Auxiliary control method of vibratory roller

Abstract

The invention relates to the technical field of engineering machinery and automation, in particular to an auxiliary control method of a vibratory roller, which comprises the following steps: determining the spatial position and first attitude information of the roller through Beidou positioning; analyzing and calculating the driving position and second attitude information of the road roller during unmanned driving; comparing and analyzing the spatial position and the first attitude information with the corresponding driving position and the second attitude information to obtain a position error and an attitude error; the position error and the attitude error are compared and analyzed with corresponding error thresholds, and if the position error and the attitude error do not exceed the threshold range, the spatial position, the first attitude information, the driving position and the second attitude information are stored in a storage module to serve as a data set of the automatic driving system; otherwise, error positioning analysis is carried out. The position and the posture of the road roller can be accurately determined, so that the control precision of the unmanned vibratory road roller can be subsequently improved.

Description

Technical Field

[0001] The present invention relates to the technical fields of construction machinery and automation technology, and particularly relates to an auxiliary control method for a vibratory roller. Background Art

[0002] With the rapid development of road construction automation technology, the driverless system of rollers has shown significant advantages in improving construction efficiency and operation safety. However, in the prior art, there are still significant technical bottlenecks in the real-time spatial positioning and attitude detection of rollers in complex construction environments. Especially when the roller operates under dynamic working conditions such as continuous amplitude-changing rolling, ramp steering, and vibratory compaction, the traditional positioning system is prone to positioning data drift due to factors such as multipath effects, mechanical vibration interference, and satellite signal occlusion. And the independent attitude solution method based on the inertial measurement unit has the inherent problem that the cumulative error is difficult to eliminate. At the same time, the prior art lacks a real-time collaborative verification mechanism for multi-source positioning information and the expected trajectory of autonomous driving, resulting in the system being unable to effectively identify abnormal positioning states, and it is even more difficult to construct a high-confidence data set for the iterative optimization of the autonomous driving model. Such technical defects not only affect the accuracy of the roller path tracking, but may also cause construction quality defects and even equipment operation safety problems due to uncorrected positioning deviations. Summary of the Invention

[0003] The purpose of the present invention is to provide an auxiliary control method for a vibratory roller, which can accurately determine the position and attitude of the roller, so as to improve the control accuracy of the driverless vibratory roller subsequently.

[0004] To achieve the above purpose, an auxiliary control method for a vibratory roller is provided, including the following steps:

[0005] Determine the spatial position and the first attitude information of the roller through Beidou positioning;

[0006] Analyze and calculate the driving position and the second attitude information of the roller when driving without a driver;

[0007] Compare and analyze the spatial position and the first attitude information with the corresponding driving position and the second attitude information respectively to obtain a position error and an attitude error; and compare and analyze the position error and the attitude error with the corresponding error thresholds respectively. If neither exceeds the threshold range, store the spatial position, the first attitude information, the driving position, and the second attitude information in the storage module as a data set for the autonomous driving system; otherwise, perform error positioning analysis.

[0008] Furthermore, the road roller is a double-drum vibratory road roller with a double-hinged structure, which consists of three parts: the front vehicle body composed of the front frame and the front vibrating drum, the rear vehicle body composed of the rear frame and the rear vibrating drum, and the middle hinge frame; the calculation formulas for the driving position and the second attitude information are as follows:

[0009]

[0010] Among them, the centers of the two vehicle bodies can be respectively expressed as P f (x f , y f ) and P r (x r , yr), l1 and l3 respectively represent the lengths from the center of gravity positions of the rear vehicle body to the hinge point 1 and from the center of gravity position of the front vehicle body to the hinge point 2, l2 represents the length of the hinge frame, γ1 and γ2 respectively represent the rotation angles of the two hinge points, θ f and θ r are the angles between the velocity directions of the front and rear vehicle bodies and the X-axis, representing the headings of the front and rear vehicle bodies. According to the generalized coordinates q T = [x f , y f , θ, γ1, γ2], the driving position and the second attitude information of the double-drum vibratory road roller are determined.

[0011] Furthermore, the following steps are also included:

[0012] After the information of the compacted material is collected, the compactness of the current rolling path is judged according to the information of the compacted material. When the compactness reaches the standard, the road roller shuts off the vibration and starts the lane-changing mode; the lane-changing mode is: making the road roller plan the traveling path in an S-shaped curve. After changing lanes, the body of the road roller is straightened, the articulated steering angle is zero, and the entire lane-changing path needs to be continuous and smooth.

[0013] Furthermore, the S-shaped curve includes a double-arc tangent curve, a polynomial curve, and a Bezier curve.

[0014] Furthermore, the following steps are also included:

[0015] Construct a road roller lane-changing model according to the polynomial curve. The state constraints of the two end points of the starting point and the ending point of the road roller in the road roller lane-changing model are as follows:

[0016]

[0017] Among them, the coordinates at the starting point and the ending point are respectively (0, 0) and (x e , y e ), the 0 moment represents the start of lane-changing, and the T moment represents the end of lane-changing.

[0018] Furthermore, during the data acquisition process of the storage module, based on the real-time vibration frequency and amplitude parameters of the roller vibration wheel, the compaction degree feedback signal of the construction road surface and the deflection angle data of the steering hydraulic cylinder are synchronously collected. The frequency-domain wavelet transform is performed on the compaction degree feedback signal to separate mechanical vibration noise, and a pose change gradient matrix associated with spatial positions is constructed by combining the deflection angle data. After spatio-temporal alignment of the gradient matrix with the processed position error and attitude error, a multi-dimensional training data set with working condition labels is generated.

[0019] Furthermore, during the error location analysis, a rigid body kinematic model of the roller is established through the heading angle data output by the biaxial inclination sensor and GNSS receiver installed at the steering pivot of the roller. The angular velocity and acceleration information output by the inertial measurement unit and the Beidou positioning data are multi-source fused to real-time calculate the three-dimensional coordinates of the roller's center of mass and the pitch-roll-yaw three-degree-of-freedom attitude angles, and the boundary conditions of the error threshold are dynamically adjusted according to the terrain undulation characteristics of the current construction section.

[0020] Furthermore, the position error and attitude error are generated into a synthetic error vector through vector space mapping. Based on the vector synthesis result of the position error and attitude error and the parameter database, the collaborative adjustment parameters of the roller steering system and vibration system are generated. The real-time trajectory is dynamically corrected by controlling the front wheel steering angle rate and the vibration wheel excitation force direction to obtain a corrected trajectory that continuously matches the curvature derivative of the planned path, and the corrected trajectory is iteratively verified with the geometric constraint conditions of the planned path to form a pose closed-loop control link.

[0021] Principle and advantages:

[0022] 1. By integrating the multi-source pose data comparison mechanism of Beidou positioning and the automatic driving system, the signal distortion problem of traditional single positioning methods under dynamic working conditions is effectively overcome, the confidence level of the roller's spatial coordinates and attitude angle detection is significantly improved. At the same time, by setting the error threshold to trigger the self-check process, not only the spatio-temporal consistency of the stored data set is ensured, but also the real-time diagnosis ability of the positioning abnormal state is constructed, providing a high-reliability environmental perception basis for automatic driving decision-making. It is beneficial to the subsequent auxiliary control of the vibratory roller.

[0023] 2. Through the coupling analysis of vibration parameters and compaction degree signals and noise suppression processing, the in-depth fusion characterization of the mechanical properties and pose states during the construction process is realized. Using the spatio-temporal alignment gradient matrix construction method, the physical relevance of the multi-dimensional data set under complex compaction conditions is strengthened, providing a global working condition sample covering mechanical dynamic responses and road surface interaction characteristics for automatic driving model training, and significantly improving the adaptability of the path planning algorithm to the non-linear compaction process.

[0024] 3. Based on the rigid body kinematic model and multi-sensor fusion algorithm, break through the error accumulation limitation of the inertial navigation system in a continuous vibration environment. By dynamically adjusting the boundary conditions of the error threshold, enable the positioning verification system to adapt to different terrain curvatures and changes in operating loads. At the same time, ensure that the three-dimensional pose calculation results are strictly matched with the geometric constraint conditions of the actual construction scenario, and improve the traceability accuracy and fault tolerance control efficiency of anomaly positioning from the mechanism level.

[0025] 4. Through the collaborative adjustment parameter generation mechanism of the steering system and the vibration system, effectively integrate the multi-dimensional control requirements of the pose error vector, and achieve the dynamic balance adjustment of the front wheel steering angle rate and the vibration force direction. Make the real-time trajectory correction process take into account both steering sensitivity and vibration compaction stability. At the same time, with the help of the continuous matching mechanism of the planned path curvature derivative, ensure the spatial consistency between the corrected trajectory and the preset geometric constraint conditions. And through the adaptive feedback link constructed by the iterative verification algorithm, significantly enhance the trajectory tracking robustness of the unmanned system under variable curvature paths. Finally, form a pose control system with the ability to eliminate real-time errors and ensure the closed-loop guarantee of construction quality, comprehensively improving the trajectory accuracy and pavement forming uniformity of rolling operations under complex working conditions, and providing good auxiliary control for the vibratory roller. Brief Description of the Drawings

[0026] Figure 1 It is a flowchart of an auxiliary control method for a vibratory roller according to an embodiment of the present invention;

[0027] Figure 2 It is a structural schematic diagram of an unmanned roller;

[0028] Figure 3 It is a schematic diagram of a lane change model. Detailed Embodiment

[0029] The following is a more detailed description through specific embodiments:

[0030] Embodiment

[0031] An auxiliary control method for a vibratory roller is basically as Figure 1 、 Figure 2 、 Figure 3 shown, and includes the following steps:

[0032] S1. Determine the spatial position and the first attitude information of the roller through Beidou positioning; integrate a Beidou triple-frequency receiver and an anti-vibration IMU module on the top of the roller cockpit. The IMU module is built with a three-axis MEMS gyroscope and an accelerometer, and transmits the first attitude information (pitch angle, roll angle, heading angle, etc.) output in real time to the vehicle-mounted controller through the CAN bus. At the same time, the controller obtains the spatial position (X, Y) of the roller centroid in the construction coordinate system through the RTK solution algorithm.

[0033] S2. Analyze and calculate the driving position and the second attitude information of the roller when it is unmanned; the roller is a double-drum vibratory roller, which adopts a double-hinged structure and consists of three parts: the front vehicle body composed of the front frame and the front vibratory drum, the rear vehicle body composed of the rear frame and the rear vibratory drum, and the middle hinge frame; the calculation formulas for the driving position and the second attitude information are as follows:

[0034]

[0035] Among them, the centers of the two vehicle bodies can be respectively represented as P f (x f , y f ) and P r (x r , y r ), l1 and l3 respectively represent the lengths from the center of gravity position of the rear vehicle body to hinge point 1 and from the center of gravity position of the front vehicle body to hinge point 2, l2 represents the length of the hinge frame, γ1 and γ2 respectively represent the rotation angles of the two hinge points, θ f and θ r are the angles between the velocity directions of the front and rear vehicle bodies and the X-axis, representing the headings of the front and rear vehicle bodies. According to the generalized coordinates q T = [x f , y f , θ, γ1, γ2], the driving position and the second attitude information of the double-drum vibratory roller are determined.

[0036] S3. Compare and analyze the spatial position and the first attitude information with the corresponding driving position and the second attitude information respectively to obtain the position error and the attitude error; and compare and analyze the position error and the attitude error with the corresponding error thresholds respectively. If both do not exceed the threshold range, store the spatial position, the first attitude information, the driving position and the second attitude information in the storage module as the data set of the automatic driving system; otherwise, perform error location analysis. In this embodiment, the Euclidean distance algorithm is used to calculate the planar projection deviation between the spatial position and the driving position as the position error, and the Euler angle deviation between the first attitude information and the second attitude information is solved by the quaternion difference method as the attitude error; when the position error ≤ 0.15 m and the attitude error ≤ 2°, the storage module binds and stores the pose data with the corresponding timestamp and working condition code as a structured data packet, otherwise triggers error location analysis.

[0037] S4. After collecting the information of the compacted material, the controller judges the compactness of the current rolling pass according to the compacted material information. When the compactness meets the standard, the vibrator of the roller is turned off and the lane-changing mode is activated. The lane-changing mode is as follows: the roller plans its traveling path in an S-shaped curve. After changing lanes, the body of the roller is straightened, the articulated steering angle is zero, and the entire lane-changing path needs to be continuous and smooth. The S-shaped curve includes a double-arc tangent curve, a polynomial curve, and a Bezier curve. Among them, the double-arc tangent curve is not very suitable for the lane-changing requirements of a tandem vibrating roller. The polynomial curve satisfies the requirements that the curvature at the starting point and the ending point is zero and the curvature of the entire curve changes continuously, and is more concise and has less computational complexity compared to the Bezier curve. Assuming that the width of the vibrating wheel of the tandem vibrating roller is B and the overlapping distance between adjacent rolling bands is l, the lateral lane-changing distance is D1 = B - l, and the longitudinal lane-changing distance is D2.

[0038] S5. Construct a lane-changing model for the roller according to the polynomial curve. Establish a coordinate system with the starting point of the curve as the origin of coordinates. Then the lane-changing model of the tandem vibrating roller is as shown in Figure 3 . The state constraints of the two endpoints of the starting point and the ending point of the roller in the lane-changing model of the roller are as follows:

[0039]

[0040] Among them, the coordinates at the starting point and the ending point are (0, 0) and (x e , y e ), respectively. The moment of 0 represents the start of lane-changing, and the moment of T represents the end of lane-changing.

[0041] S6. During the data acquisition process of the storage module, based on the real-time vibration frequency and amplitude parameters of the vibrating wheel of the roller, synchronously collect the compactness feedback signal of the construction road surface and the deflection angle data of the steering hydraulic cylinder. Perform frequency-domain wavelet transform on the compactness feedback signal to separate mechanical vibration noise, and combine the deflection angle data to construct a pose change gradient matrix associated with the spatial position. After spatio-temporal alignment of the gradient matrix with the position error and attitude error processed by the Kalman filter, generate a multi-dimensional training data set with working condition labels.

[0042] S7. In the error positioning and analysis stage, establish a kinematic model of the rigid body of the roller through the heading angle data output by the dual-axis inclinometer and GNSS receiver installed at the steering pivot of the roller. Use the extended Kalman filter algorithm to perform multi-source fusion on the angular velocity, acceleration information output by the inertial measurement unit and the Beidou positioning data, and real-time calculate the three-dimensional coordinates of the centroid of the roller and the pitch-roll-yaw three-degree-of-freedom attitude angles, and dynamically adjust the boundary conditions of the error threshold according to the terrain undulation characteristics of the current construction section.

[0043] S8. Generate a composite error vector by mapping the position error and the attitude error through a vector space. Based on the vector synthesis results of the position error and the attitude error and the parameter database, generate the collaborative adjustment parameters for the steering system and the vibration system of the roller compactor (generate the collaborative adjustment parameters by combining the two types of errors in a preset weight manner). Dynamically correct the real-time trajectory by controlling the steering angle rate of the front wheels and the direction of the excitation force of the vibrating wheels, obtain a corrected trajectory that continuously matches the curvature derivative of the planned path, and perform iterative verification on the geometric constraint conditions between the corrected trajectory and the planned path to form a pose closed-loop control link. Steering control: Send a speed adjustment command to the steering hydraulic motor according to the magnitude of the lateral position error, so that the front wheels adjust the steering angle at a rate of 2 - 5 degrees per minute; Vibration adjustment: Control the rotation phase of the eccentric block inside the vibrating wheel according to the slope direction of the attitude error, so that the direction of the excitation force is always perpendicular to the current rolling surface; Trajectory verification: Collect the corrected actual trajectory coordinates once every 10 seconds, and perform a shape matching check with the preset path. When the path curvature difference is detected to exceed the safety range continuously for 3 times, automatically send a path replanning request to the navigation system; Closed-loop update: Transmit the corrected trajectory data that passes the verification back to the storage module as the optimization basis for subsequent autonomous driving decisions.

[0044] The above are only embodiments of the present invention. Specific structures and common knowledge such as characteristics well known in the art are not described in detail here. Those of ordinary skill in the art know all the common technical knowledge in the technical field to which the invention belongs before the application date or the priority date, can know all the prior arts in this field, and have the ability to apply the conventional experimental means before this date. Those of ordinary skill in the art can, under the inspiration given in this application and combined with their own abilities, improve and implement this solution. Some typical well-known structures or well-known methods should not become an obstacle for those of ordinary skill in the art to implement this application. It should be noted that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can still be made, and these should also be regarded as the protection scope of the present invention, and these will not affect the implementation effect of the present invention and the practicality of the patent. The protection scope required by this application should be subject to the content of its claims, and the specific implementation manners and the like described in the specification can be used to explain the content of the claims.

Claims

1. An auxiliary control method for a vibratory roller, characterized in that, Including the following steps: Determine the spatial position and the first attitude information of the roller through Beidou positioning; Analyze and calculate the driving position and the second attitude information of the roller during unmanned driving; Compare and analyze the spatial position and the first attitude information with the corresponding driving position and the second attitude information respectively to obtain the position error and the attitude error; and compare and analyze the position error and the attitude error with the corresponding error thresholds respectively. If neither exceeds the threshold range, store the spatial position, the first attitude information, the driving position and the second attitude information in the storage module as the data set of the automatic driving system; otherwise, conduct error location analysis.

2. The auxiliary control method of a vibratory roller according to claim 1, characterized in that: The roller is a double-drum vibratory roller, adopting a double-hinged structure, and consists of three parts: the front vehicle body composed of the front frame and the front vibrating drum, the rear vehicle body composed of the rear frame and the rear vibrating drum, and the middle hinge frame; the calculation formulas for the driving position and the second attitude information are as follows: Among them, the centers of the two car bodies can be respectively expressed as P f (x f , y f ) and P r (x r , y r ). l1 and l3 respectively represent the lengths from the center-of-gravity positions of the rear car body to hinge point 1 and from the center-of-gravity position of the front car body to hinge point 2. l2 represents the length of the hinge frame. γ1 and γ2 respectively represent the rotation angles of the two hinge points. θ f and θ r are the angles between the velocity directions of the front and rear car bodies and the X-axis, representing the headings of the front and rear car bodies. According to the generalized coordinates q T = [x f , y f , θ, γ1, γ2], the traveling position and the second attitude information of the tandem vibrating roller are determined.

3. The auxiliary control method of a vibratory roller according to claim 2, characterized in that: Also including the following steps: After collecting the information of the compacted material, judge the compactness of the current rolling lane according to the compacted material information. When the compactness meets the standard, the roller shuts off the vibration and starts the lane-changing mode; the lane-changing mode is: make the roller plan the traveling path in an S-shaped curve. After changing lanes, the body of the roller is straightened, the articulated steering angle is zero, and the entire lane-changing path needs to be continuous and smooth.

4. The auxiliary control method of a vibratory roller according to claim 3, characterized in that: The S-shaped curve includes a double-arc tangent curve, a polynomial curve and a Bezier curve.

5. The auxiliary control method of a vibratory roller according to claim 4, characterized in that: Also including the following steps: Construct a lane-changing model for the roller according to the polynomial curve. The state constraints of the two endpoints of the starting point and the ending point of the roller in the lane-changing model of the roller are as follows: Among them, the coordinates at the starting point and the ending point are (0, 0) and (x e , y e ), respectively. The moment 0 represents the start of lane change, and the moment T represents the end of lane change.

6. The auxiliary control method of a vibratory roller according to claim 1, characterized in that: Also including the following steps: During the data acquisition process of the storage module, based on the real-time vibration frequency and amplitude parameters of the vibrating drum of the roller, synchronously collect the compactness feedback signal of the construction road surface and the deflection angle data of the steering hydraulic cylinder. Conduct frequency-domain wavelet transform on the compactness feedback signal to separate mechanical vibration noise, and construct a pose change gradient matrix associated with the spatial position in combination with the deflection angle data. After spatio-temporal alignment of the gradient matrix with the processed position error and attitude error, generate a multi-dimensional training data set with working condition labels.

7. The auxiliary control method of a vibratory roller according to claim 6, characterized in that: Also including the following steps: During error location analysis, establish a kinematic model of the rigid body of the roller through the heading angle data output by the dual-axis inclinometer and the GNSS receiver installed at the steering pivot of the roller, conduct multi-source fusion on the angular velocity and acceleration information output by the inertial measurement unit and the Beidou positioning data, and real-time calculate the three-dimensional coordinates of the centroid of the roller and the three-degree-of-freedom attitude angles of pitch-roll-yaw, and dynamically adjust the boundary conditions of the error threshold according to the terrain undulation characteristics of the current construction section.

8. An auxiliary control method for a vibratory roller according to claim 7, characterized in that: Also including the following steps: Generate a composite error vector by mapping the position error and the attitude error through a vector space. Based on the vector synthesis result of the position error and the attitude error and the parameter database, generate the collaborative adjustment parameters for the steering system and the vibration system of the roller. Dynamically correct the real-time trajectory by controlling the front-wheel steering angle rate and the vibration-wheel excitation force direction, obtain a corrected trajectory that continuously matches the curvature derivative of the planned path, and perform iterative verification on the geometric constraint conditions between the corrected trajectory and the planned path to form a pose closed-loop control link.

Images