Intelligent road roller trajectory planning method, electronic equipment and storage medium

Abstract

The invention discloses an intelligent road roller trajectory planning method, electronic equipment and a storage medium. The method comprises the following steps: acquiring a road boundary detection result and the category, speed and distance information of an obstacle by adopting a target detection and semantic segmentation network; calculating the relative pose of the road roller and the road boundary based on the road boundary detection result; using a multi-sensor data fusion method to fuse the obstacle information, the road boundary information and the road roller global positioning information obtained from the integrated navigation equipment to obtain the global positioning information of the road roller and the obstacle; establishing a global path planning model; constructing a multi-objective function considering the compaction quality and a multi-constraint condition considering a construction time window, and optimally solving the working speed and the compaction times of the road roller based on a dynamic programming method; coupling of the speed and the path of the road roller is achieved through a spline curve interpolation method, and an intelligent compaction global track is generated. The optimal compaction quality can be achieved under complex construction, and meanwhile high efficiency and economical efficiency of operation are guaranteed.

Description

Technical Field

[0001] The present invention relates to automatic driving technology for a road roller, and in particular to an intelligent road roller trajectory planning method, electronic equipment and storage medium. Background Art

[0002] In the context of increasingly high quality requirements for modern infrastructure construction and increasingly complex and harsh working environments, the intelligent roller path planning method is one of the important technologies to improve road construction quality and operating efficiency. Through precise trajectory planning and control, the roller can operate efficiently, ensuring the uniformity and stability of construction quality, thereby improving the overall quality of road construction and reducing later maintenance costs.

[0003] Traditional manually driven rollers face many challenges in actual operation. First, drivers are easily affected by factors such as fatigue and weather in a long working environment, resulting in unstable operation. Secondly, the compaction process of manual driving is often uneven, prone to over-pressure, under-pressure, or failure to compact the edge of the road. These problems not only reduce the construction quality, but may also cause early damage to the road during use and increase maintenance costs. Finally, the harsh working environment poses a great health hazard to the driver. In order to solve these problems, unmanned rollers have gradually become an important development direction. Among them, trajectory planning is the core of unmanned roller technology and the key to ensuring compaction quality and efficiency.

[0004] Invention patent CN116449848A discloses a path planning system for an unmanned roller. After determining the construction area through two-point or four-point acquisition, the method plans the roller's operating path according to the set compaction width, number of rolling passes and other parameters. However, this method only plans the path based on the compaction area, and does not consider obstacle information. It can only operate in areas without obstacles, and its application scenarios are not wide enough. Invention patent CN113406954 discloses a path planning method for an unmanned roller. After obtaining the boundary of the working area and the location and size information of obstacles, the method determines the length of the continuous obstacle-free area on each rolling road, and determines the path for continued rolling based on the comparison result of the length of the continuous obstacle-free area and the preset length. However, this method does not take into account the real-time temperature of the road surface, and cannot dynamically optimize the roller path according to the quality of the construction process. Invention patent CN1 1 5951680A discloses a control method, device, equipment and system for an unmanned intelligent roller. The method determines the compaction completion degree within a preset time period by obtaining multiple sets of compaction data of a preset roller within a preset time period, and then controls the compaction trajectory of the next time period according to the compaction completion degree of the preset time period. However, the method does not take vehicle speed into consideration, and it is difficult to take fuel economy into account while ensuring compaction quality.

[0005] In summary, the existing roller path planning methods usually only consider the coverage of the compaction area, but fail to consider dynamic factors such as the real-time road surface temperature and vehicle speed, making it difficult to balance the compaction quality and efficiency.

[0006] Therefore, there is an urgent need for a roller path planning method that comprehensively considers pavement compaction quality, operating efficiency, and fuel economy to further improve the overall quality and economy of road construction. Summary of the invention

[0007] Purpose of the invention: In order to overcome the defects of the prior art such as inaccurate compaction range, low compaction quality and efficiency, the present invention provides an intelligent roller trajectory planning method, equipment and storage medium, which can achieve the best compaction quality in a complex construction environment while ensuring the high efficiency and economy of the operation.

[0008] Technical solution: The first aspect of the present invention provides a trajectory planning method for an intelligent road roller, the method comprising the following steps:

[0009] Step 1: Based on the camera and millimeter wave radar multi-sensor data, the target detection and semantic segmentation network is used to obtain the road boundary detection results and the category, speed and distance information of obstacles;

[0010] Step 2: Calculate the relative position of the roller and the road boundary based on the road boundary detection result;

[0011] Step 3: Use multi-sensor data fusion methods such as Kalman filter / factor graph to fuse obstacle information, road boundary information and the global positioning information of the roller obtained from the integrated navigation device to obtain the global positioning information of the roller and the obstacle;

[0012] Step 4: Establish a global path planning model to plan the roller operation path based on the road boundary information and vehicle kinematic model;

[0013] Step 5: Construct a multi-objective function that considers compaction quality and multiple constraints that consider the construction time window, and optimize the roller operating speed and compaction passes based on the dynamic programming method;

[0014] Step 6: Use the spline curve interpolation method to realize the coupling of roller speed and path and generate the global trajectory of intelligent compaction.

[0015] Preferably, step 1 includes the following sub-steps:

[0016] Step 1-1: Perform distortion correction on the original image input by the fisheye camera to obtain a dedistorted image;

[0017] Step 1-2: Use the target detection and semantic segmentation network to obtain the category, location and road pixel coordinates of the obstacles in the distortion-corrected image;

[0018] Step 1-3: Use millimeter wave radar to obtain the distance, speed and azimuth information of the obstacle;

[0019] Preferably, step 2 comprises the following sub-steps:

[0020] Step 2-1: Based on the known camera internal and external parameters, the image is converted to the vehicle coordinate system through inverse perspective transformation, and the model expression of the road boundary line in the vehicle coordinate system is obtained by polynomial fitting;

[0021] Step 2-2: Calculate the lateral offset and movement direction between the vehicle and the road boundary through the road boundary line model in the vehicle coordinate system;

[0022] Preferably, step 3 includes the following sub-steps:

[0023] Step 3-1: Project the target point detected by the millimeter wave radar to the vehicle coordinate system;

[0024] Step 3-2: Correlate the outputs of the camera and the millimeter wave radar in time and space;

[0025] Step 3-3: Use methods such as Kalman filtering / factor graph to fuse the detection information of the camera and millimeter-wave radar to obtain accurate information such as obstacle location, speed and category;

[0026] Step 3-4: The lateral distance offset of the road boundary line obtained in step 2-2 and the obstacle information obtained in step 3-3 are integrated with the vehicle global positioning information obtained from the integrated navigation device to obtain the corrected global position information of the roller and the obstacle.

[0027] Preferably, the kinematic model of the road roller in step 4 is specifically:

[0028]

[0029] Among them, x f ,y f is the coordinate of the front steel wheel mass point, x r ,y r is the coordinate of the rear steel wheel mass point, θ f ,θ r are the heading angles of the front and rear steel wheels, respectively. The model assumes that there is no lateral slip during driving and that the front and rear wheels are rigid bodies. Assume that the distances from the front and rear steel wheel masses to the hinge point are l f l r , the hinge angle is β, then the position and heading angle of the rear steel wheel mass point can be expressed as:

[0030]

[0031] From this, we can get the front wheel speed v of the rollerf Steering angle velocity at the hinge point

[0032]

[0033] Preferably, step 4 includes the following sub-steps:

[0034] Step 4-1: According to the road boundary information, obtain the coordinate vertex position of the rectangular compaction area and the road width d s And the length of the compaction area L. According to the width of the vehicle's roller d w Determine the overlap width d according to the conditions at the construction site r , calculate the number of paths n for the roller's straight-line operation:

[0035]

[0036] Step 4-2: Calculate the coordinate information (x ij ,y ij ), the longitudinal spacing of each path point on each path is l meter, and the value of l is determined according to the diameter of the compacting wheel of the roller and the length of contact with the ground, the requirements of different materials and construction specifications for compaction uniformity, and the vibration frequency of the roller. In specific implementation, the longitudinal spacing is gradually optimized according to the specific compaction effect from half of the compacting wheel diameter to about the compacting wheel diameter:

[0037]

[0038] Step 4-3: Based on the path point information obtained in step 4-2, the starting point and end point of the offset lane change are selected, and based on the fifth-order polynomial and the vehicle kinematic model, the constraint conditions are constructed to plan the offset path of the roller. The fifth-order polynomial fitting formula is as follows:

[0039] y=a0+a1x+a2x 2 +a3x 3 +a4x 4 +a5x 5

[0040] Among them, a0, a1, a2, a3, a4, and a5 are polynomial coefficients.

[0041] The intelligent roller trajectory planning method according to claim 1 is characterized in that step 5 comprises the following sub-steps:

[0042] Step 5-1: Use vehicle speed, rolling passes and sensor information to quantitatively evaluate the compaction quality, operating efficiency and fuel economy of the roller construction, and use this to construct a multi-objective optimization function:

[0043] J=λ1E fuel +λ2E time +λ3E quality

[0044] Among them, E fuel is the fuel economy index, E time is the operating efficiency index, E quality is the compaction quality index, λ1, λ2, λ3 are the weight coefficients of each objective respectively.

[0045] Step 5-2: Construct the construction time window constraint based on the road surface temperature field decay model, construct the dynamic constraint based on the vehicle dynamics characteristics, and determine the obstacle avoidance constraint based on the perception positioning information. Combine the various constraints to determine the overall constraint:

[0046] Step 5-3: Based on dynamic programming, considering the constraints proposed in step 5-2, solve the objective function obtained in step 5-1 to obtain global speed information and compaction passes.

[0047] Preferably, step 6 comprises the following sub-steps:

[0048] Step 6-1: Input the path and speed information obtained in step 4 and step 5, determine the planned speed of each path point after coupling, and form the global trajectory information;

[0049] Step 6-2: According to the current position of the vehicle, determine the optimal planning speed at the current position based on cubic spline interpolation.

[0050] According to a second aspect of the present invention, there is provided an electronic device, comprising a memory and a processor, wherein a computer program is stored in the memory, and the method described above is implemented when the processor executes the program.

[0051] According to a third aspect of the present invention, there is provided a computer-readable storage medium having a computer program stored thereon, wherein the program implements the method described above when executed by a processor.

[0052] Compared with the prior art, the present invention has the following advantages:

[0053] (1) Accurate compaction range: The path planning algorithm proposed in the present invention detects the road boundary in real time based on camera input and dynamically adjusts the compaction path, effectively solving the problems of over-compaction, under-compaction, and failure to compact the road edge that are prone to occur with traditional rollers.

[0054] (2) Taking into account both compaction quality and efficiency: The present invention introduces multiple dynamic factors such as real-time road surface temperature and vehicle speed, constructs a multi-objective function that considers compaction quality and multiple constraints that consider the construction time window, to ensure that the best compaction quality can be achieved under different construction conditions and to ensure good fuel economy. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a schematic diagram of the process of the intelligent roller path planning method in the present invention;

[0056] Figure 2 Schematic diagram of the relative position relationship between the road roller and the road edge in an embodiment of the present invention;

[0057] Figure 3 It is a schematic diagram of the process of feature-level fusion of camera and millimeter-wave radar in an embodiment of the present invention;

[0058] Figure 4 Schematic diagram of the process of a multi-sensor fusion method in an embodiment of the present invention;

[0059] Figure 5 A schematic diagram of the working path of the road roller in an embodiment of the present invention;

[0060] Figure 6 Schematic diagram of the kinematic model structure of a road roller in an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0062] The intelligent roller trajectory planning method of this embodiment has the following process: Figure 1 As shown, including:

[0063] Step 1: Based on the camera and millimeter wave radar multi-sensor data, the target detection and semantic segmentation network is used to obtain the road boundary detection results and the category, speed and distance information of obstacles;

[0064] Step 2: Calculate the relative position of the roller and the road boundary based on the road boundary detection result;

[0065] Step 3: Use multi-sensor data fusion methods such as Kalman filter / factor graph to fuse obstacle information, road boundary information and the global positioning information of the roller obtained from the integrated navigation device to obtain the global positioning information of the roller and the obstacle;

[0066] Step 4: Establish a global path planning model to plan the roller operation path based on the road boundary information and vehicle kinematic model;

[0067] Step 5: Construct a multi-objective function that considers compaction quality and multiple constraints that consider the construction time window, and optimize the roller operating speed and compaction passes based on the dynamic programming method;

[0068] Step 6: Based on the spline curve interpolation method, the roller speed and path are coupled to generate the global trajectory of intelligent compaction.

[0069] Step 1 is as follows:

[0070] Step 1-1: Perform distortion correction on the original image input by the fisheye camera to obtain a dedistorted image;

[0071] The specific distortion correction method in step 1-1 is:

[0072] There are two types of distortion in fisheye cameras: radial distortion and tangential distortion. Radial distortion can be seen as a change in the coordinate point along the length direction, while tangential distortion is a change in the coordinate point along the tangent direction.

[0073] Radial distortion can be corrected using quadratic and higher-order polynomial functions related to the distance from the center:

[0074] x dist =x(1+k1r 2 +k2r 4 +k3r 6 )

[0075] y dist =y(1+k1r 2 +k2r 4 +k3r 6 )

[0076] Among them, x dist ,y dist are the normalized coordinates of the distorted point, x and y are the coordinates of the corrected point (dedistorted), is the distance between the correction point and the optical center point. k1, k2, k3 are radial distortion parameters. For tangential distortion, tangential distortion parameters p1 and p2 can be used for correction:

[0077] x dist =x+2p1xy+p2(r 2 +2x 2 )

[0078] y dist =y+p1(r 2 +2y 2 )+2p2xy

[0079] After obtaining the intrinsic parameters and distortion parameters of the camera through the checkerboard calibration method, the above relationship can be used to calculate the corrected coordinates of the pixel points to achieve image distortion correction.

[0080] Step 1-2: After obtaining the distortion-corrected image, first use the UNet semantic segmentation network to obtain the pixel coordinate range of the road, and use the ResNet backbone network to extract features for obstacle detection. Then perform multi-scale fusion on the feature map. Finally, regress the category of the obstacle and the position and size of the bounding box;

[0081] Step 1-3: Use millimeter wave radar to obtain the distance, speed and azimuth information of the obstacle;

[0082] Step 2 is as follows:

[0083] Step 2-1: Based on the known camera internal and external parameters, the image is converted to the vehicle coordinate system through coordinate transformation, and the geometric expression of the road boundary line in the vehicle coordinate system is obtained by applying polynomial fitting;

[0084] Step 2-2: Calculate the lateral offset and direction between the vehicle and the road boundary through the road boundary line model in the vehicle coordinate system, where the lateral offset and direction between the vehicle and the road boundary are as follows: Figure 2 As shown;

[0085] Step 3 is as follows:

[0086] Step 3-1: Project the target point detected by the millimeter wave radar to the vehicle coordinate system;

[0087] Step 3-2: Associate the camera and millimeter wave radar outputs in time and space.

[0088] Step 3-3: Use the Kalman filter algorithm to fuse the detection information of the camera and the millimeter-wave radar to obtain accurate information such as the obstacle position, speed, and category. The fusion process is as follows: Figure 3 As shown, specifically:

[0089] (1) Before the entire process begins, initialize the filter and set the initial state estimate and covariance matrix;

[0090] (2) Camera processing:

[0091] According to the state estimation at time k-1 and the prediction covariance matrix Predict the current state Kalman Gain And the covariance matrix Then update the state estimate and the covariance matrix

[0092] (3) Processing of millimeter wave radar:

[0093] Similarly, according to the state estimation at time k-1 and the prediction covariance matrix Predict the current state Kalman Gain and the covariance matrix

[0094] (4) Joint tracking and association:

[0095] The predicted camera and radar measurements are compared using the ellipsoid tracking gate rule v T (k)S -1 (k)v(k)≤γ 2 Determine whether the measured value is valid. If the conditions are met, it is judged as a valid echo, where v(k) is the eigenvector residual, S(k) is the covariance matrix, and γ satisfies the χ of the observation dimension. 2 distributed;

[0096] For valid echoes, calculate their association probability Where Z k represents the cumulative confirmed measurement set up to time k; Represents the cumulative confirmed event set;

[0097] (5) Based on the association probability of the camera and radar data, the updated Kalman gain is integrated to obtain a more accurate target state estimate. The fused state estimate will be used as the output of the system for subsequent control or decision-making.

[0098] Step 3-4: According to the factor graph fusion algorithm, the lateral distance offset of the road boundary line obtained in step 2-2, the obstacle information obtained in step 3-3 and the vehicle global positioning information obtained from the integrated navigation device are fused to obtain the corrected global position information of the roller and the obstacle. The overall fusion process is as follows: Figure 4 shown.

[0099] Step 4 is as follows:

[0100] Step 4-1: According to the road boundary information, obtain the coordinate vertex position of the rectangular compaction area and the road width d s And the length of the compaction area L. According to the width of the vehicle's roller d w Determine the overlap width d according to the conditions at the construction site r , calculate the number of paths n for the roller's straight-line operation, such as Figure 5 As shown:

[0101]

[0102] Step 4-2: Calculate the coordinate information (x ij ,y ij ), the longitudinal spacing between path points on each path is 1 meter:

[0103]

[0104] Step 4-3: Based on the path point information obtained in step 4-2, the starting point and end point of the offset lane change are selected, and based on the fifth-order polynomial and the vehicle kinematic model, the constraint conditions are constructed to plan the offset path of the roller. The fifth-order polynomial fitting formula is as follows:

[0105] y=a0+a1x+a2x 2 +a3x 3 +a4x 4 +a5x 5

[0106] Among them, a0, a1, a2, a3, a4, and a5 are polynomial coefficients. The structure of the roller kinematic model is as follows: Figure 6 As shown, specifically:

[0107]

[0108] Among them, x f ,y f is the coordinate of the front steel wheel mass point, x r ,y r is the coordinate of the rear steel wheel mass point, θ f ,θ r are the heading angles of the front and rear steel wheels, respectively. The model assumes that there is no lateral slip during driving and that the front and rear wheels are rigid bodies. Assume that the distances from the front and rear steel wheel masses to the hinge point are l f , l r , the hinge angle is β, then the position and heading angle of the rear steel wheel mass point can be expressed as:

[0109]

[0110] From this, we can get the front wheel speed v of the roller f Steering angle velocity at the hinge point

[0111]

[0112] Step 5 is as follows:

[0113] Step 5-1: Use vehicle speed, rolling passes and sensor information to quantitatively evaluate the compaction quality, operating efficiency and fuel economy of the roller construction, and use this to construct a multi-objective optimization function:

[0114] J=λ1E fuel +λ2E time +λ3E quality

[0115] Among them, E fuel is the fuel economy index, E time is the operating efficiency index, E quality is the compaction quality index, λ1, λ2, λ3 are the weight coefficients of each objective respectively.

[0116] Step 5-2: Construct the construction time window constraint based on the road surface temperature field decay model, construct the dynamic constraint based on the vehicle dynamics characteristics, and determine the obstacle avoidance constraint based on the perception positioning information. Combine the various constraints to determine the overall constraint:

[0117] Let t represent the total time of the job, and the time window constraint can be expressed as:

[0118] t min ≤t≤t max

[0119] From the vehicle kinematic model, we can determine the constraints that the vehicle needs to meet when changing lanes. The dynamic constraints can be expressed as:

[0120]

[0121] Where r is the radius of curvature of the lane change path, R min is the minimum turning radius of the roller, ω is the turning angular velocity, ω max is the maximum steering angular velocity, and a is the longitudinal acceleration of the vehicle.

[0122] The specific location of the obstacle and the road boundary information can be determined from the perception and positioning information, and the obstacle avoidance constraints can be determined. Let d represent the straight-line distance to the obstacle, and x and y represent the current position of the vehicle. The obstacle avoidance constraints can be expressed as:

[0123]

[0124] Among them, x min , x max ,y min ,y max Determined by the coordinates of the boundary position of the compacted area.

[0125] Step 5-3: Based on dynamic programming, considering the constraints proposed in step 5-2, solve the objective function obtained in step 5-1 to obtain global speed information and compaction passes:

[0126] The vehicle motion state is described in the ST coordinate system, the time is discretized, and dynamic programming is performed to determine the position of the vehicle at different times.

[0127] Taking the interval between each path point as 1 meter and the time interval as 1 second, based on the path length and time window constraints, the maximum length of the S axis and the maximum time of the T axis can be determined as:

[0128]

[0129] After that, let the node be node and establish the dimension as S end OK, T end The matrix DPCOST with columns is used to store the cost from the starting point to each node; a matrix DPINDEX with the same dimension is established to store the optimal node before each node. Let S0 be the starting point of the plan, node i，j For the node in row i, column i, cost(node i，k , node i，k+1 ) is the state transfer cost of the nodes between two adjacent columns, and a dynamic programming solution process can be established:

[0130] First, initialize the matrix and calculate the cost value from the planning starting point to each path point in the first column:

[0131]

[0132] After that, the shortest cost from each node in the previous sequence to the next sequence between two adjacent columns is calculated. end -1], traverse and calculate each node node k，j Transfer to the nodes in the next column i，j+1 The cost value cost(node k，j , node i，j+1 ), update the DPCOST and DPINDEX matrices:

[0133]

[0134] Afterwards, according to DPINDEX(S end , T end ), traverse forward, record the position information of each optimal node from the end point to the starting point, and perform differential calculations on each point to determine the speed between adjacent columns, and then determine the global speed information for each operation.

[0135] After each global operation is completed, determine whether it is necessary to roll again based on the real-time compaction quality information. If necessary, repeat steps 5-3 and step 6 until the compaction quality meets the requirements. Step 6 is as follows:

[0136] Step 6-1: Input the path and speed information obtained in step 4 and step 5, determine the planned speed of each path point after coupling, and form the global trajectory information;

[0137] Step 6-2: According to the current position of the vehicle, determine the optimal planning speed at the current position based on cubic spline interpolation.

[0138] The electronic device of the present invention includes a central processing unit (CPU), which can perform various appropriate actions and processes according to computer program instructions stored in a read-only memory (ROM) or loaded from a storage unit into a random access memory (RAM). In the RAM, various programs and data required for device operation can also be stored. The CPU, ROM and RAM are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus.

[0139] Multiple components in the device are connected to the I / O interface, including: input units, such as keyboards, mice, etc.; output units, such as various types of displays, speakers, etc.; storage units, such as disks, optical disks, etc.; and communication units, such as network cards, modems, wireless communication transceivers, etc. The communication unit allows the device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunication networks.

[0140] The processing unit performs the various methods and processes described above, such as methods S1 to S6. For example, in some embodiments, methods S1 to S6 may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded and / or installed on the device via a ROM and / or a communication unit. When the computer program is loaded into the RAM and executed by the CPU, one or more steps of methods S1 to S6 described above may be performed. Alternatively, in other embodiments, the CPU may be configured to perform methods S1 to S6 in any other appropriate manner (e.g., by means of firmware).

[0141] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chip systems (SOCs), load programmable logic devices (CPLDs), and the like.

[0142] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0143] In the context of the present invention, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0144] The above is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any technician in the relevant technical field, without departing from the scope of the technical solution of the present invention, makes any form of equivalent replacement or modification to the technical solution and technical content disclosed in the present invention, which does not depart from the content of the technical solution of the present invention and still falls within the protection scope of the present invention.

Claims

1. A trajectory planning method for an intelligent road roller, characterized in that: The following steps are involved: Step 1: Use the target detection and semantic segmentation network to obtain road boundary detection results and obstacle information; Step 2: Calculate the relative position of the roller and the road boundary based on the road boundary detection result; Step 3: Use a multi-sensor data fusion method to fuse obstacle information, road boundary information, and the global positioning information of the roller obtained from the integrated navigation device to obtain the global positioning information of the roller and the obstacle; Step 4: Establish a global path planning model to plan the roller operation path based on the road boundary information and vehicle kinematic model; Step 5: Construct a multi-objective function that considers compaction quality and multiple constraints that consider the construction time window, and use the dynamic programming method to optimize the roller operating speed and compaction passes; Step 6: Use the spline curve interpolation method to realize the coupling of roller speed and path and generate the global trajectory of intelligent compaction.

2. The intelligent roller trajectory planning method according to claim 1 is characterized in that: The step 1 includes the following sub-steps: Step 1-1: Perform distortion correction on the original image to obtain a dedistorted image; Step 1-2: Use the target detection and semantic segmentation network to obtain the category, location and road pixel coordinates of the obstacles in the distortion-corrected image; Step 1-3: Use millimeter wave radar to obtain the distance, speed and azimuth information of the obstacle.

3. The intelligent roller trajectory planning method according to claim 2 is characterized in that: The step 2 includes the following sub-steps: Step 2-1: Based on the known camera internal and external parameters, the image is converted to the vehicle coordinate system through inverse perspective transformation, and the geometric expression of the road boundary line in the vehicle coordinate system is obtained by polynomial fitting; Step 2-2: Calculate the lateral offset and moving direction between the vehicle and the road boundary through the geometric expression of the road boundary line in the vehicle coordinate system.

4. The intelligent roller trajectory planning method according to claim 3 is characterized in that: The step 3 includes the following sub-steps: Step 3-1: Project the target point detected by the millimeter wave radar into the vehicle coordinate system; Step 3-2: Correlate the outputs of the camera and the millimeter wave radar in time and space; Step 3-3: Use Kalman filtering or factor graph method to fuse the detection information of the camera and millimeter wave radar to obtain accurate information such as obstacle location, speed and category; Step 3-4: The lateral distance offset of the road boundary line obtained in step 2-2 and the obstacle information obtained in step 3-3 are integrated with the vehicle global positioning information obtained from the integrated navigation device to obtain the corrected global position information of the roller and the obstacle.

5. The intelligent roller trajectory planning method according to claim 4 is characterized in that: The kinematic model of the road roller in step 4 is specifically: Among them, x f ,y f is the coordinate of the front steel wheel mass point, x r ,y r is the coordinate of the rear steel wheel mass point, θ f ,θ r are the heading angles of the front and rear steel wheels, respectively. The model assumes that there is no lateral slip during driving and that the front and rear wheels are rigid bodies. Assume that the distances from the front and rear steel wheel masses to the hinge point are l f , l r , the hinge angle is β, then the position and heading angle of the rear steel wheel mass point can be expressed as: From this, we can get the front wheel speed v of the roller f Steering angle velocity at the hinge point 6. The intelligent roller trajectory planning method according to claim 5, characterized in that: The step 4 includes the following sub-steps: Step 4-1: According to the road boundary information, obtain the coordinate vertex position of the rectangular compaction area and the road width d s and the length of the compaction area L, according to the width of the vehicle's roller d w Determine the overlap width d according to the conditions at the construction site r , calculate the number of paths n for the roller's straight-line operation: Step 4-2: Calculate the coordinate information (x ij ,y ij ), the longitudinal spacing of each path point on each path is l meter, and the value of l is determined according to the diameter of the compacting wheel of the roller and the length of contact with the ground, the requirements of different materials and construction specifications for compaction uniformity, and the vibration frequency of the roller. In specific implementation, the longitudinal spacing is gradually optimized according to the specific compaction effect from half of the compacting wheel diameter to about the compacting wheel diameter: Step 4-3: Based on the path point information obtained in step 4-2, the starting point and end point of the offset lane change are selected, and based on the fifth-order polynomial and the vehicle kinematic model, the constraint conditions are constructed to plan the offset path of the roller. The fifth-order polynomial fitting formula is as follows: y=a0+a1x+a2x 2 +a3x 3 +a4x 4 +a5x 5 Among them, a0, a1, a2, a3, a4, and a5 are polynomial coefficients.

7. The intelligent roller trajectory planning method according to claim 6, characterized in that: The step 5 comprises the following sub-steps: Step 5-1: Use vehicle speed, rolling passes and sensor information to quantitatively evaluate the compaction quality, operating efficiency and fuel economy of the roller construction, and use this to construct a multi-objective optimization function: J=λ1E fuel +λ2E time +λ3E quality Among them, E fuel is the fuel economy index, E time is the operating efficiency index, E quality is the compaction quality index, λ1, λ2, and λ3 are the weight coefficients of each objective; Step 5-2: Construct the construction time window constraint based on the road surface temperature field decay model, construct the dynamic constraint based on the vehicle dynamics characteristics, determine the obstacle avoidance constraint based on the perception positioning information, combine the various constraints, and determine the overall constraint condition; Step 5-3: Based on dynamic programming, considering the constraints proposed in step 5-2, solve the objective function obtained in step 5-1 to obtain global speed information and compaction passes.

8. The intelligent roller trajectory planning method according to claim 7, characterized in that: The step 6 comprises the following sub-steps: Step 6-1: Input the path and speed information obtained in step 4 and step 5, determine the planned speed of each path point after coupling, and form the global trajectory information; Step 6-2: According to the current position of the vehicle, determine the optimal planning speed at the current position based on cubic spline interpolation.

9. An electronic device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the program, the method according to any one of claims 1 to 8 is implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images