CN119984266B - Navigation planning methods for wheeled robots in non-flat environments - Google Patents
Navigation planning methods for wheeled robots in non-flat environmentsInfo
- Publication number
- CN119984266B CN119984266B CN202510066030.0A CN202510066030A CN119984266B CN 119984266 B CN119984266 B CN 119984266B CN 202510066030 A CN202510066030 A CN 202510066030A CN 119984266 B CN119984266 B CN 119984266B
- Authority
- CN
- China
- Prior art keywords
- surfel
- point
- pose
- track
- map
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention provides a wheeled robot navigation planning method in a non-flat environment, which combines local optimization with global optimization, disassembles a long-distance track planning problem in a complex environment, solves a global distance optimal track to provide constraint, solves a navigation problem locally, approximates the global optimization with the local optimization, and greatly improves the solving efficiency. In the optimization process, the algorithm reduces the dimension of the loss function of the track by utilizing the geometric features of the ground surface, considers the landing characteristic of the track, can be well adapted to various complex and extreme ground surface topological structures, and ensures that the track is approximately optimal and simultaneously gives consideration to real-time performance and the landing constraint of the wheeled robot.
Description
Technical Field
The invention belongs to the field of wheeled robots, and particularly relates to a wheeled robot navigation planning method in a non-flat environment.
Background
Automatic navigation planning is one of the key problems in the field of wheeled robots. In a flat environment, the wheeled robot can use a two-dimensional grid to represent the environment, and only the kinematic constraint of the robot needs to be considered. However, in complex terrain environments with undulations, the geometric features of the earth's surface itself become non-negligible. In recent years, a few efforts have focused on studying the navigation planning of wheeled robots in non-flat environments, but most of the efforts only consider the collision-free nature and safety of trajectories on the local ground surface, simplifying the processing of environmental information and kinematic constraints. However, ignoring the trajectory of the surface geometry information and the kinematic constraints cannot guarantee that the kinematic constraints of the robot are met. Moreover, most methods do not perform back-end optimization on the trajectory, and therefore are only suboptimal in terms of safety, smoothness, distance.
The disadvantages of the prior art methods are summarized below:
1. The terrain is characterized by adopting complex three-dimensional structures such as point clouds, three-dimensional grids and the like, and the characteristic of sparse Z-axis earth surface of the motion environment of the wheeled robot is not considered, so that a large amount of memory space is wasted.
2. In the navigation planning process of the non-flat environment, the four-wheel landing constraint of the wheeled robot is not considered, only the discrete path is solved, in the sampling solving process, the kinematic constraint of the robot is not considered, and the robot is used as a particle to solve.
3. The track of the wheeled robot is directly processed by adopting a space curve optimization mode, and track landing constraint is not considered in the track optimization process.
Disclosure of Invention
Aiming at the problems existing in the prior art, the invention provides a wheeled robot navigation planning method in a non-flat environment, which combines local optimization with global optimization, disassembles a long-distance track planning problem in a complex environment, solves the global distance optimal track to provide constraint, solves the navigation problem locally, approximates the global optimization with the local optimization, and greatly improves the solving efficiency. In the optimization process, the algorithm reduces the dimension of the loss function of the track by utilizing the geometric features of the ground surface, considers the landing characteristic of the track, can be well adapted to various complex and extreme ground surface topological structures, and ensures that the track is approximately optimal and simultaneously gives consideration to real-time performance and the landing constraint of the wheeled robot.
The technical scheme of the invention is as follows:
the wheeled robot navigation planning method in the non-flat environment comprises the following steps:
step one, calculating road mark points in non-flat environment
Searching a relay point of a navigation track in a special non-flat environment represented by a Surfel map, and dismantling a long-distance navigation problem into a navigation problem among the relay points;
step two, sampling non-flat environment track
Track sampling conforming to kinematic constraint is carried out in Surfel map, and the sampling process rapidly approaches to the target point, so that an initial track is solved;
Step three, non-flat trajectory optimization
The initial trajectory is optimized for safety/smoothness and kinetic feasibility.
The safe smooth kinematics feasible items are different optimization items in a unified optimization function.
These three losses can be calculated for each control point. The surfel map is provided, so that local information of the earth surface can be directly indexed, and constraints in the calculation process of the three loss terms are provided. That is, the surface geometry information needed in the process of calculating the three loss terms can be directly indexed in the map.
Preferably, in the method for planning navigation of the wheeled robot in the non-flat environment,
And in the first step Surfel, the map consists of a hash table and a linked list. The map region on the XOY coordinate is indexed through a hash table, and the linked list structure represents the sparse earth surface region in the Z direction.
Preferably, in the method for planning navigation of a wheeled robot in a non-flat environment, the solving of the relay point in the first step is divided into the following parts:
And solving a minimum distance cost path Traj coarse＝{S1,S2,...,Sn from the departure pose to the target pose by using an RRT (remote radio unit) algorithm based on Surfel map. Wherein S i represents one Surfel, which can be written as Si＝{xi,yi,zi,ai,bi,ci,di,σi,γi,si,φi};, recording the index and local table characteristics of the current Surfel in the hash table;
The method comprises the steps of calculating the pose of a relay point, uniformly dividing Traj coarse＝{S1,S2,...,Sn into m-1 segments to obtain a series of dividing points Milestone= { S 1,S1+k,...,S1+mk}.S1+tk as corresponding Surfel of one dividing point in a map, and extending paths near the dividing points to obtain Traj subi＝{Si-k,Si-k+1,...,Si,...,Si+k. And selecting the middle point of the path point, and calculating the pose by using a PCA algorithm to obtain a relay point T i, so as to calculate a relay point sequence Milestone= { T 1,T2,...,Tm }.
Preferably, in the method for planning navigation of a wheeled robot in a non-flat environment, the second step includes:
the heuristic function of the sampling point is that c=g+h;
g represents the moving cost of the current sampling point N i = { x, y, z, yaw, g, h }, h represents the cost of the current heuristic point reaching the target pose, h = max { astar, dubin }, i.e. the maximum value of the distance astar and the distance dubin on the Surfel map;
Sampling state points, taking left turning as an example, assuming the state of the current node The landing ground surface Surfel S i obtained by index has pose T wi, the z axis of the robot is parallel to the Surfel normal vector, the x axis is consistent with the direction of the locomotive, and the coordinates of the vehicle after the state expansion in the horizontal plane are assumed to be poseThe pose obtained by state expansion at Surfel is
Related to the state of sampling, taking forward left turn as an example
In order to ensure that the robot is attached to the ground surface as much as possible, the robot is alignedCorrecting;
Corresponding space coordinates P j1＝{xj1,yj1,zj1 }, S j can be obtained by indexing in Surfel map, and the robot can be approximately considered to be always attached to the ground surface to move, so as to construct a rotating shaft For a pair ofPerforming rotation correction, and enabling the z axis of the corrected pose to be parallel to the S j normal vector:
considering that the wheeled robot needs to move on the ground, we will do Performing translation correction, and adjusting the projection distance from the pose to S j to be the height h of the robot mass center from the ground surface;
Twj＝Twj2·Ttran,
Calculating the yaw angle and the space position of the pose T wj to obtain the next state
Preferably, in the method for planning navigation of a wheeled robot in a non-flat environment, the third step includes:
non-planar surface trace elastic band loss:
ek＝akxk+bkyk+ckzk+dk
q k is a trace point in the trace obtained from the previous sample. For Q k, on the Surfel map, corresponds to a projection on the average plane pi k:akx+bky+ckz+dk =0 near S k. In calculating the gradient, only the gradient to the control point Q i is considered for the loss term f si;
Non-planar surface trace safety loss:
To optimize the track point Q i, we extend d ext＝min(dext,dobs),dobs along the track normal direction as the distance from the track point to the nearest unsafe Surfel in the extending process, so that two child nodes can be found at two sides of the control point
For control node Q i, we consider all child nodes Q in the fixed length window of the initial sampling trace;
The set Q is projected on the xoy plane of T wi to obtain And then will beIn the transition to the state of T iw,
Due to the aggregationSince the z-coordinate of (2) is 0, only the x-y coordinates are considered to obtain a set
The problem can be converted into a gradient calculation problem on the xoy plane under T iw at this time;
Pair aggregation Constructing a convex polygon:
wherein n zi is represented by The number of hyperplanes constructed by the collection, A iz and b iz are a pair of descriptors of the hyperplane;
Construction of safety constraints
The calculation scheme of the obstacle loss term f c is that
Transforming the actual gradient to world coordinate system
Non-planar surface trace kinematics loss
The invention has the following beneficial effects:
1. The present invention proposes a new map representation Surfel Link. Surfel Link divide the surface in the non-flat environment uniformly in two-dimensional space and adopts linked list to distinguish multi-layer surface. Each node in the linked list stores the parameterized result of the current local surface geometric feature, efficiently utilizes the memory space, and is beneficial to quick index of map information.
2. A complete scheme is formulated for the navigation task of the wheeled robot in a multilayer non-flat environment. Firstly, hybridA is promoted to a non-flat environment to obtain an initialization path which accords with the kinematic constraint of a vehicle and four-wheel landing conditions, and secondly, B-spline is adopted to parameterize the track, so that the safety, smoothness and dynamic feasibility of the track are optimized. In the optimization process, the loss of the objective function is mapped into the function representation of the local surface, and the dimension reduction of the optimization problem is realized while the four-wheel landing assumption is approximately satisfied.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention.
Drawings
FIG. 1 is a diagram showing a data structure of Surfel maps in one embodiment of a method for planning navigation of a wheeled robot in a non-flat environment in accordance with the present invention;
FIG. 2 is a flow chart of one embodiment of a method for planning navigation of a wheeled robot in a non-flat environment in accordance with the present invention;
Fig. 3 is a schematic diagram of non-planar environment track sampling in an embodiment of a method for planning navigation of a wheeled robot in a non-planar environment according to the present invention.
Detailed Description
The present invention is described in further detail below with reference to the drawings to enable those skilled in the art to practice the invention by referring to the description.
As shown in fig. 1 and 2, the invention provides a wheeled robot navigation planning method in a non-flat environment, comprising the following steps:
1. non-flat environment waypoint solution:
And searching a feasible path with the optimal distance in Surfel spaces. The RRT algorithm in space given the departure and destination poses T star,Tend, surfel is shown in the table below.
Generating relay points using discrete trajectories
2. Non-planar environmental trace sampling
Taking a left turn as an example, the next state after sampling is shown in fig. 3. Assuming the state of the current nodeThe indexed landing surface Surfel S i has a pose T iw. Wherein the z axis of the robot is parallel to Surfel normal vector, and the x axis is consistent with the direction of the headstock. Assuming that coordinates of the vehicle after the state expansion in the horizontal plane are poseThe pose obtained by state expansion at Surfel isRelated to the state of sampling, taking forward left turn as an example
Position and orientationAnd (5) performing correction.Corresponding space coordinatesThe index may be SurfelS j. The robot can be considered to move while being always attached to the ground surface. Construction of a rotating shaftFor a pair ofPerforming rotation correction, and enabling the z axis of the corrected pose to be parallel to the S j normal vector:
considering that the wheeled robot needs to move on the ground surface, the method is to And carrying out translation correction, and adjusting the projection distance from the pose to S j to be the height h of the robot mass center from the ground surface.
Calculating the yaw angle and the space position of the pose T wj to obtain the next state
In addition to the nodes extended in the above manner, the optimal Dubin trace is calculated for some nodes. The start and end points of the trajectory are planar projections of the current state and the end state. In the trajectory generation process, we do not consider obstacles and non-traversable areas in the environment. Then, we discretize the generated trajectory and expand the nodes in Surfel map according to the action of the discrete points. If the current node and the terminal point are positioned on the same surface level, the node expansion mode can greatly improve the track searching efficiency.
For state nodesAnd pose T wi, the spatial coordinates of the four-wheel landing points and the projections P wpro1,Pwpro2,Pwpro3,Pwpro4 of the four-wheel landing points on the corresponding Surfel can be calculated. Fitting P wpro1,Pwpro2,Pwpro3 to obtain a plane
π:apjx+bpjy+cpjz+dpj=0
If P wpro4 is a distance d π＜dth to plane pi, the current state node is a safe node, otherwise it is an unsafe node. Of course, if the four index-derived Surfel or S j are not secure themselves, the state node is an unsafe state node and is not adopted.
If the state node is a safe sampling point, a heuristic function needs to be calculated. Assume that the current node isThe target relay point isThe heuristic function is
hueven(Xs,Xg)＝max(dubin(Xs,Xg),astar(Xs,Xg))
Wherein dubin curves have a ready-made calculation scheme, astar is astar path length in a Surfel map. The algorithm is as follows
CaculateCost in Line2 is used for calculating the movement cost g and the heuristic cost h of the current node, and the heuristic cost is directly calculated by using Manhattan distance. key i＝{xi,yi,zi, i.e., the discrete index of the current node in the Surfel map. z i represents the index position in the linked list. The priority queue PriorityQ is realized by adopting a large root heap, and the element with smaller total cost g+h is prioritized, so that the search is facilitated to be carried out in the target direction. Line8 performs neighbor node search on each element, unlike direct expansion in flat environments. Because { x i,yi } corresponds to a linked list in Surfel's map, the node's ordinate needs to be filtered in the linked list.
3. Non-planar trajectory optimization
The optimization function may be defined as:
f=λ1fs+λ2fc+λ3(fv+fa)
Where f s and f c are the smoothing term and the obstacle term, respectively, and f v and f a are the constraints of the velocity and acceleration terms. Lambda 1,λ2,λ3 is the optimized weight of the smoothing term, the safety term and the dynamics term, respectively.
The smooth term function expression is as follows:
ek＝akxk+bkyk+ckzk+dk
wherein the method comprises the steps of Is the projection of Q k onto the average plane pi k:akx+bky+ckz+dk =0. The average plane pi k is the xoy plane of the pose T kw.
In calculating the gradient, only the gradient to control point Q i is considered for the loss term f si, as the loss term is considered at the local surface near control point Q i.
In the actual optimization, since we constrain the optimization range of the control points, pi k is enough to approximate the surrounding local surface information in the constraint range, and the calculation is only performed once in the whole optimization process.
In order to ensure that the trajectory is as far away from the obstacle as possible, we need to extract the semantic information of the environment. For each control point Q i, we extend along the normal direction of the track by d ext＝min(dext,dobs), where d obs is the distance of the track point to the nearest unsafe Surfel during extension. So that two child nodes can be found on both sides of the control pointFrom the trajectory control points and the corresponding sub-nodes, we can construct a corridor on a non-flat environment, the boundary of which is an unsafe area or an expanded edge. For control node Q i, we consider a set Q of all child nodes in a fixed length window.
The set Q is projected on the xoy plane of T iw to obtainAnd then will beIn the transition to the state of T iw,
Due to the aggregationSince the z-coordinate of (2) is 0, only the x-y coordinates are considered to obtain a set
The problem at this point can be converted into a gradient computation problem on the xoy plane at T iw. Centering on the origin of the coordinate system T iw (i.e., the control point Q i on the path), for a collectionConstructing a convex polygon:
wherein n zi is represented by The number of hyperplanes constructed by the set, A iz and b iz, is a hyperplane pair of descriptors. Thus, a sufficient requirement for the safety of the wheeled robot at this control point can be expressed as that the projection of the robot's boundary point p 1 onto the T iw xoy plane all fall within a convex polygon.
The obstacle loss term f c can be written as
Since the obstacle loss gradient at control point Q i is calculated on T iw, the actual gradient needs to be transformed into world coordinate system with
To limit the speed and acceleration of the robot, we will penalize excessive speed and acceleration terms to guarantee V i∈[-vmax,vmax],Ai∈[-amax,amax. The speed control point and the acceleration control point may be calculated as follows:
the speed loss function and the acceleration loss function are calculated as follows:
To ensure uniformity of gradient direction, we have a pair of And (5) performing approximation processing. We project the gradient on the T iw xoy plane to obtainI.e. the gradient actually used to update the control point, as shown. Although this is an approximate process, since we will guarantee the threshold of the z-axis drop of two adjacent nodes Q i+1,Qi under the coordinate system T iw (if the threshold is too large to indicate that there is a cut-off, it is unsafe) when we are node-expanding at the front end, while the roll and pitch angles of the safety node Q i are both limited (we design ± 20 ° in the experiment), thereforeThe angle with the xoy plane of T iw will be small, therefore, it is
The invention provides a wheeled robot navigation method in a non-flat environment, which is based on Surfel and a map structure SurfelMap of a linked list. The map adopts Surfel to represent the safety and geometric characteristics of the local surface, and the chain table is utilized to index the multi-layer surface efficiently, so that the method is suitable for the three-dimensional task scene of the wheeled robot under the complex topological topography. On the basis of map representation, a track planning and optimizing algorithm meeting kinematic constraint in a non-flat environment is provided, and the track planning and optimizing algorithm comprises a non-flat environment road mark point calculating module, a non-flat environment track sampling module and a non-flat track optimizing module. The algorithm adopts a time-space decoupling mode to optimize the smoothness, safety and kinematic feasibility of the track. In the optimization problem, the algorithm maps the loss of each control point into the function representation of the terrain in a local function approximation mode, so that the track is ensured to meet the land surface landing constraint, and the dimension reduction of the problem is realized while the optimization precision is ensured.
Although embodiments of the present invention have been disclosed above, it is not limited to the use of the description and embodiments, it is well suited to various fields of use for the invention, and further modifications may be readily apparent to those skilled in the art, and accordingly, the invention is not limited to the particular details without departing from the general concepts defined in the claims and the equivalents thereof.
Claims (3)
1. The wheeled robot navigation planning method in the non-flat environment is characterized by comprising the following steps:
step one, calculating road mark points in non-flat environment
Searching a relay point of a navigation track in a special non-flat environment represented by a Surfel map, and dismantling a long-distance navigation problem into a navigation problem among the relay points;
step two, sampling non-flat environment track
Track sampling conforming to kinematic constraint is carried out in Surfel map, and the sampling process rapidly approaches to the target point, so that an initial track is solved;
Step three, non-flat trajectory optimization
Optimizing the initial trajectory for safety/smoothness and kinetic feasibility;
The solving of the relay point in the first step is divided into the following parts:
Solving the minimum distance cost path Traj coarse＝{S1,S2,...,Sn from the departure pose to the target pose using Surfel map-based RRT algorithm, where S i represents one Surfel, which can be written as
Si＝{xi,yi,zi,ai,bi,ci,di,σi,γi,si,φi}; Record the current Surfel
Index and local surface features in the hash table;
{ x i,yi } represents the corresponding linked list, z i represents the index position in the linked list,
Calculating the pose of a relay point, uniformly dividing Traj coarse＝{S1,S2,...,Sn into m-1 segments to obtain a series of dividing points Milestone= { S 1,S1+k,...,S1+mk},S1+tk as a corresponding Surfel of the dividing point in a map, wherein t is an integer, the range is 1-m, and extending the path near the dividing point to obtain
Traj subi＝{Si-k,Si-k+1,...,Si,...,Si+k, selecting the middle point of the path point of the section, and calculating the pose by using a PCA algorithm to obtain a relay point T i, and further calculating a relay point sequence Milestone= { T 1,T2,...,Tm };
The second step comprises the following steps:
the heuristic function of the sampling point is that c=g+h;
g represents the moving cost of the current sampling point N i＝{xi,yi,zi, yaw, g and h, wherein x i,yi,zi is the vehicle three-dimensional coordinate, yaw is the vehicle yaw angle, h represents the cost of the current heuristic point reaching the target pose, and h=max { astar, dubin }, namely the maximum value of the astar distance and the dubin distance on the Surfel map;
Sampling state points, taking left turning as an example, assuming the state of the current node The landing ground surface Surfel S i obtained by index has a pose T wi, the z axis of the robot is parallel to the Surfel normal vector, the x axis is consistent with the direction of the locomotive, and the pose of the coordinate after the vehicle is subjected to state expansion in the horizontal plane is assumed to beThe pose obtained by state expansion at Surfel is Representing the coordinate pose of the vehicle after the state expansion in the horizontal plane,
Related to the state of sampling, taking forward left turn as an example
In order to ensure that the robot is attached to the ground surface as much as possible, the robot is alignedCorrecting; corresponding space coordinates The index in Surfel map can be obtained S j, the robot can be considered to be always attached to the ground surface to move, and a rotating shaft is constructed The vector of the construction is set up, Is used for the z-axis of (c),Surface z-axis pairPerforming rotation correction, and enabling the z axis of the corrected pose to be parallel to the S j normal vector:
k 1、k2、k3 is Θ is a yaw angle calculated from the current attitude,
Considering that the wheeled robot needs to move on the ground, we will doPerforming translation correction, and adjusting the projection distance from the pose to S j to be the height h of the robot mass center from the ground surface;
Calculating the yaw angle and the space position of the pose T wj to obtain the next state
2. The method for planning navigation of a wheeled robot in a non-flat environment according to claim 1, wherein the map Surfel in the first step is composed of a hash table and a linked list, the map region on XOY coordinates is indexed by the hash table, and the linked list structure represents the sparse surface region in the Z direction.
3. The method for planning navigation of a wheeled robot in a non-flat environment according to claim 1, wherein the third step comprises:
non-planar surface trace elastic band loss:
ei＝aixi+biyi+cizi+di
Q i is a trace point in the trace obtained from the previous sample, For the projection of Q i onto the Surfel map on the average plane pi i:aix+biy+ciz+di =0 in the vicinity of S i, only the gradient to the control point Q i is considered for the loss term f si in calculating the gradient;
Non-planar surface trace safety loss:
To the optimized track point Q i, we extend d ext＝min(dext,dobs),dobs along the track normal direction as the distance from the track point to the nearest unsafe Surfel in the extending process, so that two child nodes Q isub1,Qisub2 can be found at two sides of the control point;
For control node Q i, we consider all child nodes Q in the fixed length window of the initial sampling trace;
The set Q is projected on the xoy plane of T wi to obtain And then will beTransition to T iw in T iw represents a transformation of the current pose to a world pose,
Due to the aggregationSince the z-coordinate of (2) is 0, only the x-y coordinates are considered to obtain a set
The problem can be converted into a gradient calculation problem on the xoy plane under T iw at this time;
Pair aggregation Constructing a convex polygon:
Pi H＝{q∈R2:Aiq≤bi},
wherein n zi is represented by The number of hyperplanes obtained by set construction, A iz and b iz are a pair of hyperplane descriptors, and safety constraint conditions are constructed
The calculation scheme of the obstacle loss term f c is that
Transforming the actual gradient to world coordinate system
Non-planar surface trace kinematics loss
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202510066030.0A CN119984266B (en) | 2025-01-16 | 2025-01-16 | Navigation planning methods for wheeled robots in non-flat environments |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202510066030.0A CN119984266B (en) | 2025-01-16 | 2025-01-16 | Navigation planning methods for wheeled robots in non-flat environments |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN119984266A CN119984266A (en) | 2025-05-13 |
| CN119984266B true CN119984266B (en) | 2025-10-31 |
Family
ID=95623487
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202510066030.0A Active CN119984266B (en) | 2025-01-16 | 2025-01-16 | Navigation planning methods for wheeled robots in non-flat environments |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119984266B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112432648A (en) * | 2020-11-17 | 2021-03-02 | 中山大学 | Real-time planning method for safe motion trail of mobile robot |
| CN115752474A (en) * | 2022-11-23 | 2023-03-07 | 浙江大学 | Robot navigation planning method and device under non-flat ground environment and robot |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8798840B2 (en) * | 2011-09-30 | 2014-08-05 | Irobot Corporation | Adaptive mapping with spatial summaries of sensor data |
| US10553026B2 (en) * | 2017-07-20 | 2020-02-04 | Robert Bosch Gmbh | Dense visual SLAM with probabilistic surfel map |
| KR102449279B1 (en) * | 2020-06-30 | 2022-10-04 | 한국과학기술원 | Mobile robot control apparatus for three dimensional modeling, three dimensional modeling system having the same and method of three dimensional modeling using the same |
| CN117029844A (en) * | 2023-05-19 | 2023-11-10 | 昆明理工大学 | Path planning method in non-flat environment based on improved A-Star algorithm |
| CN117470241A (en) * | 2023-10-30 | 2024-01-30 | 山东大学 | An adaptive mapless navigation method and system for uneven terrain |
| CN118314305A (en) * | 2024-04-10 | 2024-07-09 | 东北大学 | A cross-modal visual localization method based on voxel map surface constraints |
-
2025
- 2025-01-16 CN CN202510066030.0A patent/CN119984266B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112432648A (en) * | 2020-11-17 | 2021-03-02 | 中山大学 | Real-time planning method for safe motion trail of mobile robot |
| CN115752474A (en) * | 2022-11-23 | 2023-03-07 | 浙江大学 | Robot navigation planning method and device under non-flat ground environment and robot |
Also Published As
| Publication number | Publication date |
|---|---|
| CN119984266A (en) | 2025-05-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113359757B (en) | A method for path planning and trajectory tracking of unmanned vehicles | |
| CN112378408B (en) | A Path Planning Method for Real-time Obstacle Avoidance of Wheeled Mobile Robot | |
| CN110962130B (en) | Heuristic RRT mechanical arm motion planning method based on target deviation optimization | |
| WO2023197092A1 (en) | Unmanned aerial vehicle path planning method based on improved rrt algorithm | |
| CN112432648A (en) | Real-time planning method for safe motion trail of mobile robot | |
| CN115639823A (en) | Terrain sensing and movement control method and system for robot under rugged and undulating terrain | |
| CN111307158A (en) | A kind of AUV three-dimensional route planning method | |
| CN114428499A (en) | Astar and DWA algorithm fused mobile trolley path planning method | |
| CN116136417B (en) | Unmanned vehicle local path planning method facing off-road environment | |
| Ma et al. | A Fast path re-planning method for UAV based on improved A* algorithm | |
| CN110954124A (en) | An adaptive path planning method and system based on A*-PSO algorithm | |
| CN116009558A (en) | Mobile robot path planning method combined with kinematic constraint | |
| CN116147653A (en) | A 3D reference path planning method for unmanned vehicles | |
| CN117705123B (en) | Trajectory planning method, device, equipment and storage medium | |
| CN119984306A (en) | Path planning method integrating improved A* algorithm and DWA algorithm | |
| CN118655885A (en) | Path planning system and method for cooperative operation of special vehicles in dynamic environment | |
| CN116817913B (en) | New path planning method utilizing turning penalty factors and twin road network improvement | |
| CN117193315A (en) | A TWA*-DWA global optimal path planning method based on improved A* algorithm and DWA algorithm | |
| Li et al. | UAV 3D trajectory planning and optimization using A* algorithm based on free space method | |
| CN115729260B (en) | Unmanned ship route planning method based on particle swarm algorithm and time potential field | |
| CN117193308A (en) | Smart vehicle obstacle avoidance path planning method based on improved RRT and back-end optimization strategy | |
| CN119984266B (en) | Navigation planning methods for wheeled robots in non-flat environments | |
| CN120196109A (en) | A trajectory planning method for urban logistics UAV based on improved Hybrid A* algorithm | |
| CN119065384B (en) | Unmanned aerial vehicle optimal path planning method based on complete expansion diagram | |
| CN120445201A (en) | Ground path planning method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |