CN114877905A - Inform-RRT path planning method for bidirectional dynamic growth - Google Patents
Inform-RRT path planning method for bidirectional dynamic growth Download PDFInfo
- Publication number
- CN114877905A CN114877905A CN202210503076.0A CN202210503076A CN114877905A CN 114877905 A CN114877905 A CN 114877905A CN 202210503076 A CN202210503076 A CN 202210503076A CN 114877905 A CN114877905 A CN 114877905A
- Authority
- CN
- China
- Prior art keywords
- node
- new
- sampling
- expansion
- dynamic
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
- G01C21/34—Route searching; Route guidance
- G01C21/3446—Details of route searching algorithms, e.g. Dijkstra, A*, arc-flags, using precalculated routes
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a bidirectional dynamic growth Inform-RRT path planning method, which comprises the following steps: modeling and binarizing according to the barrier map, determining a starting point and a target point, and setting initial parameters; initializing two random trees according to the barrier map; generating sampling points by utilizing bidirectional search fusion P probability fan-shaped constraint; determining the node expansion direction by using the growth corner offset; determining the dynamic expansion step length of the node by utilizing variable step length growth to generate a new expansion node; performing collision detection, updating the main tree graph set and the auxiliary tree graph set, and outputting an initial path; determining a dynamic ellipse sampling range; generating a new expansion node in the dynamic elliptical region by using dynamic growth; and (5) searching the father node for the new expansion node again within the specified range, reconnecting and searching the path with the minimum cost. The method can improve the generation efficiency of the initial path, increase the target guidance and the path planning fault tolerance rate, accelerate the convergence speed of the algorithm and solve the problem of low path planning efficiency caused by complex environment.
Description
Technical Field
The invention relates to the field of robot path planning, in particular to an Inform-RRT path planning method for bidirectional dynamic growth.
Background
With the rapid development of society, intelligent robots are more and more widely applied in fields such as smart factories, intelligent routing inspection, home services, space exploration and the like. However, there are still shortcomings in robot path planning, such as low search efficiency, non-optimal planned path, and the like.
Common motion planning methods are classified into three categories, namely searching, intelligence and sampling. Common Dijkstra and A-x algorithms belong to search path planning, ant colony algorithms, genetic algorithms and the like belong to intelligent path planning algorithms, and the two algorithms need to establish a space environment barrier model and are not suitable for path planning in a space with a complex environment. The sampling-based path planning algorithm Rapidly explores Random Trees (RRT), RRT, informationed-RRT and the like, can be widely applied to complex high-dimensional space, does not need to describe a configuration space obstacle, and is suitable for path planning of the multi-degree-of-freedom robot in complex environment and dynamic environment.
The traditional inform-RRT algorithm was developed based on RRT, RRT. The RRT algorithm is directed to the blank area through random sampling of the state space, so as to quickly find a feasible planned path, but the generated path is not optimal and the environmental complexity affects the convergence speed. In view of the above disadvantages of the RRT algorithm, RRT generates an optimal path by two steps of re-searching parent nodes for the current node within a specified range and re-connecting nodes within the range, but the planning efficiency is low due to the large range of redundant branches and leaves. In view of the above disadvantages of RRT, informationed-RRT establishes a gradual ellipse to define the sampling space when the branches and leaves are finally pruned based on the initial path generated, thereby improving the planning efficiency. However, the algorithm does not constrain the sampling space when generating the initial path, and does not optimize the node expansion mode when generating the initial path and pruning the redundant branches and leaves, so that the planning efficiency does not meet the timeliness requirement of the robot in the complex environment. Therefore, for the above problems, it is urgently needed to provide a path planning method capable of guiding the informationed-RRT algorithm to quickly find a piece of asymptotically optimal path.
Disclosure of Invention
The invention aims to provide an inform-RRT path planning method for bidirectional dynamic growth. The generation efficiency of the initial path is effectively improved by adopting a bidirectional search mode of the initial path; by adopting a P probability fan-shaped constraint sampling method, the target guidance and the path planning fault-tolerant rate are increased; growth corner offset is adopted during node expansion, so that the convergence speed of the algorithm is effectively increased; the problem of low path planning efficiency caused by complex environment in the path planning process is solved by adopting an expansion mode of variable step growth.
In order to achieve the purpose, the invention adopts the following technical scheme:
the invention aims to provide a bidirectional dynamically-grown information-RRT path planning method, which aims to improve the planning efficiency of path planning in the prior art under a complex environment and comprises the following steps:
s1, modeling and binarizing according to the barrier map, determining a starting point and a target point, and setting initial parameters;
s2, initializing two random trees according to the barrier map;
s3, sampling by using a sampling strategy of bidirectional search fusion P probability fan-shaped constraint to generate sampling points;
s4, determining the direction of the node during expansion by using a growth corner bias strategy;
s5, determining the dynamic expansion step length when the node is expanded by using a variable step length growth strategy to generate the latest expansion node;
s6, performing collision detection, updating the primary tree graph set and the secondary tree graph set, and outputting an initial path;
s7, determining a dynamic ellipse sampling range;
s8, generating a new expansion node in the dynamic elliptical region by using a dynamic growth strategy;
and S9, searching the parent node and reconnecting within the specified range of the new extended node, and searching the path with the minimum cost.
The invention also discloses a method for planning an Inform-RRT path with bidirectional dynamic growth, which is characterized in that the step S1 comprises the following steps:
the barrier map comprises barriers and blank areas in the map, binarization processing is carried out on the barriers and the blank areas, and an initial point q is determined init And target point q goal 。
The step S2 includes:
initializing a set of vertices V in two random trees according to an obstacle map 1 、V 2 And edge set E 1 、E 2 And respectively collect the vertex points V 1 、V 2 And edge set E 1 、E 2 Atlas G to join Primary and Secondary Trees 1 、G 2 In (1).
The step S3 includes:
and S31, generating a P probability fan-shaped constraint sampling random value and comparing the P probability fan-shaped constraint sampling random value with a specified threshold B, adopting fan-shaped sampling if the P probability fan-shaped constraint sampling random value is smaller than the threshold B, and selecting common random sampling if the P probability fan-shaped constraint sampling random value is not smaller than the threshold B. Generating a sampling point q rand ；
S311, sector sampling refers to connecting the update node q once new(i-1) And target point q goal As the central line of the sector, if it is the initial sampling, the initial point q is used init And target point q goal As the midline of the sector. Taking 2 theta as a sector included angle, wherein a sector area can be determined by a central line and the sector included angle, and a sampling area of sector sampling can be determined by an intersection area of the sector area and a map boundary;
s312, determining a sampling area of ordinary random sampling by a map boundary;
s32, judgment Tree Tree1 vertex set V 1 Whether the number of medium vertexes numv1 is less than or equal to the Tree Tree2 vertex set V 2 If the number of the medium vertexes is numv2, the main Tree is Tree1, and the auxiliary Tree is Tree 2. Searching a node q closest to a sampling point on a main tree nearst Completing new expansion node q by growing corner offset and variable step growth new The expansion of (2).
The step S4 includes:
the growing corner bias strategy is to change the direction of the expansion of a target node, and the included angle alpha is a straight line and is the nearest node q nearst And a target node q goal The resulting straight line L (q) nearst ,q goal ) And q is near1 And a sampling node q rand The straight line L (q) formed nearst ,q rand ) The angle between the lines, L (q) of the nodes grown in principle if there is no growth direction bias nearst ,q goal ) By adding an adjustable bias coefficient mu (0)<μ<1) The change may change the bias direction.
The step S5 includes:
s51, setting the initial fixed step length as r, if the sampling node q is known rand And nearest node q nearst With q nearst Is the center of a semicircle with a diameter of 2M (M > r), a direction and a q nearst To q rand Is vertical;
s52, the semicircular area is q nearst In the range considered by the environment complexity, all nodes in the range are made into circles with the radius of N, and the environment complexity index is defined as eta:
in the formula, S is the number of intersected circles formed by nodes in the range and barriers, and U and V are two threshold levels (U is more than 0 and less than V);
s53, obtaining the size of S through each sampling, and defining the dynamic expansion step length r according to the environment complexity d The following were used:
the expansion direction may be determined through step S4, and the dynamic expansion step may be determined through step S5Long r d So far, a new extension node q can be completed new The expansion of (2).
The step S7 includes:
the major axis of the dynamic ellipse is the length c of the currently generated path best Starting point q star And target point q goal Two foci of the ellipse, respectively, the distance of which is set as c min The minor axis length of the ellipse is
The elliptical sampling area is determined from the two foci and the known major and minor axes.
The step S8 includes:
after the sampling of the elliptical area is determined in step S7, P probability is generated in step S31, a sampling mode is selected and sampled, the expansion direction is determined by the growth corner offset in step S4, and the dynamic expansion step r is determined by the variable step growth in step S5 d Finally, a new extension node q is generated new 。
The step S9 includes:
s91, generating new expansion node q in step S8 new Reselecting parent node q within a prescribed range father ；
S911, setting the node q closest to the current sampling point nearest Temporary storage of least costly variable q min Performing the following steps;
s912, according to the current new extension node q new Defining a range, i.e. a range for pruning branches and leaves, which is generally defined by the current new extension node q new Circle C with 1.5 times of basic step length r as radius as center near . Traverse the specified range C near All nodes in the node are new extension nodes q new Re-finding parent node q father ；
S913, expanding the new node q new And each traversed node is used for collision detection, if the node passes the collision detection, the point q near Plus the cost of the new extension node q new And q is near The cost between q 'is temporarily stored in c', if the parent node q is newly constructed father Time q new The cost c' is higher than that of the original q new If the cost is small, the current traversal point q is set near Storing the current new expansion node q new Of a minimum cost parent node q min In this way, the current new extension node q can be found after the traversal new In a predetermined range C near Inner minimum cost parent node q min Updating edge set E 1 、E 2 ；
S92, in the current prescribed range C near Reconnecting all nodes therein;
s921, in the prescribed Range C near Inner (except for the least cost parent node q) min Other nodes than the other) of the nodes will expand the new node q new As a new parent q father Comparing the previous costs;
s922, the traversed points and the new extension node q new If the original path cost of the traversal point is higher than the cost of new connection through collision detection, updating the father node q of the current traversal point father And updating the total atlas G.
Compared with the prior art, the invention has the following beneficial effects:
the invention aims to provide a bidirectional dynamic growth Inform-RRT path planning method. Firstly, a mode of bidirectional search of an initial path is adopted, so that the generation efficiency of the initial path is effectively improved; secondly, a P probability fan-shaped constraint sampling method is adopted, and target guidance and path planning fault-tolerant rate are increased; then, growth corner offset is adopted during node expansion, so that the convergence speed of the algorithm is effectively increased; and finally, an expansion mode of variable step growth is adopted, so that the problem of low path planning efficiency caused by complex environment in the path planning process is solved.
Drawings
FIG. 1 is a general flow chart of an inform-RRT path planning method for bidirectional dynamic growth according to the present invention;
FIG. 2 is a schematic diagram of the P probability sector constraint sampling principle of the present invention;
FIG. 3 is a schematic view of the growth corner offset principle of the present invention;
FIG. 4 is a schematic diagram of the growth principle of the present invention with variable step size;
FIG. 5 is a schematic view of an elliptical sampling space of the present invention;
FIG. 6 is a diagram illustrating the path planning effect of the present invention for testing different algorithms;
FIG. 7 is a plot of the algorithm of the present invention versus run time.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. The specific embodiments described herein are merely illustrative of the relevant invention and are not intended to be limiting of the invention. It should be noted that, for convenience of description, only the portions related to the present invention are shown in the drawings.
The invention provides a bidirectional dynamic growth Inform-RRT path planning method, which comprises the following steps as shown in figure 1:
step one, modeling and binaryzation are carried out according to an obstacle map, the obstacle map comprises obstacles and blank areas in the map, and an initial point q is determined init And target point q goal 。
Step two, initializing a vertex set V in two random trees according to the barrier map 1 、V 2 And edge set E 1 、E 2 And respectively collect the vertex points V 1 、 V 2 And edge set E 1 、E 2 Atlas G to join Primary and Secondary Trees 1 、G 2 In (1).
And step three, sampling in the initialized map space by using a sampling strategy of bidirectional search fusion P probability fan-shaped constraint to generate sampling points. And generating a P probability fan-shaped constraint sampling random value and comparing the P probability fan-shaped constraint sampling random value with a specified threshold B, adopting fan-shaped sampling if the P probability fan-shaped constraint sampling random value is smaller than the threshold B, and selecting common random sampling if the P probability fan-shaped constraint sampling random value is not smaller than the threshold B. As shown in fig. 2, a sampling point q is generated rand . Sector sampling refers to connecting a primary update node q new(i-1) And target point q goal As the central line of the sector, if it is the initial sampling, the initial point q is used init And target point q goal As the midline of the sector. Taking 2 theta as the sector angle (theta is 60 degrees), the sector area can be determined by the midline and the sector angle, and the sampling area of sector sampling isMay be determined by the area where the sector area intersects the map boundary. The sampling area of the ordinary random sampling is determined by the map boundary. Judgment Tree Tree1 vertex set V 1 Whether the number of medium vertexes numv1 is less than or equal to the Tree Tree2 vertex set V 2 If the number of the medium vertexes is numv2, the main Tree is Tree1, and the auxiliary Tree is Tree 2. Searching a node q closest to a sampling point on a main tree nearst Completing new expansion node q by growing corner offset and variable step growth new The expansion of (2).
And 4, step 4: and determining the direction of the node expansion by using a growing corner offset strategy based on the newly generated sampling point. The growing corner bias strategy is to change the direction of the target node expansion, as shown in FIG. 3, the included angle α is the nearest node q nearst And a target node q goal The resulting straight line L (q) near1 ,q goal ) And q is near1 And a sampling node q rand The straight line L (q) formed near1 ,q rand ) The angle between the lines, L (q) of the nodes grown in principle if there is no growth direction bias near1 ,q goal ) By adding an adjustable bias coefficient mu (0)<μ<1) The change may change the bias direction.
And 5: method for determining dynamic expansion step length r during node expansion by using variable step length growth strategy d Generating the latest extension node q new . As shown in fig. 4, an initial fixed step size is set to r, if a sampling node q is known rand And nearest node q nearst With q netsra Is the center of a semicircle with a diameter of 2M (M > r), a direction and a q nearst To q rand Is vertical. Semicircular region of q nearst In the range considered by the environment complexity, all nodes in the range are made into circles with the radius of N, and the environment complexity index is defined as eta:
in the above formula, S is the number of intersections between the circle formed by the nodes in the range and the barrier, and U and V are two threshold levels (0 < U <)V). The size of S can be obtained by sampling each time, and a dynamic expansion step length r is defined according to the environment complexity d The following were used:
the extension direction can be determined by step 4 and the dynamic extension step r can be determined by step 5 d So far, a new extension node q can be completed new The expansion of (2).
Step 6: for new extension node q new Performing collision detection and updating the main and auxiliary tree graph sets G 1 、G 2 And outputting an initial path if the latest expansion node threshold condition of the primary and secondary trees is met after iteration.
And 7: a dynamic elliptical sampling range is generated based on the initial point, the target point, and the planned initial path. The major axis of the dynamic ellipse in FIG. 5 is the length c of the current generation path best Starting point q star And target point q goal Two foci of the ellipse, respectively, the distance of which is set as c min The minor axis length of the ellipse is
The elliptical sampling area is determined from the two foci and the known major and minor axes.
And 8: generating a new extended node q in a dynamic elliptical region by utilizing a P probability fan-shaped constraint sampling strategy, a growing corner bias strategy and a variable step length growing strategy new . After the sampling of the elliptical area is determined in step 7, P probability is generated in step 3, a sampling mode is selected and sampled, the expansion direction is determined by growth corner offset in step 4, and the dynamic expansion step r is determined by variable step growth in step 5 d Finally, a new extension node q is generated new 。
And step 9: for the new extended node q generated in step 8 new Reselecting parent node q within a prescribed range father The node q closest to the current sampling point is determined nearest Temporary storage of least costly variable q min In (1),according to the current new extension node q new Defining a range of pruning branches and leaves, the defined range C near Typically with the current new extension node q new Circle C with 1.5 times of basic step length r as radius as center near . Traverse the specified range C near All nodes in the node are new extension nodes q new Re-finding parent node q father New extension node q new And each traversed node is used for collision detection, if the node passes the collision detection, the point q near Plus the cost of the new extension node q new And q is near The cost between them is temporarily stored in c', if the parent node q is newly constructed father Time q is new The cost c' is higher than that of the original q new If the cost is small, the current traversal point q is set near Storing the current new expansion node q new q new Of a minimum cost parent node q min In this way, the current new extension node q can be found after the traversal new In a predetermined range C near Inner minimum cost parent node q min Updating edge set E 1 、E 2 . In the currently specified range C near Reconnecting all nodes within a specified range C near Inner (except for the least cost parent node q) min Other nodes than the other) of the nodes will expand the new node q new q new As a new parent q father Comparing the previous cost, the traversed points and the new extension node q new If the collision detection is passed, if the cost of the original path of the traversal point is larger than that of the new connection, updating the father node q of the current traversal point father And updating the total atlas G. And outputting the final planned path after the specified iteration times are finished.
In order to verify the effectiveness of the method in solving the path planning problem, two groups of test experiments are designed, wherein one group of test experiments is an effect comparison experiment aiming at different path planning algorithms under the same environment complexity; the second group is test experiments for algorithm performance indicators. The system environment is Windows 10 operating system and Pycharm 2020.3.3(Community Edition).
The experiment is that RRT, RRT and Inform-RRT under the same environment complexity are compared with the path quality planned by the bidirectional dynamic growth Inform-RRT path planning method under the same basic step 2 and the same iteration times 300. The map size is 25 x 25, and the obstacles are shown in the black part in fig. 6. The comparison result is shown in fig. 6, the planned path of the improved algorithm is smaller than other path lengths and turn widths, and compared with other planned paths, the improved algorithm is more adaptive to the environment with higher complexity under the condition of limited and same sampling times.
And the second experiment is to carry out quantitative test on the time consumed by the algorithm. Firstly, the traditional Inform-RRT algorithm and the improved algorithm after introducing the growth corner offset are compared under the same environment and basic step 2 for 20 times of experiments, the average planning time and the total planning time of the initial path are tested, the size of a map is 25 × 25, and obstacles are black parts as shown in figure 6. It can be seen that the overall performance is improved to a certain extent both in the planning process of the initial path and in the total planning process, and the improvement of the total planning performance is more obvious compared with the initial path, and the specific experimental results are shown in table 1 below:
table 1 comparative table of α -bias performance of introduced growth rotation angle
Then, compared with the traditional Inform-RRT algorithm and the improved algorithm of the invention, 20 times of experiments are carried out under the same environment and the basic step length 2, as shown in FIG. 7, it can be seen that the time consumption difference of the two algorithms is enlarged along with the increase of the sampling iteration times, and the efficiency of the path planning algorithm of the invention is higher.
Aiming at the simulation experiment, the invention adopts indexes such as path length, planning time, planning success rate, initial path iteration times and the like. Path length
Where m trees have planned edge sets E 1 、E 2 Number of middle edges, L i Representing the length of the planned ith segment edge; the planning time T is expressed as an algorithmTotal length of time of operation in the simulation environment; planning success rate
Wherein C represents the number of times the planning has been successful in a known iteration round; the initial path iteration number F represents the iteration number when the first planning is successful. In addition to the four indexes, an Average Turning Index (ATI) algorithm cost Index is also provided. The average turning index refers to an average value of angles of all nodes connected on the path after the path is planned. The calculation method is shown in formula 3.
N in the above formula (3) is the total number of nodes of the planned path on the tree, node i q i Has the coordinates of (x) i y i ) And ATI is the average turn index. The performance comparison results obtained by performing 20 sets of experiments respectively under the same iteration times in different environments are shown in table 2 below.
The parameter performances such as average planning path length, average planning time, initial path iteration times, planning success rate, ATI and the like are improved no matter in a simple environment or a complex environment; the performance of the initial path average iteration times is improved almost in a complex environment and a simple environment; except for the average iteration times and the planning success rate of the initial path, the complex environment with improved parameter performance of other parameters is better than the simple environment because the improved algorithm completes the optimization of the planning process after the initial path is generated. For ATI parameters, because the improved algorithm applies a step length changing thought, the improved algorithm of the invention has more obvious performance improvement on the complex environment and is more suitable for path planning of the complex environment. The overall improved algorithm has great improvement on planning efficiency, planning success rate, quality of a planned path and environmental applicability.
TABLE 2 comparison of algorithmic Properties
The above-mentioned embodiments are merely preferred embodiments to fully illustrate the present invention, and the scope of the present invention is not limited thereto. The equivalents and changes made by those skilled in the art based on the present invention are all within the scope of the present invention. The protection scope of the invention is subject to the claims.
Claims (9)
1. An inform-RRT path planning method for bidirectional dynamic growth is characterized by comprising the following steps:
s1, modeling and binarizing according to the barrier map, determining a starting point and a target point, and setting initial parameters;
s2, initializing two random trees according to the barrier map;
s3, sampling by using a sampling strategy of bidirectional search fusion P probability sector constraint to generate sampling points;
s4, determining the direction of the node during expansion by using a growth corner bias strategy;
s5, determining the dynamic expansion step length when the node is expanded by using a variable step length growth strategy to generate the latest expansion node;
s6, performing collision detection, updating the primary tree graph set and the secondary tree graph set, and outputting an initial path;
s7, determining a dynamic ellipse sampling range;
s8, generating a new expansion node in the dynamic elliptical region by using a dynamic growth strategy;
and S9, searching the parent node and reconnecting within the specified range of the new extended node, and searching the path with the minimum cost.
2. The method for bidirectional dynamically-grown inform-RRT path planning as claimed in claim 1, wherein said step S1 comprises:
the barrier map comprises barriers and blank areas in the map, binarization processing is carried out on the barriers and the blank areas, and an initial point q is determined init And target point q goal 。
3. The method for bidirectional dynamically-grown inform-RRT path planning as claimed in claim 1, wherein said step S2 comprises:
initializing a set of vertices V in two random trees according to an obstacle map 1 、V 2 And edge set E 1 、E 2 And respectively collect the vertex points V 1 、V 2 And edge set E 1 、E 2 Atlas G of additions to Primary and Secondary Trees 1 、G 2 In (1).
4. The method for bidirectional dynamically-grown inform-RRT path planning as claimed in claim 1, wherein said step S3 comprises:
and S31, generating a P probability fan-shaped constraint sampling random value and comparing the P probability fan-shaped constraint sampling random value with a specified threshold B, adopting fan-shaped sampling if the P probability fan-shaped constraint sampling random value is smaller than the threshold B, and selecting common random sampling if the P probability fan-shaped constraint sampling random value is not smaller than the threshold B. Generating a sampling point q rand ；
S311, sector sampling refers to connecting the update node q once new(i-1) And target point q goal As the central line of the sector, if it is the initial sampling, the initial point q is used init And target point q goal Taking 2 theta as a central line of a sector as a sector included angle, wherein a sector area can be determined by the central line and the sector included angle, and a sector sampling area can be determined by an intersection area of the sector area and a map boundary;
s312, determining a sampling area of ordinary random sampling by a map boundary;
s32, judgment Tree Tree1 vertex set V 1 Whether the number of medium vertexes numv1 is less than or equal to the Tree Tree2 vertex set V 2 If the number of the medium vertexes is numv2, the main Tree is Tree1, and the auxiliary Tree is Tree 2. Searching a node q closest to a sampling point on a main tree nearst Completing new expansion node q by growing corner offset and variable step growth new The expansion of (2).
5. The method for bidirectional dynamically-grown inform-RRT path planning as claimed in claim 1, wherein said step S4 comprises:
the growing corner bias strategy is to change the direction of the expansion of a target node, and the included angle alpha is a straight line and is the nearest node q nearst And a target node q goal The resulting straight line L (q) nearst ,q goal ) And q is near1 And a sampling node q rand The straight line L (q) formed nearst ,q rand ) The angle between the lines, L (q) of the nodes grown in principle if there is no growth direction bias nearst ,q goal ) By adding an adjustable bias coefficient mu (0)<μ<1) The change may change the bias direction.
6. The method for bidirectional dynamically-grown inform-RRT path planning as claimed in claim 1, wherein said step S5 comprises:
s51, setting the initial fixed step length as r, if the sampling node q is known rand And nearest node q nearst With q nearst Is the center of a semicircle with a diameter of 2M (M > r), a direction and a q nearst To q rand Is vertical;
s52, the semicircular area is q nearst In the range considered by the environment complexity, all nodes in the range are made into circles with the radius of N, and the environment complexity index is defined as eta:
in the formula, S is the number of intersected circles formed by nodes in the range and barriers, and U and V are two threshold levels (U is more than 0 and less than V);
s53, obtaining the size of S through each sampling, and defining the dynamic expansion step length r according to the environment complexity d The following were used:
by passingStep S4 can determine the expansion direction, and step S5 can determine the dynamic expansion step r d So far, a new extension node q can be completed new The expansion of (2).
7. The method for bidirectional dynamically-grown inform-RRT path planning as claimed in claim 1, wherein said step S7 comprises:
the major axis of the dynamic ellipse is the length c of the currently generated path best Starting point q star And target point q goal Two foci of the ellipse, respectively, the distance of which is set as c min The minor axis length of the ellipse is
The elliptical sampling area is determined from the two foci and the known major and minor axes.
8. The method for bidirectional dynamically-grown inform-RRT path planning as claimed in claim 1, wherein said step S8 comprises:
after the sampling of the elliptical area is determined in step S7, P probability is generated in step S31, a sampling mode is selected and sampled, the expansion direction is determined by the growth corner offset in step S4, and the dynamic expansion step r is determined by the variable step growth in step S5 d Finally, a new extension node q is generated new 。
9. The method for bidirectional dynamically-grown inform-RRT path planning as claimed in claim 1, wherein said step S9 comprises:
s91, generating new expansion node q in step S8 new Reselecting parent node q within a prescribed range father ；
S911, setting the node q closest to the current sampling point nearest Temporary storage of least costly variable q min Performing the following steps;
s912, according to the current new extension node q new Defining a range within which to trim the branches and leaves, the gaugeScoping typically with the current new extension node q new Circle C with 1.5 times of basic step length r as radius as center near Go through the specified range C near All nodes in the node are new extension nodes q new Re-finding parent node q father ；
S913, expanding the new node q new And each traversed node is used for collision detection, if the node passes the collision detection, the point q near Plus the cost of the new extension node q new And q is near The cost between q 'is temporarily stored in c', if the parent node q is newly constructed father Time q new Cost c 'of q' is higher than that of the original q new If the cost is small, the current traversal point q is set near Storing the current new expansion node q new Of a minimum cost parent node q min In this way, the current new extension node q can be found after the traversal new In a predetermined range C near Inner minimum cost parent node q min Updating edge set E 1 、E 2 ；
S92, in the current prescribed range C near Reconnecting all nodes therein;
s921, in the prescribed Range C near Inner (except for the least cost parent node q) min Other nodes than the other) of the nodes will expand the new node q new As a new parent q father Comparing the previous costs;
s922, the traversed points and the new extension node q new If the original path cost of the traversal point is higher than the cost of new connection through collision detection, updating the father node q of the current traversal point father And updating the total atlas G.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210503076.0A CN114877905A (en) | 2022-05-10 | 2022-05-10 | Inform-RRT path planning method for bidirectional dynamic growth |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210503076.0A CN114877905A (en) | 2022-05-10 | 2022-05-10 | Inform-RRT path planning method for bidirectional dynamic growth |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114877905A true CN114877905A (en) | 2022-08-09 |
Family
ID=82674597
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210503076.0A Pending CN114877905A (en) | 2022-05-10 | 2022-05-10 | Inform-RRT path planning method for bidirectional dynamic growth |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114877905A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115268456A (en) * | 2022-08-10 | 2022-11-01 | 哈尔滨理工大学 | Unmanned vehicle path planning method for dynamically changing strategy informad-RRT |
| CN115683149A (en) * | 2022-11-14 | 2023-02-03 | 武汉轻工大学 | Interactive intelligent path planning method based on map information |
| CN115826591A (en) * | 2023-02-23 | 2023-03-21 | 中国人民解放军海军工程大学 | Multi-target point path planning method based on neural network estimation path cost |
| CN116817947A (en) * | 2023-05-30 | 2023-09-29 | 北京航空航天大学 | Random time path planning method based on variable step length mechanism |
-
2022
- 2022-05-10 CN CN202210503076.0A patent/CN114877905A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115268456A (en) * | 2022-08-10 | 2022-11-01 | 哈尔滨理工大学 | Unmanned vehicle path planning method for dynamically changing strategy informad-RRT |
| CN115268456B (en) * | 2022-08-10 | 2023-10-17 | 哈尔滨理工大学 | Unmanned vehicle path planning method adopting dynamic variable strategy formed-RRT |
| CN115683149A (en) * | 2022-11-14 | 2023-02-03 | 武汉轻工大学 | Interactive intelligent path planning method based on map information |
| CN115826591A (en) * | 2023-02-23 | 2023-03-21 | 中国人民解放军海军工程大学 | Multi-target point path planning method based on neural network estimation path cost |
| CN116817947A (en) * | 2023-05-30 | 2023-09-29 | 北京航空航天大学 | Random time path planning method based on variable step length mechanism |
| CN116817947B (en) * | 2023-05-30 | 2024-02-23 | 北京航空航天大学 | Random time path planning method based on variable step length mechanism |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114877905A (en) | Inform-RRT path planning method for bidirectional dynamic growth | |
| CN111610786B (en) | Mobile robot path planning method based on improved RRT algorithm | |
| CN110887484B (en) | Mobile robot path planning method based on improved genetic algorithm and storage medium | |
| CN113219998B (en) | Improved bidirectional-RRT-based vehicle path planning method | |
| CN109116841B (en) | Path planning smooth optimization method based on ant colony algorithm | |
| CN113110522B (en) | Robot autonomous exploration method based on composite boundary detection | |
| CN112985408B (en) | Path planning optimization method and system | |
| CN110220521A (en) | A kind of generation method and device of high-precision map | |
| CN115079705A (en) | Routing planning method for inspection robot based on improved A star fusion DWA optimization algorithm | |
| CN109931943B (en) | Unmanned ship global path planning method and electronic equipment | |
| CN110196602A (en) | The quick underwater robot three-dimensional path planning method of goal orientation centralized optimization | |
| CN112486178A (en) | Dynamic path planning method based on directed D (delta) algorithm | |
| CN113485369A (en) | Indoor mobile robot path planning and path optimization method for improving A-x algorithm | |
| LU102942B1 (en) | Path planning method based on improved a* algorithm in off-road environment | |
| CN114237302B (en) | Three-dimensional real-time RRT route planning method based on rolling time domain | |
| CN114545921B (en) | Unmanned vehicle path planning algorithm based on improved RRT algorithm | |
| CN113325867B (en) | Path planning method and device for searching of unmanned aircraft and unmanned aircraft | |
| Wang et al. | Target-biased informed trees: Sampling-based method for optimal motion planning in complex environments | |
| CN113701777B (en) | High-precision map lane associated trajectory line automatic generation method based on space vector | |
| CN115826591B (en) | Multi-target point path planning method based on neural network estimation path cost | |
| CN117124335A (en) | Improved RRT path planning method based on path marking backtracking strategy | |
| CN117406713A (en) | Multi-target point path planning method based on improved water wave optimization algorithm | |
| CN116793375A (en) | Mobile robot path planning method for improving RRT based on bidirectional search decision | |
| CN115741686A (en) | Robot path planning method based on variable probability constraint sampling | |
| CN116009527A (en) | Path planning algorithm based on dynamic scene structure expansion perception |
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 |