A trajectory determination method, device and medium of a slip steering wheel-legged vehicle

Abstract

The application discloses a trajectory determination method, device and medium of a slippage steering wheel-legged vehicle, and relates to the field of robot trajectory control. The method comprises the following steps: constructing an elevation grid map based on three-dimensional point cloud data of a motion area of the wheel-legged vehicle; determining a passable map of the wheel-legged vehicle according to the elevation grid map; constructing a trajectory determination model of the wheel-legged vehicle with the minimum contact force and the minimum wheel speed as the target; the constraint conditions of the trajectory determination model of the wheel-legged vehicle comprise a single rigid body dynamics model, a slippage steering kinematics model and a tire slippage model; and determining the trajectory of the wheel-legged vehicle based on the trajectory determination model of the wheel-legged vehicle according to the starting point, the ending point and the passable map of the wheel-legged vehicle. The application can improve the controllability and motion accuracy of the wheel-legged vehicle in a complex environment.

Description

Technical Field

[0001] This application relates to the field of robot trajectory control, and in particular to a method, device and medium for determining the trajectory of a skid-steer wheel-type vehicle. Background Technology

[0002] Wheel-legged vehicles are a new type of unmanned platform that integrates wheeled and legged mobility mechanisms, combining the high-speed driving capability of wheeled structures on flat surfaces with the strong adaptability of legged structures to complex terrain. In structured environments, these vehicles can perform rapid maneuvers and tasks with efficiency approaching that of traditional unmanned ground robots, while in unstructured or extremely rugged terrain, the combined wheel-legged motion mode significantly improves passability and continuous operation capability.

[0003] Despite their significant mobility potential, motion planning for wheeled vehicles still faces serious challenges. First, existing trajectory planning methods focus solely on generating the motion of the base's center of mass, neglecting the systematic consideration of wheel kinematics and foot forces. However, in real-world environments, a high degree of coordinated motion between the base and wheels is required. On many complex terrains (such as rugged rocks and soft sand), active control of wheel posture and ground contact force is necessary to coordinate with the base's motion and achieve stable traversal. Planning only the center of mass trajectory cannot generate this type of coordinated body movement, severely limiting the vehicle's intended high passability. Second, wheeled mobility mechanisms inevitably experience slippage and rotation on unstructured terrain, significantly impacting the accuracy, stability, and energy efficiency of motion. However, most existing planning methods establish kinematic and dynamic models based on the ideal assumption of pure rolling (no slippage). This assumption holds approximately true on flat, hard surfaces, but it introduces significant errors in loose, slippery, or uneven terrain. Furthermore, wheel slippage occurs when a vehicle is skidding and turning. If wheel slippage and turning are not taken into account, the planned trajectory will deviate significantly during actual execution, control will fail, and the vehicle may even be unable to move stably, severely restricting the vehicle's high-speed dynamic performance and robustness.

[0004] Meanwhile, many existing planning algorithms primarily handle obstacle avoidance at the geometric level, failing to incorporate the physical properties of the terrain (such as ground slope and roughness) into the planning decisions. This leads to planners potentially generating paths that are geometrically collision-free but require wheels to land on physically infeasible or high-risk footholds (such as unstable gravel). Because they fail to adaptively reason based on foothold quality, vehicles cannot autonomously select a stable wheel-ground contact point sequence, severely limiting their ability to traverse complex terrains and their motion reliability. Furthermore, existing methods typically rely on pre-defined gait sequences and perform trajectory optimization based on these sequences. This fixed pattern is ill-suited to dynamically changing or highly irregular terrain environments, failing to adjust gait strategies according to the terrain and severely limiting the vehicle's ability to autonomously discover optimal motion strategies. This restricts the vehicle's autonomy in real-world complex environments. Finally, since the actual motion of the base is indirectly determined by the interaction between the wheels and the ground, directly planning its trajectory often introduces physical deviations, leading to decreased control performance or even system instability. To achieve high-performance motion of wheel-legged vehicles in real-world environments, the wheel trajectory and ground reaction forces must be systematically considered at the planning level. This process is constrained by various physical constraints, including wheel-ground contact geometry, feasible region of ground reaction forces, kinematic limits of the mechanism, and anti-skid requirements. When ground adhesion is poor, the rational distribution of contact forces and maintenance of stability are particularly critical. The complex interactions between these constraints make it extremely difficult to design the trajectory and force distribution manually.

[0005] Therefore, based on the above problems, there is an urgent need to provide a trajectory determination method, equipment and medium for skid-steer wheel-leg vehicles, which can improve the controllability and motion accuracy of wheel-leg vehicles in complex environments.

[0006] Application content The purpose of this application is to provide a method, device and medium for determining the trajectory of a skid-steer wheel-leg vehicle, which can improve the controllability and motion accuracy of the wheel-leg vehicle in complex environments.

[0007] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a trajectory determination method for a skid-steer leg vehicle, the trajectory determination method for a skid-steer leg vehicle including: Construct an elevation raster map based on 3D point cloud data of the movement area of ​​wheel-legged vehicles; Based on the elevation grid map, determine the passable map for wheel-leg vehicles; A trajectory determination model for wheel-legged vehicles is constructed with the objectives of minimizing contact force and wheel speed. The constraints of the trajectory determination model for wheel-legged vehicles include: a single rigid body dynamics model, a slip steering kinematics model, and a tire slip model. Based on the starting point and ending point of the wheel-legged vehicle and the accessible map, the trajectory of the wheel-legged vehicle is determined using the trajectory determination model.

[0008] Optionally, the construction of an elevation raster map based on the 3D point cloud data of the wheel-legged vehicle's movement area specifically includes: Acquire 3D point cloud data of the motion area of ​​the wheel-legged vehicle; The three-dimensional point cloud data is sequentially subjected to denoising, filtering and coordinate transformation to obtain preprocessed three-dimensional point cloud data. The preprocessed 3D point cloud data is discretized into grid cells, and the elevation value corresponding to each grid cell is determined. An elevation raster map is determined based on the raster cells and their corresponding elevation values.

[0009] Optionally, determining the passable map for the wheel-leg vehicle based on the elevation grid map specifically includes: The terrain features of the elevation raster map are extracted to obtain the quality metric of each raster cell. The quality metric is weighted to obtain the landing point score for each grid cell; Based on the landing point score, each grid cell is binarized and classified to obtain a classification result; the classification result is a passable area and an impassable area. A passable map is determined based on the passable area.

[0010] Optionally, the weighting of the quality metric to obtain the landing point score for each grid cell specifically includes: Using formula Determine the score for the landing point of each grid cell; in, Scoring is given to the landing point of the grid cell. The coordinates of the center point of the raster cell. An index for quality metrics. For the first The weighting coefficients of each quality metric For the first A quality metric For the first The maximum permissible value for a quality metric.

[0011] Optionally, the process of constructing a trajectory determination model for a wheel-legged vehicle with the objectives of minimizing contact force and wheel speed specifically includes: Using formula Determine the objective function ; Using formula Determine the dynamic model of a single rigid body; Using formula Determine the kinematic model for slip steering; Using formula Determine the tire slip model; in, For wheel-leg vehicles One wheel The force at the tip of the foot at all times For wheel-leg vehicles One wheel The position of the wheels at any given moment. For wheel-leg vehicles One wheel The speed of the wheels at any given moment For wheel-type vehicles, Let the mass of the rigid body be... for The acceleration of the wheeled vehicle body at all times. for The angular acceleration of a wheeled vehicle at all times. for The angular velocity of the wheeled vehicle at all times. The number of feet. It is the acceleration due to gravity. Let the moment of inertia be the center of mass. for The position of the center of gravity of a wheel-type vehicle. and These are the longitudinal and lateral velocities of the wheel-legged vehicle, respectively. For the wheel radius, This represents the wheel speed of the left wheel. The slip ratio of the left wheel. This represents the wheel speed of the right-hand wheel. The slip ratio of the right wheel. For vehicle body side slip angle, Let be the angular velocity of the wheel-legged vehicle. The wheelbase is the distance between the wheels. and The first The longitudinal and lateral tire forces of each wheel Let be the longitudinal stiffness coefficient of the wheel. This is the lateral stiffness coefficient of the wheel. For the first The longitudinal slip ratio of each wheel For the first The slip angle of each wheel.

[0012] Optionally, determining the trajectory of the wheel-legged vehicle based on the starting point and ending point of the wheel-legged vehicle and the passable map, using the wheel-legged vehicle trajectory determination model, specifically includes: Based on the starting point and ending point of the wheel-leg vehicle and the accessible map, initial discrete nodes and initial gait information are generated using linear interpolation. Based on the initial discrete nodes and the initial gait information, the state variables of the wheel-legged vehicle and the interval time between adjacent discrete nodes are obtained based on the trajectory determination model of the wheel-legged vehicle. The state variables include: the position of the center of mass of the wheel-legged vehicle, the attitude of the wheel-legged vehicle, the wheel position of the wheel-legged vehicle, and the foot force of the wheel-legged vehicle. The trajectory of the wheel-leg vehicle is determined by performing polynomial curve fitting on the state variables and the time intervals between adjacent discrete nodes using the Hermite interpolation method.

[0013] Optionally, the step of obtaining the state variables of the wheel-legged vehicle and the interval time between adjacent discrete nodes based on the initial discrete nodes and the initial gait information, according to the wheel-legged vehicle trajectory determination model, specifically includes: Using formula The state variables of the wheel-legged vehicle and the time interval between adjacent discrete nodes are obtained; in, for The state variable at time t, for The derivative of the state variable at time t. for The second derivative of the state variable at time t. The time interval between adjacent discrete nodes. For the trajectory starting point state variable, The total duration of the trajectory. For the trajectory endpoint state variable, For inequality constraints, For equality constraints, for The position of the center of gravity of a wheel-type vehicle. for The posture of the wheeled vehicle at all times. For wheel-leg vehicles One wheel The position of the wheels at any given moment. For wheel-leg vehicles One wheel The force at the tip of the foot at all times.

[0014] Optionally, the step of using Hermite interpolation to perform polynomial curve fitting on the state variables and the time intervals between adjacent discrete nodes to determine the trajectory of the wheel-leg vehicle specifically includes: Using formula Determine the trajectory of the wheel-legged vehicle; in, , , and These are the coefficients of the polynomial. The time interval between adjacent discrete nodes. For the first The state variables of each node For the first The derivative of the state variables of each node, For the first The state variables of each node For the first The derivatives of the state variables of each node.

[0015] Secondly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described trajectory determination method for a skid-steer wheel-type vehicle.

[0016] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described trajectory determination method for a skid-steering wheel-type vehicle.

[0017] According to the specific embodiments provided in this application, this application has the following technical effects: This application provides a method, device, and medium for determining the trajectory of a skid-steer wheel-leg vehicle. By combining an elevation grid map and a passable map, the area where the wheels can reliably contact is determined, thus ensuring the movement safety of the wheel-leg vehicle. By constructing a trajectory determination model for the wheel-leg vehicle, the skid steering between the tires and the ground of the wheel-leg vehicle is analyzed, thereby determining the trajectory of the wheel-leg vehicle. This ensures that the wheel-leg vehicle still has good controllability and movement accuracy even under high-speed or low-adhesion conditions. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating a trajectory determination method for a skid-steering wheel-type vehicle according to an embodiment of this application. Figure 2 This is a flowchart illustrating a trajectory determination method for a skid-steering wheel-type vehicle according to an embodiment of this application. Figure 3 This is a schematic diagram of the Hermite parameterization method in one embodiment of this application. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0022] Wheel-legged vehicles are a type of ground mobile robot system that organically combines wheeled drive with legged gait. Each motion unit of a wheel-legged vehicle consists of multi-degree-of-freedom leg joints and end-drive wheels, enabling it to flexibly switch between or coordinate the use of rolling and walking locomotion modes in different terrain environments. On flat and highly traction surfaces, wheel-legged vehicles can achieve high-speed, low-energy long-distance movement using wheeled drive; while in unstructured, rugged, or obstacle-heavy terrain, the lifting, traversing, and posture adjustment capabilities of the leg joints enhance terrain adaptability and mobility.

[0023] In one exemplary embodiment, such as Figure 1 and Figure 2 As shown, a trajectory determination method for a skid-steer wheel-leg vehicle is provided. Skid-steer wheel-leg vehicles (hereinafter referred to as wheel-leg vehicles) experience wheel slippage when moving in complex terrain. Wheel slippage refers to a phenomenon where the actual motion state of a vehicle's wheel deviates from its ideal pure rolling state when the wheel is moving on the ground. That is, when the wheel is rotating, relative sliding occurs between it and the ground; this is called slippage. This slippage is mainly divided into two forms: longitudinal slippage and lateral slippage. The trajectory determination method for the skid-steer wheel-leg vehicle of this application includes the following S1-S4. Wherein: S1: Construct an elevation raster map based on the 3D point cloud data of the movement area of ​​the wheel-legged vehicle.

[0024] An elevation raster map is a type of raster map specifically designed to represent the elevation of every point on the Earth's surface. Instead of focusing on the specific types of objects on the surface (such as houses, roads, and vegetation), elevation raster maps concentrate on the undulating shape of the terrain. Its core data is the elevation value (Z-value) corresponding to each raster cell.

[0025] S1 specifically includes: S11: Acquire 3D point cloud data of the motion area of ​​the wheel-legged vehicle.

[0026] The LiDAR was mounted on the experimental vehicle to control its movement within the motion area of ​​the wheel-legged vehicle and to collect 3D point cloud data of the motion area in real time.

[0027] S12: Denoise, filter, and transform the coordinates of the 3D point cloud data sequentially to obtain the preprocessed 3D point cloud data.

[0028] Specifically, coordinate transformation refers to unifying the coordinates of 3D point cloud data to the coordinate system of the wheel-leg vehicle.

[0029] S13: Discretize the preprocessed 3D point cloud data into raster cells and determine the elevation value corresponding to each raster cell.

[0030] Specifically, the elevation value of a raster cell is calculated from the elevation information of the corresponding 3D point cloud data.

[0031] S14: Determine the elevation raster map based on the raster cells and their corresponding elevation values.

[0032] S2: Determine the passable map for wheel-leg vehicles based on the elevation grid map.

[0033] When wheeled vehicles navigate terrain with highly irregular elevations, they must rely on the external environment for decision-making. While traditional elevation raster maps can describe terrain undulations, they cannot distinguish between foothold and non-foothold areas. Therefore, the introduction of traversable maps is crucial. A traversable map is a map used for path planning and navigation that clearly identifies which areas in the environment are passable and which are obstacles. Its core data is the traversability of each area, typically a binary value (passable / non-passable), but it can also be a continuous value including the passage cost.

[0034] S2 specifically includes: S21: Extract terrain features from the elevation raster map to obtain the quality metric for each raster cell.

[0035] Specifically, terrain feature extraction refers to extracting physical attributes of the terrain such as slope and smoothness (flatness and roughness). Slope reflects the degree of terrain inclination, and smoothness characterizes the surface unevenness. These features together affect the contact stability between the wheels of wheel-leg vehicles and the ground.

[0036] Before calculating the slope angle of a grid cell, it is necessary to calculate the surface normal vector of the grid cell. Normal vector estimation is performed based on a local surface fitting method. For each point in the elevation raster map, its nearest neighbor is searched. Given a set of discrete points, we then compute the local plane in the least-squares sense for these points. Local plane It can be represented as: ; in, For local plane The normal vector, For local plane Distance to the origin For the index of the nearest discrete point, For raster cell indexes.

[0037] Depend on The plane normal vector fitted by the nearest discrete points is the normal vector of the current scan point. Local plane The normal vector can be obtained by Principal Component Analysis (PCA), and the local plane can be calculated. go through Centroid of the domain point And normal vector satisfy Local plane The formula for calculating the normal vector is as follows: ; in, Let covariance matrix be the variance matrix. This represents the number of grid cells, and also the total number of discrete points. This is a transpose.

[0038] For covariance matrix Perform eigenvalue decomposition to obtain The eigenvalues, The eigenvector corresponding to the smallest eigenvalue is the local plane. The normal vector.

[0039] When calculating the slope angle of a grid cell, a reference vector is first set. The vector is oriented positively, and the vector to be measured... dot product of reference vector ,like Then Flip it, otherwise leave it unchanged. The slope angle of a given grid cell. The calculation formula is as follows: ; in, This represents the normal vector of the current grid cell.

[0040] Smooth the elevation values ​​of the raster cells, and then smooth the elevation values. Subtract the original elevation value This refers to the smoothness of the grid cells. The calculation formula is as follows: .

[0041] The slope and smoothness of the grid cells are normalized, and the corresponding quality metrics are obtained.

[0042] S22: Weight the quality metrics to obtain the landing point score for each grid cell.

[0043] Specifically, based on the importance of quality metrics to the passability of wheel-legged vehicles in different task scenarios, corresponding weights are assigned for weighted fusion. Through linear combinations of different quality metrics, the landing point score of grid cells can be calculated effectively. The calculation formula is as follows: ; in, Scoring is given to the landing point of the grid cell. The coordinates of the center point of the raster cell. An index for quality metrics. For the first The weighting coefficients of each quality metric For the first A quality metric For the first The maximum permissible value for a quality metric.

[0044] Specifically, the foot placement score is used to comprehensively evaluate whether a particular grid cell is suitable for wheel entry; a higher score indicates that the grid cell is more conducive to stable support and reliable movement. The foot placement score represents the quality of the foot placement. When, it indicates that the grid cell is a safe landing point, when When this occurs, it indicates that the grid cell may cause the foot to slip.

[0045] S23: Based on the landing point score, perform binary classification on each grid cell to obtain the classification result; the classification result is a passable area and an impassable area.

[0046] Specifically, to simplify the selection process of landing points, an adaptive threshold is set to perform binary classification of landing point scores, creating binary landing point scores. The elevation raster map is clearly divided into two parts: passable and impassable areas; among them, The area with a value of 1 is a passable area. Areas with a value of 0 are impassable.

[0047] S24: Determine the passable map based on the passable area.

[0048] Accessibility maps can provide clear and reliable semantic information on terrain accessibility for subsequent determination of the trajectory of wheeled vehicles, ensuring the safety and stability of wheeled vehicles in complex environments.

[0049] S3: Construct a trajectory determination model for wheel-legged vehicles with the objectives of minimizing contact force and wheel speed.

[0050] The trajectory determination model for wheel-legged vehicles includes constraints and an objective function. The constraints include a single-rigid-body dynamics model, a skid steering kinematics model, a tire slip model, and other constraints. The single-rigid-body dynamics model ensures overall force balance and dynamic feasibility of the trajectory determination model. The skid steering kinematics model coordinates the consistency between wheel speed and base motion during vehicle steering, describing the motion characteristics of vehicles steering differentially, such as tracked vehicles or some four-wheel differential drive vehicles. In the skid steering kinematics model, the vehicle achieves steering by driving the left and right wheels (or tracks) at different linear velocities, rather than relying on traditional Ackermann steering geometry constraints. Therefore, lateral slippage of the tires or tracks relative to the ground typically occurs during steering, hence the name "skid steering." The tire slip model accurately describes the tangential force behavior between the tires and the ground, improving the realism of planning for wheel-legged vehicles in low-traction environments.

[0051] Specifically, the calculation formula for the single rigid body kinematic model is as follows: ; in, Let the mass of the rigid body be... for The acceleration of the wheeled vehicle body at all times. The number of feet. For wheel-type vehicles, For wheel-leg vehicles One wheel The force at the tip of the foot at all times It is the acceleration due to gravity. Let the moment of inertia be the center of mass. for The angular acceleration of a wheeled vehicle at all times. for The angular velocity of the wheel-legged vehicle at any given moment can be calculated from the optimized Euler angles and the speed. for The position of the center of gravity (COM) of a wheel-type vehicle. For wheel-leg vehicles One wheel The position of the wheels at any given moment.

[0052] The calculation formula for the skid steering kinematics model is as follows: ; in, and These are the longitudinal and lateral velocities of the wheel-legged vehicle, respectively. Let be the angular velocity of the wheel-legged vehicle. For the wheel radius, This represents the wheel speed of the left wheel. The slip ratio of the left wheel. This represents the wheel speed of the right-hand wheel. The slip ratio of the right wheel. For vehicle body side slip angle, This refers to the wheel track.

[0053] The calculation formula for the tire slip model is as follows: ; in, and The first The longitudinal and lateral tire forces of each wheel Let be the longitudinal stiffness coefficient of the wheel. This is the lateral stiffness coefficient of the wheel. For the first The longitudinal slip ratio of each wheel For the first The slip angle of each wheel.

[0054] In addition, to ensure that the trajectory is physically feasible, the following constraints are added: starting point constraint, ending point constraint, curve connection point derivative continuity constraint, wheel motion range constraint, wheel lateral constraint, contact constraint, thrust constraint, friction cone constraint, oscillating phase foot end force constraint, and total time constraint.

[0055] The formula for calculating the starting constraint is as follows: ; in, This refers to the position of the vehicle body in a wheel-legged vehicle. The posture of a wheel-legged vehicle body. For a moment, The starting point for the body design of wheeled vehicles. The starting posture for the wheel-legged vehicle body design.

[0056] The formula for calculating the endpoint constraint is as follows: ; in, The total duration of the trajectory. The endpoint of the wheeled vehicle body is planned. The final posture planned for the wheel-legged vehicle body.

[0057] The formula for calculating the derivative continuity constraint at the curve connection point is as follows: ; in, Indicates the index of the segmented trajectory. For the first The acceleration at the starting point of each segmented trajectory, For the first The acceleration at the end of the segmented trajectory, For the first The duration of each segmented trajectory.

[0058] The formula for calculating the wheel motion range constraint is as follows: ; in, For wheel-leg vehicles One wheel The position of the wheels at any given moment. This indicates that due to limitations in joints and leg length, the first The range of motion of each wheel in the vehicle body coordinate system.

[0059] The formula for calculating wheel lateral constraints is as follows: ; in, For the first The speed of each wheel in the y-direction Indicates the first The wheels are in contact.

[0060] The formula for calculating contact constraints is as follows: ; in, The elevation value is obtained from the elevation raster map. For the first The position of each wheel in the z-direction For the first The position of each wheel in the x and y directions The radius is the wheel radius.

[0061] The formula for calculating thrust constraint is as follows: ; in, This is the terrain normal vector for the current location.

[0062] The calculation formula for friction cone constraint is as follows: ; in, Indicates a friction cone. This represents the coefficient of friction of the ground.

[0063] The calculation formula for the force constraint at the foot of the swing phase is as follows: .

[0064] The formula for calculating the total time constraint is as follows: ; in, Indicates the first The number of segmented tracks per wheel For the first The first of the wheels Segmented trajectory, This represents the total duration of the trajectory.

[0065] Objective function of the wheel-leg vehicle trajectory determination model The calculation formula is as follows: ; in, For wheel-leg vehicles One wheel The speed of the wheels at any given moment.

[0066] The trajectory determination model for wheel-legged vehicles can minimize contact force and wheel speed while satisfying constraints through the objective function, thereby reducing energy consumption.

[0067] S4: Determine the trajectory of the wheel-legged vehicle based on the starting point and ending point of the wheel-legged vehicle and the accessible map, and based on the wheel-legged vehicle trajectory determination model.

[0068] S4 specifically includes: S41: Based on the starting point and ending point of the wheel-leg vehicle and the accessible map, generate initial discrete nodes and initial gait information using linear interpolation.

[0069] The initial discrete node height is set to the terrain height. The initialization phase (the phase of generating initial discrete nodes and initial gait information) provides initial guesses for the base, wheel motion, foot force, and time intervals between discrete nodes. Good initialization can significantly improve the convergence efficiency and solution quality of the wheel-legged vehicle trajectory determination model.

[0070] S42: Based on the initial discrete nodes and initial gait information, and based on the wheel-leg vehicle trajectory determination model, obtain the state variables of the wheel-leg vehicle and the interval time between adjacent discrete nodes (i.e., optimize and solve the wheel-leg vehicle trajectory determination model).

[0071] Based on the initial discrete nodes and initial gait information, the trajectory determination model of the wheel-leg vehicle is optimized to minimize the objective function while satisfying various constraints. The objective function of this application is to minimize the contact force and wheel speed. Furthermore, the objective function can also be optimized to minimize energy consumption, time, or motion smoothness, depending on the actual situation. The solution yields a series of discrete state variables and their corresponding timestamps, i.e., the interval between adjacent discrete nodes, also known as the trajectory duration. The state variables include: the position of the wheel-leg vehicle's center of mass, the vehicle's attitude, the wheel positions, and the foot force. Specifically, the solution formula is as follows: ; in, for The state variable at time t, , for The derivative of the state variable at time t. for The second derivative of the state variable at time t. The time interval between adjacent discrete nodes. For the trajectory starting point state variable, The total duration of the trajectory. For the trajectory endpoint state variable, For inequality constraints, This is an equality constraint.

[0072] S43: Use Hermite interpolation to perform polynomial curve fitting on the state variables and the time intervals between adjacent discrete nodes to determine the trajectory of the wheel-leg vehicle.

[0073] Trajectory optimization (TO), as an automated motion generation method, provides an effective solution for determining the trajectory of wheel-legged vehicles. By defining only high-level task objectives (such as target pose, gait type, or speed requirements), TO can automatically generate time-varying trajectories with state, control input, and contact forces while satisfying dynamics and various constraints, significantly enhancing the versatility and environmental adaptability of the planning system.

[0074] This application does not optimize the entire polynomial curve, but rather optimizes the starting value of the polynomial. (No. (state variables of each node), endpoint value (No. (state variables of each node) and their corresponding first derivatives , and the duration of the polynomial curve The time interval between adjacent discrete nodes is used to fully define each polynomial curve, such as... Figure 3 As shown. This Hermite interpolation method is more intuitive because the state variables directly describe the trajectory. Furthermore, the endpoint of the previous polynomial can serve as the starting point of the next polynomial, ensuring the continuous change of wheel-end velocity and force along the trajectory. The calculation formula for the Hermite interpolation method is as follows: ; in, , , and These are the coefficients of the polynomial. Let be the time interval between adjacent discrete nodes. For the first The state variables of each node For the first The derivative of the state variables of each node, For the first The state variables of each node For the first The derivatives of the state variables of each node.

[0075] The discrete state variables and the time intervals between adjacent discrete nodes are fitted using polynomial curves. A fifth-order polynomial is used to smoothly fit the base trajectory. The fifth-order polynomial curve can optimize acceleration because its six undetermined coefficients perfectly match the six boundary conditions (position, velocity, and acceleration) of the starting and ending points of a trajectory segment, thus ensuring the continuity of position, velocity, and acceleration, effectively avoiding dynamic shocks, and meeting the requirements of a single rigid body dynamics model. Simultaneously, a cubic polynomial is used to fit the wheel and foot force trajectories, ultimately generating a continuous, dynamically feasible, and multi-engineering-constrained full-body motion trajectory.

[0076] This application employs a single rigid body dynamics model to describe the overall dynamic characteristics of wheel-leg vehicles, and integrates a tire-ground contact physical model (slip steering kinematics model and tire slip model) to more accurately characterize the motion of wheel-leg vehicles under slip conditions. Based on elevation grid maps and traversable maps, the motion of the base, wheels, and foot contact forces are optimized in an integrated manner to achieve high-speed, stable, and highly passable dynamic motion in unstructured environments.

[0077] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal. The computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is connected to the system bus via the I / O interfaces. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the computer device stores trajectory data of a skid-steer leg vehicle. The I / O interfaces of the computer device are used for exchanging information between the processor and external devices. The communication interface of the computer device is used for communicating with an external terminal via a network connection. When the computer program is executed by the processor, it implements a trajectory determination method for a skid-steer leg vehicle.

[0078] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0079] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0080] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0081] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0082] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0083] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0084] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for determining the trajectory of a skid-steering wheel-type vehicle, characterized in that, The trajectory determination method for the skid-steer wheel-type vehicle includes: Construct an elevation raster map based on 3D point cloud data of the movement area of ​​wheel-legged vehicles; Based on the elevation grid map, determine the passable map for wheel-leg vehicles; A trajectory determination model for wheel-legged vehicles is constructed with the objectives of minimizing contact force and wheel speed. The constraints of the trajectory determination model for wheel-legged vehicles include: a single rigid body dynamics model, a slip steering kinematics model, and a tire slip model. Based on the starting point and ending point of the wheel-legged vehicle and the accessible map, the trajectory of the wheel-legged vehicle is determined using the trajectory determination model.

2. The trajectory determination method for a skid-steering wheel-type vehicle according to claim 1, characterized in that, The construction of an elevation raster map based on the 3D point cloud data of the wheel-legged vehicle's movement area specifically includes: Acquire 3D point cloud data of the motion area of ​​the wheel-legged vehicle; The three-dimensional point cloud data is sequentially subjected to denoising, filtering and coordinate transformation to obtain preprocessed three-dimensional point cloud data. The preprocessed 3D point cloud data is discretized into grid cells, and the elevation value corresponding to each grid cell is determined. An elevation raster map is determined based on the raster cells and their corresponding elevation values.

3. The trajectory determination method for a skid-steering wheel-type vehicle according to claim 1, characterized in that, The step of determining the passable map for wheel-leg vehicles based on the elevation grid map specifically includes: The terrain features of the elevation raster map are extracted to obtain the quality metric of each raster cell. The quality metric is weighted to obtain the landing point score for each grid cell; Based on the landing point score, each grid cell is binarized and classified to obtain a classification result; the classification result is a passable area and an impassable area. A passable map is determined based on the passable area.

4. The trajectory determination method for a skid-steering wheel-type vehicle according to claim 3, characterized in that, The weighting of the quality metric to obtain the landing point score for each grid cell specifically includes: Using formula Determine the score for the landing point of each grid cell; in, Scoring is given to the landing point of the grid cell. The coordinates of the center point of the raster cell. An index for quality metrics. For the first The weighting coefficients of each quality metric For the first A quality metric For the first The maximum permissible value for a quality metric.

5. The trajectory determination method for a skid-steering wheel-type vehicle according to claim 1, characterized in that, The method for determining the trajectory of a wheel-legged vehicle, with the objectives of minimizing contact force and wheel speed, specifically includes: Using formula Determine the objective function ; Using formula Determine the dynamic model of a single rigid body; Using formula Determine the kinematic model for slip steering; Using formula Determine the tire slip model; in, For wheel-leg vehicles One wheel The force at the tip of the foot at all times For wheel-leg vehicles One wheel The position of the wheels at any given moment. For wheel-leg vehicles One wheel The speed of the wheels at any given moment For wheel-type vehicles, Let the mass of the rigid body be... for The acceleration of the wheeled vehicle body at all times. for The angular acceleration of a wheeled vehicle at all times. for The angular velocity of the wheeled vehicle at all times. The number of feet. It is the acceleration due to gravity. Let the moment of inertia be the center of mass. for The position of the center of gravity of a wheel-type vehicle. and These are the longitudinal and lateral velocities of the wheel-legged vehicle, respectively. For the wheel radius, This represents the wheel speed of the left wheel. The slip ratio of the left wheel. This represents the wheel speed of the right-hand wheel. The slip ratio of the right wheel. For vehicle body side slip angle, Let be the angular velocity of the wheel-legged vehicle. The wheelbase is the distance between the wheels. and The first The longitudinal and lateral tire forces of each wheel Let be the longitudinal stiffness coefficient of the wheel. This is the lateral stiffness coefficient of the wheel. For the first The longitudinal slip ratio of each wheel For the first The slip angle of each wheel.

6. The trajectory determination method for a skid-steering wheel-type vehicle according to claim 1, characterized in that, The step of determining the trajectory of the wheel-legged vehicle based on its starting point, ending point, and the accessible map, using the wheel-legged vehicle trajectory determination model, specifically includes: Based on the starting point and ending point of the wheel-leg vehicle and the accessible map, initial discrete nodes and initial gait information are generated using linear interpolation. Based on the initial discrete nodes and the initial gait information, the state variables of the wheel-legged vehicle and the interval time between adjacent discrete nodes are obtained based on the trajectory determination model of the wheel-legged vehicle. The state variables include: the position of the center of mass of the wheel-legged vehicle, the attitude of the wheel-legged vehicle, the wheel position of the wheel-legged vehicle, and the foot force of the wheel-legged vehicle. The trajectory of the wheel-leg vehicle is determined by performing polynomial curve fitting on the state variables and the time intervals between adjacent discrete nodes using the Hermite interpolation method.

7. The trajectory determination method for a skid-steering wheel-type vehicle according to claim 6, characterized in that, The step of obtaining the state variables of the wheel-legged vehicle and the interval time between adjacent discrete nodes based on the initial discrete nodes and the initial gait information, according to the wheel-legged vehicle trajectory determination model, specifically includes: Using formula The state variables of the wheel-legged vehicle and the time interval between adjacent discrete nodes are obtained; in, for The state variable at time t, for The derivative of the state variable at time t. for The second derivative of the state variable at time t. The time interval between adjacent discrete nodes. For the trajectory starting point state variable, The total duration of the trajectory. For the trajectory endpoint state variable, For inequality constraints, For equality constraints, for The position of the center of gravity of a wheel-type vehicle. for The posture of the wheeled vehicle at all times. For wheel-leg vehicles One wheel The position of the wheels at any given moment. For wheel-leg vehicles One wheel The force at the tip of the foot at all times.

8. The trajectory determination method for a skid-steering wheel-type vehicle according to claim 6, characterized in that, The process of using Hermite interpolation to perform polynomial curve fitting on the state variables and the time intervals between adjacent discrete nodes to determine the trajectory of the wheel-leg vehicle specifically includes: Using formula Determine the trajectory of the wheel-legged vehicle; in, , , and These are the coefficients of the polynomial. The time interval between adjacent discrete nodes. For the first The state variables of each node For the first The derivative of the state variables of each node, For the first The state variables of each node For the first The derivatives of the state variables of each node.

9. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the trajectory determination method for a skid-steer leg vehicle as described in any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the trajectory determination method for a skid-steering leg vehicle as described in any one of claims 1-8.

Images