Motion planning method, device, robot, readable storage medium and program product

Abstract

The application relates to a motion planning method and device, a robot, a readable storage medium and a program product. A chassis corresponding kinematic model is generated based on chassis state information and each virtual wheel information corresponding to each wheel group, and the speed and angle of each virtual wheel are not related; motion constraint information of the chassis is obtained, and the chassis is subjected to motion planning according to the kinematic model and the motion constraint information. Compared with the traditional motion planning in the switching mode of double Ackerman steering mode and skew mode, the scheme removes the restriction that each wheel in the chassis must rotate at the same angle size and in the opposite direction, constructs multiple virtual wheel information, generates a kinematic model in combination with the chassis state information and the virtual wheel information, and subjects the chassis to motion planning in combination with the kinematic model and the motion constraint information of the chassis, thereby improving the operation efficiency of the chassis.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a motion planning method, apparatus, robot, computer-readable storage medium, and computer program product. Background Technology

[0002] With the advancement of technology, the demand for automation and intelligence in the service and manufacturing industries is constantly increasing, and autonomous mobile robots have been deployed in various fields. Autonomous mobile robots have a mobile chassis for movement. The chassis is equipped with drive wheels, which control the robot's movement. Currently, the chassis movement of autonomous mobile robots includes a dual Ackerman steering mode and a diagonal traverse mode. The chassis switches between these two modes to plan the trajectory of the movement path. However, at the moment of switching between the two modes, the chassis speed can still drop to zero, resulting in an inefficient planned path.

[0003] Therefore, current methods for motion planning of the chassis of autonomous mobile robots suffer from low operational efficiency. Summary of the Invention

[0004] Therefore, it is necessary to provide a motion planning method, device, robot, computer-readable storage medium, and computer program product that can improve the operating efficiency of the chassis in response to the above-mentioned technical problems.

[0005] In a first aspect, this application provides a motion planning method, including:

[0006] Obtain the kinematic model corresponding to the chassis; the kinematic model is generated based on the chassis state information, first virtual wheel information, and second virtual wheel information; the speed and angle of the first virtual wheel information and the second virtual wheel information are not related; the first virtual wheel information corresponds to the first wheel set of the chassis, and the second virtual wheel information corresponds to the second wheel set of the chassis; the perpendicular line corresponding to the wheel of the first wheel set points to the first center when turning; the perpendicular line corresponding to the wheel of the second wheel set points to the second center when turning.

[0007] Obtain the motion constraint information of the chassis;

[0008] Motion planning is performed on the chassis based on the kinematic model and the motion constraint information.

[0009] Secondly, this application also provides a motion planning device, comprising:

[0010] The first acquisition module is used to acquire the kinematic model corresponding to the chassis; the kinematic model is generated based on the chassis state information, first virtual wheel information, and second virtual wheel information of the chassis; the speed and angle of the first virtual wheel information and the second virtual wheel information are not related; the first virtual wheel information corresponds to the first wheel set of the chassis, and the second virtual wheel information corresponds to the second wheel set of the chassis; the perpendicular line corresponding to the wheel of the first wheel set points to the first center when turning; the perpendicular line corresponding to the wheel of the second wheel set points to the second center when turning.

[0011] The second acquisition module is used to acquire the motion constraint information of the chassis;

[0012] The planning module is used to perform motion planning for the chassis based on the kinematic model and the motion constraint information.

[0013] Thirdly, this application also provides a robot, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described above.

[0014] Fourthly, this application also 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 method.

[0015] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described method.

[0016] The aforementioned motion planning method, device, robot, computer-readable storage medium, and computer program product generate a kinematic model of the chassis based on chassis state information and information about each virtual wheel corresponding to each wheel set, with the speeds and angles of each virtual wheel being independent. It then acquires the chassis's motion constraint information and performs motion planning based on the kinematic model and the motion constraint information. Compared to the traditional method of switching between dual Ackerman steering and swerving modes for motion planning, this solution removes the restriction that each wheel in the chassis must rotate at the same angle but in opposite directions. It constructs information about multiple virtual wheels and combines this information with the chassis state information and the virtual wheel information to generate a kinematic model. By combining the chassis's kinematic model with the motion constraint information, it performs motion planning for the chassis, thus improving the chassis's operating efficiency. Attached Figure Description

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

[0018] Figure 1 This is a flowchart illustrating the motion planning method in one embodiment;

[0019] Figure 2 This is a schematic diagram of the kinematic model in one embodiment;

[0020] Figure 3 This is a structural block diagram of the motion planning device in one embodiment;

[0021] Figure 4 This is a diagram of the internal structure of a robot in one embodiment. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0023] In the application scenarios of autonomous mobile robots, two-wheeled differential chassis, four-wheeled Mecanum wheel chassis, and four-wheeled four-rotor chassis each have their unique advantages and disadvantages. Two-wheeled differential chassis have a simple structure, mature control algorithms, and low cost, making them suitable for flat indoor environments and enabling stationary rotation and relatively flexible movement. However, they cannot achieve omnidirectional movement, have lower motion accuracy, require high ground flatness, and have poor passability on complex or rough surfaces. Four-wheeled Mecanum wheel chassis provide omnidirectional movement capability and can achieve precise motion control, making them suitable for occasions requiring precision operation. However, they have higher manufacturing costs, limited load capacity, significant wheel wear, operating noise, and energy loss, and their complex structure makes control algorithms difficult to implement. Four-wheeled four-rotor chassis combine the advantages of structure and control, featuring independent steering, drive, and suspension designs, thus achieving omnidirectional movement. They offer extremely high control efficiency and accuracy, low failure rate, and high stability, providing superior passability and off-road capability, and can adapt to various complex terrains and scenarios. Although the four-wheel, four-rotor chassis is more expensive than the other two types of chassis, its superior performance and wide range of applications make it an ideal choice for many high-end and complex tasks.

[0024] Current motion planning methods for four-wheel, four-turn chassis typically simplify the chassis's kinematic model by converting chassis motion into different drive modes to achieve complex movements. Two common modes are the dual Ackermann steering mode and the swerve mode. In dual Ackermann steering mode, the four-wheel, four-turn chassis mimics the differential motion of two wheels, with the front and rear virtual center wheels having the same steering angle but opposite directions. By precisely calculating the steering angles of all four wheels, the centers of their travel paths are ensured to converge at a single point—the instantaneous steering center—to achieve smooth driving. In swerve mode, the rotation angles and drive speeds of all four wheels are the same, enabling the chassis to move diagonally and translate horizontally. This mode is particularly suitable for motion planning in confined spaces.

[0025] To ensure a smooth transition between these two modes and achieve efficient and safe trajectory planning, the current state of the chassis, kinematic constraints, and the surrounding environment must be comprehensively considered during the transition to generate a smooth trajectory that meets the switching requirements. This method not only requires the continuity and feasibility of the trajectory but also ensures the stability and safety of the vehicle during the transition to achieve efficient and flexible chassis control. However, at the moment of switching between the two modes, the chassis inevitably experiences a situation where its speed reaches zero. Furthermore, because the above method does not completely simplify the model, it eliminates most of the chassis's motion characteristics, failing to utilize the full motion capabilities of the chassis and thus increasing the energy loss of the planned trajectory. Therefore, how to plan an efficient motion trajectory remains a pressing technical problem to be solved.

[0026] Based on this, this application removes the restriction that each wheel in the chassis must rotate at the same angle but in opposite directions, constructs multiple virtual wheel information, and generates a kinematic model by combining chassis state information and virtual wheel information. By combining the chassis kinematic model and motion constraint information, motion planning is performed on the chassis, thereby improving the chassis's operating efficiency.

[0027] In one embodiment, such as Figure 1 As shown, a motion planning method is provided. This embodiment uses a robot executing the method as an example for illustration. The robot can be a humanoid robot or other four-wheeled autonomous mobile robot. It is understood that this method can also be applied to a system including a robot and a server, and implemented through the interaction between the robot and the server, including the following steps S202 to S206. Wherein:

[0028] Step S202: Obtain the kinematic model corresponding to the chassis; the kinematic model is generated based on the chassis state information, the first virtual wheel information, and the second virtual wheel information; the speed and angle of the first virtual wheel information and the second virtual wheel information are not related; the first virtual wheel information corresponds to the first wheel group of the chassis, and the second virtual wheel information corresponds to the second wheel group of the chassis; the perpendicular line corresponding to the wheel of the first wheel group points to the first center when turning; the perpendicular line corresponding to the wheel of the second wheel group points to the second center when turning.

[0029] The aforementioned robot can be a four-wheeled autonomous mobile robot, which integrates the autonomous navigation capabilities of a mobile robot with the flexible maneuverability of a humanoid robot. This type of robot not only possesses a mobile chassis capable of autonomous positioning and navigation, and flexible movement in complex environments, but also has a robotic arm capable of performing precision operations and transporting objects. To achieve efficient motion planning, the robot needs to combine the kinematic model of the chassis to realize motion planning.

[0030] The robot can determine the information of a first virtual wheel based on the first wheel set of the chassis, and the information of a second virtual wheel based on the second wheel set of the chassis. The first and second wheel sets can be wheels located at different positions on the robot's chassis. For example, the first wheel set could be the two front wheels, and the second wheel set could be the two rear wheels. The aforementioned first and second virtual wheel information can be information about virtual wheels whose center point is on the centerline of the chassis; in this case, the first virtual wheel information can be called virtual front wheel information, and the second virtual wheel information can be called virtual rear wheel information. Based on the chassis state information, the first virtual wheel information, and the second virtual wheel information, the robot can construct a kinematic model corresponding to the chassis, thus obtaining the kinematic model of the chassis for the robot.

[0031] In one embodiment, obtaining the kinematic model corresponding to the chassis includes: generating corresponding first virtual wheel information and second virtual wheel information based on the motion information of each wheel corresponding to the first wheel group and the second wheel group of the chassis; obtaining chassis state information based on the chassis position information and orientation information; and determining the corresponding kinematic model based on the chassis state information including position information and orientation information, the first virtual wheel information and the second virtual wheel information.

[0032] In this embodiment, when constructing the kinematic model, the robot can acquire the motion information of each wheel corresponding to the first wheel group of the chassis and the motion information of each wheel corresponding to the second wheel group. The robot then generates first virtual wheel information based on the motion information of each wheel in the first wheel group and second virtual wheel information based on the motion information of each wheel in the second wheel group. The aforementioned motion information includes wheel speed, acceleration, and rotation angle, among other things.

[0033] The robot can also determine the chassis state information based on the aforementioned position and orientation information. That is, the chassis state information includes the chassis's position and orientation information, allowing the robot to determine the corresponding kinematic model based on the chassis state information (including position and orientation information), the first virtual wheel information, and the second virtual wheel information.

[0034] In this design, the speeds and angles of the first and second virtual wheel information are independent; that is, the speeds of the first and second virtual wheel information are unrelated, and the angles of the first and second virtual wheel information are also unrelated, thus enabling more complex motions. Furthermore, the perpendicular lines corresponding to the wheels of the first wheel group point to the first center of the circle when turning, and the wheels of the second wheel group point to the second center of the circle when turning. Specifically, when the wheels of the first wheel group turn, the perpendicular lines of both wheels point to the same center of the circle, and the wheels of the second wheel group turn. The first and second centers can be the same or different centers, thus ensuring that both the first and second wheel groups satisfy the Ackermann steering pattern.

[0035] Specifically, the structure corresponding to the above kinematic model can be as follows: Figure 2 As shown, Figure 2 This is a schematic diagram of the kinematic model in one embodiment. The chassis can be a four-wheeled, four-rotor chassis. The robot simplifies the chassis to a bicycle model that can steer in both the front and rear. The two front wheels and two rear wheels of the chassis follow Ackerman steering, but there is no restriction that the rotation angles of the front and rear virtual center wheels must be the same in magnitude and opposite in direction. This allows the chassis to not only include oblique traverse motion modes but also greatly increases its motion characteristics, enabling more complex movements.

[0036] Assuming the four-wheel, four-rotation chassis is rigidly connected and wheel deformation is negligible, the chassis primarily travels on level ground. The chassis control point is the geometric center point M of the four wheels. In motion programming, point M represents the chassis, and the state variables are (x, y, θ, v). f ,v r ,δ f ,δ r This mainly includes chassis position information (x, y), chassis orientation information θ, and other chassis state information. It also includes information on the first and second virtual wheels. The first virtual wheel information can be information on the virtual front wheels, and the second virtual wheel information can be information on the virtual rear wheels, specifically including the virtual front wheel speed v. f Virtual rear wheel speed v r Virtual front wheel steering angle δ f Virtual rear wheel steering angle δ rThe control amount of the chassis is (a) f ,a r ,δ f ',δ r '), mainly including virtual front wheel acceleration a f Virtual front wheel steering speed δ f 'And the first virtual wheel information, and the virtual rear wheel acceleration a' r Virtual rear wheel steering speed δ r 'Wait for the second virtual wheel information. The kinematic model is shown below:'

[0037] ;

[0038] Where β represents the sideslip angle of control point M, v m Let a be the velocity of control point M. m Let x represent the acceleration at control point M, Δs represent the path length of the chassis between two adjacent control points M, Δθ represent the change in orientation of the chassis between two adjacent control points M, Δt represent the time taken by the chassis between two adjacent control points, Δβ represent the change in sideslip angle between two adjacent control points M, t represent time point t, and Δt represents the change in time between two adjacent control points M. t y t This indicates the position of control point M at time point t, v f,t v represents the virtual front wheel speed at time t. r,t δ represents the virtual rear wheel speed at time t. f,t δ represents the virtual front wheel steering angle at time t. r,t θ represents the virtual rear wheel steering angle at time t. t β represents the orientation of the chassis at time t. t Let M represent the sideslip angle of control point M at time t, L represent the length of the chassis, and W represent the width of the chassis. Here, two adjacent points represent two adjacent trajectory points in the motion planning.

[0039] Since the chassis is a four-wheel, four-rotation chassis, the first virtual wheel information corresponds to the two front wheels of the chassis, i.e., the first wheel set, and the second virtual wheel information corresponds to the two rear wheels of the chassis, i.e., the second wheel set. The robot needs to simplify the four-wheel, four-rotation chassis into a bicycle model that can steer in both the front and rear. When constructing the optimization problem to solve the planned trajectory, in order to consider the smoothness of the planned trajectory and the current state of the chassis, it is necessary to derive the steering drive state from the four-wheel steering drive state to the steering drive state of the front and rear virtual middle wheels, thereby converting the four wheels into the first virtual wheel information and the second virtual wheel information. The specific formula is as follows:

[0040]

[0041] Where, δ flIndicates the steering angle of the left front wheel, v fl Expressed as the left front wheel drive speed, δ fr This represents the steering angle of the right front wheel, v. fr Expressed as the right front wheel drive speed, δ rl This represents the steering angle of the left rear wheel, v. rl Represented as the left rear wheel drive speed, δ rr This represents the steering angle of the right rear wheel, v. rr This represents the speed of the right rear wheel drive.

[0042] Step S204: Obtain the motion constraint information of the chassis.

[0043] When controlling the chassis to move, due to obstacles in the environment and the chassis's own constraints, motion planning for the chassis requires combining various motion constraint information. The robot can then acquire the chassis's motion constraint information. This motion constraint information includes, but is not limited to, constraints on the movement of the wheels, constraints on the chassis's kinematic model, and constraints on the trajectory limited by obstacles. The robot can construct the chassis's motion constraints using a cost function.

[0044] Step S206: Perform motion planning for the chassis based on the kinematic model and motion constraint information.

[0045] After obtaining the kinematic model and motion constraint information, the robot can perform motion planning for the chassis based on these information. This planning process can be completed in multiple steps. For example, the robot can first obtain an initial path solution to get an initial trajectory, and then optimize the trajectory by combining the initial solution, kinematic model, and motion constraint information to obtain a smoother trajectory.

[0046] The kinematic model contains state variables that are virtual wheel information of each virtual wheel. When solving the trajectory planning, in order to ensure the smoothness of the trajectory and the smooth execution of commands, the state described by the first point of the trajectory is the current state of the robot. Therefore, it is necessary to convert the real-time state of the robot into the state of virtual wheel information of each virtual wheel and assign it to the first trajectory point to realize the determination of the trajectory point.

[0047] In the aforementioned motion planning method, a kinematic model of the chassis is generated based on the chassis state information and the information of each virtual wheel corresponding to each wheel set, with the speed and angle of each virtual wheel being independent. The motion constraint information of the chassis is then obtained, and motion planning is performed on the chassis based on the kinematic model and the motion constraint information. Compared to the traditional method of switching between dual Ackerman steering mode and swerve mode for motion planning, this solution removes the restriction that each wheel in the chassis must rotate at the same angle but in opposite directions. It constructs multiple virtual wheel information sets, combines the chassis state information and the virtual wheel information to generate a kinematic model, and then uses this model and the motion constraint information to perform motion planning on the chassis, thus improving the chassis's operating efficiency.

[0048] In one embodiment, obtaining motion constraint information of the chassis includes: generating first motion constraint information based on speed and steering limit information corresponding to the first and second wheel sets of the chassis; and / or generating second motion constraint information based on kinematic model constraints corresponding to the chassis; and / or generating third motion constraint information based on trajectory limit information; and obtaining motion constraint information based on the first motion constraint information, the second motion constraint information, and / or the third motion constraint information.

[0049] In this embodiment, during the motion planning process of the chassis, the robot can optimize the chassis's motion trajectory. The robot can combine multiple motion constraint information to plan the chassis's motion. These motion constraint information can include various types, such as first, second, and third motion constraint information. The robot can generate the first motion constraint information based on the speed and steering limitations corresponding to the first and second wheel sets of the chassis. This first motion constraint information indicates that the speed and steering of the first and second wheel sets of the chassis are subject to corresponding limitations. The robot can also generate the second motion constraint information based on the kinematic model constraints corresponding to the chassis. This second motion constraint information indicates that due to the chassis's structural volume, its own limitations need to be considered during motion planning. The robot can also generate the third motion constraint information based on trajectory constraint information. This means that when planning the trajectory, the planned trajectory should be sufficiently smooth and continuous, and have a start and end point; therefore, the robot needs to consider trajectory constraint information to form the corresponding motion constraint information.

[0050] Thus, the robot can obtain motion constraint information based on the first motion constraint information, the second motion constraint information, and / or the third motion constraint information, and optimize the trajectory when performing motion planning by combining the above-mentioned motion constraint information.

[0051] Specifically, when the robot performs path planning, it can first plan an initial solution to obtain initial path information. Each trajectory point in the initial path information contains position information in two-dimensional space. The expression for the robot to construct the aforementioned chassis trajectory optimization is as follows: f(B) = ∑ k (γ k f k (B)), Among them, f k (B) is the k-th cost function, B * It is the optimized optimal solution, i.e., the optimal path or state in motion planning; that is, the robot can define the optimization problem as a least squares problem. To better represent the motion constraint information contained in the least squares problem, a cost function corresponding to the motion constraint information can be constructed. The cost function is specifically expressed as: e τ (x,x r ,ε,S,n) asymptotically equal to {((x-(x)} r -ε)) / S) n if x > x r -ε; 0, otherwise}. Where x represents the parameter to be optimized, x r The parameters represent the threshold values ​​that need to be referenced, ε represents the tolerance parameter, S represents the scaling factor used to adjust the magnitude of the error, and n represents the exponential parameter used to control the shape of the error function. The first set of motion constraints includes one or more of the following: limitations on the driving speed of the four wheels, limitations on the driving acceleration of the four wheels, physical limits on the steering of the four wheels, limits on the maximum steering speed of the four wheels, and limitations on the matching degree between the steering and speed of the four wheels. The second set of motion constraints includes constraints on the kinematic model. The third set of motion constraints includes one or more of the following: trajectory points must not collide with obstacles and must be as close as possible to the global path; the time and distance between adjacent trajectory points must be minimized; and limitations on the start and end points of the trajectory. Each of the above motion constraints can be represented by a cost function.

[0052] In the optimization process, a maximum speed limit for the four wheels is considered. A cost function is added for each trajectory point. The cost function for each wheel can be specifically expressed as: f v(i,fl) =e τ (v i,fl ,v max ,ε,S,n)+e τ (-v i,fl ,-v min ,ε,S,n);f v(i,fr) =e τ (v i,fr ,v max ,ε,S,n)+e τ (-v i,fr,-v min ,ε,S,n);f v(i,rl) =e τ (v i,rl ,v max ,ε,S,n)+e τ (-v i,rl ,-v min ,ε,S,n);f v(i,rr) =e τ (v i,rr ,v max ,ε,S,n)+e τ (-v i,rr ,-v min ,ε,S,n). Where, v i,fl v represents the velocity of the left front wheel at point i. i,fr v represents the velocity of the right front wheel at point i. i,rl v represents the velocity of the left rear wheel at point i. i,rr v represents the velocity of the right rear wheel at point i. max v represents the maximum speed. min This represents the minimum speed.

[0053] There is a maximum limit to the driving acceleration of the four wheels. For each trajectory point, the robot adds a cost function as follows: f a(i,fl) =e τ (a i,fl ,a max ,ε,S,n)+e τ (-a i,fl ,-a min ,ε,S,n);f a(i,fr) =e τ (a i,fr ,a max ,ε,S,n)+e τ (-a i,fr ,-a min ,ε,S,n);f a(i,rl) =e τ (a i,rl ,a max ,ε,S,n)+e τ (-a i,rl ,-a min ,ε,S,n);f a(i,rr) =e τ (a i,rr ,a max ,ε,S,n)+e τ (-a i,rr ,-a min ,ε,S,n). Where, a i,fl Let a represent the acceleration of the left front wheel at point i.i,fr Let a represent the acceleration of the right front wheel at point i. i,rl Let a represent the velocity of the left rear wheel at point i. i,rr Let a represent the acceleration of the right rear wheel at point i. max a represents the maximum acceleration. min This represents the minimum acceleration.

[0054] During trajectory optimization, there is a maximum limit to the steering speed of the four wheels. Therefore, for each trajectory point, the following cost function can be added: f δ'(i,fl) =e τ (δ ' i,fl ,δ ' max ,ε,S,n)+e τ (-δ ' i,fl ,-δ ' min ,ε,S,n);f δ'(i,fr) =e τ (δ ' i,fr ,δ ' max ,ε,S,n)+e τ (-δ ' i,fr ,-δ ' min ,ε,S,n);f δ'(i,rl) =e τ (δ ' i,rl ,δ ' max ,ε,S,n)+e τ (-δ ' i,rl ,-δ ' min ,ε,S,n);f δ'(i,rr) =e τ (δ ' i,rr ,δ ' max ,ε,S,n)+e τ (-δ ' i,rr ,-δ ' min ,ε,S,n). Where, δ ' i,fl δ represents the steering speed of the left front wheel at point i. ' i,fr δ represents the steering speed of the right front wheel at point i. ' i,rlδ represents the velocity of the left rear wheel at point i. ' i,rr δ represents the steering speed of the right rear wheel at point i. ' max δ represents the maximum steering speed. ' min This indicates the minimum steering speed.

[0055] During the optimization process, the mismatch between four-wheel steering and speed is considered to affect the stability of the chassis. Therefore, the robot can add the following cost function for each trajectory point: f stab,i =(v f,i cos(δ f,i )-v r,i cos(δ r,i )) 2 Among them, f stab,i Let v represent a stable function. f,i v represents the speed of the virtual front wheel at point i. r,i δ represents the speed of the virtual rear wheel at point i. f,i δ represents the rotation angle of the virtual front wheel at point i. r,i This represents the rotation angle of the virtual rear wheel at point i.

[0056] The optimization process also includes kinematic model constraints between adjacent trajectory points. Therefore, the robot adds the following cost function to adjacent trajectory points:

[0057]

[0058] Among them, X i Represents the kinematic model of point i.

[0059] Regarding trajectory constraints, during optimization, it is considered that trajectory points cannot collide with obstacles and should be as close as possible to the global path. Therefore, the robot adds the following cost function for each trajectory point: f path,ij =e τ (d min,ij ,r pmax ,ε,S,n);f obs,ij =e τ (-d min,ij ,-r pmax ,ε,S,n). Where, f path,ij Let f represent the trajectory stabilization function. obs,ij Describes the obstacle function, d min,ij r represents the minimum distance of the trajectory from point i to point j. pmax Indicates the maximum radius of influence of the obstacle.

[0060] During optimization, the robot also needs to minimize the time and distance between adjacent trajectory points. Therefore, the following cost function is added to each pair of adjacent trajectory points: f t,i =Δt i 2 ;f s,i =Δs i 2 Among them, f t,i f represents the time function for moving to point i. s,i Let Δt represent the distance function to move to point i. i Δs represents the time taken to move to point i. i This represents the distance traveled to point i.

[0061] In addition, path planning also needs to consider the starting and ending points of the trajectory. The starting and ending points are constants and not optimization variables, but rather serve as motion constraint information limiting the trajectory. Specifically, they can be represented as: X0=[x init ,y init ,θ init ,v f,init ,v r,init ,v r,init ,δ f,init ,δ r,init ], X n =[x end ,y end ,θ end ,v f,end ,v r,end ,v r,end ,δ f,end ,δ r,end ]. Where, x init ,y init θ represents the initial position of the chassis as represented by the kinematic model. init Indicates the initial orientation of the chassis, v f,init The initial velocity of the virtual front wheel, v r,init δ represents the initial velocity of the virtual rear wheel. f,init δ represents the initial steering angle of the virtual front wheel. r,init Indicates the initial steering angle of the virtual front wheel; x end ,y end θ represents the position of the chassis represented by the kinematic model when it reaches the endpoint. end Indicates the orientation of the chassis when it reaches its destination, v f,end v represents the speed of the virtual front wheel when it reaches the destination. r,end δ represents the speed of the virtual rear wheel when it reaches the finish line. f,end δ represents the steering angle of the virtual front wheel when it reaches the endpoint. r,end This indicates the steering angle of the virtual front wheel when it reaches the endpoint.

[0062] In this embodiment, the robot constructs cost functions corresponding to the speed and steering constraints, kinematic model constraints, and trajectory constraints of the chassis, and uses the motion constraint information to plan the motion of the chassis, thereby improving the rationality of the motion planning and the effectiveness of the planned trajectory.

[0063] In one embodiment, motion planning for the chassis is performed based on a kinematic model and motion constraint information, including: obtaining initial trajectory point information; optimizing the chassis trajectory based on motion constraint information, initial trajectory point information, and kinematic model to obtain target trajectory point information; and obtaining the corresponding motion planning result for the chassis based on the target trajectory point information.

[0064] In this embodiment, the robot can perform motion planning for the chassis by combining kinematic models and motion constraint information. This planning includes initial planning and trajectory optimization. The robot can first acquire initial trajectory point information, such as generating initial trajectory point information based on the start point, end point, and obstacle information in between. The robot can also optimize the initial trajectory point information. For example, based on motion constraint information, initial trajectory point information, and the kinematic model, the robot optimizes the chassis trajectory to obtain target trajectory point information, and based on the target trajectory point information, obtains the corresponding motion planning result for the chassis. The aforementioned motion constraint information includes first motion constraint information, second motion constraint information, and third motion constraint information; the first motion constraint information represents the speed and steering restrictions on the first and second wheel sets; the second motion constraint information represents the restrictions on the kinematic model corresponding to the chassis; and the third motion constraint information represents the restrictions on the trajectory. The robot can combine the above-mentioned motion constraint information to perform the trajectory optimization process.

[0065] In one embodiment, the chassis trajectory is optimized based on motion constraint information, initial trajectory point information, and a kinematic model to obtain target trajectory point information. This includes: determining the first wheel motion information and second wheel motion information of the first wheel group and the second wheel group of the chassis corresponding to the initial trajectory point information, based on the first virtual wheel information and the second virtual wheel information corresponding to the kinematic model; determining optimized reference wheel motion information based on the first motion constraint information, the first wheel motion information, and the second wheel motion information; determining optimized reference chassis state information based on the second motion constraint information and the chassis state information corresponding to the kinematic model; determining optimized reference trajectory point information based on the third motion constraint information and the initial trajectory point information; and adjusting the initial trajectory point information based on the reference wheel motion information, the reference chassis state information, and the reference trajectory point information until the weighted sum of the reference wheel motion information, the reference chassis state information, and the reference trajectory point information is minimized, thus obtaining the target trajectory point information.

[0066] In this embodiment, the robot can optimize the motion of the chassis at different angles based on different motion constraint information. Specifically, the robot can determine the motion information of the first and second wheels of the chassis at the initial trajectory points based on the first and second virtual wheel information corresponding to the kinematic model; that is, the robot can convert the virtual wheel information into the motion information of the wheels corresponding to the wheels of the wheelset. Thus, the robot can determine the optimized reference wheel motion information based on the first motion constraint information, the first wheel motion information, and the second wheel motion information. The reference wheel motion information represents the speed and angle of the wheel obtained after optimization according to the motion constraint information.

[0067] The robot can also determine optimized reference chassis state information based on the second motion constraint information and the chassis state information corresponding to the kinematic model. The reference chassis state information refers to the position and orientation of the chassis obtained after optimization according to the motion constraint information.

[0068] The robot can also determine optimized reference trajectory point information based on the third motion constraint information and the initial trajectory point information. The reference trajectory point information represents the trajectory point information obtained after optimization according to the motion constraint information. Therefore, the robot can adjust the initial trajectory point information based on the reference wheel motion information, reference chassis state information, and reference trajectory point information until the weighted sum of these three information is minimized, thus obtaining the target trajectory point information. In other words, the robot can combine each cost function f... k Find the path B that minimizes f(B) to obtain the final target trajectory point information, and then determine the optimal path B based on the target trajectory point information. * .

[0069] Specifically, during motion planning, obstacle information around the chassis is represented using a grid map, and initial trajectory point information is determined. The robot uses the A* algorithm to perform a global search on the created grid map to obtain initial solution information consisting of a sequence of points from the starting point to the ending point, which serves as the initial trajectory point information. Each trajectory point in the initial trajectory point information contains its position information in two-dimensional space. Based on these premises and the cost functions corresponding to the various motion constraints, the robot can solve the optimization problem, thereby obtaining a smooth planned trajectory for the four-wheel, four-rotation chassis from the starting point to the ending point, which serves as the target trajectory point information.

[0070] Through the above embodiments, the robot performs initial trajectory planning on the chassis, constructs speed and steering constraints, kinematic model constraints, and trajectory constraints for the chassis, and constructs cost functions corresponding to the motion constraint information for each. The initial trajectory of the chassis is optimized using the motion constraint information, which improves the rationality and smoothness of the planned trajectory point information, thereby improving the robot's travel efficiency.

[0071] In one embodiment, based on the first and second virtual wheel information corresponding to the kinematic model, the motion information of the first wheel group and the second wheel group of the chassis at the initial trajectory point information are determined, respectively. This includes: determining the first wheel rotation angle corresponding to the first wheel group and the second wheel rotation angle corresponding to the second wheel group based on the first virtual wheel rotation angle corresponding to the first virtual wheel information at the initial trajectory point information and the second virtual wheel rotation angle corresponding to the second virtual wheel information; determining the first wheel speed corresponding to the first wheel group based on the first virtual wheel rotation angle and the first wheel rotation angle; determining the second wheel speed corresponding to the second wheel group based on the second virtual wheel rotation angle and the second wheel rotation angle; determining the first wheel motion information of the first wheel group at the initial trajectory point information based on the first wheel rotation angle and the first wheel speed; and determining the second wheel motion information of the second wheel group at the initial trajectory point information based on the second wheel rotation angle and the second wheel speed.

[0072] In this embodiment, during planning, it is necessary to construct first and second virtual wheel information corresponding to the chassis. After optimization, it is necessary to determine the control commands for each wheel. Therefore, the robot needs to convert the virtual wheel information into the corresponding wheel motion information. Specifically, the robot can determine the first wheel rotation angle corresponding to the first wheel group and the second wheel rotation angle corresponding to the second wheel group based on the first virtual wheel rotation angle corresponding to the initial trajectory point information and the second virtual wheel rotation angle corresponding to the first virtual wheel information. That is, the robot determines the first wheel rotation angle corresponding to the first wheel group based on the first virtual wheel rotation angle corresponding to the initial trajectory point information, where the first wheel group can be the front wheel group of the chassis; the robot determines the second wheel rotation angle corresponding to the second wheel group based on the second virtual wheel rotation angle corresponding to the initial trajectory point information, where the second wheel group can be the rear wheel group of the chassis.

[0073] The robot can also determine the speed of the first wheel corresponding to the first wheel group based on the rotation angle of the first virtual wheel and the rotation angle of the first wheel; and determine the speed of the second wheel corresponding to the second wheel group based on the rotation angle of the second virtual wheel and the rotation angle of the second wheel.

[0074] Thus, the robot can determine the motion information of the first wheel assembly at the initial trajectory point based on the rotation angle and speed of the first wheel. Similarly, it can determine the motion information of the second wheel assembly at the initial trajectory point based on the rotation angle and speed of the second wheel. At each trajectory point, the robot can use the above method to convert virtual wheel information into motion information for all four wheels.

[0075] Specifically, after the robot solves the optimization problem, it selects a trajectory point close to the current position of the chassis from the planned trajectory. Based on the kinematic model state of the trajectory point, it calculates the four-wheel four-turn chassis control command. Thus, the robot needs to transition from the steering drive state of the front and rear virtual wheels to the four-wheel steering drive state, that is, to realize the conversion of virtual wheel information to the motion information of the four wheels. This can be specifically represented as:

[0076]

[0077] Among them, v f Represents the virtual front wheel speed, v r Represents the virtual rear wheel speed, δ f Represents the virtual front wheel steering angle, δ r δ represents the virtual rear wheel steering angle. fl Indicates the steering angle of the left front wheel, v fl Expressed as the left front wheel drive speed, δ fr This represents the steering angle of the right front wheel, v. fr Expressed as the right front wheel drive speed, δ rl This represents the steering angle of the left rear wheel, v. rl Represented as the left rear wheel drive speed, δ rr This represents the steering angle of the right rear wheel, v. rr This represents the speed of the right rear wheel drive.

[0078] Through this embodiment, the robot can convert the motion information of the corresponding wheelset into the motion information of the virtual wheels based on the relevant information of the virtual wheels, thereby controlling the chassis motion with control commands for the actual wheel motion, which improves the operating efficiency of the chassis during motion planning.

[0079] In related technologies, the chassis achieves motion by switching between two operating modes. However, at the moment of switching between the two modes, the chassis inevitably experiences a situation where its speed is zero. Furthermore, because the above method does not completely simplify the model, it eliminates most of the chassis's motion characteristics, failing to utilize the chassis's full motion capabilities, thus increasing the energy loss of the planned trajectory.

[0080] In one exemplary embodiment, another motion planning method is provided. In this embodiment, to simplify the kinematic model of a four-wheel, four-turn chassis, the simplification method for the four-wheel, four-turn chassis is similar to the double Ackerman steering mode. The robot simplifies the chassis to a bicycle model that can steer in both the front and rear, such as... Figure 2 As shown, the two front wheels and two rear wheels each follow Ackerman steering, but there is no restriction that the rotation angles of the front and rear virtual center wheels must be the same in magnitude and opposite in direction. Compared with the dual Ackerman steering mode, this simplified method not only includes the slant motion mode, but also greatly increases the chassis's motion characteristics, enabling more complex motions. However, at the same time, the complexity of the kinematic model is also significantly increased.

[0081] Assuming the four-wheel, four-rotation chassis is rigidly connected and wheel deformation is negligible, the chassis primarily travels on level ground. The chassis control point is the geometric center point M of the four wheels. In motion programming, point M represents the chassis, and the state variables are (x, y, θ, v). f ,v r ,δ f ,δ r This mainly includes chassis position information (x, y), chassis orientation information θ, and other chassis state information. It also includes information on the first and second virtual wheels. The first virtual wheel information can be information on the virtual front wheels, and the second virtual wheel information can be information on the virtual rear wheels, specifically including the virtual front wheel speed v. f Virtual rear wheel speed v r Virtual front wheel steering angle δ f Virtual rear wheel steering angle δ r The control amount of the chassis is (a) f ,a r ,δ f ',δ r '), mainly including virtual front wheel acceleration a f Virtual front wheel steering speed δ f 'And the first virtual wheel information, and the virtual rear wheel acceleration a' r Virtual rear wheel steering speed δ r 'Wait for the second virtual wheel information.'

[0082] Furthermore, the robot can also switch from a four-wheel steering drive state to a front and rear virtual wheel information drive state. Because the four-wheel, four-turn chassis was simplified to a bicycle model where both wheels can turn in the front and rear during the chassis kinematic model construction, trajectory optimization requires deriving a transition from a four-wheel steering drive state to a front and rear virtual wheel steering drive state, thus converting the four wheels into first and second virtual wheel information, to ensure the smoothness of the planned trajectory and the current chassis state. After completing the optimization, the robot needs to actually control the movement of each wheel. Therefore, the robot selects a trajectory point close to the current chassis position from the planned trajectory and calculates the four-wheel, four-turn chassis control command based on the kinematic model state at that point. This requires the robot to switch from a front and rear virtual wheel steering drive state to a four-wheel steering drive state, i.e., converting virtual wheel information into four-wheel motion information.

[0083] In the optimization process, it is necessary to construct various cost functions corresponding to the motion constraints. The obstacle information around the chassis is represented in the form of a grid map, and the initial trajectory point information is determined. Specifically, the robot can use the A* algorithm to perform a global search on the created grid map to obtain initial solution information consisting of a sequence of points from the starting point to the ending point, which serves as the initial trajectory point information. Each trajectory point in the initial trajectory point information contains its position information in two-dimensional space. Therefore, based on the above premises and the various cost functions corresponding to the aforementioned motion constraints, the robot can solve the optimization problem, thereby obtaining smooth planned trajectory information for the four-wheel, four-rotation chassis from the starting point to the ending point, which serves as the target trajectory point information.

[0084] The chassis trajectory optimization for four-wheel, four-turn operation can be represented by a least squares problem. To better describe the motion constraint information contained in the least squares problem, a cost function corresponding to the motion constraint information can be constructed. The cost function is specifically expressed as: e τ (x,x r ,ε,S,n).

[0085] Specifically, during optimization, a maximum limit is considered for the driving speed of the four wheels, and a cost function is added for each trajectory point; a maximum limit is also considered for the driving acceleration of the four wheels, and a cost function is added for each trajectory point; a maximum limit exists for the steering speed of the four wheels during trajectory optimization, so a cost function can be added for each trajectory point; the mismatch between the steering and speed of the four wheels is considered to affect the stability of the chassis driving, so the robot can add a cost function for each trajectory point; the optimization process also includes kinematic model constraints between adjacent trajectory points, so the robot adds a cost function for adjacent trajectory points.

[0086] Regarding trajectory constraints, during optimization, the robot considers that trajectory points must not collide with obstacles and should be as close as possible to the global path. Therefore, a cost function is added for each trajectory point. During optimization, the robot also needs to minimize the time and distance between adjacent trajectory points, so a cost function is added for every two adjacent trajectory points. Furthermore, the starting and ending points of the trajectory must be considered in path planning. These starting and ending points are constants and not optimization variables, but rather serve as motion constraint information for trajectory constraints. Thus, by solving the optimization problem constructed above, the robot can obtain smooth planned trajectory information for the four-wheel, four-rotor chassis from the starting point to the ending point, which serves as the target trajectory point information.

[0087] Through the above embodiments, the robot removes the restriction that each wheel in the chassis must rotate at the same angle but in opposite directions, constructs multiple virtual wheel information, and generates a kinematic model by combining chassis state information and virtual wheel information. By combining the chassis's kinematic model and motion constraint information, the robot performs motion planning for the chassis, thereby improving the chassis's operating efficiency.

[0088] Furthermore, a complete kinematic model can more comprehensively and accurately describe robot motion, including the steering angle and drive speed of each wheel, as well as the overall linear and angular velocities of the chassis, thus improving the accuracy and stability of motion planning. Secondly, it exhibits stronger smoothness and adaptability, dynamically adjusting motion strategies according to different working environments and task requirements without relying on preset mode switching. This reduces potential abrupt changes and discontinuities during mode switching, as well as energy loss in the planned trajectory, enabling the chassis to better cope with complex and ever-changing environments and tasks. Motion planning methods based on complete kinematic models have significant advantages in accuracy, smoothness, and adaptability, making them particularly suitable for applications requiring high motion accuracy and stability.

[0089] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0090] Based on the same inventive concept, this application also provides a motion planning device for implementing the motion planning method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more motion planning device embodiments provided below can be found in the limitations of the motion planning method described above, and will not be repeated here.

[0091] In one exemplary embodiment, such as Figure 3 As shown, a motion planning device is provided, including: a first acquisition module 500, a second acquisition module 502, and a planning module 504, wherein:

[0092] The first acquisition module 500 is used to acquire the kinematic model corresponding to the chassis. The kinematic model is generated based on the chassis state information, the first virtual wheel information, and the second virtual wheel information. The speed and angle of the first virtual wheel information and the second virtual wheel information are not related. The first virtual wheel information corresponds to the first wheel group of the chassis, and the second virtual wheel information corresponds to the second wheel group of the chassis. The perpendicular line corresponding to the wheel of the first wheel group points to the first center when turning. The perpendicular line corresponding to the wheel of the second wheel group points to the second center when turning.

[0093] The second acquisition module 502 is used to acquire the motion constraint information of the chassis.

[0094] Planning module 504 is used to perform motion planning for the chassis based on the kinematic model and motion constraint information.

[0095] In one embodiment, the first acquisition module 500 is configured to generate corresponding first virtual wheel information and second virtual wheel information based on the motion information of each wheel corresponding to the first wheel group and the second wheel group of the chassis; obtain chassis state information based on the chassis position information and orientation information; and determine the corresponding kinematic model based on the chassis state information including position information and orientation information, the first virtual wheel information and the second virtual wheel information.

[0096] In one embodiment, the second acquisition module 502 is configured to generate first motion constraint information based on the speed and steering limit information corresponding to the first and second wheel sets of the chassis; and / or generate second motion constraint information based on the kinematic model limit of the chassis; and / or generate third motion constraint information based on the trajectory limit information; and obtain motion constraint information based on the first motion constraint information, the second motion constraint information, and / or the third motion constraint information.

[0097] In one embodiment, the second acquisition module 502 is used to acquire initial trajectory point information; optimize the chassis trajectory based on motion constraint information, initial trajectory point information and kinematic model to obtain target trajectory point information; and obtain the motion planning result corresponding to the chassis based on the target trajectory point information.

[0098] In one embodiment, the second acquisition module 502 is configured to: determine the first wheel motion information and the second wheel motion information of the first wheel group and the second wheel group of the chassis respectively at the initial trajectory point information based on the first virtual wheel information and the second virtual wheel information corresponding to the kinematic model; determine the optimized reference wheel motion information based on the first motion constraint information, the first wheel motion information and the second wheel motion information; determine the optimized reference chassis state information based on the second motion constraint information and the chassis state information corresponding to the kinematic model; determine the optimized reference trajectory point information based on the third motion constraint information and the initial trajectory point information; and adjust the initial trajectory point information based on the reference wheel motion information, the reference chassis state information and the reference trajectory point information until the weighted sum of the reference wheel motion information, the reference chassis state information and the reference trajectory point information is minimized, thereby obtaining the target trajectory point information.

[0099] In one embodiment, the second acquisition module 502 is configured to: determine the first wheel rotation angle corresponding to the first wheel group and the second wheel rotation angle corresponding to the second wheel group based on the first virtual wheel rotation angle corresponding to the first virtual wheel information at the initial trajectory point information and the second virtual wheel rotation angle corresponding to the second virtual wheel information; determine the first wheel speed corresponding to the first wheel group based on the first virtual wheel rotation angle and the first wheel rotation angle; determine the second wheel speed corresponding to the second wheel group based on the second virtual wheel rotation angle and the second wheel rotation angle; determine the first wheel motion information of the first wheel group at the initial trajectory point information based on the first wheel rotation angle and the first wheel speed; and determine the second wheel motion information of the second wheel group at the initial trajectory point information based on the second wheel rotation angle and the second wheel speed.

[0100] Each module in the aforementioned motion planning device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the robot's processor in hardware form or independent of it, or stored in the robot's memory in software form, so that the processor can call and execute the operations corresponding to each module.

[0101] In one exemplary embodiment, a robot is provided, which may be an autonomous mobile robot, and its internal structure diagram may be as follows: Figure 4As shown, the robot includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The robot's processor provides computational and control capabilities. The robot's memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The robot's input / output interface is used for exchanging information between the processor and external devices. The robot's communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When the computer program is executed by the processor, it implements a motion planning method. The robot's display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The robot's input device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the robot's shell, or external keyboards, touchpads, or mice, etc.

[0102] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the robot to which the present application is applied. A specific robot may include more or fewer parts than shown in the figure, or combine certain parts, or have different part arrangements.

[0103] In one exemplary embodiment, a robot 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 motion planning method described above.

[0104] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the motion planning method described above.

[0105] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the motion planning method described above.

[0106] 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.

[0107] Those skilled in the art will understand that all or part of the processes in the methods of 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, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory 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). 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, artificial intelligence (AI) processors, etc., and are not limited to these.

[0108] 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 application.

[0109] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A motion planning method, characterized in that, The method includes: Obtain the kinematic model corresponding to the chassis; the kinematic model is generated based on the chassis state information, first virtual wheel information, and second virtual wheel information; the first virtual wheel information and the second virtual wheel information are generated based on the motion information of each wheel corresponding to the first wheel group and the second wheel group of the chassis, respectively; the chassis state information includes the position information and orientation information corresponding to the chassis; the speed and angle of the first virtual wheel information and the second virtual wheel information are not related; the first virtual wheel information corresponds to the first wheel group of the chassis, and the second virtual wheel information corresponds to the second wheel group of the chassis; the perpendicular line corresponding to the wheel of the first wheel group points to the first center when turning; the perpendicular line corresponding to the wheel of the second wheel group points to the second center when turning. The motion constraint information of the chassis is obtained based on the first motion constraint information, the second motion constraint information, and / or the third motion constraint information; the first motion constraint information is generated based on the speed and steering limit information corresponding to the first and second wheel sets of the chassis; the second motion constraint information is generated based on the kinematic model constraints corresponding to the chassis; and the third motion constraint information is generated based on the trajectory limit information. Based on the kinematic model, the motion constraint information, and the initial trajectory point information, the chassis trajectory is optimized to obtain the motion planning result corresponding to the chassis.

2. The method according to claim 1, characterized in that, The step of optimizing the chassis trajectory based on the kinematic model, the motion constraint information, and the initial trajectory point information to obtain the motion planning result corresponding to the chassis includes: Based on the motion constraint information, the initial trajectory point information, and the kinematic model, the chassis trajectory is optimized to obtain the target trajectory point information; Based on the target trajectory point information, the motion planning result corresponding to the chassis is obtained.

3. The method according to claim 2, characterized in that, The motion constraint information includes first motion constraint information, second motion constraint information, and third motion constraint information; the first motion constraint information represents the speed and steering restrictions on the first and second wheel sets; the second motion constraint information represents the kinematic model restrictions corresponding to the chassis; and the third motion constraint information represents the trajectory restrictions. The step of optimizing the chassis trajectory based on the motion constraint information, the initial trajectory point information, and the kinematic model to obtain target trajectory point information includes: Based on the first virtual wheel information and the second virtual wheel information corresponding to the kinematic model, the first wheel motion information and the second wheel motion information of the first wheel group and the second wheel group of the chassis corresponding to the initial trajectory point information are determined respectively. Based on the first motion constraint information, the first wheel motion information, and the second wheel motion information, the optimized reference wheel motion information is determined. Based on the second motion constraint information and the chassis state information corresponding to the kinematic model, the optimized reference chassis state information is determined; Based on the third motion constraint information and the initial trajectory point information, the optimized reference trajectory point information is determined; Based on the reference wheel motion information, the reference chassis state information, and the reference trajectory point information, the initial trajectory point information is adjusted until the weighted sum of the reference wheel motion information, the reference chassis state information, and the reference trajectory point information is minimized, thus obtaining the target trajectory point information.

4. The method according to claim 3, characterized in that, The step of determining the first wheel motion information and the second wheel motion information of the chassis's first wheel group and second wheel group corresponding to the initial trajectory point information based on the first virtual wheel information and the second virtual wheel information corresponding to the kinematic model includes: Based on the first virtual wheel rotation angle corresponding to the initial trajectory point information and the second virtual wheel rotation angle corresponding to the second virtual wheel information, determine the first wheel rotation angle corresponding to the first wheel set and the second wheel rotation angle corresponding to the second wheel set. Based on the rotation angle of the first virtual wheel and the rotation angle of the first wheel, determine the speed of the first wheel corresponding to the first wheel set; Based on the rotation angle of the second virtual wheel and the rotation angle of the second wheel, determine the speed of the second wheel corresponding to the second wheel set; Based on the rotation angle of the first wheel and the speed of the first wheel, determine the motion information of the first wheel of the first wheel assembly corresponding to the initial trajectory point information; Based on the rotation angle and speed of the second wheel, determine the motion information of the second wheel corresponding to the initial trajectory point information of the second wheel set.

5. A motion planning device, characterized in that, The device includes: The first acquisition module is used to acquire the kinematic model corresponding to the chassis. The kinematic model is generated based on the chassis state information, first virtual wheel information, and second virtual wheel information. The first virtual wheel information and the second virtual wheel information are generated based on the motion information of each wheel corresponding to the first wheel group and the second wheel group of the chassis, respectively. The chassis state information includes the position information and orientation information corresponding to the chassis. The speed and angle of the first virtual wheel information and the second virtual wheel information are not related. The first virtual wheel information corresponds to the first wheel group of the chassis, and the second virtual wheel information corresponds to the second wheel group of the chassis. The perpendicular line corresponding to the wheel of the first wheel group points to the first center when turning. The perpendicular line corresponding to the wheel of the second wheel group points to the second center when turning. The second acquisition module is used to obtain the motion constraint information of the chassis based on the first motion constraint information, the second motion constraint information, and / or the third motion constraint information; the first motion constraint information is generated based on the speed and steering limit information corresponding to the first and second wheel sets of the chassis; the second motion constraint information is generated based on the kinematic model constraints corresponding to the chassis; and the third motion constraint information is generated based on the trajectory limit information. The planning module is used to optimize the trajectory of the chassis based on the kinematic model, the motion constraint information, and the initial trajectory point information, so as to obtain the motion planning result corresponding to the chassis.

6. The apparatus according to claim 5, characterized in that, The planning module is used for: Based on the motion constraint information, the initial trajectory point information, and the kinematic model, the chassis trajectory is optimized to obtain the target trajectory point information; Based on the target trajectory point information, the motion planning result corresponding to the chassis is obtained.

7. The apparatus according to claim 6, characterized in that, The motion constraint information includes first motion constraint information, second motion constraint information, and third motion constraint information; the first motion constraint information represents the speed and steering restrictions on the first and second wheel sets; the second motion constraint information represents the kinematic model restrictions corresponding to the chassis; and the third motion constraint information represents the trajectory restrictions. The planning module is used for: Based on the first virtual wheel information and the second virtual wheel information corresponding to the kinematic model, the first wheel motion information and the second wheel motion information of the first wheel group and the second wheel group of the chassis corresponding to the initial trajectory point information are determined respectively. Based on the first motion constraint information, the first wheel motion information, and the second wheel motion information, the optimized reference wheel motion information is determined. Based on the second motion constraint information and the chassis state information corresponding to the kinematic model, the optimized reference chassis state information is determined; Based on the third motion constraint information and the initial trajectory point information, the optimized reference trajectory point information is determined; Based on the reference wheel motion information, the reference chassis state information, and the reference trajectory point information, the initial trajectory point information is adjusted until the weighted sum of the reference wheel motion information, the reference chassis state information, and the reference trajectory point information is minimized, thus obtaining the target trajectory point information.

8. A robot comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 4.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

Images