CN110487279A - A kind of paths planning method based on improvement A* algorithm - Google Patents

Abstract

The invention discloses a path planning method based on the improved A* algorithm. In view of the fact that there are many turning points in the path obtained by the A* algorithm, which causes the path to be not optimal, firstly, the key turning points in the path are merged on the basis of the A* algorithm, which not only reduces the turning points in the path, but also reduces the three-time sample rate. The number of interpolation points is improved to improve the interpolation efficiency, and then according to the path nodes after the inflection points are merged, the cubic spline interpolation is used to achieve the purpose of smoothing the path. Through the improved algorithm, the path length is shorter, the overall smoother, and more in line with the non-holonomic robot motion.

Description

A Path Planning Method Based on Improved A* Algorithm

technical field

The invention belongs to the field of intelligent robots, in particular to a path planning method based on an improved A* algorithm.

Background technique

With the continuous development of robot technology, more and more robots are used to provide various services and even replace human work. Autonomous navigation is the key for robots to realize intelligence and play a role, and path planning is an important component of robot automatic navigation capabilities. Path planning has always been a hot research topic in the field of intelligent mobile robots. Path planning means that in the prior map, the mobile robot can automatically plan a collision-free path from the starting point to the ending point according to the surrounding environment and using sensors to obtain information.

There are many classifications of path planning algorithms. According to whether the external environment information is known, it can be divided into global path planning algorithm and local path planning algorithm; and according to the search method of the algorithm, it can also be divided into blind search algorithm and heuristic search algorithm. Blind search focuses on the search process rather than the search target, and is often accompanied by a huge search space, resulting in the consumption of a large amount of memory resources and low efficiency. Specifically, there are breadth-first algorithms, depth-first algorithms, and Dijkstra algorithms. Heuristic search In the process of searching, according to the heuristic information related to the problem, the search is carried out in a favorable direction, which can avoid many meaningless search paths, greatly reduce the search scope, and reduce the complexity of the problem. The common one is greedy Algorithm and A* Algorithm. As a widely used heuristic search algorithm, the A* algorithm takes into account the advantages of the Dijkstra algorithm and the greedy algorithm. It can not only guarantee to find an optimal path, but also make the search direction clearer, so that the search space is smaller. , Search faster.

Contents of the invention

Purpose of the invention: When the traditional A* algorithm expands the adjacent nodes during the search process, it only expands one layer outward with the current node as the center, that is, the 8 nodes adjacent to the current node. At this time, the angle of the robot's motion direction will be limited. Integer multiples of 45 degrees, the direction of action is limited. If the traditional A* algorithm is expanded to 8 neighbor nodes, the final path obtained by robot path planning in the actual environment may not be optimal. In view of the above problems, the present invention proposes a path planning method based on the improved A* algorithm. The path length planned by the improved algorithm is shorter, the path is smoother, and the path search efficiency is higher.

Technical scheme: in order to realize the purpose of the present invention, the technical scheme adopted in the present invention is: a kind of path planning method based on improved A* algorithm, comprises the following steps:

Step 1: Express the environment of the robot as a grid map, and use the A* algorithm to search for an initial path from the starting point to the target point in the grid map;

Step 2: Extract all the nodes in the initial path, start from the starting point, use the area method to judge whether the three adjacent nodes are collinear in turn, so as to find all the turning points on the path, that is, the inflection points;

Step 3: Starting from the starting point, use a straight line to connect the inflection point to the target point in turn to obtain the updated path; record each straight line as L(i, j), i, j represent the starting point and end point of the straight line segment, i=0, 1, 2, ..., n-1, j=1, 2, ..., n, where n is the number of inflection points;

Step 4: By judging whether the straight line segment in the path updated in step 3 passes through obstacles, redundant inflection points are eliminated, and the path is updated;

Step 5: For the path after removing redundant inflection points in step 4, smoothing is performed by cubic spline interpolation to obtain the final planned path.

Further, in step 4, since the environment where the robot is located is represented as a grid map, when searching whether there is an obstacle between any two nodes in the path with a fixed step in a fixed direction (horizontal or vertical), it may be The situation of missing detection of obstacles occurs; the present invention judges whether the straight line segment passes through the obstacle by dynamically selecting the search direction, and removes the redundant inflection point between i and j when the straight line segment L(i, j) does not cross the obstacle.

Further, the method of dynamically selecting the search direction is used to determine whether the straight line segment passes through obstacles and remove redundant inflection points. The specific process is as follows:

4-1. Select any two nodes in the updated path in step 3 as the search starting point A and the target point B. The coordinates of the two points A and B are the center of the grid map where they are located, and they are respectively recorded as (x ₁ , y ₁ ) and (x ₂ , y ₂ );

4-2, calculate the straight line equation y=kx+b of AB, k is the slope of the straight line, and b is the ordinate of the intersection point of the straight line and the vertical axis;

4-3. Determine the size of abs(y ₂ -y ₁ ) and abs(x ₂ -x ₁ ) of two points AB, if abs(y ₂ -y ₁ )>abs(x ₂ -x ₁ ), use horizontal search , go to step 4-4; otherwise, use vertical search and go to step 4-5;

4-4. Search whether there is an obstacle between the two nodes of AB with a fixed step in the horizontal direction; if there is an obstacle, the inflection point between the two nodes of AB cannot be removed, and the corresponding path cannot be updated; if there is no obstacle, Eliminate the inflection point between the two nodes of AB, and update the corresponding path;

4-5. Search whether there is an obstacle between the two nodes of AB with a fixed step length in the longitudinal direction; if there is an obstacle, the inflection point between the two nodes of AB cannot be removed, and the corresponding path cannot be updated; if there is no obstacle, Eliminate the inflection point between the two nodes of AB, and update the corresponding path;

4-6. Traverse all nodes in the path according to the steps 4-1 to 4-5, remove redundant inflection points, and update the path.

Further, the step 4-4 searches whether there is an obstacle between the two nodes AB with a fixed step in the horizontal direction, specifically as follows:

4-4-a, record all intersections of the longitudinal grid lines and line segment AB on the grid map as a(1), a(2),..., a(m), m is the number of intersections; Use the straight line equation of AB to calculate the ordinate of all the intersection points, and through the ordinate, you can analyze whether there are obstacles around the intersection point;

4-4-b, determine the intersection point a (i), i=1, 2, ..., m as the number of grids of common intersection points; if intersection point a (i) is the common intersection point of four grids, enter step 4- 4-c; if the intersection point a(i) is the common intersection point of two grids, enter step 4-4-d;

4-4-c, judge whether the slope k of the line segment AB is positive or negative. If the slope is positive, check whether there are obstacles in the three adjacent grids above the line segment AB; if the slope is negative, check whether the three adjacent grids below the line segment AB are There are obstacles;

If there is an obstacle, the inflection point between the two nodes of AB cannot be removed, the corresponding path cannot be updated, and the obstacle search between the two nodes of AB ends; if there is no obstacle, i=i+1, go to step 4-4-b , continue to judge the next intersection point;

4-4-d, judge whether the left and right grid adjacent to the intersection point is an obstacle; if it is an obstacle, the obstacle search between the two nodes AB ends; if it is not an obstacle, i=i+1, go to step 4- 4-b, continue to judge the next intersection point;

4-4-e, after all intersection traversals are completed, the obstacle search between the two nodes AB ends.

Further, the steps 4-5 search whether there is an obstacle between the two nodes AB with a fixed step length in the longitudinal direction, specifically as follows:

4-5-a, record all intersections of the horizontal grid lines and line segment AB on the grid map as b(1), b(2),..., b(m), m is the number of intersections; Use the straight line equation of AB to calculate the abscissa of all intersections, and analyze whether there are obstacles around the intersection through the abscissa;

4-5-b, determine the intersection point b(j), j=1, 2,..., m as the number of grids of common intersection points; if intersection point b(j) is the common intersection point of four grids, enter step 4- 5-c; if the intersection point b(j) is the common intersection point of two grids, go to step 4-5-d;

4-5-c, judge whether the slope k of the line segment AB is positive or negative. If the slope is positive, check whether there are obstacles in the three adjacent grids above the line segment AB; if the slope is negative, check whether the three adjacent grids below the line segment AB are There are obstacles;

If there is an obstacle, the inflection point between the two nodes AB cannot be removed, the corresponding path cannot be updated, and the obstacle search between the two nodes AB ends; if there is no obstacle, i=i+1, go to step 4-5-b , continue to judge the next intersection point;

4-5-d, judge whether the left and right grid adjacent to the intersection point is an obstacle, if it is an obstacle, the obstacle search between the two nodes AB ends; if it is not an obstacle, i=i+1, go to step 4- 5-b, continue to judge the next intersection point;

4-5-e, after all intersections are traversed, the obstacle search between nodes A and B ends.

Further, in the step 5, the path after removing redundant inflection points is smoothed by cubic spline interpolation, and the specific process is as follows:

5-1. Let the path have n+1 data nodes, and the node coordinates are (x ₀ , y ₀ ), (x ₁ , y ₁ ), (x ₂ , y ₂ ),..., (x _n , y _n ); in each subinterval x _i ≤ x ≤ x _i+1 , create a spline difference equation:

g _i (x)＝a _i +b _i (xx _i )+c _i (xx _i ) ² +d _i (xx _i ) ³

Among them, a _i , b _i , c _i , d _i represent the coefficients of the spline curve;

5-2, calculation step size h _i = _xi+1 -xi , _i =0, 1,..., n-1;

5-3. Put the data node and endpoint conditions M ₀ =0, M _n =0 into the following matrix equation to get:

Wherein, M _i , i=0, 1, ..., n represents the quadratic differential value of the spline difference equation;

5-4. Solve the matrix equation to obtain the second differential value M _i of the spline difference equation, i=0, 1, ..., n;

5-5. Calculate the coefficients a _i , b _i , c _i , d _i of the spline curve, the formula is as follows:

a _i =y _i

Wherein, i=0, 1, ..., n-1;

5-6. Solve the coefficients in the difference equation of each segment of spline to get the specific expression of each segment of the curve.

Beneficial effects: Compared with the prior art, the technical solution of the present invention has the following beneficial technical effects:

Compared with the traditional cubic spline interpolation smooth path algorithm, the fusion improved A* algorithm and cubic spline interpolation smooth trajectory planning algorithm proposed by the present invention not only reduces the turning points of the path, but also reduces the number of interpolation points, and also reduces the total number of paths. length. The introduction of cubic spline interpolation makes the overall path smoother, thereby avoiding the rapid acceleration and deceleration of the robot at the turn, making its motion form more coherent and more in line with the dynamic control of the non-holonomic robot.

Description of drawings

Fig. 1 is the overall flowchart of the inventive method;

Fig. 2 is the schematic diagram of some three points of the A* algorithm planning road section of the embodiment of the present invention;

Fig. 3 is a schematic diagram of the process of merging key inflection points according to an embodiment of the present invention;

Fig. 4 is a schematic diagram of horizontal traversal according to an embodiment of the present invention;

Fig. 5 is a schematic diagram of longitudinal traversal according to an embodiment of the present invention;

Fig. 6 is a traditional A* algorithm planning path simulation diagram;

Fig. 7 is a simulation diagram of path planning by the improved A* algorithm according to the embodiment of the present invention.

Detailed ways

The technical solutions of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.

A kind of path planning method based on improved A* algorithm described in the present invention, overall process as shown in Figure 1, comprises the following steps:

Step 1: Express the robot's environment as a grid map, and use the A* algorithm to search for an initial path from the starting point to the target point in the grid map.

Step 2: Extract all the nodes in the initial path, start from the starting point, use the area method to judge whether the three adjacent nodes are collinear in turn, so as to find all the turning points on the path, that is, the inflection points; the details are as follows:

As shown in Figure 2, A, B, and C represent three adjacent nodes in the path planned by the A* algorithm, and the area S _ABC of the triangle ABC is calculated. When S _ABC is not 0, the three points A, B, and C are not Collinear, point B is the inflection point in the path;

The formula for calculating the area S _ABC of triangle ABC is as follows:

In the formula, the coordinates of point A are (x _A , y _A ), the coordinates of point B are (x _B , y _B ), and the coordinates of point C are (x _C , y _C ).

Step 3: Starting from the starting point, use a straight line to connect the inflection point to the target point in turn to obtain the updated path; record each straight line as L(i, j), i, j represent the starting point and end point of the straight line segment, i=0, 1, 2, ..., n-1, j=1, 2, ..., n, where n is the number of inflection points;

Step 4: By judging whether the straight line segment in the path updated in step 3 passes through obstacles, redundant inflection points are eliminated, and the path is updated;

The present invention judges whether the straight line segment passes through the obstacle by dynamically selecting the search direction, and removes the redundant inflection point between i and j when the straight line segment L(i, j) does not pass through the obstacle. The specific process is as follows:

4-1. Select any two nodes in the updated path in step 3 as the search starting point A and the target point B. The coordinates of the two points A and B are the center of the grid map where they are located, and they are respectively recorded as (x ₁ , y ₁ ) and (x ₂ , y ₂ ).

4-2. Calculate the straight line equation of AB y=kx+b, k is the slope of the straight line, and b is the ordinate of the intersection point of the straight line and the vertical axis.

4-3. Determine the size of abs(y ₂ -y ₁ ) and abs(x ₂ -x ₁ ) of two points AB, if abs(y ₂ -y ₁ )>abs(x ₂ -x ₁ ), use horizontal search , go to step 4-4; otherwise, use vertical search and go to step 4-5.

4-4. Search whether there is an obstacle between the two nodes of AB with a fixed step in the horizontal direction; as shown in Figure 4, the black grid indicates the obstacle, the white grid indicates the passable area, and the straight line indicates any Inflection point of the line segment, the black point represents the intersection of the line and the grid;

If there is an obstacle, the inflection point between the two nodes of AB cannot be removed, and the corresponding path cannot be updated; if there is no obstacle, the inflection point between the two nodes of AB is removed, and the corresponding path is updated;

The method of searching whether there is an obstacle between the two nodes of AB with a fixed step in the horizontal direction is as follows:

4-4-a, record all intersections of the longitudinal grid lines and line segment AB on the grid map as a(1), a(2),..., a(m), m is the number of intersections; Use the straight line equation of AB to calculate the ordinate of all the intersection points, and through the ordinate, you can analyze whether there are obstacles around the intersection point;

4-4-b, determine the intersection point a (i), i=1, 2, ..., m as the number of grids of common intersection points; if intersection point a (i) is the common intersection point of four grids, enter step 4- 4-c; if the intersection point a(i) is the common intersection point of two grids, enter step 4-4-d;

4-4-c, judge whether the slope k of the line segment AB is positive or negative. If the slope is positive, check whether there are obstacles in the three adjacent grids above the line segment AB; if the slope is negative, check whether the three adjacent grids below the line segment AB are There are obstacles;

If there is an obstacle, the inflection point between the two nodes of AB cannot be removed, the corresponding path cannot be updated, and the obstacle search between the two nodes of AB ends; if there is no obstacle, i=i+1, go to step 4-4-b , continue to judge the next intersection point;

4-4-d, judge whether the left and right grid adjacent to the intersection point is an obstacle, if it is an obstacle, the obstacle search between the two nodes AB ends; if it is not an obstacle, i=i+1, go to step 4- 4-b, continue to judge the next intersection point;

4-4-e, after all intersection traversals are completed, the obstacle search between the two nodes AB ends.

4-5. Search whether there is an obstacle between the two nodes of AB with a fixed step length in the longitudinal direction, as shown in Figure 5; if there is an obstacle, the inflection point between the two nodes of AB cannot be removed, and the corresponding path cannot be updated ; If there is no obstacle, remove the inflection point between the two nodes of AB, and update the corresponding path;

The method of searching whether there is an obstacle between the two nodes of AB with a fixed step length in the longitudinal direction is as follows:

4-5-a, record all intersections of the horizontal grid lines and line segment AB on the grid map as b(1), b(2),..., b(m), m is the number of intersections; Use the straight line equation of AB to calculate the abscissa of all intersections, and analyze whether there are obstacles around the intersection through the abscissa;

4-5-b, determine the intersection point b(j), j=1, 2,..., m as the number of grids of common intersection points; if intersection point b(j) is the common intersection point of four grids, enter step 4- 5-c; if the intersection point b(j) is the common intersection point of two grids, go to step 4-5-d;

4-5-c, judge whether the slope k of the line segment AB is positive or negative. If the slope is positive, check whether there are obstacles in the three adjacent grids above the line segment AB; if the slope is negative, check whether the three adjacent grids below the line segment AB are There are obstacles;

If there is an obstacle, the inflection point between the two nodes AB cannot be removed, the corresponding path cannot be updated, and the obstacle search between the two nodes AB ends; if there is no obstacle, i=i+1, go to step 4-5-b , continue to judge the next intersection point;

4-5-d, judge whether the left and right grid adjacent to the intersection point is an obstacle, if it is an obstacle, the obstacle search between the two nodes AB ends; if it is not an obstacle, i=i+1, go to step 4- 5-b, continue to judge the next intersection point;

4-5-e, after all intersections are traversed, the obstacle search between nodes A and B ends.

4-6. Traverse all nodes in the path according to the steps 4-1 to 4-5, remove redundant inflection points, and update the path.

As shown in Figure 3, where A, B, C, D, E represent the extracted adjacent path inflection points. Starting from point A, there is no obstacle when connecting AC and AD, then the two redundant inflection points of B and C can be eliminated, and the update path is A-D-E. There are obstacles in the middle of connecting AE, so point D cannot be removed, and the path cannot be updated. Then take point D as the starting point, connect it with the following inflection points in turn to judge whether the path can be updated until the end point ends.

Step 5: For the path after removing redundant inflection points in step 4, smoothing is performed by cubic spline interpolation to obtain the final planned path. The specific process is as follows:

5-1. Let the path have n+1 data nodes, and the node coordinates are (x ₀ , y ₀ ), (x ₁ , y ₁ ), (x ₂ , y ₂ ),..., (x _n , y _n ); in each subinterval x _i ≤ x ≤ x _i+1 , create a spline difference equation:

g _i (x)＝a _i +b _i (xx _i )+c _i (xx _i ) ² +d _i (xx _i ) ³

Among them, a _i , b _i , c _i , d _i represent the coefficients of the spline curve;

5-2, calculation step size h _i = _xi+1 -xi , _i =0, 1,..., n-1;

5-3. Put the data node and endpoint conditions M ₀ =0, M _n =0 into the following matrix equation to get:

Wherein, M _i , i=0, 1, ..., n represents the quadratic differential value of the spline difference equation;

5-4. Solve the matrix equation to obtain the second differential value M _i of the spline difference equation, i=0, 1, ..., n;

5-5. Calculate the coefficients a _i , b _i , c _i , d _i of the spline curve, the formula is as follows:

a _i =y _i

Wherein, i=0, 1, ..., n-1;

5-6. Solve the coefficients in the difference equation of each segment of spline to get the specific expression of each segment of the curve.

Fig. 6 is a planned path diagram of the traditional A* algorithm, and Fig. 7 is a planned path diagram after the A* algorithm is improved and smoothed according to the embodiment of the present invention, S represents the starting point, and E represents the target point. It can be seen from Figure 7 that after merging key inflection points, the number of inflection points in the path is significantly reduced. In addition to the redundant inflection points of the traditional A* algorithm, the total length of the path is also reduced, and the path gradually becomes smoother. However, there are many peaks at the turn of the path, and then the cubic spline interpolation is introduced to make the path smooth enough, so that the speed and acceleration of the robot at the turn remain continuous.

Claims (6)

1. A path planning method based on the improved A* algorithm, characterized in that: the method may further comprise the steps: Step 1: Express the environment of the robot as a grid map, and use the A* algorithm to search for an initial path from the starting point to the target point in the grid map; Step 2: Extract all the nodes in the initial path, start from the starting point, use the area method to judge whether the three adjacent nodes are collinear in turn, so as to find all the turning points on the path, that is, the inflection points; Step 3: Starting from the starting point, use a straight line to connect the inflection point to the target point in order to obtain the updated path; record each straight line as L(i, j), i, j represent the starting point and end point of the straight line segment, i=0, 1,2,...,n-1,j=1,2,...,n, where n is the number of inflection points; Step 4: By judging whether the straight line segment in the path updated in step 3 passes through obstacles, redundant inflection points are eliminated, and the path is updated; Step 5: For the path after removing redundant inflection points in step 4, smoothing is performed by cubic spline interpolation to obtain the final planned path. 2. a kind of path planning method based on improved A* algorithm according to claim 1, it is characterized in that: in described step 4, judge whether straight line segment crosses obstacle by the method for dynamically selecting search direction, when straight line segment L( When i, j) does not cross obstacles, remove redundant inflection points between i and j. 3. a kind of path planning method based on improved A* algorithm according to claim 2, it is characterized in that: judge whether straight line segment crosses obstacle by the method for dynamically selecting search direction, remove redundant inflection point, concrete process is as follows: 4-1. Select any two nodes in the updated path in step 3 as the search starting point A and the target point B. The coordinates of the two points A and B are the center of the grid map where they are located, and they are respectively recorded as (x ₁ , y ₁ ) and (x ₂ , y ₂ ); 4-2, calculate the straight line equation y=kx+b of AB, k is the slope of the straight line, and b is the ordinate of the intersection point of the straight line and the vertical axis; 4-3. Determine the size of abs(y ₂ -y ₁ ) and abs(x ₂ -x ₁ ) of two points AB, if abs(y ₂ -y ₁ )>abs(x ₂ -x ₁ ), use horizontal search , go to step 4-4; otherwise, use vertical search and go to step 4-5; 4-4. Search whether there is an obstacle between the two nodes of AB with a fixed step in the horizontal direction; if there is an obstacle, the inflection point between the two nodes of AB will not be removed, and the corresponding path will not be updated; if there is no obstacle, Eliminate the inflection point between the two nodes of AB, and update the corresponding path; 4-5. Search whether there is an obstacle between the two nodes of AB with a fixed step length in the longitudinal direction; if there is an obstacle, the inflection point between the two nodes of AB will not be removed, and the corresponding path will not be updated; if there is no obstacle, Eliminate the inflection point between the two nodes of AB, and update the corresponding path; 4-6. Traverse all nodes in the path according to the steps 4-1 to 4-5, remove redundant inflection points, and update the path. 4. a kind of path planning method based on improved A* algorithm according to claim 3, it is characterized in that: described step 4-4 searches whether there is obstacle between two nodes of AB with fixed step length in lateral direction ,details as follows: 4-4-a, record all the intersection points of the longitudinal grid lines and the line segment AB on the grid map as a(1), a(2),...,a(m), m is the number of intersection points; Use the straight line equation of AB to calculate the ordinate of all intersection points, and analyze whether there are obstacles around the intersection point through ordinate analysis; 4-4-b, determine the number of grids that use the intersection point a(i), i=1, 2,..., m as the common intersection point; if the intersection point a(i) is the common intersection point of four grids, enter step 4- 4-c; if the intersection point a(i) is the common intersection point of two grids, enter step 4-4-d; 4-4-c, judge whether the slope k of the line segment AB is positive or negative. If the slope is positive, check whether there are obstacles in the three adjacent grids above the line segment AB; if the slope is negative, check whether the three adjacent grids below the line segment AB are There are obstacles; If there is an obstacle, the inflection point between the two nodes of AB will not be removed, the corresponding path will not be updated, and the obstacle search between the two nodes of AB will end; if there is no obstacle, i=i+1, go to step 4-4-b , continue to judge the next intersection point; 4-4-d, judge whether the left and right grid adjacent to the intersection point is an obstacle; if it is an obstacle, the obstacle search between the two nodes AB ends; if it is not an obstacle, i=i+1, go to step 4- 4-b, continue to judge the next intersection point; 4-4-e, after all intersection traversals are completed, the obstacle search between the two nodes AB ends. 5. a kind of path planning method based on improved A* algorithm according to claim 3, it is characterized in that: said step 4-5 searches whether there is obstacle between two nodes of AB with fixed step length in longitudinal direction ,details as follows: 4-5-a, mark all the intersection points of the horizontal grid lines and the line segment AB on the grid map as b(1), b(2),...,b(m), m is the number of intersection points; Use the straight line equation of AB to calculate the abscissa of all intersections, and analyze whether there are obstacles around the intersection through abscissa analysis; 4-5-b, determine the number of grids that use the intersection point b(j), j=1, 2,..., m as the common intersection point; if the intersection point b(j) is the common intersection point of the four grids, go to step 4- 5-c; if the intersection point b(j) is the common intersection point of two grids, go to step 4-5-d; 4-5-c, judge whether the slope k of the line segment AB is positive or negative. If the slope is positive, check whether there are obstacles in the three adjacent grids above the line segment AB; if the slope is negative, check whether the three adjacent grids below the line segment AB are There are obstacles; If there is an obstacle, the inflection point between the two nodes of AB will not be removed, the corresponding path will not be updated, and the obstacle search between the two nodes of AB will end; if there is no obstacle, i=i+1, go to step 4-5-b , continue to judge the next intersection point; 4-5-d, judge whether the left and right grid adjacent to the intersection point is an obstacle, if it is an obstacle, the obstacle search between the two nodes AB ends; if it is not an obstacle, i=i+1, go to step 4- 5-b, continue to judge the next intersection point; 4-5-e, after all intersections are traversed, the obstacle search between nodes A and B ends. 6. A kind of path planning method based on the improved A* algorithm according to any one of claims 1-5, characterized in that: in the step 5, the path after removing redundant inflection points is carried out by cubic spline interpolation Smoothing, the specific process is as follows: 5-1. Let the path have n+1 data nodes, and the node coordinates are (x ₀ ,y ₀ ),(x ₁ ,y ₁ ),(x ₂ ,y ₂ ),...,(x _n , y _n ); in each subinterval x _i ≤ x ≤ x _i+1 , create a spline difference equation: g _i (x)＝a _i +b _i (xx _i )+c _i (xx _i ) ² +d _i (xx _i ) ³ Among them, a _i , b _i , c _i , d _i represent the coefficients of the spline curve; 5-2, calculation step size h _i = _xi+1 -xi , _i =0,1,...,n-1; 5-3. Put the data node and endpoint conditions M ₀ =0, M _n =0 into the following matrix equation to get: Wherein, M _i , i=0,1,...,n represents the quadratic differential value of the spline difference equation; 5-4. Solve the matrix equation to obtain the second differential value M _i of the spline difference equation, i=0,1,...,n; 5-5. Calculate the coefficients a _i , b _i , c _i , d _i of the spline curve, the formula is as follows: a _i =y _i Among them, i=0,1,...,n-1; 5-6. Solve the coefficients in the difference equation of each segment of spline to get the specific expression of each segment of the curve.