the ironing board. This indicates the actual elevation of the screed on the right side. This indicates the actual elevation of the screed on the left. This indicates the lateral distance between the left and right adjustment points of the ironing board. Indicates the target cross slope angle of the ironing plate. This indicates the target elevation of the right-side ironing plate. This indicates the target elevation of the left-side ironing plate.

[0047] By adjusting the screed height in advance based on the deviation information at the base level, the impact of the actuator's response lag is reduced. Real-time correction is made based on the deviation between the actual elevation of the screed and the target elevation, thereby improving the accuracy and robustness of elevation control.

[0048] In one embodiment of the present invention, the formula for calculating the feedback control output is: in, For feedback control output quantity, For elevation tracking error, This is the proportionality coefficient. The integral coefficient is... To control the cycle, Index for historical moments.

[0049] The present invention also provides a 3D intelligent paving control system, which applies the 3D intelligent paving control method described above, including: The deviation calculation module is used to receive the building information model file containing road design information, generate the base layer design elevation surface function and the surface layer design thickness function based on the building information model file, and obtain the spatial position coordinates and attitude angle of the paver to calculate the deviation between the measured elevation of the base layer and the design elevation. After gridding and smoothing filtering, the smoothed base layer deviation field is obtained. The thickness compensation module is used to calculate the initial target paving thickness based on the smoothed base course deviation field and the surface course design thickness function, using a negative feedback compensation strategy. It then constrains the initial target paving thickness to obtain the constrained target paving thickness. Based on the constrained target paving thickness and the compaction coefficient, it calculates the loose paving thickness. Finally, it calculates the target elevation of the screed based on the base course design elevation, the smoothed base course deviation, and the loose paving thickness. The deviation analysis module is used to calculate the longitudinal rate of change sequence of the smoothed base deviation field for each lateral position within the scanning area in front of the current position of the paver, and to normalize it based on the standard deviation of the longitudinal rate of change sequence to obtain the deviation change intensity index. The parameter adjustment module is used to dynamically adjust the prediction distance and feedforward compensation coefficient of the ironing plate feedforward control according to the deviation change severity index using a nonlinear mapping function, and to perform lateral smoothing on the adjusted prediction distance and feedforward compensation coefficient of the ironing plate to obtain the processed prediction distance and processed feedforward compensation coefficient. The control output module is used to calculate the feedforward control output based on the smoothed predicted distance and the target elevation of the screed, combined with the smoothed feedforward gain and the current screed elevation; it uses a proportional-integral control law to calculate the feedback control output based on the deviation between the actual screed elevation and the target elevation; it then superimposes the feedforward control output and the feedback control output to obtain the screed height adjustment amount, and adjusts the screed height of the paver according to the screed height adjustment amount.

[0050] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A 3D intelligent paving control method, characterized in that: Includes the following steps: S1. Receive the building information model file containing road design information, generate the base layer design elevation surface function and the surface layer design thickness function based on the building information model file, and obtain the spatial position coordinates and attitude angle of the paver to calculate the deviation between the measured elevation and the design elevation of the base layer. After gridding and smoothing filtering, the smoothed base layer deviation field is obtained. S2. Based on the smoothed base course deviation field and the surface course design thickness function, the initial target paving thickness is calculated using a negative feedback compensation strategy. The initial target paving thickness is then constrained to obtain the constrained target paving thickness. The loose paving thickness is calculated based on the constrained target paving thickness and compaction coefficient, and the target elevation of the screed is calculated based on the base course design elevation, the smoothed base course deviation, and the loose paving thickness. S3. Within the scanning area in front of the current position of the paver, for each lateral position, calculate the longitudinal change rate sequence of the smoothed base deviation field, and normalize it based on the standard deviation of the longitudinal change rate sequence to obtain the deviation change intensity index. S4. Based on the deviation change severity index, the anticipation distance and feedforward compensation coefficient of the ironing plate feedforward control are dynamically adjusted using a nonlinear mapping function, and the anticipation distance and feedforward compensation coefficient of the adjusted ironing plate are then subjected to lateral smoothing to obtain the processed anticipation distance and processed feedforward compensation coefficient. S5. Based on the smoothed predicted distance and the target elevation of the screed, the feedforward control output is calculated by combining the smoothed feedforward gain and the current screed elevation. The proportional-integral control law is used to calculate the feedback control output based on the deviation between the actual screed elevation and the target elevation. The feedforward control output and the feedback control output are superimposed to obtain the screed height adjustment amount. The screed height adjustment amount is coordinated and corrected according to the cross slope angle of the screed. The screed height of the paver is adjusted according to the coordinated and corrected screed height adjustment amount.

2. The 3D intelligent paving control method as described in claim 1, characterized in that: Step S1 specifically includes: The spatial position coordinates and attitude angles of the paver are obtained through multiple sensors; The three-dimensional geometry of the base surface in front of the paving operation is scanned by a front laser contour scanner to obtain the three-dimensional point coordinates in the local coordinate system of the sensor. The coordinates are then transformed into the engineering coordinate system and the road alignment coordinate system in turn through a homogeneous coordinate transformation matrix to obtain the mileage value and lateral offset of each point in the point cloud. For any point in the point cloud, the corresponding design elevation is queried from the base design elevation surface function, and the base deviation is obtained by subtracting the design elevation from the measured base elevation. The base deviation of each point in the point cloud is processed into a grid. The paving area is divided into regular rectangular grids in the road linear coordinate system. For each grid cell, all points falling within the grid range are searched, and the median of the elevation deviation of all these points is calculated as the representative value of the grid, thus obtaining the gridded base deviation field. Gaussian smoothing filtering is applied to the gridded base deviation field to obtain the smoothed base deviation field.

3. The 3D intelligent paving control method as described in claim 1, characterized in that: The negative feedback compensation strategy specifically includes: When the actual elevation of the base course is lower than the design elevation, the paving thickness is increased to compensate for the base course deficit; when the actual elevation of the base course is higher than the design elevation, the paving thickness is decreased to correct the base course superelevation. The compensation direction is opposite to the base course deviation direction. The preliminary target paving thickness is calculated by subtracting the compensation coefficient from the product of the designed paving thickness and the smoothed base course deviation. The calculation formula is as follows: in, The initial target paving thickness, To design the paving thickness, For compensation coefficient, To smooth the base layer deviation field, and These are the grid coordinates for longitudinal mileage and lateral offset, respectively.

4. The 3D intelligent paving control method as described in claim 3, characterized in that: The constraints include absolute thickness constraints and thickness gradient constraints, wherein, The absolute thickness constraint is set according to the material compaction process requirements, specifying the minimum and maximum thicknesses. The thickness gradient constraint indirectly controls the attitude angle of the ironing plate by limiting the thickness variation gradient between adjacent mesh cells, wherein the thickness gradient constraint includes longitudinal thickness gradient constraint and lateral thickness gradient constraint; When the initial target paving thickness is less than the minimum thickness or greater than the maximum thickness, or when the thickness variation gradient between adjacent grid cells exceeds the allowable thickness variation rate, the gradient projection method is used to correct the initial target paving thickness so that the corrected constrained target paving thickness satisfies the absolute thickness constraint and the thickness gradient constraint.

5. The 3D intelligent paving control method as described in claim 1, characterized in that: Step S3 specifically includes: For each lateral position within the scanning area in front of the current position of the paver, extract the smoothed base deviation value distributed along the longitudinal direction at that lateral position to form a longitudinal deviation sequence; For any forward grid point in the longitudinal deviation sequence, the rate of change of the longitudinal deviation at that grid point is calculated using the central difference method. Extract the longitudinal deviation rate of change sequence within the forward scanning range and calculate the standard deviation of the rate of change sequence; The standard deviation of the rate of change series is normalized to obtain the index of the severity of deviation change.

6. The 3D intelligent paving control method as described in claim 1, characterized in that: The method of dynamically adjusting the prediction distance and feedforward compensation coefficient of the ironing plate feedforward control using a nonlinear mapping function specifically includes: The baseline foreseeability distance is calculated by multiplying the paving speed by the response time of the screed height adjustment unit. Based on the deviation change severity index, the hyperbolic tangent function is used to adaptively adjust the baseline prediction distance. Upper and lower limit constraints are applied to the adjusted prediction distance to obtain the adjusted prediction distance: When the deviation change severity index is higher than the severity threshold, the hyperbolic tangent function outputs a positive value, which extends the prediction distance to read the elevation information of the target ahead in advance. When the deviation change severity index is lower than the severity threshold, the hyperbolic tangent function outputs a negative value, which shortens the prediction distance; Set a baseline feedforward compensation coefficient. When the intensity index is below the intensity threshold, maintain the baseline feedforward compensation coefficient. When the intensity index is above the intensity threshold, decrease the baseline feedforward compensation coefficient according to the degree to which the intensity index exceeds the threshold.

7. The 3D intelligent paving control method as described in claim 6, characterized in that: The formula for calculating the adjusted forecast distance is as follows: ； in, This is the adjusted forecast distance. Indicates the baseline foresight distance. To adjust the amplitude coefficient, Represents the hyperbolic tangent function. The intensity threshold, This is the soft threshold transition width. These are the grid coordinates for the horizontal position. Indicates horizontal position The index of the degree of deviation change at a certain point.

8. The 3D intelligent paving control method as described in claim 1, characterized in that: Step S5 specifically includes: The predicted position is determined based on the smoothed predicted distance. The predicted position is equal to the current position of the paver plus the smoothed predicted distance. The target elevation of the screed at the predicted position is calculated using a linear interpolation method. The difference between the target elevation of the ironing plate at the predicted location and the current elevation of the ironing plate is multiplied by the smoothed feedforward compensation coefficient to calculate the feedforward control output. The difference between the actual elevation of the ironing board and the target elevation of the ironing board at the current position is defined as the elevation tracking error, and the feedback control output is calculated using a proportional-integral control law. The height adjustment of the ironing board is obtained by superimposing the output of the feedforward control and the output of the feedback control. Calculate the actual and target cross slope angles of the screed, and adjust the screed height adjustment amount accordingly. Adjust the screed height of the paver based on the adjusted screed height adjustment amount.

9. The 3D intelligent paving control method as described in claim 8, characterized in that: The formula for calculating the feedback control output is: ； in, For feedback control output quantity, For elevation tracking error, This is the proportionality coefficient. The integral coefficient is... To control the cycle, Index for historical moments.

10. A 3D intelligent paving control system, characterized in that: The application of the 3D intelligent paving control method as described in any one of claims 1-9 includes: The deviation calculation module is used to receive the building information model file containing road design information, generate the base layer design elevation surface function and the surface layer design thickness function based on the building information model file, and obtain the spatial position coordinates and attitude angle of the paver to calculate the deviation between the measured elevation of the base layer and the design elevation. After gridding and smoothing filtering, the smoothed base layer deviation field is obtained. The thickness compensation module is used to calculate the initial target paving thickness based on the smoothed base course deviation field and the surface course design thickness function, using a negative feedback compensation strategy. It then constrains the initial target paving thickness to obtain the constrained target paving thickness. Based on the constrained target paving thickness and the compaction coefficient, it calculates the loose paving thickness. Finally, it calculates the target elevation of the screed based on the base course design elevation, the smoothed base course deviation, and the loose paving thickness. The deviation analysis module is used to calculate the longitudinal rate of change sequence of the smoothed base deviation field for each lateral position within the scanning area in front of the current position of the paver, and to normalize it based on the standard deviation of the longitudinal rate of change sequence to obtain the deviation change intensity index. The parameter adjustment module is used to dynamically adjust the prediction distance and feedforward compensation coefficient of the ironing plate feedforward control according to the deviation change severity index using a nonlinear mapping function, and to perform lateral smoothing on the adjusted prediction distance and feedforward compensation coefficient of the ironing plate to obtain the processed prediction distance and processed feedforward compensation coefficient. The control output module is used to calculate the feedforward control output based on the smoothed predicted distance and the target elevation of the screed, combined with the smoothed feedforward gain and the current screed elevation; it uses a proportional-integral control law to calculate the feedback control output based on the deviation between the actual screed elevation and the target elevation; it then superimposes the feedforward control output and the feedback control output to obtain the screed height adjustment amount, and adjusts the screed height of the paver according to the screed height adjustment amount.

Images