CN114740898A - Three-dimensional flight path planning method based on free space and A-star algorithm - Google Patents

Abstract

The invention discloses a three-dimensional flight path planning method based on free space and A-x algorithm. And then, identifying and distinguishing each obstacle area in the task flight curved surface by using a binary image connected domain detection algorithm, and converting the obstacle into a polygon by using an obstacle rectangular segmentation and combination method. Finally, an improved a-algorithm based on graph preprocessing is used to find the optimal path. Compared with the existing method, the method has the advantages of high planning efficiency, and high flight path flyability and trackability of the planned flight path.

Description

Three-dimensional flight path planning method based on free space and A-star algorithm

Technical Field

The invention relates to the field of flight path planning, in particular to a flight path planning method based on graphics and heuristic search, and particularly relates to a three-dimensional flight path planning method based on free space and an A-star algorithm.

Background

Unmanned aerial vehicles are being used in a large number of military and civilian fields with their advantages of small size, low cost, high flexibility, and the like, including battlefield reconnaissance, disaster monitoring, environmental monitoring, and the like. And wanting to realize these functions, unmanned aerial vehicle's flight path planning function is indispensable. The flight path planning of the unmanned aerial vehicle is to autonomously find an optimal barrier-free travel route from a starting position to a target position in a known or unknown working environment according to certain constraints and criteria. Because the complexity of a three-dimensional map is higher than that of a two-dimensional map, the track planning algorithm of the three-dimensional map is often more complex and more time-consuming. The algorithm a is a classic track planning algorithm, and can realize better track planning, but as the map becomes complex, the calculation amount and the search redundancy of the algorithm a are increased by times, so that the track searching efficiency is low, and the searched track is not optimal.

On the basis of the a-algorithm, there are many improvements made to the a-algorithm, including modifying heuristic functions in the a-algorithm, changing methods for finding nodes, increasing the search range of the neighborhood, and so on. The method avoids redundant search to a certain extent, and improves the track search efficiency of the A-star algorithm and the smoothness of the track. However, the improved a-star algorithm still has the problems of slow planning speed and non-optimal flight path.

Disclosure of Invention

The invention provides a three-dimensional flight path planning method based on free space and an A-star algorithm, aiming at solving the problem of complex three-dimensional flight path planning.

The invention relates to a three-dimensional flight path planning method based on free space and A-star algorithm, which comprises the following specific steps:

step 1: and extracting all nodes with corresponding heights from the terrain elevation model according to the given planning space height dimension to form a task flight curved surface.

Step 2: and identifying and distinguishing each obstacle area in the task flight curved surface by using a binary image connected domain detection algorithm, and further converting the obstacles on the task flight curved surface into polygons.

And step 3: the free space on the task flight curved surface can be regarded as a multi-connected concave polygon, and the multi-connected concave polygon is converted into a single-connected concave polygon; the single connected concave polygon is then decomposed into several sub-convex polygons.

And 4, step 4: and finding the optimal track by using an A-x algorithm based on graph preprocessing.

Firstly, processing each sub-free space on a task flight curved surface into a node used by an A-x algorithm; the specific method comprises the following steps:

and adding node serial numbers, node coordinates, adjacent node serial numbers, father node serial numbers and heuristic information into the free space of the children.

Subsequently, the optimal regional channels are searched using the a-algorithm.

After obtaining the optimal area channel, generating an initial feasible track connecting a starting point and a terminal point in the optimal area channel; further optimizing the obtained flight path; the optimization method comprises the following steps:

setting the global search frequency i to be 0;

let local search number j equal to 0;

thirdly, the midpoint M on the common dividing line of each pair of adjacent areas in the area channel is taken₁,M₂,…M_nVia midpoint { M }₁,M₂,…M_nConnect out the initial feasible track { X }_current}:X_start→M₁→M₂...→M_n→X_end；

Fourthly, randomly selecting a parting line Q_kQ_k+1Adjusting track node M thereon_k→R_k，R_kAnd obtaining a new flight path for another point on the parting line after reconnection:

{X_new}:X_start→M₁→M₂...→R_k→...→M_n→X_end

and calculating the length thereof;

new track { X_newIs shorter than the current track { X }_currentAnd replacing the current track with the new track: { X_current}←{X_newJ is made 0; if the new track is longer than the current track, making j equal to j + 1;

sixthly, repeating the step four and the step five until j is larger than the upper limit t of the local search, and enabling i to be i + 1;

and seventhly, repeating the step III to the step VI until the value i is larger than the global search upper limit T.

Compared with the prior art, the three-dimensional flight path planning method based on the free space and the A-star algorithm has the advantages of high planning efficiency, and high flight path flyability and trackability obtained by planning.

Drawings

FIG. 1 is a flow chart of a three-dimensional flight path planning method based on free space and A-x algorithm according to the present invention;

FIG. 2 is a schematic diagram of the results of extracting a mission flight surface according to the present invention;

FIG. 3 is a schematic diagram of the recognition of the distinction of obstacles according to the present invention;

FIG. 4 is a schematic view of an obstacle-translating polygon of the present invention;

FIG. 5 is a schematic diagram of a single connected polygon constructed in accordance with the present invention;

FIG. 6 is a diagram illustrating the result of map segmentation according to the present invention;

FIG. 7 is a schematic view of the optimal visibility point in the present invention;

FIG. 8 is a schematic diagram of a two-dimensional track planning result according to the present invention;

FIG. 9 is a diagram illustrating the result of flight path planning for a curved flight surface according to the present invention;

FIG. 10 is a schematic diagram of the three-dimensional flight path planning result of the present invention.

Detailed Description

The present invention will be described in further detail below with reference to the accompanying drawings.

The invention relates to a three-dimensional flight path planning method based on free space and A-star algorithm, as shown in figure 1, the specific steps are designed as follows:

step 1: extracting all nodes with corresponding heights from a terrain elevation model according to a given planning space height dimension to form a task flight curved surface;

in order to reduce the complexity of flight path planning, a task flight curved surface is extracted based on a planar binary map, and the three-dimensional flight path planning problem is reduced into a two-dimensional flight path planning problem. The specific method comprises the following steps:

traversing all nodes in the terrain elevation model, and if the corresponding terrain height h at the node A is high_gSatisfy h_g≤h_max-h_mIf so, regarding the node A as a safety node on the task flight curved surface, and taking 0 as the corresponding node of the node A on the binary map; otherwise, the node A is regarded as a barrier node, and the node A is in twoThe corresponding node on the value map takes 1. H above_maxRepresents the maximum flying sea height of the unmanned aerial vehicle; h is a total of_mIndicating the relative ground level specified for the mission. The mission flight profile extracted by the above method is shown in fig. 2.

Step 2: transforming obstacles on the curved surface of the mission flight into polygons

In order to use the a-x algorithm based on graphic preprocessing on a binary map, the obstacles on the mission flight surface need to be converted into polygons. For this purpose, firstly, a binary image connected domain detection algorithm is used to identify and distinguish each obstacle region in the mission flight curved surface, as shown in fig. 3. Making m rows and n columns of a binary map matrix, and the specific process of obstacle area identification is as follows:

1) access node x₁₁If the node is an obstacle node, recording the class order of the node as 1;

2) traverse node x_1j(j 2.. n), if it is an obstacle node, its left node x is detected_1(j-1)Whether the node is an obstacle node; if so, then x_1jUsing x_1(j-1)The class order of (2); otherwise, x_1jThe class order of (2) is the total current class number (the total number of the barrier node class orders) plus 1; for example: let the first row have 10 points, x respectively₁₁，x₁₂，x₁₃，...，x₁₁₀Wherein x is₁₁，x₁₃，x₁₅，x₁₆Is an obstacle point. x is a radical of a fluorine atom₁₁Has a class order of 1; x is the number of₁₂Points are non-obstacle points without class sequences; x is the number of₁₃If it is an obstacle point, its left point x is detected₁₂Whether it is an obstacle point, and the result is no, so x₁₃The sequence of (1 + 1) is 2; x is a radical of a fluorine atom₁₄Is a non-obstacle point without class order; x is a radical of a fluorine atom₁₅Is a barrier point, the class order of which is 2+1 ═ 3; x is the number of₁₆As a barrier point, its left side point x is detected₁₅Is an obstacle point, so x is used₁₅Class order of (3).

3) For a node on the ith row (i ═ 2, …, m), the following is performed:

a) access node x_i1If it is a barrier node, then the upper node x is detected_(i-1)1And upper right node x_(i-1)2: if x_(i-1)1And x_(i-1)2Are not classified, then x_i1The class order of (1) is the current total class number plus 1; if x_(i-1)1Has been classified into x_(i-1)2Not classified, then x_i1Using x_(i-1)1The class order of (2); if x_(i-1)2Has been classified into_(i-1)1Not classified, then x_i1Using x_(i-1)2The class order of (2); if x_(i-1)1And x_(i-1)2All have been classified and have the same class order, then x_i1Using x_(i-1)1And x_(i-1)2The class order of (2); if x_(i-1)1And x_(i-1)2All have been classified and have different class order, then x_i1Using x_(i-1)1And x is_(i-1)1And x_(i-1)2Is marked as an equivalent pair;

b) access node x_ij(j 2.., n-1), if it is an obstacle node, its left node x is sequentially detected_i(j-1)Upper left node x_(i-1)(j-1)Upper node x_(i-1)jAnd upper right node x_(i-1)(j+1): if none of the four points are obstacle-free nodes, then x_ijThe class order of (1) is the current total class number plus 1; if the four points have only one obstacle node, x_ijFollowing the class order of the obstacle point; if there are 2-4 obstacle nodes in the four points, x_ijThe class order of the point detected as the obstacle node first from the four points is used, and the point having the different class order from the four points is regarded as an equivalent pair.

c) Access node x_inIf it is a barrier node, its left node x is detected in turn_i(j-1)Upper left node x_(i-1)(j-1)And an upper node x_(i-1)j: if none of the three points are obstacle-free nodes, then x_ijThe class order of (1) is the current total class number plus 1; if the three points have only one obstacle node, then x_ijFollowing the class order of the barrier node; if there are 2-3 obstacle nodes in the three points, then x_ijUsing the class sequence of the point detected as the barrier node from the first three points, and recording the point pairs with different class sequences as an equivalent pair;

d) unifying the class order of the equivalent categories according to the obtained equivalent pair; all the barrier points with the same class sequence form a barrier area on the binary map.

Subsequently, the identified obstacle area is converted into a polygon, as shown in fig. 4, the specific method is as follows:

finding the maximum and minimum values in the horizontal and vertical coordinates of all points on the obstacle area: x_max，X_max，Y_maxAnd Y_minTo (X)_min,Y_min)，(X_min,Y_max)，(X_max,Y_min)，(X_max,Y_max) For the vertices, a rectangular area is constructed that covers the entire obstacle area. Then, with a certain precision acc (abbreviation for accurate), the rectangular region is divided into a × b rectangular subregions of the same shape, where a ≈ (Y ≈ b)_max-Y_min)/acc，b≈(X_max-X_max) And/acc. And traversing all the sub-regions, and if no barrier node exists on the rectangular sub-region, discarding the sub-region. And finally, connecting the outer vertexes in the reserved sub-regions to obtain an approximate polygon of the original obstacle.

Step 3, decomposing the free space on the task flight curved surface into a plurality of convex polygons

After the obstacles in the mission flight curved surface are converted into polygons, the free space on the mission flight curved surface can be regarded as a multiply-connected concave polygon. To use the a-x algorithm based on graph preprocessing, it is also necessary to decompose the free space in the mission flight surface into a plurality of convex polygons. The process firstly needs to convert a plurality of connected concave polygons into a single connected concave polygon, and the specific principle process is as follows:

when point M is shown in FIG. 5₁M₂…M_nOne barrier region (each barrier vertex O) exists in the enclosed concave polygonal region₁O₂…O_nA city barrier region), one vertex O of the barrier region is selected_iWith one vertex M of the concave polygonal area_iConnecting line, is marked as O_iM_i(ii) a The vector of which can have two directions

And

let two vectors

And with

A distance delta D → 0 between the two points, and the top point O of the obstacle_iAnd the vertex M of the map_iBy connecting two vectors, the map becomes a single connected domain map M₁M₂…M_iO_iO_i+1…O_nO₁…O_iM_i…M_n. If there are other obstacles in the concave polygonal area, each obstacle needs to be directly or indirectly connected to the map boundary vertex.

If another obstacle O 'is present'₁O′₂…O′_nDirect connection is the connection of one vertex O 'of the obstacle, also according to the method described above'_iWith a certain vertex M of the map boundary_jThe following single connectivity region is denoted:

M₁M₂…M_iO_iO_i+1…O_nO₁…O_iM_i…M_jO′_iO′_i+1…O′_nO′₁…O′_iM_j…M_n

indirect connection means that the obstacles can connect the concave polygonal areas by connecting the obstacles that have been transformed. An obstacle is connected to the concave polygonal area before, then the boundary of the obstacle can be regarded as a part of the boundary of the concave polygonal area, and then a new obstacle can be connected to the vertex of the converted obstacle, thereby being equivalent to being connected to the concave polygonal area. O 'to New Barrier'_iO point-connected to first obstacle_mPoint, then the single connected region after connection is represented as:

M₁M₂…M_iO_iO_i+1…O_mO′_iO′_i+1…O′_nO′₁…O′_iO_m…O_nO₁…O_iM_i…M_n

after converting the multi-connected concave polygon into the single-connected concave polygon, the single-connected concave polygon needs to be decomposed into a plurality of sub-convex polygons, as shown in fig. 6. The specific process is as follows:

a. judging the positive and negative of the single connected concave polygon, and if the polygon is a negative polygon, collecting the vertexes thereof to form a set { P }_nB, sequencing vertexes in the sequence in a reverse direction, and if the vertexes are regular polygons, entering the step b;

b. from { P_nChoose a pit P_kSearching all visible points of the point to form a visible point set { V }_k}；

c. If { V_kThere are pits, P is selected among these pits_kIs given as the optimal visibility point of if { V }_kThere is no pit in, then in { V }_kSelect P in_kThe optimal visibility point of (a);

d. connection P_kDecomposing the map with the optimal visible point, and adding a new sub-polygon;

e. repeating steps b) to d) until P_nThere are no pits in the polygon, all the sub-polygons of the polygon are convex polygons.

The positive and negative polarities of the polygon and the optimal visible point are defined as follows:

the positive and negative of the polygon: if the set of vertices of the polygon { P }_nThe vertex sequences in the polygon are arranged in a counterclockwise order, then the polygon is a positive polygon, otherwise, the polygon is a negative polygon. The specific judgment method comprises the following steps: arbitrarily take a convex point P on the polygon_iIf it is adjacent to two points P_i-1And P_i+1Formed vector of

And

satisfy the relation:

then the polygon is positive, otherwise the polygon is negative.

Optimal visibility points: as shown in fig. 7, a pit M is taken_iTwo adjacent points M_i-1And M_i+1Vector of introduction

And

angle bisector of the formed angle

Get M_iSet of visible points of { V }_kOne visible point M in_k，

And vector

Forming an included angle alpha, and calculating the rest chord values

{V_kOne visible point of (i) which maximizes the value is called M_iThe optimal visibility point of (c).

And 4, step 4: finding optimal track using a-x algorithm based on graph preprocessing

First, each sub-free space on the mission flight surface (the mission flight surface is processed by using the method of processing the polygon before, and the obtained sub-polygon which is not the obstacle is called sub-free space) needs to be processed into the node used by the a-x algorithm. The specific method comprises the following steps:

and adding node serial numbers, node coordinates, adjacent node serial numbers, father node serial numbers and heuristic information into the free space of the children. For a certain sub-free space, the above information has the following meaning:

the node sequence number is the sequence number of the free space in all the free spaces;

the node coordinates are the convex polygon corresponding to the sub-free space (the vertex is P)₁,P₂,...,P_n) The centroid coordinates of (a);

the adjacent node sequence number represents the corresponding node sequence number of other sub-areas adjacent to the area pointed by the node;

the parent node sequence number is the sequence number of the node with the minimum cost reaching the free space in the free space adjacent to the free space;

the heuristic information is the euclidean distance between the centroids of the two sub-free spaces, also known as the track cost.

Then, the optimal regional channel is searched by using the a-star algorithm, and the method is as follows:

(1) starting point X_startNode S of the region_startAdding a sequence to be retrieved Open List;

(2) if the Open List is empty, no feasible flight path exists, and the algorithm is ended; otherwise, performing the step (3).

(3) Go through the Open List, find the F value therein (F value represents from the starting point through the current point S_curentNode S with minimum estimated track cost to the end point_curentMove it into Close List, move it out of Open List;

(4) for node S_curentEach of which is not in a Close List_neighbour: if node S_neighbourIf not, it is moved into Open List, and node S is moved into Open List_curentSet as node S_neighbourThe parent node of (2) updates its G value (G value is the actual track cost from the starting point to the current point) and F value; if node S_neighbourIn Open List, determine to go through node S_curentTo node S_neighbourThe resulting node S_neighbourWhether the G value of (a) is less than the original G value; if yes, the node S is connected_curentSet as node S_neighbourThe parent node of (1) updating its G value and F value;

(5) repeating the steps (2) to (4) until the corresponding node of the area where the end point is located is in the Open List;

(6) starting from the node corresponding to the area where the end point is located, connecting each node S_i(not relevant. here each node refers to the node in the process of generating the track, i.e. the end point, the parent of the end point, …, up to the start point) in turn with its parent node S_{i_parent}Connect to form a local channel { S_i}。

In obtaining the regional channel { S_iAfter that, can be at { S }_iAn initial feasible track connecting the starting point and the end point is generated in the station. At this time, the obtained flight path is not optimal, and further optimization of the flight path is needed. The optimization method specifically comprises the following steps:

setting the global search frequency i to be 0;

let local search number of times j equal to 0;

taking the middle point { M ] on the common dividing line of each pair of adjacent areas in the area channel₁,M₂,…M_nConnect the initial feasible track { X } through the middle points_current}:X_start→M₁→M₂...→M_n→X_end。

Fourthly, randomly selecting a parting line Q_kQ_k+1Adjusting track node M thereon_k→R_k(R_kAs another point on the division line), and a new track is obtained after reconnection

{X_new}:X_start→M₁→M₂...→R_k→...→M_n→X_end

And calculates its length.

New track { X_newIs shorter than the current track { X }_currentAnd replacing the current track with the new track: { X_current}←{X_newJ is 0; if the new track is longer than the current track, j is equal to j + 1;

sixthly, repeating the step four and the step five until j is larger than the upper limit t of the local search, and enabling i to be i + 1;

and seventhly, repeating the step three to the step six until i is larger than the global search upper limit T.

The local search upper limit T and the global search upper limit T represent the maximum values of the local optimization times and the global optimization times, respectively.

The resulting track is shown in FIG. 8. This flight path is shown on the original mission flight surface as shown in fig. 9. The height dimension information is added to the obtained flight path, and the finally obtained three-dimensional flight path is shown in fig. 10.

Claims (5)

1. A three-dimensional flight path planning method based on free space and A-star algorithm is characterized by comprising the following steps: the method comprises the following specific steps:

step 1: extracting all nodes with corresponding heights from a terrain elevation model according to a given planning space height dimension to form a task flight curved surface;

step 2: identifying and distinguishing each obstacle area in the task flight curved surface by using a binary image connected domain detection algorithm, and further converting the obstacles on the task flight curved surface into polygons;

and step 3: the free space on the task flight curved surface can be regarded as a multi-connected concave polygon, and the multi-connected concave polygon is converted into a single-connected concave polygon; then decomposing the single connected concave polygon into a plurality of sub convex polygons;

and 4, step 4: finding an optimal flight path by using an A-star algorithm based on graphic preprocessing;

firstly, processing each sub-free space on a task flight curved surface into a node used by an A-x algorithm; the specific method comprises the following steps:

adding node serial numbers, node coordinates, adjacent node serial numbers, father node serial numbers and heuristic information into the child free space;

then, searching an optimal region channel by using an A-star algorithm;

after obtaining the optimal area channel, generating an initial feasible track connecting a starting point and a terminal point in the optimal area channel; further optimizing the obtained flight path; the optimization method comprises the following steps:

setting the global search frequency i to be 0;

let local search number j equal to 0;

thirdly, the midpoint M on the common dividing line of each pair of adjacent areas in the area channel is taken₁,M₂,…M_nVia midpoint { M }₁,M₂,…M_nConnect out the initial feasible track { X }_current}:X_start→M₁→M₂...→M_n→X_end；

Fourthly, randomly selecting a parting line Q_kQ_k+1Adjusting track node M thereon_k→R_k，R_kFor another point on the dividing line, reconnecting to obtain a new track

{X_new}:X_start→M₁→M₂...→R_k→...→M_n→X_end

And calculating the length thereof;

new track { X_newIs shorter than the current track { X }_currentAnd replacing the current track with the new track: { X_current}←{X_newJ is 0; if the new track is longer than the current track, making j equal to j + 1;

sixthly, repeating the step four and the step five until j is larger than the upper limit t of the local search, and enabling i to be i + 1;

and seventhly, repeating the step three to the step six until i is larger than the global search upper limit T.

2. The three-dimensional flight path planning method based on the free space and the a-x algorithm as claimed in claim 1, characterized in that: in step 2, making the binary map matrix have m rows and n columns, and the specific method for identifying and distinguishing the obstacle areas comprises the following steps:

1) access node x₁₁If the node is an obstacle node, recording the class order of the node as 1;

2) traverse node x_1j(j 2.. n), if it is an obstacle node, its left node x is detected_1(j-1)Whether the node is an obstacle node; if so, x_1jUsing x_1(j-1)The class order of (2); otherwise, x_1jThe class order of (2) is the total current class number (the total number of the barrier node class orders) plus 1;

3) for a node on the ith row (i ═ 2, …, m), the following is performed:

a) access node x_i1If it is a barrier node, then detect its upper node x_(i-1)1And upper right node x_(i-1)2: if x_(i-1)1And x_(i-1)2Are not classified, then x_i1The class order of (2) is that the current total class number is added with 1; if x_(i-1)1Has been classified into_(i-1)2Not classified, then x_i1Using x_(i-1)1The class order of (2); if x_(i-1)2Has been classified into x_(i-1)1Not classified, then x_i1Using x_(i-1)2The class order of (2); if x_(i-1)1And x_(i-1)2All have been classified and have the same class order, then x_i1Using x_(i-1)1And x_(i-1)2The class order of (2); if x_(i-1)1And x_(i-1)2All have been classified and have different class order, then x_i1Using x_(i-1)1And x is_(i-1)1And x_(i-1)2Is marked as an equivalent pair;

b) access node x_ij(j 2.., n-1), if it is an obstacle node, its left node x is sequentially detected_i(j-1)Upper left node x_(i-1)(j-1)Upper node x_(i-1)jAnd upper right node x_(i-1)(j+1): if none of the four points are obstacle-free nodes, then x_ijThe class order of (2) is that the current total class number is added with 1; if the four points have only one obstacle node, x_ijFollowing the class order of the obstacle point; if there are 2-4 obstacle nodes in the four points, x_ijUsing the class sequence of the point detected as the barrier node from the first four points, and recording the point pairs with different class sequences from the four points as an equivalent pair;

c) access node x_inIf it is a barrier node, then the left node x is detected in turn_i(j-1)Upper left node x_(i-1)(j-1)And an upper node x_(i-1)j: if none of the three points are obstacle-free nodes, then x_ijThe class order of (1) is the current total class number plus 1; if the three points have only one obstacle node, then x_ijFollowing the class order of the barrier node; if there are 2-3 obstacle nodes in the three points, then x_ijFollow the first of the three pointsDetecting the class order of the points which are the barrier nodes, and recording the points with different class orders as an equivalent pair;

d) unifying the class order of the equivalent categories according to the obtained equivalent pair; all the barrier points with the same class sequence form a barrier area on the binary map;

and then, converting the identified obstacle area into a polygon, wherein the specific method comprises the following steps:

find the maximum and minimum in the abscissa and ordinate of all points on the obstacle area: x_max，X_max，Y_maxAnd Y_minIn the order of (X)_min,Y_min)，(X_min,Y_max)，(X_max,Y_min)，(X_max,Y_max) Constructing a rectangular area covering the whole barrier area for the vertex; then, with a certain accuracy acc, the rectangular area is divided into a × b rectangular sub-areas of the same shape, where a ≈ (Y)_max-Y_min)/acc，b≈(X_max-X_max) A/acc; traversing all the sub-regions, and if no barrier node exists on the rectangular sub-region, discarding the sub-region;

and finally, connecting the outer vertexes in the reserved sub-regions to obtain an approximate polygon of the original obstacle.

3. The three-dimensional flight path planning method based on the free space and the A-x algorithm as claimed in claim 1, characterized in that: in step 3, the specific method for converting the multi-connected concave polygon into the single-connected concave polygon is as follows:

when point M₁M₂…M_nSelecting a vertex O of an obstacle area when the obstacle area exists in the enclosed concave polygonal area_iWith a certain vertex M of the concave polygonal area_iConnecting line, is marked as O_iM_i(ii) a Its vector has two directions

And

let two vectors

And

a distance delta D → 0 between the two points, and the top point O of the obstacle_iAnd the vertex M of the map_iBy connecting two vectors, the map becomes a single connected domain map M₁M₂…M_iO_iO_i+1…O_nO₁…O_iM_i…M_n；

If there are other obstacles in the concave polygonal area, each obstacle needs to be directly or indirectly connected to the map boundary vertex.

4. The three-dimensional flight path planning method based on the free space and the A-x algorithm as claimed in claim 1, characterized in that: in step 3, the specific method for decomposing the single connected concave polygon into a plurality of sub convex polygons comprises the following steps:

a. judging the positive and negative of the single-connection concave polygon, and if the polygon is a negative polygon, collecting vertexes thereof { P }_nB, reversely ordering the vertexes, and if the vertexes are regular polygons, entering the step b;

b. from { P_nChoose a pit P_kSearching all visible points of the point to form a visible point set { V }_k}；

c. If { V_kHaving pits in it, P is selected among the pits_kIf { V } is the optimal visibility point_kThere is no pit in { V }, then_kSelect P in_kThe optimal visibility point of (a);

d. connection P_kDecomposing the map with the optimal visible point, and adding a new sub-polygon;

e. repeating steps b) to d) until P_nThere are no pits in the polygon, then all the sub-polygons of the polygon are convex polygons.

5. The three-dimensional flight path planning method based on the free space and the A-x algorithm as claimed in claim 1, characterized in that: in step 4, the method for searching the optimal area channel by the a-star algorithm comprises the following steps:

(1) starting point X_startNode S of the region_startAdding a sequence to be retrieved Open List;

(2) if the Open List is empty, no feasible flight path exists, and the algorithm is ended; otherwise, performing the step (3);

(3) traversing an Open List, and searching a node S with the minimum F value_curentMove it into Close List, move it out of Open List; f is from the starting point through the current point S_curentEstimated track cost to endpoint;

(4) for node S_curentEach of which is not in a Close List_neighbour: if node S_neighbourIf not, it is moved into the Open List, and the node S is moved_curentSet as node S_neighbourThe parent node of (1) updating its G value and F value; the G value is the actual track cost from the starting point to the current point; if node S_neighbourIn Open List, determine to go through node S_curentTo node S_neighbourThe resulting node S_neighbourWhether the G value of (a) is less than the original G value; if yes, the node S is connected_curentSet as node S_neighbourThe parent node of (1) updating its G value and F value;

(5) repeating the steps (2) to (4) until the corresponding node of the area where the end point is located is in the Open List;

(6) starting from the node corresponding to the area where the end point is located, each node S_iIn turn with its parent node S_{i_parent}Connect to form the optimal area channel { S_i}。

Images